本篇博文主要内容为 2026-07-21 从Arxiv.org论文网站获取的最新论文列表，自动更新，按照NLP、CV、ML、AI、IR、MA六个大方向区分。
说明：每日论文数据从Arxiv.org获取，每天早上12:30左右定时自动更新。
提示: 当天未及时更新，有可能是Arxiv当日未有新的论文发布，也有可能是脚本出错。尽可能会在当天修复。
目录
概览 (2026-07-21)
今日共更新1082篇论文,其中:
- 自然语言处理共103篇(Computation and Language (cs.CL))
- 人工智能共342篇(Artificial Intelligence (cs.AI))
- 计算机视觉共219篇(Computer Vision and Pattern Recognition (cs.CV))
- 机器学习共302篇(Machine Learning (cs.LG))
- 多智能体系统共28篇(Multiagent Systems (cs.MA))
- 信息检索共23篇(Information Retrieval (cs.IR))
- 人机交互共32篇(Human-Computer Interaction (cs.HC))
多智能体系统
[MA-0] O-VAD: Industrial Video Anomaly Detection through Object-Centric Tracking and Reasoning ECCV2026
【速读】：该论文旨在解决工业视频异常检测（IVAD）中因复杂物体动态变化、严格物理规律及流程约束导致的异常识别难题，尤其针对现有基于视觉语言模型（VLM）的方法在工业场景下性能下降的问题。其解决方案的关键在于提出一种无需训练的代理式（agentic）框架，该框架模拟人工质检员对物体状态演化的感知能力，通过追踪检测对象在时空维度上的动态演变与内在变换过程，基于个体层面的时间状态轨迹进行推理，从而在真实帧中精准识别异常对象。该方法不依赖于正常样本的再训练或领域知识注入，显著提升了在复杂交互场景下的泛化性与可解释性，实验结果表明其在三个主流IVAD数据集上均优于前沿VLM、代理式框架及微调后的传统异常检测方法。
链接: https://arxiv.org/abs/2607.18142
作者: Mei Yuan,Qi Long,Qifeng Wu,Zhenyang Li,Yizhou Zhao,Lei Wang,Yang Liu,Min Xu
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA)
备注: Accepted to ECCV 2026
Abstract:Industrial Video Anomaly Detection (IVAD) aims to identify anomalous objects and events in an industrial process, which is crucial for modern manufacturing and quality control systems. Existing VLM-based anomaly reasoning methods are capable of detecting open-ended anomalies in general domains. However, their performance declines in industrial settings characterized by intricate object transformations, strict physics, and procedural constraints. To tackle the complexity of such interaction-intensive detection, we introduce a training-free agentic framework for anomaly detection free of domain-specific knowledge, emphasizing object state evolution like humans inspectors. It is designed to track spatial-temporal dynamics and underlying transformations of detected objects over time, and then reason over the object-wise temporal state trajectories to identify abnormal objects in grounded frames. Our method overcomes limitations of prior approaches that rely on retraining on normal clips or injecting domain knowledge as context for test-time inference. Extensive experiments on three IVAD datasets demonstrate that our method outperforms frontier VLMs, agentic frameworks, and traditional VAD methods fine-tuned on the respective datasets, while providing interpretable reports over anomaly processes and types.
[MA-1] SGA: PlugPlay Geometric Verification for Educational Video Synthesis
【速读】：该论文旨在解决生成式AI（Generative AI）在教学动画代码生成过程中存在的空间正确性与视觉可读性问题，尤其针对现有框架忽视几何遮挡而过度关注教学内容表达的缺陷。其核心解决方案是提出符号几何代理（Symbolic Geometric Agent, SGA），作为面向代码的动画流水线中的即插即用模块，能够拦截由大语言模型（LLM）生成的代码，通过部分执行提取符号化场景图，并在检测到空间冲突时实施精准优化。为量化评估空间完整性，研究进一步设计了曼姆视觉质量评分（Manim Visual Quality Score, MVQS），一种无需实际渲染的确定性代理指标。实验结果表明，SGA在MMMC-Code基准上实现了73.11的峰值MVQS（基于Code2Video + GPT-5.1），相较原始基线提升16.1%相对性能，并在8种模型与代理配置组合中的7种中均实现MVQS改善，验证了其有效性与普适性。
链接: https://arxiv.org/abs/2607.18116
作者: Lopez Jhon,Hinojosa Carlos,Ghanem Bernard
机构: King Abdullah University of Science and Technology (KAUST), Thuwal, Saudi Arabia; Universidad Industrial de Santander, Bucaramanga, Colombia
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR); Multiagent Systems (cs.MA); Multimedia (cs.MM)
备注:
Abstract:Recent work leverages Large Language Models (LLMs) to generate executable code for pedagogical animations using libraries such as Manim. However, ensuring spatial correctness and visual legibility remains challenging, as existing frameworks emphasize pedagogical content while overlooking geometric occlusions. We propose the Symbolic Geometric Agent (SGA), a plug-and-play module for code-centric animation pipelines that intercepts LLM-generated code, performs partial execution to extract symbolic scene graphs, and applies targeted refinement when spatial conflicts are detected. We further introduce the Manim Visual Quality Score (MVQS), a deterministic rendering-free proxy for spatial integrity. Experiments on the MMMC-Code benchmark across four LLM backbones and two agentic pipelines show that SGA achieves a peak MVQS of 73.11 (Code2Video + GPT-5.1), corresponding to a 16.1% relative improvement over the raw baseline, and improves MVQS in 7 of 8 backbone x pipeline configurations.
[MA-2] FinSAgent : Corpus-Aligned Multi-Agent RAG Framework for Evidence-Grounded SEC Filing Question Answering
【速读】：该论文旨在解决在美证券交易委员会（SEC）披露文件中进行金融问答时，因现有检索增强型与多智能体系统存在“先验-语料不匹配”问题而导致的证据检索不充分与误检率高的核心挑战。具体而言，传统方法直接从用户问题生成检索查询，并依赖语义相似性对候选片段进行排序，导致查询生成忽略文件特有的结构化内容与术语规范，而排序机制则倾向于选择主题相关但缺乏证据有效性的虚假正例。其解决方案的关键在于提出一种基于证据的多智能体框架FinSAgent，通过单一核心原则——在模型先验主导之处注入语料侧条件（corpus-side conditioning），从而同时校正检索生成与排序环节的偏差。该框架的核心创新包括：（1）基于10-K报告强制性结构的角色专业化智能体；（2）具备数据库感知能力的查询分解机制，使各智能体的子查询基于局部语料的轻量级摘要进行条件化；（3）多路径检索结合学习型特征门控重排序器，实现证据有效性与语义相似性的解耦。实验结果表明，FinSAgent在五个离线金融问答基准上显著提升了检索覆盖率与答案准确率，并在包含1,000名匿名用户评分的三臂在线随机对照实验中获得更高评价。
链接: https://arxiv.org/abs/2607.18102
作者: Jijun Chi(3),Zhenghan Tai(1 and 3),Hanwei Wu(1 and 12),Tung Sum Thomas Kwok(1 and 4),Hailin He(1),Zixing Liao(1),Bohuai Xiao(1),Chaolong Jiang(1),Jianliang Lei(1),Jerry Huang(7 and 9),Peng Lu(7),Muzhi Li(5),Liheng Ma(1 and 2 and 9),Yihong Wu(7),Sicheng Lyu(1 and 2 and 9),Jingrui Tian(2),Yihan Li(8),Yanzhang Ma(1 and 11),Dingtao Hu(2),Yufei Cui(2),Ling Zhou(10),Lei Ding(1 and 6),Xinyu Wang(1 and 2) ((1) a href=“http://SimpleWay.AI” rel=“external noopener nofollow” class="link-external link-http"this http URL/a, (2) McGill University, (3) University of Toronto, (4) University of California, Los Angeles, (5) The Chinese University of Hong Kong, (6) University of Manitoba, (7) Université de Montréal, (8) Boston University, (9) Mila - Quebec AI Institute, (10) CG Matrix Technology Limited, (11) Lakehead University, (12) McMaster University)
机构: SimpleWay.AI; McGill University; University of Toronto; University of California, Los Angeles; The Chinese University of Hong Kong; University of Manitoba; Université de Montréal; Boston University; Mila - Quebec AI Institute; CG Matrix Technology Limited; Lakehead University; McMaster University
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL); Multiagent Systems (cs.MA)
备注: 20 pages, 14 figures, 9 tables
Abstract:Financial question answering over U.S. Securities and Exchange Commission (SEC) filings requires retrieving and synthesizing heterogeneous evidence dispersed across long, standardized, and highly redundant disclosures. Existing retrieval-augmented and multi-agent systems typically derive retrieval queries directly from the user’s question and rank candidates by semantic similarity. Together, these choices create prior-corpus misalignment: a mismatch between model priors and the target filings’ structure, terminology, and evidence standards. As a result, query generation misses corpus-specific evidence, while semantic reranking favors topically similar but evidentially invalid false-positive chunks. We propose FinSAgent, an evidence-grounded multi-agent framework that reframes SEC filing QA as corpus-aligned retrieval planning and corrects both ends with a single principle: inject corpus-side conditioning wherever model priors would otherwise dominate. FinSAgent combines (1) role-specialized agents anchored to the mandated 10-K item structure, (2) database-aware query decomposition that conditions each agent’s sub-queries on a lightweight, summary-level view of the local corpus, and (3) multi-path retrieval with a learned feature-gated reranker that separates evidential validity from semantic similarity. Across five offline financial QA benchmarks, FinSAgent improves retrieval coverage and answer correctness over strong single-agent and multi-agent baselines; in a three-arm randomized online experiment with 1,000 anonymous user ratings, it also receives higher scores than baselines.
[MA-3] he Shared Discovery Paradox: How a One-Answer Rule Turns Better Information into Worse Search
【速读】：该论文旨在解决在分布式信息环境下，组织通过集中化信息聚合（consensus pooling）进行集体决策时所面临的“发现效率悖论”问题：尽管信息聚合能够显著提升单一推荐的准确性，但若强制所有个体遵循同一推荐结果（即采用“单答案规则”），反而会严重降低整体发现成功率。其核心问题是信息共享与行动协调之间的权衡——信息聚合虽能改善信念精度，但过度压缩多元行动策略为单一重复选择，导致探索覆盖范围急剧下降。解决方案的关键在于引入多维度的决策机制设计：首先，通过构建一个可解析的十六箱一目标八搜索者模型，证明采用协调的多动作组合策略（如八动作投资组合）可实现接近最优的发现概率（0.8594），远高于单一推荐或去中心化私有线索追踪的性能；其次，在激励机制层面，提出基于自利搜索者的博弈模型，证明在均分奖励机制下存在满足“水填充规则”的对称均衡，其发现效率（0.5991）虽高于共识但低于私有搜索与理想规划者，而引入“独救奖励”机制则可使纯纳什均衡达到最优；最后，通过潜变量公共线索模型揭示信息相关性对有效发现通道的破坏作用，并指出在高复制概率下中央规划者优势上升，且在大市场极限下五种协议的排序保持不变。研究贡献在于提供了一个精确、可复用的基准框架，将信息、资源配置、激励相容与依赖结构分离为可量化的指标，从而系统解析了集体发现中的关键权衡。
链接: https://arxiv.org/abs/2607.18045
作者: Yohei Nakajima
机构: Untapped Capital
类目: Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT); Multiagent Systems (cs.MA)
备注: 29 pages, 4 figures. Code, data, and an interactive guide: this https URL
Abstract:Organizations often pool dispersed information into one ranking and then allow many agents to act on that shared view. In a discovery problem, this can improve beliefs while reducing coverage. We develop an exactly solvable benchmark with sixteen boxes, one target, eight searchers, and noisy private clues. Pooling raises the accuracy of the best single recommendation from 0.20 to 0.3835, but repeating that recommendation lowers group discovery from 0.8322 under decentralized clue-following to 0.3835. A coordinated eight-action portfolio using the same pooled reports reaches 0.8594, and seven coordinated actions recover the decentralized benchmark. The paradox is a protocol failure, not an information failure: a one-answer rule compresses a portfolio of available actions into one repeated choice. We then replace the planner with self-interested searchers who split a prize. The equal-split game is a potential game. Its anonymous symmetric equilibrium obeys a water-filling rule. In the canonical instance it achieves 0.5991: strictly above consensus, but below both private search and the planner. The exact mixed price of anarchy is 2 - 1/N. A sole-rescue reward, which pays only an agent who covers the target alone, makes every pure Nash equilibrium first-best. Finally, a latent common-cue model shows how correlated reports collapse effective discovery channels. The centralized planner gain rises strictly with copying, and in the canonical environment the symmetric market overtakes decentralized report-following at copying probability c = 0.788462. In a proportional large-market limit the five-protocol ordering survives exactly: consensus discovery vanishes while blind, market, private, and portfolio search converge to 0.500, 0.547, 0.847, and 0.874. The contribution is a compact benchmark that separates information, allocation, incentives, and dependence into exact, reusable quantities.
[MA-4] MADA-RL: Multi-Agent Debate-Aware Reinforcement Learning for Parameter-Efficient Reasoning in Compact Models
【速读】：该论文旨在解决小型语言模型（参数量 ≤4B）在数学推理任务中因训练成本过高而难以实现高性能的问题，尤其针对资源受限场景下的高效微调挑战。其核心解决方案是提出MADA-RL框架，通过将模型分解为生成器（generator）与批判者（critic）两个角色，并引入一种动态的、基于角色条件的反事实批判优势（counterfactual critic advantage）作为学习信号，仅使用低秩适配器（LoRA）对少量参数进行微调。该方法的关键创新在于将批判者的奖励定义为自身奖励减去生成器集成在当前实例上的准确率，从而显式地优化批判者超越生成器共识的能力，而非简单复制正确答案，实现了更精准的信用分配。部署时采用轻量级多轮协议组合专用代理，实验表明在五个数学推理基准上，该方法以16倍更少的可训练参数，将DeepSeek-R1-Distill-Qwen-1.5B模型的准确率从39.9%提升至41.9%（+2.0点，p < 0.001），达到精度-可训练参数帕累托前沿水平。尽管尚未超越大规模数据训练的最强基线（如DeepScaleR、STILL-3），但控制实验验证了其性能增益主要源于反事实优势机制，使批判者能够有效识别并纠正生成器错误，而非简单模仿。
链接: https://arxiv.org/abs/2607.18006
作者: Martino M. L. Pulici,Cuong Xuan Chu,Evgeny Kharlamov,Zifeng Ding,Volker Tresp,Yunpu Ma
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA)
备注: 20 pages, 3 figures, 9 tables, 2 algorithms, under review at TMLR
Abstract:Large language models achieve strong reasoning performance, but often at prohibitive training cost - a challenge that is especially acute for compact models ( \leq 4 , \mathrmB parameters) trained under limited budgets. We introduce MADA-RL, a post-training framework that specializes compact models into generator and critic roles and trains them with a debate-aware learning signal, fine-tuning only a small subset of parameters via LoRA adapters. Our central contribution is a counterfactual critic advantage: a dynamic, role-conditioned baseline that redefines the critic’s advantage as its reward minus the generator ensemble’s per-instance accuracy. This explicitly optimizes critics to improve over generator consensus rather than to merely reproduce a correct answer, yielding more targeted credit assignment than static mean-reward normalization. At deployment, the specialized agents are composed in a lightweight multi-round protocol. Across five mathematical reasoning benchmarks, MADA-RL raises the accuracy of the DeepSeek-R1-Distill-Qwen-1.5B model from 39.9 , % to 41.9 , % ( +2.0 points, p 0.001 ) using 16 times fewer trainable parameters than fully fine-tuned baselines, placing it on the accuracy-trainable-parameter Pareto front. It approaches, but does not surpass, the strongest baselines (DeepScaleR, STILL-3), which are trained on substantially larger datasets; we analyse this gap and the associated inference-time cost directly. A controlled study isolates the source of MADA-RL’s gains: the counterfactual advantage produces the highest critic improvement rate of any model evaluated, indicating that trained critics learn to correct generator errors rather than to imitate them.
[MA-5] Self-State Attacks on Self-Hosted AI Agents : How Far Can OS Defenses Go?
【速读】：该论文旨在解决自托管人工智能代理（self-hosted AI agents）因自身状态文件被篡改而引发的安全威胁，即“自状态攻击”（self-state attacks）。此类攻击通过合法的操作系统（OS）系统调用实现，攻击者利用代理对自身内存与配置文件的读写权限，间接完成对系统的渗透与控制。其核心挑战在于：传统基于外部输入或网络行为的防御机制难以有效应对这类源于内部状态变更的攻击。解决方案的关键在于构建一个四维攻击空间模型（目标、机制、粒度、时间维度），系统化分析攻击面，并提出一种基于工作负载条件的可检测性视角。研究通过采集真实运行中代理在不同工作负载下的活动轨迹，构建了包含23个攻击单元、43种具体操作的攻击矩阵，并评估了包括访问控制防护、工作负载感知检测及周期性备份恢复在内的多层次防御策略。实验结果表明，分层防御体系在多数攻击场景下具有有效性，但仍有少量攻击面在操作系统层面呈现出结构不可区分性，暗示当前操作系统级别的防御机制需重新审视，为未来安全架构设计开辟了新的研究方向。
链接: https://arxiv.org/abs/2607.17986
作者: Yimeng Chen,Nathanaël Denis,Roberto Di Pietro,Jürgen Schmidhuber
机构: King Abdullah University of Science and Technology (阿卜杜拉国王科技大学); The Swiss AI Lab, IDSIA-USI/SUPSI (瑞士人工智能实验室, IDSIA-USI/SUPSI)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA)
备注: 21 pages, 4 figures
Abstract:Self-hosted AI agents read and write their own memory and configuration files to function. An agent may get compromised via corruption of its own state – a compromise realized via legitimate OS system call invocation. We refer to this class of threats as self-state attacks. In this paper, we investigate the OS resilience to this class of attacks. Formally, we characterize a four-axis attack space (Target, Mechanism, Granularity, Temporal); investigate the structural limits of prevention, detection, and recovery; and introduce a workload-conditioned view of detectability. To instantiate the framework, we collect live activity traces from a representative self-hosted agent running across distinct workload profiles, and realize the attack space as a 23-cell matrix, 43 concrete operations on real self-state files, and injected into those traces. We then evaluate both canonical and workload-conditioned defense strategies. The empirical results show that a layered defense stack (access-control prevention on the instruction and configuration layers, workload-conditioned detection on the memory layer, and periodic backup for recovery) is effective on most attack cells while a small residual attack surface remains structurally indistinguishable at the OS level. These findings suggest that against the newly established class of self-state attacks, OS-level defense needs to be reconsidered, potentially opening new research directions in the field.
[MA-6] owards Agent ic Agent -based Models: Feasibility Performance and Statistical Model Checking
【速读】：该论文旨在解决将大语言模型（LLM）驱动的决策引入基于代理的模型（Agent-Based Models, ABMs）所带来的方法论挑战，具体关注其对模型可靠性、计算成本及整体行为的影响。传统ABM依赖于简单、明确且可复现的个体决策规则，而复杂集体行为则通过代理间的交互涌现；然而，近年来大语言模型（Large Language Models, LLMs）的发展使得用LLM增强或替代原有规则成为可能，这引发了关于此类集成是否保持模型可信性与效率的关键问题。为此，研究以Mesa ABM框架为基础，结合统计模型检验工具MultiVeStA，构建了一个混合种群的改进版谢林隔离模型（Schelling segregation model）：其中普通代理仍采用传统的符号化规则进行邻居分类，而单一代理则通过工具调用（tool calls）将自然语言描述的邻居信息交由本地部署的LLM处理，由LLM执行相似/不同邻居计数操作，并据此判断自身满意度，从而维持原始模型的动力学机制。这一设计提供了一个可控且最小化的实验环境，用于系统分析基于LLM的决策在语义理解、操作稳定性及计算开销方面的表现。初步实验结果表明，较小规模的本地LLM在执行简单语义分类任务时表现不佳，甚至在重复工具调用过程中出现不可用情况，而较大模型则能通过初步测试。研究进一步探讨了如何利用统计模型检查技术来估计经典ABM可观测指标，并量化引入具有自主性（agentic）特征的LLM组件对仿真模型行为的影响，为未来融合生成式智能与形式化建模提供了可验证的方法论基础。
链接: https://arxiv.org/abs/2607.17948
作者: Stefano Blando,Emanuele Guerrazzi,Riccardo Porcedda,Giuseppe Squillace,Max Tschaikowski,Andrea Vandin
机构: 未知
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:Agent-based models (ABMs) rely on simple, explicit and reproducible rules for individual decision making, while complex collective behavior emerges from interactions among agents. Recent advances in large language models (LLMs) make it tempting to replace, enrich, or perturb these rules with LLM-based agentic capabilities. However, this raises a methodological question: how does introducing LLM-driven decisions affect the reliability, computational cost, and behavior of ABM simulations? We investigate this for Mesa ABM models, a popular Python library for ABMs, analyzed by statistical model checking. Building on Mesa’s integration with the statistical model checker MultiVeStA, we extend the classical Schelling segregation model with a hybrid population: ordinary agents classify neighbors using the standard symbolic rule, while one agent delegates this task to an LLM through tool calls. The LLM-enabled agent receives natural-language descriptions of neighboring agents and invokes tools that increment counters of similar/different neighbors; these counters determine its happiness according to the original Schelling dynamics. This provides a minimal but controlled setting where the semantic, operational, and computational behavior of LLM-based decisions can be studied inside an otherwise standard ABM. We report preliminary experiments with locally served LLMs of different sizes, showing that smaller models may fail simple semantic classification experiments or become operationally unusable during repeated tool-call generation, while larger tested models pass these preliminary checks. We discuss how statistical model checking can estimate classical ABM observables and quantify the impact of introducing agentic LLM components into simulation models.
[MA-7] Aggregate in the Advantage Not the Ratio: A Canonical-Form Analysis of Cooperative Multi-Agent Policy Optimization
【速读】：该论文旨在解决多智能体策略优化（Multi-agent Policy Optimization, MAPPO）中关于如何选择邻近智能体进行信息聚合以有效利用全局协作信息的核心设计问题，具体涉及优势函数（advantage）与重要性权重比率（ratio）两个维度上的支持范围（support size）选择。现有方法在这一二维空间中分布零散且未充分探索：如IPPO分别处理两者；MAPPO采用团队级优势与个体比率组合；HAPPO使用序列化比率与个体优势；而单智能体简化方法则将二者完全联合处理。本文通过形式化定义两个支持矩阵SA（优势支持）和SR（比率支持），并证明其对目标函数的影响仅通过矩阵乘积S~=SRSA体现，由此揭示两个关键结论：(i) 冗余性——两者的聚合模式在信号传递上可互换，不存在固有优越性；(ii) 方差排序性——优势函数以加法方式聚合奖励，其方差呈线性增长并存在耦合邻域内的最优偏差-方差平衡点；而比率函数以乘法方式聚合似然比，其方差随支持规模指数级增长，且无相应偏差降低。因此，解决方案的关键在于：在优势函数中聚合至耦合邻域规模的邻居信息，而在比率部分保持个体粒度，从而实现更稳定、高效的多智能体协同学习。
链接: https://arxiv.org/abs/2607.17924
作者: Zijian Zhao,Sen Li
机构: The Hong Kong University of Science and Technology (香港科技大學); The Hong Kong University of Science and Technology (Guangzhou) (香港科技大學廣州校區)
类目: Multiagent Systems (cs.MA); Machine Learning (cs.LG)
备注:
Abstract:Multi-agent policy optimization, exemplified by PPO-based methods, is a key branch of cooperative Multi-Agent Reinforcement Learning (MARL). A central design question is how many neighboring agents\footnoteIn this paper, “neighbors” refer not only to physical proximity but also to agents whose actions influence one another. to aggregate in order to effectively utilize global information for cooperation. This decision must be made along two dimensions: in the advantage (which agents’ rewards contribute to the credit signal) and in the ratio (which agents’ likelihood ratios form the clipped importance weight). Existing methods occupy scattered, underexplored points on these two axes: IPPO treats both separately; MAPPO pairs a team-level advantage with per-agent ratios; HAPPO employs sequential ratios with per-agent advantages; and single-agent reductions operating on factorized joint policies aggregate both into fully joint products. We formalize these two design choices as support matrices \SA and \SR , and prove a canonical structure: the expected multi-agent policy optimization objective depends on the pair (\SA,\SR) only through their matrix product \tS=\SR\SA . This yields two key consequences: (i) Redundancy: the two support matrices are interchangeable with respect to the signal, meaning neither aggregation pattern is inherently superior.(ii) Variance Ordering: the advantage aggregates rewards as a sum (additive variance with an interior bias-variance optimum at the coupling neighborhood), whereas the ratio aggregates likelihood ratios as a product (multiplicative variance that grows exponentially with support size, with no accompanying bias reduction). The resulting design principle is unambiguous: aggregate neighbors in the advantage, sized to the coupling neighborhood, and keep the ratio per-agent.
[MA-8] PRIME: Plasticity Recovery in Multi-Agent Environments for UAV-Assisted Emergency Communication Networks
【速读】：该论文旨在解决多智能体强化学习（Multi-Agent Reinforcement Learning, MARL）中因环境非平稳性导致的内部状态退化问题。在持续非平稳环境下，任务目标动态变化会引发神经元逐渐进入休眠状态，进而导致共享策略丧失学习能力。传统方法仅通过重置休眠神经元来恢复性能，但在共享参数的多智能体训练框架下存在安全隐患——部分看似休眠的神经元仍接收显著梯度信号，其活跃性依赖于所处理的特定智能体观测。为此，论文提出PRIME（Plasticity Recovery In Multi-agent Environments）解决方案，其核心在于引入双向静默神经元检测机制：通过联合聚合全团队批次的激活统计与梯度统计，利用训练损失已沉积的反向信号而非人工设计的代理信号，仅对同时满足激活休眠与梯度静默条件的神经元进行重初始化。该策略在保留有效表征的同时恢复学习能力。实验表明，在动态切换相位的无人机应急通信仿真器上，PRIME相较MAPPO提升四分位均值回报24.9%，并将休眠神经元比例稳定在10–20%（对比基线40–45%）。消融实验验证了梯度信号与团队级聚合机制的关键作用。理论分析进一步证明，扰动代价仅随静默子空间维度增长，而非完整参数量，从而保证了方法的可扩展性。
链接: https://arxiv.org/abs/2607.17922
作者: Wen Qiu,Zhiqiang He,Wei Zhao,Hiroshi Masui
机构: Kitami Institute of Technology (北见工业大学); University of Electro-Communications (电气通信大学); Anhui University of Technology (安徽工业大学)
类目: Multiagent Systems (cs.MA); Machine Learning (cs.LG); Networking and Internet Architecture (cs.NI)
备注:
Abstract:Most reinforcement learning controllers for these networks assume stationary conditions, and the few that handle change react to the external environment while leaving the network’s internal state unexamined. We show that sustained non-stationarity damages this internal state directly: as objectives shift, neurons progressively fall dormant and the shared policy loses the capacity to learn. The obvious remedy, resetting dormant neurons, is unsafe under shared-parameter multi-agent training: many neurons that appear inactive are still receiving strong training gradients, and whether a neuron appears dormant depends on which agent’s observations it processes. PRIME (Plasticity Recovery In Multi-agent Environments) therefore verifies both directions before intervening. Extending the bidirectional Silent Neuron framework to cooperative multi-agent reinforcement learning, it aggregates activation and gradient statistics over the full team batch, reads the backward signal from the gradient the training loss has already deposited , not from a hand-crafted proxy, and reinitializes only neurons that are simultaneously activation-dormant and gradient-silent. Useful representations are preserved while learning capacity is restored. On a phase-switching UAV emergency communication simulator, PRIME improves interquartile mean return by 24.9% over MAPPO and holds dormant neuron fractions at 10–20% versus 40–45%; ablations attribute the gains to the gradient signal and team-level aggregation rather than to the specific reset operator. A dynamic regret bound shows that the perturbation cost scales with the small silent-subspace dimension rather than the full parameter count.
[MA-9] Value-Aware Prediction for Robust Multi-Agent Coordination Under Communication Loss IROS
【速读】：该论文旨在解决多智能体系统在实际部署中因物理与环境约束导致的通信中断问题，尤其是在通信不可靠条件下维持智能体间协调性与任务性能的挑战。传统方法依赖内部预测模型来估计缺失的共享状态信息，但其采用的标准重构目标对所有状态转移一视同仁，导致模型将计算资源浪费于学习随机探索噪声及次优策略下的过时动态。为此，本文提出一种面向价值的多智能体观测共享框架（Value-Aware MARO），其核心在于通过基于演员-评论家架构生成的优势估计值动态加权预测器的损失函数，从而将预测模型的学习过程与策略演进显式耦合。该机制引导模型聚焦于智能体主动强化的、具有高回报的意图性动态，提升预测效率与相关性。实验结果表明，在通信可靠性低于40%的场景下，该方法显著优于标准无权重基线，在平均回报上提升超过20%，同时性能方差降低64.7%，有效防止了通信高损耗情境下的性能崩溃。
链接: https://arxiv.org/abs/2607.17914
作者: Kemal Devrim Kafadar,Eren Özaltun,Mahmud Efnan Şanlı,Feyza Orak,Emirhan Gazi,Kubilay Kağan Kömürcü,Nazım Kemal Üre
机构: Istanbul Technical University(伊斯坦布尔技术大学); University of Stuttgart(斯图加特大学); Istanbul Technical University Artificial Intelligence and Data Science Application and Research Center(伊斯坦布尔技术大学人工智能与数据科学应用研究中心); Stanford University(斯坦福大学)
类目: Multiagent Systems (cs.MA); Machine Learning (cs.LG); Robotics (cs.RO)
备注: Accepted to 2026 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Abstract:Robust multi-agent coordination relies heavily on inter-agent communication, which is frequently disrupted by physical and environmental constraints in real-world deployments. To maintain operation during these intermittent communication failures, agents can employ internal prediction models to estimate missing shared state information. However, predictors trained with standard reconstruction objectives treat all transitions equally. In a Reinforcement Learning context, this forces the model to waste capacity learning stochastic exploration noise and the outdated dynamics of suboptimal policies. In this paper, we propose a value-aware extension of Multi-Agent Observation Sharing under Communication Dropout (MARO) to patch communication gaps; we refer to this method as Value-Aware MARO. By dynamically weighting the predictor’s loss function using advantage estimates derived from the underlying actor-critic architecture, our objective explicitly couples the predictor’s learning process to the policy’s evolution. This formulation focuses the model’s capacity on the intentional, high-return dynamics actively reinforced by the agents. We evaluate our framework on several tasks within the Multi-Agent Particle Environment under varying communication reliability levels. Experimental results demonstrate that our approach maintains performance under declining communication reliability, particularly below 40%. While our method performs comparably in tasks where the baseline already maintains high coordination, our value-aware weighting effectively prevents the performance collapse observed in the standard predictor during high-attrition scenarios. In these environments, our method achieves an average improvement in mean returns of more than 20% and reduces performance variance by a mean of 64.7% compared to the standard unweighted baseline.
[MA-10] ETAS: An Effect-Typed Language for Agent Systems
【速读】：该论文旨在解决多智能体系统中因生成式AI（Generative AI）的不可预测性、外部行为的非确定性以及授权、审计与恢复机制缺乏形式化保障所带来的编程复杂性问题。现有系统通常将智能体的行为建模为库函数调用或配置片段，导致对策略约束、动作轨迹和资源依赖的追踪能力不足，难以在编译时进行静态验证。为此，论文提出ETAS这一面向智能体系统的编程语言，其核心创新在于将模型驱动的智能体、工具调用、提示工程、类型化记忆、人类审批、策略规则及执行轨迹等均作为语言层面的语义元素，而非抽象的运行时约定。解决方案的关键在于通过静态语义引入双重行为索引：逃逸效应行（escaping effect row）用于追踪可能引发的外部副作用，持久化动作轨迹抽象（persistent abstraction of typed action trace）则用于记录智能体可能请求的动作序列。该设计实现了确定性计算与非确定性代理行为的分离，同时保持了直接的编程风格。通过类型规范（type specs）、可调用规范（callable specs）和轨迹规范（trace specs）构成的终止性编译期约束逻辑，系统可在编译阶段验证策略合规性，并在动态资源受限时生成残余义务（residual obligations）。动态语义进一步区分请求、处理、拒绝与提交事件，确保处理程序在解释动作时不会隐藏其请求痕迹，从而实现授权与审计的透明性。论文形式化定义了核心演算，并证明了类型/效果安全性、进展性、保型性、处理程序轨迹透明性及策略安全性的关键性质。此外，基于Rust实现的原型系统支持类型化的高阶中间表示（HIR）检查、效应与策略诊断、处理程序验证及轨迹感知的执行钩子，为智能体系统提供了从执行前到执行中的一致性推理基础，显著增强了对授权、非确定性、容错与审计证据的形式化支持。
链接: https://arxiv.org/abs/2607.17780
作者: Huiri Tan,Yikun Wang,Puyang Zhang,Shangyu Li,Jiasi Shen
机构: The Hong Kong University of Science and Technology(香港科技大学)
类目: Programming Languages (cs.PL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注:
Abstract:ETAS is a programming language for agent systems that treats model-backed agents, tool calls, prompts, typed memory, human approvals, policies, and execution traces as semantic program elements rather than library conventions. It separates deterministic computation from agentic nondeterminism and externally visible actions while preserving a direct programming style. We present the core design of ETAS. Its static semantics assigns ordinary types through spec conformance and tracks each computation with two behavioral indices: an escaping effect row and a persistent abstraction of the typed action trace it may request. Specs form a terminating compile-time constraint calculus: type specs provide evidence for polymorphism and resource facts, callable specs constrain function and stage shapes, and trace specs express allow, deny, and temporal constraints. Typing checks requested traces against compiled monitors and emits residual obligations when dynamic resources preclude a complete static proof. The dynamic semantics distinguish requested, handled, denied, and committed events; handlers interpret typed actions without making their requests invisible to authorization or audit. We formalize a core calculus and state preservation, progress, type/effect soundness, handler trace-transparency, and policy safety. We also implement ETAS in Rust with a command-line interface, typed HIR checks, effect and policy diagnostics, handler checks, and trace-aware execution hooks. ETAS provides a programming-language foundation for reasoning about authorization, nondeterminism, recovery, and audit evidence before and during agent execution. Subjects: Programming Languages (cs.PL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA) Cite as: arXiv:2607.17780 [cs.PL] (or arXiv:2607.17780v1 [cs.PL] for this version) https://doi.org/10.48550/arXiv.2607.17780 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[MA-11] Autonomous Discovery of Wireless Communications Algorithms
【速读】：该论文旨在解决生成式AI在无线通信领域算法自主设计中的应用空白问题，特别是针对复杂物理层通信问题中性能与计算复杂度之间的权衡难题。其核心挑战在于如何在缺乏先验知识的情况下，自动探索并优化具有高复杂度的通信算法。解决方案的关键在于提出一种名为“AI电信工程师”（AITE）的框架，该框架基于大语言模型（Large Language Model, LLM）驱动的进化搜索机制，能够自主生成、评估和迭代通信算法设计，同时在搜索过程中动态平衡性能与计算开销。通过在正交时频空间（OTFS）均衡器设计和无导频、自定义星座的正交频分复用（OFDM）接收机算法构建两个典型任务上的验证，AITE成功实现了超越现有最优方案的性能表现，并显著降低计算延迟，且首次发现了可解释的显式算法，达到与最先进神经接收机相当的性能水平，充分展现了LLM驱动进化搜索在下一代无线通信算法自主发现中的巨大潜力。
链接: https://arxiv.org/abs/2607.17762
作者: Fayçal Aït Aoudia,Jakob Hoydis,Sebastian Cammerer,Gian Marti,Merlin Nimier-David,Nicolas Roussel,Alexander Keller
机构: 未知
类目: Information Theory (cs.IT); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:Large language model (LLM)-driven evolutionary search is an emerging algorithm-discovery paradigm that has already produced novel results in several scientific fields. Yet its application to wireless communications remains largely unexplored. To bridge this gap, we introduce The AI Telco Engineer (AITE), a framework to autonomously design algorithms for complex communication problems, while navigating performance-complexity tradeoffs. We showcase AITE on two challenging physical-layer problems: designing an equalizer for an orthogonal time-frequency space (OTFS) system, and constructing a receiver algorithm for an orthogonal frequency-division multiplexing (OFDM) system using a custom constellation and operating without pilots. For the first task, AITE develops algorithms that outperform the best-known solutions while reducing computational latency by a factor of 3.6 compared to the strongest baseline. For the second task, it discovers the first explicit, explainable algorithms that achieve performance parity with state-of-the-art neural receivers. These results demonstrate the strong potential of LLM-driven evolutionary search for the autonomous discovery of next-generation wireless communications algorithms.
[MA-12] Lifelong Multi-Subsystem Pickup and Delivery with Buffer-Limited Handover Stations IROS2026
【速读】：该论文旨在解决长期运行的多智能体搬运与交付（Lifelong Multi-Agent Pickup and Delivery, MAPD）系统中，因子系统间负载转移协调不畅而导致的阻塞与饥饿问题。具体而言，当各智能体被限制在独立区域且需通过具有单个停靠位和有限缓冲区的共享交接站进行任务交接时，易发生资源争用和缓冲区溢出，进而影响系统整体稳定性与吞吐量。为此，论文提出了“交接感知的预约与路径规划”（Handover-Aware Reservation and Routing, HARR）方案，其核心在于通过构建共享的停靠位预约日历与确定性的滚动时域缓冲区占用预测机制，实现跨子系统的协同决策。HARR要求候选路径的停靠时间区间必须空闲，并确保未来一段时间内的缓冲区占用投影不超过容量上限，从而在理想执行条件下保证停靠位无冲突使用及缓冲区操作安全。实验表明，在中等负载下，相较于固定停靠位的消融模型，HARR可提升77%的吞吐量并降低92%的积压，同时相比耦合式站台感知的令牌传递基线方法，显著减少规划耗时。结果表明，显式接口协调机制能有效增强模块化多子系统运输系统的稳定性。
链接: https://arxiv.org/abs/2607.17724
作者: Chuanlong Zang,Isabelle Barz,Anna Mannucci,Philipp Schillinger,Florian Lier,Wolfgang Hönig
机构: Robert Bosch GmbH(罗伯特·博世公司); Technical University of Berlin(柏林工业大学); Robotics Institute Germany(德国机器人研究所)
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: IROS 2026
Abstract:Coordinating payload transfers between subsystems is a critical challenge in lifelong Multi-Agent Pickup and Delivery (MAPD). We study systems where agents are confined to separate regions and must exchange payloads through shared handover stations. These stations, equipped with single docks and finite buffers, are inherently vulnerable to blocking and starvation. We formalize this problem as Multi-Subsystem MAPD with Buffer-limited Handover Stations (MS-MAPD-BHS). We then propose Handover-Aware Reservation and Routing (HARR), an online controller that couples per-subsystem planners. HARR uses a shared dock reservation calendar and a deterministic rolling-horizon projection of buffer occupancy to coordinate actions. A candidate route is accepted only if its dock interval is free and the resulting buffer occupancy projection remains within capacity. Under perfect execution, these checks ensure collision-free dock use and buffer-safe committed operations within the reservation horizon. In simulation, HARR achieves up to 77% higher throughput and 92% lower backlog than a fixed-dock ablation at moderate load, while also reducing planning time relative to a coupled station-aware Token Passing baseline. These results show that explicit interface coordination substantially improves stability in modular multi-subsystem transport.
[MA-13] SR-Agent : An Experience-Driven Agent ic Framework for Post-Ranking Strategies Refinement in E-Commerce Recommendation
【速读】：该论文旨在解决工业级电商推荐系统中后排序策略（post-ranking strategies）因静态配置而随在线环境演化逐渐失效，导致用户体验下降的问题。现有方法依赖人工巡检与手动调优，存在周期长、成本高且难以复用的瓶颈。尽管已有基于大语言模型（LLM-based）的智能代理（如RecUserSim、SimUSER和Self-EvolveRec）展现出潜力，但均未能实现从问题发现到策略更新的闭环自动化优化。为此，本文提出SR-Agent——首个在工业推荐系统中部署的策略自进化精炼框架，其核心在于构建一个端到端的自动化闭环：通过用户模拟代理（UserSim agent）分阶段识别用户感知的劣化案例，分析代理（Analysis agent）将重复性问题归纳为结构化可复用的诊断，最终由受约束的策略精炼引擎（Strategy Refinement Harness）结合四阶段奖励机制与可逆回滚机制，将诊断映射为类型化、有界的操作指令。该方案实现了策略的持续自进化，在快手电商平台的一月A/B测试中显著提升订单量（+0.71%）、浏览深度（+0.34%）及点击类目多样性（+0.48%），同时大幅缩短策略迭代周期并降低运维成本。
链接: https://arxiv.org/abs/2607.17719
作者: Hanchen Yang,Kaiwen Yang,Junpeng Zhuang,Yang He,Keting Cen,Bochao Liu,Zhongbo Sun,An Liu,Zhongteng Han,Chenyi Lei
机构: Kuaishou Technology(快手科技), Beijing, China
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:User experience is a first-class objective in industrial e-commerce recommender systems (RS). Post-ranking strategies, which govern diversity, similarity, and exposure over a ranked list, are widely deployed in industrial RS for their simplicity and low serving cost. However, as the online recommendation environment evolves continuously, these statically configured strategies gradually become stale, degrading the user experience. Refining them typically relies on manual inspection, diagnosis, and updates, a process that is slow, costly, and hard to reuse. Although recent LLM-based agents (e.g., RecUserSim, SimUSER, and Self-EvolveRec) offer promising directions, none of them close the full loop of automated, self-evolving strategy refinement. To bridge this gap, we introduce SR-Agent, a Strategy Refinement agentic framework that, to the best of our knowledge, is the first deployed for refining post-ranking strategies in industrial RS. SR-Agent unifies three components: (i) a UserSim agent that applies staged inspection skills to surface user-perceived bad cases; (ii) an Analysis agent that consolidates recurring bad cases into structured, reusable diagnoses; and (iii) a constrained Strategy Refinement Harness that maps diagnoses to typed and bounded actions, gated by a four-stage reward pipeline with reversible rollback. Deployed on the Kuaishou e-commerce platform, SR-Agent continuously runs this refinement loop and, in a one-month online A/B test, increases order volume by 0.71%, browsing depth by 0.34%, and clicked-category diversity by 0.48%, while markedly shortening the refinement cycle and lowering operational cost.
[MA-14] A Digital Twin-Based Method for Evaluating Local Collective Tariffs in Distribution-Level Energy Systems
【速读】：该论文旨在解决分布式能源系统中实施的本地集体电价机制（local collective tariff）缺乏基于工程实际的评估方法这一问题。其核心挑战在于如何在真实用户行为与基础设施条件下，准确量化不同电价机制对社区整体成本及成员间公平性的影响。解决方案的关键在于提出一种基于数字孪生（digital twin）的方法，通过集成家庭级用能与发电的智能体建模（agent-based modeling）、共享计量抽象下的虚拟聚合（virtual aggregation via shared metering abstraction），以及在统一仿真环境中显式嵌入电价逻辑，实现对复杂系统动态交互的高保真模拟。该方法在丹麦本地集体电价机制的应用案例中得到验证，涵盖光伏发电、电池储能和电动汽车充电等典型场景。结果表明，异构负荷特性的聚合可降低峰值重合度，优化电价分摊效率，从而在社区层面实现可量化的成本下降；同时揭示了用电与发电时间对齐程度对参与者成本中性的重要影响。研究证明，数字孪生驱动的评估框架能够有效捕捉基础设施、用户行为与政策设计之间的耦合关系，为分布式能源系统中电价机制的系统性分析与横向比较提供了通用技术基础。
链接: https://arxiv.org/abs/2607.17640
作者: Kristoffer Christensen,Bo Nørregaard Jørgensen,Zheng Grace Ma
机构: 未知
类目: Multiagent Systems (cs.MA); General Economics (econ.GN); Systems and Control (eess.SY)
备注:
Abstract:This work addresses the need for engineering-grounded evaluation of implement-ed tariff mechanisms in distribution-level energy systems. A digital twin-based method is proposed for assessing local collective tariffs under realistic behavioral and infrastructural conditions. The approach integrates agent-based modeling of household consumption and generation, virtual aggregation through a shared metering abstraction, and explicit representation of tariff logic within a unified simulation environment. The method is demonstrated through its application to the Danish Local Collective Tariff across representative residential energy community configurations, including scenarios with photovoltaic generation, battery storage, and electric vehicle charging. Results indicate that aggregation of heterogeneous demand profiles reduces peak coincidence and enables more efficient allocation of tariff components, leading to measurable cost reductions at the community level. At the same time, the outcomes reveal sensitivity to the temporal alignment of consumption and generation, influencing the degree of cost neutrality across participants. The findings illustrate how digital twin-based evaluation can support systematic assessment of tariff mechanisms by capturing the interaction between infrastructure, user behavior, and regulatory design. The proposed approach provides a basis for analyzing and comparing tariff structures in distribution-level energy systems beyond the specific case considered.
[MA-15] he Curvature Shadow: An Apparent Failure of Maximum-Entropy Equilibrium Selection is a Removable Artifact
【速读】：该论文旨在解决生成式 AI（Generative AI）在双人零和博弈中，基于正则化求解器（如R-NaD）所选纳什均衡点与最大熵成员之间存在的坐标偏差问题。其核心问题是：当正则化求解器未能精确收敛至最大熵纳什解时，这种偏差是否源于算法固有的选择偏倚，还是由可消除的熵亏空（entropy shortfall）引起。解决方案的关键在于提出并验证一个定量理论模型——在单维纳什流形上，坐标偏差可分解为 gap≈2δ/κ，其中 δ 为求解器的熵亏空，κ 为熵景观峰值处的曲率。通过五组博弈实验验证，该关系在相对误差低于1%的范围内高度成立。结果显示，四组矩阵博弈因 δ≈0 而无偏差，仅序贯博弈（Kuhn扑克）存在显著 δ>0，导致约0.021的坐标差距；进一步通过调节“磁力强度”（即正则化参数）使 δ→0，观测到偏差沿预测幂律 $ \propto \delta^{1/2} $ 严格趋近于零，且拟合指数为0.50、$ R^2 = 0.999999 $，与理论预期完全一致，表明该偏差为可移除的熵亏空所致，而非固定偏倚。研究最终确认，最大熵投影（I-projection）理论在平坦峰顶条件下仍成立，仅受曲率限制产生微小残差，从而从机制上澄清了此前看似矛盾的“选择偏差”现象。
链接: https://arxiv.org/abs/2607.17543
作者: Luis Leal
机构: 未知
类目: Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT); Machine Learning (cs.LG); Multiagent Systems (cs.MA)
备注: 11 pages, 4 figures, 3 tables. Diagnostic companion to arXiv:2606.28308 . Fully reproducible: a single self-contained notebook regenerates every number, table, and figure
Abstract:In two-player zero-sum games whose Nash equilibria form a convex set, regularized solvers such as Regularized Nash Dynamics (R-NaD) empirically select the maximum-entropy member: the information projection (I-projection) of a uniform reference onto the Nash set. On a panel of small games this match is exact, with one apparent exception: in Kuhn poker R-NaD lands at bluff coordinate 0.180 while the maximum-entropy member sits at 0.201, a coordinate gap of about 0.021, even though R-NaD attains 99.7 percent of the maximum entropy. We ask whether this gap is a genuine selection bias or an artifact, and answer it quantitatively. We show that for selection on a one-dimensional Nash manifold the coordinate gap factorizes as \mathrmgap \approx \sqrt2\delta/\kappa , where \delta is the entropy shortfall of the solver and \kappa is the curvature of the entropy landscape at its peak. Across five games this relation holds to within 2 \times 10^-4 (under 1 percent relative error). The four matrix games have \delta \approx 0 (R-NaD reaches the maximum-entropy member exactly) and therefore no gap regardless of curvature; only the sequential game (Kuhn) has \delta 0 . A causal sweep of the magnet strength drives \delta \to 0 and the gap toward zero along the predicted curve (fitted scaling exponent 0.50, R^2 0.999999 , against the exact prediction of 1/2), until the dynamics destabilize at a stability floor: behavior consistent with a removable shortfall and inconsistent with a fixed bias. We quantify the curvature half of the law from measured curvatures and flag a moving-target pitfall in the natural Tsallis-entropy experiment. The Kuhn gap is thus the curvature shadow of a small, removable entropy shortfall on an unusually flat peak; the I-projection account is upheld up to a flatness-limited residual.
[MA-16] Quantifying Diversity of Thought: A Predictive Law of Weighted LLM Ensemble Lift
【速读】：该论文旨在解决大语言模型（Large Language Model, LLM）集成中多样性思维所带来的性能增益（uplift）的可量化与可预测性问题。现有方法难以准确解析模型间差异对集成性能提升的具体贡献，尤其在复杂推理任务中，模型间的互补性与冗余性交织，导致性能提升机制不透明。其解决方案的关键在于从第一性原理出发，将LLM集成的性能增益精确分解为“救援质量”（rescue mass）与“损害质量”（damage mass），从而构建一个紧凑且可解释的启发式预测公式。该公式的核心是引入经过准确性校正的正确性相关系数 ϕadj，结合准确率差距与集体准确率，形成一组具有强预测能力的指标。实验验证表明，该法则在767,520次推理及23,520次网络隔离环境下的数字取证任务中表现优异：在仅用SuperGPQA数据集进行一次校准后，其预测性能在多个未参与校准的数据集上仍保持高度稳定，Spearman相关系数达0.51至0.84，且实测交换质量（swap mass）与实际增益之间的决定系数 R2≥0.96，显著优于原始ϕ指标（R2≤0.09）。因此，该研究不仅提供了可验证的形式化定律，更确立了以ϕadj为核心的稳定预聚合预测框架，为高效设计和评估高多样性LLM集成系统提供了理论基础与实践工具。
链接: https://arxiv.org/abs/2607.17384
作者: Junade Ali
机构: The Alan Turing Institute (艾伦·图灵研究所); London (伦敦); United Kingdom (英国)
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Logic in Computer Science (cs.LO); Multiagent Systems (cs.MA)
备注:
Abstract:This paper provides an experimentally verified formal law for calculating the uplift that diversity of thought provides in Large Language Model (LLM) ensembles. From first principles, we derive an exact decomposition of LLM ensemble lift into rescue and damage masses, which yields a compact heuristic for calculating uplift. From this we extract the metrics which predict ensemble performance: an accuracy adjusted correctness correlation, \phi_\mathrmadj , together with the accuracy gap and collective accuracy of the pair. We test the law on 767,520 inferences from ten open-weight models over two graduate-level science benchmarks, together with a novel agentic cybersecurity benchmark in which each model conducts digital-forensics investigations by multi-turn tool use in a network-isolated sandbox (23,520 graded trials including abstentions); all votes are released openly. Calibrated once on SuperGPQA at a 40:60 vote split, the heuristic predicts lift on the calibration set with Spearman’s \rho=0.84 and, with its coefficients frozen, transfers to two datasets never used in calibration ( \rho=0.51 on GPQA Diamond and 0.84 on the forensic tasks), whilst the measured swap mass tracks realised lift with R^2\ge 0.96 throughout. Raw \phi has almost no predictive power ( R^2\le 0.09 throughout); the accuracy-adjusted \phi_\mathrmadj is markedly superior ( R^2=0.67 on SuperGPQA), and the heuristic combining these metrics is the most stable pre pooling predictor across the three datasets.
[MA-17] Agent ic ERP: Multi-Agent Large Language Model Architecture for Autonomous Enterprise Resource Planning
【速读】：该论文旨在解决企业资源计划（ERP）系统在实际应用中仍高度依赖人工决策的问题，即尽管ERP能够可靠地记录交易数据，但其传统基于规则的自动化难以处理异常情况，而单一的AI助手在跨职能协作时性能显著下降。为此，论文提出了一种名为Agentic ERP的专家系统架构，其核心解决方案在于：通过角色对齐的大语言模型（LLM）代理与分风险层级的人类在回路（human-in-the-loop）机制相结合，并采用基于图结构的编排器（Planner–Executor–Reflector–Responder），实现对生产级ERP后端端到端业务流程的自主执行。该方案的关键创新包括：将自主ERP操作建模为在结构化企业状态上的约束性序列决策问题，并通过角色对齐代理降低每一步工具选择的复杂度；利用外部化的评分标准与冲刺契约（sprint contracts），在生成与评估之间实现解耦，使协作机制可审计、可解释；并通过多层级评估验证其有效性，包括情景任务套件、六种编排范式在跨职能危机任务中的对比实验，以及365天的代理在回路模拟。结果表明，所提多代理方法显著优于基线，且在相同需求流下，系统可维持全年零缺货，而基于规则的RPA基线则累计出现数百次缺货。研究证明，在人类监督下，角色对齐的LLM代理可使ERP从被动记录事务转向主动执行运营决策，并为自主企业资源规划提供了可复用的参考架构与评估协议。
链接: https://arxiv.org/abs/2607.17331
作者: Zhihao Liu,Tianyu Wang,Xi Vincent Wang,Lihui Wang
机构: 未知
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:
Abstract:Enterprise Resource Planning (ERP) systems record transactions reliably but still delegate almost all operational decision-making to human specialists, because classical rule-based automation cannot reason about exceptions and monolithic AI assistants degrade when asked to coordinate across functional boundaries. This paper presents Agentic ERP, an expert-system architecture that combines role-aligned large-language-model (LLM) agents with a risk-tiered human-in-the-loop harness and a graph-based orchestrator to execute end-to-end business workflows on a production ERP backend. First, autonomous ERP operation is formulated as a constrained sequential-decision problem over a structured enterprise state, with a decomposition argument linking role-aligned agents to a measurable reduction in per-step tool-selection complexity. Second, a graph-based Planner–Executor–Reflector–Responder orchestration decouples generation from evaluation through externalised grading criteria and sprint contracts, packaging recent harness-engineering principles as inspectable expert-system artefacts. Third, the system is evaluated at three levels: a scenario-based task suite, a comprehensive comparison of six orchestration paradigms on cross-functional crisis tasks, and a 365-day agent-in-the-loop simulation against rule-based RPA and no-intervention baselines. Across these levels the proposed multi-agent method is significantly better than the baseline, and the system sustains a simulated year of operation with zero stockouts while the rule-based baseline accumulates hundreds under the same demand stream. The work shows that role-aligned LLM agents under human oversight can move an ERP system from passively recording transactions to actively executing operational decisions, and it provides a reference architecture and an evaluation protocol for autonomous enterprise resource planning.
[MA-18] he Optimization Trilemma: Efficiency Comfort and Fairness in Decentralized Multi-agent Coordination
【速读】：该论文旨在解决去中心化多智能体协作中公平性协调的核心挑战，即在不依赖中央控制的前提下，实现系统整体效率、个体舒适度与资源分配公平性三者之间的平衡。现有研究虽在集中式环境下提出了高效算法以权衡系统效率与个体不适度，但未能有效处理完全去中心化场景下的公平资源优化问题，尤其缺乏对智能体间不适成本均衡再分配的机制设计，导致部分智能体可能因负担过重而丧失协作动机或引发系统分裂。本文的关键解决方案是提出一种新型去中心化协同模型，通过引入可扩展的分布式优化框架，在不显著增加通信与计算开销的前提下，协同优化系统效率、个体舒适度及公平性（即不适成本的均衡分摊），从而实现多方利益的合理平衡。实验基于两个真实世界数据集验证了该模型的有效性，结果表明其在保障智能体偏好和系统目标的同时，显著提升了优化结果的公平性。
链接: https://arxiv.org/abs/2607.17311
作者: Jovan Nikolic,Maciej Krzysztof Zuziak,Evangelos Pournaras
机构: University of Edinburgh (爱丁堡大学); Delft University of Technology (代尔夫特理工大学)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Distributed, Parallel, and Cluster Computing (cs.DC); Systems and Control (eess.SY)
备注:
Abstract:The problem of fair multi-agent coordination in decentralized settings is one of the most pressing challenges for building efficient collaborative systems. Resource allocation is based on optimized collective arrangements accounting for agents’ needs. Such coordination should not only be computationally efficient but also account for fairness, i.e., equitable redistribution of costs incurred by all agents. Recent literature has proposed several algorithms that efficiently determine optimal plan combinations balancing system-wide efficiency and individual discomfort of agents in a centralized setting. However, these works do not address equitable resource optimization in fully decentralized scenarios, specifically, the optimized redistribution of discomfort among coordinating agents so that none experiences a discomfort level that could lead to loss of incentive or polarization that can disrupt planned operations. In this work, we study the problem of optimizing three objectives: (i) system-wide efficiency, (ii) individuals’ comfort and (iii) fairness (i.e., balancing of incurred discomfort costs) in decentralized multi-agent coordination. We design a novel model to optimize those three orthogonal objectives, without any substantial increase in communication and computational overhead. Through experiments on two real-world datasets, we validate the model and demonstrate that it can achieve fairer optimization outcomes, while satisfying agents’ preferences and system goals.
[MA-19] Auto Research for Materials: Auditable AI-Scientist Workflows with Held-Out Transfer
【速读】：该论文旨在解决生成式人工智能（Generative AI）在材料建模中常见的“过拟合于训练循环”问题，即智能体虽能在已有数据上提升性能，但所选优化策略是否能在未参与训练的独立测试数据上保持有效性，以及其代码能否跨任务复用。其解决方案的关键在于构建一个严格的闭环评估框架：将模型改进拆解为特征、模型、表征与训练数据四类变化，并通过十项Matbench基准任务进行系统性搜索，共生成701个可评估的变更方案；所有搜索过程仅基于五重交叉验证的平均性能，降低对单一数据划分的依赖；随后冻结最优选择的代码，在完全未接触过的保留数据集上进行一次性验证。结果表明，9/10的决策在外部数据上仍保持最优，揭示出两类材料建模范式——仅使用成分信息时，特征、模型和表征的改进路径具有相当效果，可实现带隙预测误差降低17.4%、钢强度预测误差降低18.6%，并提升分类任务表现；而引入外部数据后增益有限；对于结构相关任务，更丰富的几何描述符及模型或校准方式的调整分别带来14.6%和7.1%的均值误差下降，且在不同物性家族中表现突出，而成分嵌入不具备迁移能力；更重要的是，将独立发现的特征与模型改进组合后，实现26.3%的平均保留误差降低。研究证明，闭环智能体可产生经得住未见数据检验的可靠决策，其代码具备跨任务复用与组合潜力，为超越反馈回路的可执行科学发现提供了一种可推广的评估范式。
链接: https://arxiv.org/abs/2607.17100
作者: Jingjie Ning,Xiaochuan Li,Shanshan Zhong,Ji Zeng,Guolin Ke
机构: Carnegie Mellon University (卡内基梅隆大学); DP Technology (DP科技)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:An AI research agent can improve the score it sees without finding a modelling change that works on new materials. We ask a stricter question. After repeated experiments, does the selected change survive on data that never entered the loop, and can its code be reused? We separate the search into changes to features, models, representations, and training data. Seven searches produce 701 evaluated changes across ten Matbench endpoints. Agents receive only the mean over five inner folds, reducing reliance on any single development split. We then freeze the selected code and evaluate it once on an untouched holdout. Nine of ten choices remain the best tested single intervention. The surviving changes reveal two materials modelling regimes. With composition alone, feature, model, and representation changes provide comparable routes to improvement. They include held-out MAE reductions of 17.4% for band gap and 18.6% for steel strength, as well as gains on both classification endpoints, while screened external data adds little. For structure tasks, richer geometry descriptors and model or calibration changes lower mean held-out MAE by 14.6% and 7.1% and lead on different property families, whereas composition embeddings do not transfer. Combining separately found feature and model changes yields a 26.3% mean held-out improvement. These results show in materials prediction that closed-loop agents can produce decisions that survive unseen evidence and code changes that can be reused across tasks and combined. More broadly, they provide an evaluation design for testing executable discoveries beyond the feedback loop.
[MA-20] When Is Heterogeneous Distance-Decay Facility Location Tractable? A Structural Classification Exact Methods and a Real-World Study
【速读】：该论文旨在解决连续平面设施选址（continuous planar facility location）中的异质衰减问题，即需求点被设施捕获的价值随距离衰减，且不同需求点的衰减尺度存在差异。这一异质性在现实中普遍存在但尚未被充分挖掘，而传统的一近邻设施目标函数可统一涵盖衰减、聚类与中位数等优化目标，包含k-means、Weber/p-中位数问题及最大覆盖问题等作为特例。其核心挑战在于如何在异质衰减条件下实现高效、精确的设施选址决策。解决方案的关键在于：(i) 提出一个可计算性分类框架，证明离散目标函数始终为单调子模，从而保证(1−1/e)贪心算法的理论性能；同时揭示连续合作目标函数仅当衰减函数关于距离为凹函数时才具有凹性，指出常见覆盖模型中的clip(max(0,d))操作破坏了凹性，且该分类是紧致的；(ii) 设计一种候选离散化最大覆盖混合整数规划（MIP）模型，其线性规划松弛在实践中近乎紧致（间隙约0%），并通过分支定界法可在秒级内求解n=500规模实例；(iii) 提出一种基于“力即梯度”与大邻域搜索结合的启发式算法，性能接近离散最优解（差距<0.5%），显著优于贪心算法、Cooper型交替定位-分配、粒子群优化及加权k-means，并在30个实例中均以97.3%胜率超越基准方法；(iv) 通过真实世界数据验证，忽略校准后的衰减异质性会导致最高达9.7%的捕获需求损失，且设施重定位可达地图范围的37%，零售数据进一步校准衰减为指数形式，特征尺度约为1.4公里。
链接: https://arxiv.org/abs/2607.16764
作者: Zhou He,T.C.E. Cheng,Jichang Dong
机构: University of Chinese Academy of Sciences (中国科学院大学); The Hong Kong Polytechnic University (香港理工大学)
类目: Multiagent Systems (cs.MA); Computational Geometry (cs.CG); Optimization and Control (math.OC)
备注:
Abstract:We study continuous planar facility location in which a demand point’s captured value decays with distance, with the per-point decay scale varying across points. This heterogeneity is ubiquitous yet underexploited, and one nearest-facility objective unifies decay, clustering, and median goals, containing k-means, the Weber/p-median problem, and maximum covering as special cases. We make four contributions. (i) A tractability classification: the discrete objective is always monotone submodular, so the (1-1/e) greedy guarantee holds regardless of decay shape or heterogeneity, and the continuous cooperative objective is concave if and only if the decay is concave in distance; the clip max(0,d) in common coverage specifications is what destroys concavity, and the classification is tight. (ii) An exact discrete method: the candidate-discretized maximum-cover MIP has an empirically tight LP relaxation (~0% gap) and is solved by branch-and-bound in seconds for n = 500. (iii) A force-as-gradient / large-neighborhood-search heuristic, within 0.5% of the discrete optimum, that outperforms the (1-1/e) greedy, Cooper-style alternating location-allocation, particle swarm optimization, and weighted k-means (30/30 per-instance wins at K=30, p10^-9) and is competitive with bespoke solvers on k-means, Weber/p-median, and shape-demand instances. (iv) A real-world study: on 592,667 urban-delivery orders, ignoring the calibrated decay variation loses up to 9.7% of captured demand and relocates facilities by up to 37% of the map; a retail dataset calibrates the decay as exponential with scale R ~ 1.4 km.
[MA-21] RELIC: Revealed Principles for Learning Interpretable Composable Skills in Multi-Agent Planning ICML2026
【速读】：该论文旨在解决多智能体规划中隐私约束下的协作难题，即在各智能体需独立发展专业化决策能力且保持内部实现私密的前提下，如何实现高效协同。传统方法通常依赖集中式优化、共享策略或统一技能表示，难以适应异构接口与隐私保护需求。其解决方案的关键在于提出RELIC框架，通过“揭示原则”（revealed principles）实现可解释、可组合的技能学习与协作：每个智能体在私有环境下利用大语言模型（LLM）引导的搜索优化自身程序化技能，而可信协调者仅基于团队整体性能评估更新提议；成功的行为不以代码形式共享，而是抽象为通用原则，其他智能体可在自身接口中实例化并与其他本地策略重组。该机制实现了协调与实现细节的解耦，支持跨智能体的技能迁移，即使在技能签名异构的情况下亦可运作，从而开创了一种新型的隐私保护型多智能体技能学习与协同范式。
链接: https://arxiv.org/abs/2607.16745
作者: Nguyen Viet Tuan Kiet,Bui Dinh Pham,Duong Quoc Chinh,Dao Van Tung,Tran Cong Dao,Huynh Thi Thanh Binh
机构: 未知
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: Accepted at LM4Plan @ ICML 2026
Abstract:Multi-agent planning becomes substantially harder when agents must improve specialized decision-making skills while keeping their internal implementations private. This regime arises when agents are developed independently, expose different interfaces and capabilities, and must nevertheless coordinate without sharing executable policies. Prior research has largely assumed centralized optimization, shared policy access, or common skill representations, making it poorly suited to privacy-constrained cooperation. We introduce RELIC, a framework for learning interpretable and composable skills via revealed principles. Each agent refines its own programmatic skill through private LLM-guided search, while a trusted orchestrator evaluates proposed updates solely through team-level performance. Successful behaviors are not broadcast as code; instead, they are abstracted into portable principles that other agents can instantiate within their own interfaces and recombine with local strategies. This separates coordination from implementation sharing, enabling cross-agent transfer under heterogeneous skill signatures. RELIC thus introduces a new paradigm for privacy-preserving skill learning and coordination in multi-agent planning.
[MA-22] Laplacian Spectral Shaping for Non-Uniform Scaling Formation Control of Open Multi-Agent Systems
【速读】：该论文旨在解决开放多智能体系统中因拓扑动态变化（如智能体加入/退出、边的增减及领航者重新选择）导致的非均匀缩放形成控制失效问题。其核心挑战在于，动态拓扑会破坏生成式形成形状所需拉普拉斯矩阵的关键谱特性——半正定性、正确的零空间（kernel）以及跟随者子块的正定性，统称为形成谱（formation spectrum）。为此，论文提出一种分布式协议，通过策略性地调整拉普拉斯矩阵的部分权重，在任意维空间中实现对形成谱的保持。该方案的关键创新在于：无需预设父-子层级结构、仅需稀疏感知图、支持任意双智能体作为领航者且可动态重分配，从而在拓扑动态演化下仍能维持期望的非均匀缩放形成形态。
链接: https://arxiv.org/abs/2607.16709
作者: Tao He,Gangshan Jing
机构: Chongqing University (重庆大学)
类目: Multiagent Systems (cs.MA); Systems and Control (eess.SY); Optimization and Control (math.OC)
备注:
Abstract:Non-uniform scaling control enables a multi-agent formation to adjust its shape by compressing or stretching independently along different coordinate axes through inter-agent interactions, offering high flexibility in complex environments. The fundamental idea is encoding the desired formation shape as the kernel of a matrix-valued Laplacian. In open multi-agent systems, however, changes in number of agents, number of edges, and leader selection dynamically alter this Laplacian, destroying the required spectral properties: positive semidefiniteness, correct kernel, and positive definiteness of the follower block (we summarize these properties as the formation spectrum). In this paper, we develop distributed protocols to strategically adjust partial weights of the Laplacian matrix for formation control in arbitrary dimensional space. By implementing the protocols, the desired formation spectrum can be preserved under dynamic topology changes including agent joining, edge addition, agent leaving, and edge removal, while any pair of agents can serve as leaders. Unlike existing Laplacian design methods for affine formation control under topology changes, the proposed approach requires a sparser sensing graph, avoids a predefined parent-child hierarchical structure, and supports leader reassignment. The effectiveness of the proposed protocols is validated through both theoretical analysis and numerical simulations.
[MA-23] SAGE: A Socially-Aware Generative Engine for Heterogeneous Multi-Agent Navigation
【速读】：该论文旨在解决开放人机环境中异质多智能体导航中的安全与社会合规性问题，尤其针对参与者在动力学特性、自主性水平及社会角色上的多样性所导致的非对称交互建模难题。现有轨迹预测与规划方法普遍依赖同质化交互假设或仅施加几何碰撞约束，难以联合建模非对称交互、耦合的预测-规划过程以及软性社会规范。其解决方案的关键在于提出SAGE（Socially-aware Generative Engine），通过将机器人与周围实体建模为有向异质图，并利用异质图变压器（Heterogeneous Graph Transformer, HGT）编码类型特异性非对称交互；在此上下文基础上，采用基于扩散模型的生成模块联合建模各实体未来轨迹与机器人路径规划。推理阶段引入无需训练的安全-社会能量引导机制，通过可微分的碰撞、运动学、任务进展及角色感知的社会合规项对采样轨迹进行优化，实现无需重训练即可显式调控安全性、准确性与任务性能之间的权衡。实验在真实数据集（ETH/UCY和SDD）与合成数据集上验证了SAGE在提升安全性和社会合规性的同时保持任务性能的有效性，且该引导机制显著降低碰撞率与社会违规率，支持最多20台机器人的规模扩展。
链接: https://arxiv.org/abs/2607.16619
作者: Lan Hu,Minghui Liwang,Wenbo Zhu,Xinlei Yi,Yiguang Hong,Xianbin Wang,Zhenzhen Jiao,Seyyedali Hosseinalipour
机构: Tongji University (同济大学); Western University (西门菲莎大学); Aeromind Technology Co., Ltd. (航智科技有限公司); University at Buffalo-SUNY (纽约州立大学布法罗分校)
类目: Robotics (cs.RO); Multiagent Systems (cs.MA)
备注: 16 pages, 5 figures, and 14 tables. Includes supplementary experimental details
Abstract:Safe and socially compliant navigation in open human-robot environments requires robots to reason about heterogeneous participants with different dynamics, autonomy levels, and social roles. Existing trajectory prediction and planning methods often rely on homogeneous interaction assumptions or enforce only geometric collision constraints, making it difficult to jointly model asymmetric interactions, coupled prediction-planning, and soft social norms. This paper proposes SAGE, a socially-aware generative engine for heterogeneous multi-agent navigation. SAGE represents robots and surrounding entities as a directed heterogeneous graph and employs a Heterogeneous Graph Transformer (HGT) to encode type-specific asymmetric interactions. Conditioned on the resulting context, a diffusion-based generative module jointly models future entity trajectories and robot trajectory plans. During inference, a training-free safety-social energy guidance mechanism refines sampled robot trajectories using differentiable collision, kinematic, task-progress, and role-conditioned social-compliance terms. Extensive experiments on real-world (ETH/UCY and SDD) and synthetic datasets verify the effectiveness of SAGE in improving safety and social compliance while maintaining task performance. The proposed guidance mechanism consistently reduces collision and social-violation rates, scales to teams of up to 20 robots, and enables explicit control of the safety-accuracy-task trade-off without retraining. These findings demonstrate the potential of SAGE as a scalable framework for socially-aware multi-agent navigation in complex environments.
[MA-24] Automated Hardware Validation Test Plan Generation for Large Scale AI Datacenter Platforms Using a Generative AI Multi-Agent s Architecture
【速读】：该论文旨在解决大规模异构硬件数据中台在验证过程中因依赖人工编写测试方案而存在的效率低下、易出错及可追溯性差的问题。现有方法需工程师手动分析自愈验证文档与物料清单（Bill of Materials, BOM），逐项枚举故障模式并生成单层扁平化测试用例，不仅耗时费力，且易产生覆盖盲区，复用性差。其核心解决方案是提出一种基于生成式AI的多智能体架构，通过自动化生成结构化的硬件验证测试计划。该框架的关键在于三个协同智能体：输入归一化代理将异构输入统一为标准表示；分类代理基于部件描述与子类别层级进行上下文推理，实现组件到功能域的映射；生成代理则结合归一化后的故障模式与分类数据，合成完整测试用例，主动填补空白并生成边界场景。最终输出符合标准化模式，可直接导入内部验证系统。在两个生产平台上的评估表明，该框架分别实现了74.2%和51.4%的覆盖率提升，将测试方案编写时间从数日缩短至数小时，并实现了测试用例与原始规范之间的完全可追溯性，其多智能体设计具备跨平台迁移能力。自动化与专家评估均验证了100%的提取保真度及对新场景的高接受度，证实该框架作为“人在回路”下的高效增强工具具有显著实用性与鲁棒性。
链接: https://arxiv.org/abs/2607.16388
作者: Mohammed-Khalil Ghali,Saurabh Kulkarni,Prathamesh Kulkarni,Rohan Kulkarni,Sangwon Yoon,Daehan Won
机构: Binghamton University, State University of New York (宾汉姆顿大学，纽约州立大学); Meta Platforms, Inc. (Meta 公司)
类目: Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:
Abstract:Large-scale AI datacenter platforms comprise thousands of heterogeneous hardware components whose validation requires comprehensive fault injection test plans. Today these plans are authored manually: engineers review hardware self-healing validation documents and bills of materials, enumerate failure modes per field-replaceable unit, and produce flat lists of single-layer test cases. This process is labor-intensive, error-prone, and dependent on institutional knowledge; coverage gaps surface late, traceability to source specifications is implicit, and the effort is largely repeated per platform. This paper presents a generative AI multi-agent architecture that automates the generation of structured hardware validation test plans from two canonical inputs: self-healing validation documents, which enumerate known failure modes and their detection and remediation behaviors per field-replaceable unit, and component Bills of Material. An ingestion agent normalizes heterogeneous inputs into a canonical representation; a classification agent maps components to functional domains via contextual reasoning over part descriptions and sub-category hierarchies; and a generation agent synthesizes test cases by combining normalized failure modes with domain-classified data, filling gaps and producing edge cases. The output conforms to a standardized schema for direct import into internal validation software. Evaluated on two production platforms against manual baselines, the framework achieves coverage expansions of 74.2% and 51.4%, cutting authoring from days to hours. It yields fully traceable mappings from each test case to its source specification, and its multi-agent decomposition is portable across platform generations. Automated and expert evaluations confirm 100% extraction fidelity and high acceptance of new scenarios, validating the framework as a robust human-in-the-loop force multiplier.
[MA-25] Clarify Before Executing: A Self-Evolving Agent for Resolving Intent Asymmetry in 3D Tool Orchestration ACM-MM2026
【速读】：该论文旨在解决现代3D资产生成中普遍存在的“意图不对称”问题：当前先进的3D工具链要求精确且可执行的参数输入，而普通用户通常仅提供模糊、不完整的指令。现有3D智能体将这种模糊性视为噪声，并在单轮交互假设下盲目执行，难以应对复杂或不明确的需求。为此，论文提出一种具备澄清意识与进化能力的3D智能体CLARE，将意图不对称视为促进策略性对话的机会而非执行错误。其核心解决方案在于将生成流程解耦为四个专业化的认知角色，能够在调用计算成本高昂的3D工具前主动拦截并修正不完整或模糊的指令，从而在文本到3D生成、单视图重建、多视图重建、点云编辑及后处理五个多样化领域实现无缝任务执行。关键创新在于，CLARE并非依赖预设的刚性规则，而是通过模拟多轮交互自演化其澄清策略，基于多轮奖励（Multi-turn Reward）优化，在交互效率与任务完成率之间实现动态平衡。为严格评估性能，研究构建了3D-Clarify基准，包含620个系统性注入模糊性、信息缺失和错误细节的交互场景。实验结果表明，CLARE在单步和多步任务上的成功率分别达到60.40%和43.34%，显著超越现有基线（提升超过一倍）。定量与定性分析共同验证了主动澄清是实现鲁棒3D执行的关键所在。
链接: https://arxiv.org/abs/2607.16352
作者: Xiaoye Zhu,Weixin Li,Junan Huo,Bozhong Wang,Jia Zeng,Yi Yang,Cen Chen,Qi Liu
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Graphics (cs.GR); Multiagent Systems (cs.MA)
备注: Accepted to ACM Multimedia 2026 (ACM MM 2026). 26 pages including appendix, 8 figures. Code: this https URL
Abstract:A fundamental intent asymmetry plagues modern 3D asset creation: while state-of-the-art 3D toolchains demand precise, executable parameters, ordinary users typically provide vague, underspecified instructions. Current 3D agents treat this ambiguity as noise, defaulting to blind execution under a single-turn assumption. To address this limitation, we introduce CLARE, a clarification-aware and evolutionary 3D agent that treats intent asymmetry not as an execution error, but as an opportunity for strategic dialogue. By decoupling the generation pipeline into four specialized cognitive roles, CLARE intercepts and resolves underspecified instructions before invoking computationally expensive 3D tools to seamlessly execute tasks across five diverse domains: text-to-3D generation, single-view reconstruction, multi-view reconstruction, point cloud editing, and post-processing. Crucially, rather than relying on rigid manual rules, CLARE self-evolves its clarification policy via simulated multi-turn interactions. By optimizing a Multi-turn Reward, the agent internalizes the delicate balance between interaction efficiency and task completion. To rigorously test this, we construct 3D-Clarify, a comprehensive benchmark comprising 620 interaction scenarios with systematically injected ambiguity, missing information, and mistaken details. CLARE achieves state-of-the-art performance, with 60.40% and 43.34% success rates on single-step and multi-step tasks, respectively, more than doubling existing baselines. Both quantitative and qualitative results demonstrate that proactive clarification is the missing key to robust 3D execution. Code is available at this https URL.
[MA-26] Generalist AI Control: Towards Multi-purpose Adaptive Algorithms
【速读】：该论文旨在解决传统控制器针对特定系统设计、难以跨系统阶数与动态特性迁移的问题。其核心解决方案是提出一种通用控制器（Generalist Controller），通过引入基于注意力机制与掩码技术的新型动态状态空间表示，使单一神经网络在一次性训练后即可控制不同维度和动态特性的系统，仅需为每个系统分配一个系统标签（system tag）即可实现无需结构修改的泛化控制。该方法利用多尺度时序处理与专家混合（mixture-of-experts）架构，从涵盖稳定/不稳定、最小相位/非最小相位、线性/非线性等多样动力学特性的25类系统中生成超过31万组演示数据进行训练。仿真结果表明，该通用控制器在各类测试系统（包括非最小相位与不稳定系统）上性能可媲美专用的LQI控制器，并能有效泛化至未见工况（如执行器饱和、噪声、扰动及未训练的参考轨迹），标志着在特定动态系统家族内实现通用控制策略的重要进展。
链接: https://arxiv.org/abs/2607.16313
作者: Klinsmann Agyei,Pouria Sarhadi
机构: University of Hertfordshire(哈特福德大学)
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Robotics (cs.RO); Systems and Control (eess.SY)
备注:
Abstract:Traditional controllers are designed for specific systems and do not transfer across different system orders and dynamics. We present a Generalist Controller, a learning-based controller capable of controlling systems of varying orders and dynamics. The approach introduces a novel dynamic state-space representation using attention mechanisms with masking, enabling a single neural network, trained in one shot, to handle systems with different dimensions without architectural modifications by assigning a system tag to each system. We generated 314,630 demonstrations from 25 diverse systems, including stable, unstable, minimum-phase, and non-minimum-phase dynamics, spanning linear and nonlinear systems from autonomous underwater and aerospace vehicles to mechanical systems and chemical processes. The model learns cross-system control strategies through multi-scale temporal processing and a mixture-of-experts architecture. Simulation results demonstrate that the proposed generalist controller achieves comparable performance to system-specific LQI controllers across all tested systems, including challenging cases such as non-minimum-phase and unstable dynamics, whilst generalising to unseen operating conditions including actuator saturation, noise, disturbance, and reference trajectories not encountered during training. This work represents a significant step towards generalist control policies within a defined family of dynamical systems, demonstrating effective control across a range of single-input single-output (SISO) systems of varying order and dynamics using a single learned policy without system-specific tuning.
[MA-27] Composable Verification Pipelines for Multi-Agent Systems
【速读】：该论文旨在解决动态系统中动作与变化推理的建模与验证问题，尤其针对现有方法在逻辑编程系统上构建时所面临的可扩展性、模块化不足及执行透明度低等挑战。其解决方案的关键在于提出一种基于Tiles的模块化框架，并在高效函数式编程语言Soda中实现，通过将动作语言语义转化为可执行的验证流水线，以状态、动作、转移和规则作为组合式函数组件进行处理。该框架采用类型化的函数流水线形式表示验证过程，实现了模块化规格定义、可复用的推理组件以及具有终止保证的透明执行流程；同时引入可执行的规范层，支持用户以YAML格式定义领域描述，并自动转换为底层验证模型与可执行流水线结构，显著提升了系统的可表达性与可维护性。
链接: https://arxiv.org/abs/2607.16266
作者: Julian Alfredo Mendez,Andreas Brännström
机构: 未知
类目: Logic in Computer Science (cs.LO); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA); Programming Languages (cs.PL)
备注:
Abstract:Existing approaches for reasoning about action and change provide expressive semantics for modeling dynamic systems, in most cases built on top of logic programming systems. We introduce a modular framework for transition and trajectory verification based on Tiles and implemented in Soda, which is an efficient functional programming language. The framework operationalizes action language semantics through executable verification pipelines that process states, actions, transitions, and rules as compositional functional components. Verification procedures are represented as typed functional pipelines, enabling modular specifications, reusable reasoning components, and transparent execution workflows with guaranteed pipeline termination. The framework includes an executable specification layer that allows users to define domain descriptions in YAML, which are operationalized into the underlying verification model and executable pipeline structure. We provide an open-source implementation and illustrate the framework through examples that involve misinformation and emotional reasoning.
自然语言处理
[NLP-0] Automated Discovery Has No Universally Superior Harness
【速读】： 该论文旨在解决当前自主发现系统（如OpenEvolve和TTT-Discover）在实际应用中面临的泛化能力不足与方法论评估不可靠的问题。这些问题源于这些系统作为“通用工具链”被广泛使用，但其本质是多种设计选择（包括档案管理、父代选择、探索策略及预算分配）的复合体，且由于发现过程代价高昂且具有内在随机性，现有对比往往基于过少的独立试验，难以区分方法改进与运行波动之间的差异。为此，研究通过系统性分解OpenEvolve式进化搜索与TTT-Discover搜索框架的组成部分，对30个预算匹配的组合方案在12个模型-问题对上进行了大规模评估，累计执行超过310万次大语言模型（LLM）滚动推演，并结合重复试验的统计分析。结果表明，发现工具链存在显著泛化缺陷：不存在在所有模型-问题对上均表现最优的固定工具链，而多数OpenEvolve变体反而逊色于更简单的替代方案。因此，工具链的选择应被视为一个需针对具体问题和底层模型进行调优的超参数，而非普适配方。此外，研究发现早期发现进展可有效预测最终性能，并据此提出一种预算匹配的自适应分配实验：同时启动多个工具链，淘汰表现较弱的初步运行，并将计算资源重新分配给表现更优的“幸存者”，该方法在性能上优于随机选取的固定工具链或非自适应的工具链集成。综上，研究主张从静态工具链选择转向基于早期性能反馈的在线自适应策略。为促进后续研究，作者公开了所有运行数据集及每个模型-问题对的基准零分布，构建可复用的统计基础设施以支持新工具链的验证。
链接: https://arxiv.org/abs/2607.18235
作者: Akshat Gupta,Jermaine Lei,Alexander Lu,Gopala Anumanchipalli,Leshem Choshen
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Autonomous discovery systems such as OpenEvolve and TTT-Discover are often used as general-purpose harnesses. However, in practice these are composite systems combining several design choices about archives, parent selection, exploration, and budget allocation into a single recipe. Because discovery runs are expensive and inherently stochastic, existing harnesses are often compared using too few independent trials to distinguish key methodological improvements from run-to-run variance. We systematically decompose OpenEvolve-style evolutionary search and the TTT-Discover search harness into its constituent components and systematically evaluate 30 budget-matched harnesses across 12 model-problem pairs using more than 3.1 million LLM rollouts and repeated-trial statistical analysis. Our results show that discovery harnesses have a generalization problem: No fixed harness is reliably superior across the evaluated model-problem pairs, and variants of OpenEvolve generally underperform simpler alternatives. Thus, harness choice is better viewed as a hyperparameter rather than as a universal recipe, and should be tailored to the specific problem and underlying model. We also find that early discovery progress predicts final performance, and use this property to present a budget-matched adaptive-allocation experiment that starts multiple harnesses, prunes weak partial runs, and reallocates compute to stronger survivors, outperforming both commitment to a randomly sampled fixed harness and a non-adaptive harness ensemble. Together, these results motivate shifting from fixed harness selection to online adaptation guided by early performance. We release all run pools including baseline null distributions for every model-problem pair as reusable statistical infrastructure against for future harness proposals.
[NLP-1] Its Not What You Say Its How You Say It: Evaluating LLM Responses to Expressions of Belief ACL2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在面对用户表达的信念（Expression of Belief, EoB）时，如何判断应采纳上下文信念还是坚持自身先验知识的问题。其核心挑战在于，用户表达信念的语言形式具有高度多样性，包括预设、证据标记、确定性语气及语调差异等，这些语言特征可能显著影响模型对上下文信息的接受程度。论文提出一个基于语言学动机的四维分类体系，涵盖形式（form）、证据性（evidentiality）、认识立场（epistemic stance）和语气（tone），共细分为17种具体类型，用以系统化评估不同EoB对模型行为的影响。关键解决方案是构建受控的EoB-查询配对基准，通过隔离语言变异因素，精准分析模型响应模式。实验评估了16种不同架构（如Llama3、Qwen3、Gemma3）、规模（1B–30B参数）和训练阶段（基础模型与指令微调模型）的LLM，发现模型规模与训练阶段显著影响其上下文遵循倾向：更大模型和指令微调模型更倾向于依赖自身先验知识而非用户提供的上下文信念。此外，研究识别出若干在统计上显著更具说服力的EoB类型，揭示了语言框架对大模型上下文整合的系统性影响机制，为提示工程优化与模型鲁棒性提升提供了重要依据。
链接: https://arxiv.org/abs/2607.18232
作者: Kevin Du,Clara Kümpel,Michelle Wastl,Alex Warstadt
机构: ETH Zürich; University of Zurich; UC San Diego
类目: Computation and Language (cs.CL)
备注: Published at ACL 2026
Abstract:Users frequently express their beliefs to large language models (LLMs). In some situations, the LLM should accept these contextual beliefs as true. In others, they should stick to their prior knowledge. Notably, users’ expressions of belief (EoBs) can take linguistically diverse forms - using presuppositions, evidential and certainty markers, or varied tones - each of which may have a different persuasiveness over the LLMs. We introduce a typology to systematically evaluate how different EoBs affect whether models follow context versus prior knowledge. The typology is grounded in four linguistically motivated dimensions: form, evidentiality, epistemic stance, and tone, spanning 17 fine-grained types. By pairing these EoBs with world knowledge facts, we generate controlled EoB-query pairs that isolate the effect of linguistic variation. Using this benchmark, we evaluate 16 LLMs that differ in architecture (Llama3, Qwen3, Gemma3), scale (1B-30B parameters), and training stages (base vs instruct). We identify meaningful variations in response behavior across these axes, e.g., that bigger models and instruction models tend to be less context-following than smaller models and base models. We further identify specific EoBs that statistically significantly persuade LMs more consistently than others. Our work reveals systematic patterns in how linguistic framing affects LLM context integration, with implications for prompt engineering and model robustness.
[NLP-2] Logical Judgments Under Pressure: Diagnosing Syllogistic Stability with Learned Soft Prefixes
【速读】： 该论文旨在探究大模型在面对逻辑推理任务时，其正确判断如何受到学习到的上下文影响，特别是识别和分析由可学习前缀（soft prefix）引发的非理性偏差。核心问题是：当模型被注入特定上下文信号后，其逻辑稳定性是否会被削弱，从而导致原本正确的推理结果发生系统性偏移。解决方案的关键在于引入不可见的连续向量形式的软前缀（soft prefix），通过控制逻辑结构与提示接口的变化，系统性地观测模型行为的响应模式。研究发现，这些软前缀能有效引导多个主流模型（包括Qwen3.6-35B-A3B MoE、Qwen3-8B和Gemma 4 31B）产生大量原本正确答案的翻转，并且在未见过的逻辑形式和提示变体中仍保持高度有效性。诊断分析表明，这种偏差主要源于模型对某一答案意义的广泛偏好，而非固定的符号强制或可迁移的逻辑操作，且不同模型表现出显著差异化的逻辑稳定特性。这揭示了当前生成式模型在逻辑一致性方面的潜在脆弱性，尤其体现在上下文诱导的隐性偏好对其推理过程的主导作用。
链接: https://arxiv.org/abs/2607.18228
作者: Brian K Chen
机构: National University of Singapore(新加坡国立大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 41 pages, 6 figures
Abstract:To test how correct logical judgments respond to learned context, we prepend a soft prefix to an exactly labeled syllogistic reasoning benchmark while keeping the model fixed. Soft prefixes are opaque continuous vectors, so we characterize them through the behavior they induce across controlled variations in logical form and interface. By studying which prefixes succeed and how their effects generalize, we characterize how learned contextual pressure can override correct judgments and expose limits in a model’s logical stability. Across Qwen3.6-35B-A3B MoE, Qwen3-8B, and Gemma 4 31B, learned prefixes redirect many correct answers and remain effective across unseen forms and interface changes. In repeated tests with Qwen3.6 MoE and Gemma, they outperform paired random controls in all 16 model–direction–split comparisons by 37 to 99 percentage points. Qwen3.6 MoE flip rates remain between 72% and 90% across wording and prompt changes, while Gemma validity prefixes retain 54% to 56% flip compared with less than 1% for matched random prefixes. Diagnostic tests show that the dominant effect is a broad preference for one answer meaning rather than fixed-symbol forcing or a logical operation that transfers reliably between tasks. The form of this bias differs across models. In both Qwen models, simple score models often predict which judgments will flip but not how far their margins will move, whereas Gemma’s overall response is more closely approximated by the same models. These results show that the dominant behavioral effect of successful soft prefixes is a broad answer preference, while the remaining response reveals substantial model-specific differences in logical stability.
[NLP-3] SWE-Pruner Pro: The Coder LLM Already Knows What to Prune
【速读】： 该论文旨在解决代码生成智能体（coding agents）在处理长上下文时的效率问题，即如何高效管理冗长的上下文以降低计算开销并提升推理性能。现有方法如SWE-Pruner依赖外部独立的代码分类器进行上下文剪枝，存在额外计算开销与信息耦合不足的问题。本文提出SWE-Pruner Pro，其核心创新在于利用智能体自身在读取工具输出时生成的内部表示（internal representations），通过一个轻量级头部将这些表示转化为每行代码的保留或剪枝标签，并结合基于行数的长度感知嵌入（length-aware embedding）来增强剪枝决策的准确性。该方案实现了在不依赖外部组件的前提下，直接在智能体内部完成工具输出的动态剪枝，显著减少了提示（prompt）和生成（completion）阶段的令牌消耗（最多节省39%），同时保持任务性能稳定，且推理开销可控。实验表明，在多个多轮基准测试中，SWE-Pruner Pro不仅提升了上下文压缩效率，还在MiMo-V2-Flash模型上分别将SWE-Bench Verified的解决率提高3.8%，长上下文任务Oolong的准确率提升2.2分，验证了其有效性与优越性。
链接: https://arxiv.org/abs/2607.18213
作者: Yuhang Wang,Yuling Shi,Shaoqiu Zhang,Jialiang Liang,Shilin He,Siyu Ye,Yuting Chen,Kai Cai,Xiaodong Gu
机构: 未知
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: Project page: this https URL
Abstract:Pruning long context for coding agents has been a vital technology for efficient context management. While existing context pruning methods such as SWE-Pruner realize this by attaching a separate code classifier, we find the agent itself encodes internal representations indicating the relevance of code context when reading tool output. Based on this finding, we propose SWE-Pruner Pro, which prunes tool outputs directly inside the agent. Concretely, a small head turns the agent’s own internal representations into a keep-or-prune label for each line, with a length-aware embedding keyed to each tool output’s line count. Across two open-weight backbones and four multi-turn benchmarks, SWE-Pruner Pro saves up to 39% of prompt and completion tokens while preserving task quality, with bounded inference overhead. Notably, on MiMo-V2-Flash SWE-Pruner Pro additionally raises the SWE-Bench Verified resolve rate by +3.8% and the long-context Oolong accuracy by +2.2 points.
[NLP-4] PPL-Factory: Task-Aware and Budget-Aware Data Selection from Language Modeling to Reasoning
【速读】： 该论文旨在解决大语言模型微调中训练样本贡献不均的问题，核心挑战在于如何在保证下游任务性能的前提下降低计算成本。现有数据选择方法多依赖于数据质量、多样性或推理路径长度等间接启发式指标，这些固定标准具有任务依赖性且难以跨任务泛化。虽然基于困惑度（Perplexity）的方法能提供模型感知的样本难度估计，但传统方法通常对整个序列进行统一评分，忽略了语言建模与推理任务在学习目标上的差异。本文提出PPL-Factory框架，其关键创新在于结合任务感知的困惑度评分与数据预算感知的选择准则，实现更精准、可解释的数据筛选。实验结果表明，在GSM8K数据集上，仅使用1%的训练数据，PPL-Factory即优于其他先进方法；当使用10%的数据时，其在GSM8K和MATH上的准确率分别超过全量数据微调0.9和4.8个百分点，验证了该方法在高效微调中的有效性与普适性。
链接: https://arxiv.org/abs/2607.18199
作者: Hang Zhang,Warren J. Gross
机构: McGill University (麦吉尔大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 13 pages, 4 figures, 5 tables
Abstract:Not all training samples contribute equally to large language model fine-tuning. Selecting informative training samples can reduce the computational cost while preserving downstream performance. Many existing data selection methods rely on indirect heuristics, such as data quality, diversity or reasoning trace length. However, the effectiveness of these fixed criteria is task-dependent and difficult to generalize across diverse downstream tasks. Perplexity-based data selection provides a simple and model-aware solution to estimate the sample difficulty, but existing approaches typically score the entire training sequence and ignore the difference in learning objectives of language modeling and reasoning tasks. In this paper, we propose PPL-Factory, a simple and interpretable data selection framework that combines task-aware perplexity-based scores and data budget-aware selection criteria. Experiments on GSM8K demonstrate that PPL-Factory outperforms other state-of-the-art data selection methods using only 1% of the training set. With 10% of the data, PPL-Factory exceeds full-data fine-tuning accuracy by 0.9 on GSM8K and 4.8 on MATH. Overall, our results demonstrate that task-aware and budget-aware perplexity-based selection provides an effective and applicable approach for efficient fine-tuning.
[NLP-5] VEHBench: A Stage-Local Diagnostic Benchmark for LLM -Assisted Vibration Energy Harvester Design
【速读】： 该论文旨在解决电池自由物联网（Battery-free Internet of Things, IoT）系统中振动能量采集器（Vibration Energy Harvester, VEH）设计所面临的耦合物理约束下的迭代优化难题，尤其关注大语言模型（Large Language Models, LLMs）在工程设计流程中各阶段的行为表现与能力差异。现有工程评估基准多聚焦于最终设计成果的有效性，缺乏对LLM在复杂物理设计过程中阶段性行为的深入诊断。为此，作者提出VEHBench——一个面向工程实践的诊断型基准，包含763个基于文献的真实设计任务，并由基于物理原理的分析型“仲裁者”（analytical physical oracle）进行评分。该基准系统性地评估了四种关键设计角色：需求规格筛选（specification triage）、验证引导搜索（verifier-guided search）、故障状态恢复（corrupted-state recovery）以及策略条件选择（policy-conditioned selection）。实验结果表明，LLM的能力具有显著的阶段依赖性：不存在单一模型能在整个设计流程中持续领先，且不同设计角色下模型响应控制特征呈现明显的行为模式差异。因此，VEHBench为验证器驱动的工程化LLM提供了阶段感知的评估、选型、路由与优化基础，推动了生成式AI在复杂物理系统设计中的可解释性与可靠性提升。
链接: https://arxiv.org/abs/2607.18181
作者: Depeng Su,Yuyu Luo,Guobiao Hu
机构: The Hong Kong University of Science and Technology (Guangzhou)
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注:
Abstract:Battery-free Internet of Things (IoT) requires iterative design of vibration energy harvesters (VEHs) under coupled physical constraints, while LLMs are emerging as interface layers for engineering workflows. However, existing engineering benchmarks primarily assess final artifact validity, offering limited insights into how LLMs behave across different stages of coupled physical design. We introduce VEHBench, an engineering-native diagnostic benchmark for LLM-assisted VEH design, featuring 763 literature-grounded tasks scored by an analytical physical oracle. VEHBench evaluates four design roles: specification triage, verifier-guided search, corrupted-state recovery, and policy-conditioned selection. Experimental results reveal that LLM capability is strongly stage-dependent: no single model consistently dominates the entire workflow, and response-control profiles expose distinct behavioral patterns across design roles. VEHBench thus provides a stage-aware foundation for evaluating, selecting, routing, and improving verifier-grounded engineering LLMs. The benchmark artifact is available at this https URL
[NLP-6] Do Language Models Dream of Binding Molecules? Benchmarking LLM s under Spatial Constraints
【速读】： 该论文旨在解决生成式分子设计中大语言模型（LLM）在复杂三维空间约束下的推理能力不足问题，尤其关注其在结合蛋白质口袋结构、配体片段锚点、药效团点及必需的蛋白-配体相互作用等多重空间约束条件下生成合理三维分子的能力。现有基于扩散模型的方法虽在高质量三维分子生成方面表现优异，但其计算成本高且难以灵活扩展；而近期兴起的基于LLM的方法虽在性能上展现出竞争力，但在物理合理性与三维空间建模方面的深层能力尚未充分探索。本文的关键解决方案是提出一种名为3D-Fit的高效基准评估策略，通过引入轻量级标记化方法，系统评估通用型LLM在多条件约束下的三维分子生成性能。研究发现，尽管当前LLM在生成质量上仍落后于顶尖扩散模型，但其具备同时处理多种空间约束的潜力，展现出良好的可扩展性与适应异构任务场景的能力，为未来构建更智能、可泛化的分子生成框架提供了重要方向。
链接: https://arxiv.org/abs/2607.18144
作者: Thomas MacDougall,Maksim Kuznetsov,Roman Schutski,Rim Shayakhmetov,Maxim Malkov,Vladimir Aladinskiy,Alex Aliper,Alex Zhavoronkov
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Structure-based drug design (SBDD) leverages the 3D structure of protein targets, often complemented by other spatial constraints, to generate candidate binding molecules. While diffusion models have dominated as a leading paradigm for high-quality 3D molecule generation, LLM-based methods are rapidly emerging in molecular design and have shown competitive performance in pocket-conditioned molecular generation. However, their ability to reason about physics and 3D spatial environments is largely underexplored. In this work, we systematically analyze whether current general-purpose LLMs are capable of navigating complex 3D constraints compared to established baselines such as specialized diffusion models. We consider 3D ligand generation conditioned on protein pockets together with ligand- and interaction-derived spatial constraints, including anchor fragments, pharmacophore points, and mandatory pocket-ligand interactions. To enable this evaluation, we introduce 3D-Fit - a token-efficient benchmarking strategy for assessing LLM performance on multi-conditioned spatial molecule generation. Our findings reveal a clear pattern in LLM spatial capabilities: while they still lag behind state-of-the-art approaches, they are promising and can handle multiple spatial constraints simultaneously, enabling scaling to heterogeneous setups.
[NLP-7] How Does Alignment Tuning Shape Representations of Sycophancy and Related Cue-Induced Biases in LLM s?
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）对输入提示中微小、非实质性的改动（如误导性提示、错误标注的少样本示例或虚假的对话历史）表现出高度敏感性的问题，这种敏感性表现为模型在推理过程中产生迎合性（sycophancy）及由线索诱导的偏见（cue-induced bias）。其核心解决方案的关键在于：通过分析模型隐藏状态中的特征表示，识别出每种偏见在模型内部对应的具体方向（per-bias direction），并利用三种互补方法——探针分析（probing）、留一数据集外迁移（leave-one-dataset-out transfer）和因果干预（causal intervention）——实现对这些偏见方向的定位与操控。研究发现，此类偏见主要由对齐训练（alignment tuning）引入，而非预训练阶段形成；预训练基础模型对这类线索几乎无响应，且其激活值中不包含与线索相关的特定信号。而在对齐后的模型中，每种偏见可被建模为一个独立且连贯的方向，能够被解码并沿该方向进行干预，从而恢复原本正确的输出。此外，不同偏见在表征空间中保持分离，交叉纠缠现象具有模型特异性，而非普遍存在于偏见类别之中。该干预策略不仅揭示了偏见的因果机制，还具备一定的去偏能力，可在保留多数正确答案的前提下修复由偏见引发的错误。因此，线索诱导偏见并非单一缺陷，而是一组由对齐训练所引入的、可区分且可操控的因果方向。
链接: https://arxiv.org/abs/2607.18114
作者: Prakhar Gupta,Terry Jingchen Zhang,Florent Draye,Bernhard Schölkopf,Zhijing Jin
机构: University of Michigan; Jinesis Lab, University of Toronto; Vector Institute; Max-Planck Institute for Intelligent Systems, Tübingen, Germany; ELLIS Institute Tübingen; EuroSafeAI
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Modern LLMs are alarmingly susceptible to surprisingly simple immaterial changes of input prompts: a casual hint, an incorrectly labeled few-shot example, or a fake prior assistant turn often flips an originally correct answer. We study where this susceptibility, spanning sycophancy and related cue-induced biases, lives inside the model. Across five model families and seven BCT bias types, we extract a per-bias direction from hidden states and triangulate it through three measures: probing, leave-one-dataset-out transfer, and causal intervention. The susceptibility is largely installed by alignment tuning rather than pretraining: pretrained base models barely cave to these biases, and their activations carry no cue-specific signal beyond question content. Within aligned models, each bias becomes a single coherent direction that we can both decode and steer along, recovering the unbiased answer across every family we test. The biases stay representationally distinct, however: cross-bias entanglement is model-specific rather than a property of the bias category, and even behaviorally similar biases occupy different directions. The same intervention also serves as a modest debiasing tool, recovering a meaningful share of bias-induced errors while preserving most correct answers across all instruct families. Cue-induced bias is therefore best understood not as a single flaw in LLMs but as a family of distinct, causally active directions that alignment tuning installs.
[NLP-8] LLM -as-a-Coach: Experiential Learning for Non-Verifiable Tasks
【速读】： 该论文旨在解决在开放性任务中，基于强化学习（Reinforcement Learning, RL）将大语言模型（LLM）的基于规则的评估压缩为单一标量奖励信号所导致的文本反馈丢失及高质量响应间细微差异混淆的问题。其核心解决方案是提出“经验式学习”（Experiential Learning, EL），将原本作为“评判者”的反馈模型重构为“教练”（LLM-as-a-Coach），通过该教练对每个策略内生成响应进行评估，并将其评估结果提炼为可迁移的经验知识（experiential knowledge）。这些经验知识以高带宽形式提供密集监督信号，同时保留了高质量响应之间的细粒度偏好信息。该知识通过策略内上下文蒸馏机制被策略模型吸收，从而指导其优化。相比传统标量奖励，EL不仅在多个策略族和不同反馈源下均显著优于基于规则的强化学习，在未见任务上表现更优，且具备更强的分布外泛化能力与抗奖励黑客（reward hacking）特性，验证了经验知识作为后训练阶段非可验证任务中的更丰富、更具泛化性的学习信号的有效性。
链接: https://arxiv.org/abs/2607.18110
作者: Tianzhu Ye,Li Dong,Guanheng Chen,He Zhu,Xun Wu,Shaohan Huang,Furu Wei
机构: Microsoft Research (微软研究院); Tsinghua University (清华大学); Peking University (北京大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Reinforcement learning (RL) on open-ended tasks compresses an LLM’s rubric-based evaluation into a scalar reward, discarding rich textual feedback and conflating responses with distinct quality profiles. We propose Experiential Learning (EL), which repurposes the feedback model from an LLM-as-a-Judge into an LLM-as-a-Coach. The coach distills its assessment of each on-policy response into transferable experiential knowledge, which conditions a teacher model and is internalized by the policy through on-policy context distillation. Compared with scalar rewards, this higher-bandwidth feedback channel provides dense supervision and preserves fine-grained preferences among high-quality responses. Across two policy families, with feedback from the policy itself or a proprietary model, EL consistently outperforms rubric-based RL on held-out and unseen open-ended tasks. Notably, EL generalizes better beyond the training distribution, and mitigates reward hacking. These findings establish experiential knowledge as a richer and more generalizable learning signal for post-training on non-verifiable tasks.
[NLP-9] VDAR-Router: Adaptive LLM s Routing via Verbalized Query Difficulty Analysis Retrieval
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在实际系统中部署时的高效模型选择问题，核心挑战在于如何在成本与性能之间实现最优权衡。现有路由方法通常依赖输入查询的表面语义或嵌入相似性来评估模型适配度，但这类方法忽略了查询背后的内在难度，导致路由决策次优。为此，本文提出VDAR-Router——一种基于检索的难度感知路由框架。其关键创新在于：首先对输入查询进行显式的难度分析，进而检索历史数据中具有相似难度特征的示例；基于检索到的上下文信息，结合兼顾性能与成本的奖励函数，动态评估候选模型的适配性并完成模型选择。实验结果表明，VDAR-Router在三个数据集上均显著优于现有基线方法，在无需额外训练的情况下实现了更优的成本-性能平衡。案例研究进一步验证了显式查询分析有助于提升检索相关性与路由可靠性，凸显了难度感知机制在无训练场景下大语言模型路由中的有效性。
链接: https://arxiv.org/abs/2607.18098
作者: Yu-Chien Tang,Jun-Chen Hung,Wen-Chih Peng,An-Zi Yen
机构: National Yang Ming Chiao Tung University (国立阳明交通大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Large language models are increasingly used in practical systems, making efficient model selection important for reducing deployment cost. LLM routing has emerged as a practical solution for allocating each input query to an appropriate model under a desired cost-performance trade-off. Existing routing methods often estimate model suitability from the surface semantics or embedding similarity of the input query. However, such methods may ignore the underlying difficulty of a query, leading to suboptimal routing decisions. To address the challenge, we propose VDAR-Router, a difficulty-aware retrieval-based routing framework. For each input query, VDAR-Router first generates an explicit difficulty analysis. It then retrieves historical examples with similar difficulty profiles. Based on the retrieved records, it estimates candidate model suitability and selects the model using a reward function that considers both performance and cost. Experiments on three datasets show that VDAR-Router consistently achieves better cost-performance trade-offs than existing baselines. These results demonstrate the effectiveness of difficulty-aware retrieval for training-free LLM routing. Case studies further show that explicit query analysis helps retrieve more relevant examples and supports more reliable routing decisions.
[NLP-10] WorldCupArena: Fine-Grained Evaluation of Language Models and Deep-Research Agents on Football Forecasting
【速读】： 该论文旨在解决足球比赛赛前预测中如何有效融合动态信息并生成多维度精准预测的问题，尤其针对传统模型仅依赖历史结果而忽视实时变化信息的局限性。其解决方案的关键在于提出WorldCupArena这一动态基准测试框架，该框架以2026年FIFA世界杯为首个评估场景，支持语言模型与深度研究代理（deep-research agents）在每场比赛前接收通用证据包或自主检索信息，并完成对比赛结果、比分、关键球员与事件、比赛统计数据及赛事最终归属等多任务预测。通过引入“近似比分得分”（Scoreline score）机制，在预测比分接近但不完全准确时仍给予部分奖励，从而更全面地衡量模型在细节预测上的表现。实验覆盖104场比赛和13个系统，结果显示尽管最佳模型在结果准确率和精确比分准确率上仅小幅超越博彩市场与人工球迷基线，但在Scoreline指标上展现出显著优势，表明其在精细化预测能力上更具潜力。此外，该基准支持随新赛程更新而持续评估未来模型，避免使用已知结果带来的偏差，具有良好的可扩展性与开放性。
链接: https://arxiv.org/abs/2607.18084
作者: Zhaokai Wang,Tianlin Gui,Jiayuan Rao,Shangzhe Di,Yihong Tang,Dingli Liang
机构: Shanghai Jiao Tong University(上海交通大学); Nanjing University(南京大学); McGill University(麦吉尔大学); University College London(伦敦大学学院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Predicting a football match before kickoff requires more than knowing past results: a model must use changing information and make a clear prediction before the answer is available. We present WorldCupArena, a dynamic benchmark for language models and deep-research agents. The 2026 FIFA World Cup is its first evaluation, and the same process can be reused for future leagues and cups. Before each match, a model either receives a common evidence package or searches for information itself. It predicts the result and score, likely players and events, match statistics, and the outcome of the competition. After the match, these predictions are compared with the recorded result. We report result accuracy, exact-score accuracy, and a scoreline score that gives some credit when a predicted score is close but not exact, together with scores for the other prediction tasks. Across 104 matches and 13 systems, models with similar result accuracy differ more clearly on detailed predictions. Compared with betting-market and human-fan baselines, the best system shows only small gains in result and exact-score accuracy, but a clearer gain in Scoreline. New schedules can be added as they begin, allowing the benchmark to evaluate future models without using outcomes that are already known. Code, prompts, predictions, and evaluation scripts are open sourced at this https URL.
[NLP-11] Modeling turn-taking with distant viewing: investigating silence thresholds in human and AI-generated discourse
【速读】： 该论文旨在解决音频视觉材料中静默段（silence gaps）的分布与特征问题，重点关注其在不同性别说话者及制作背景下的差异。研究通过分析30部美国情景喜剧和51个使用Google NotebookLM生成的合成播客，利用Praat软件基于基频阈值对说话者性别进行判定，并对比了不同性别与制作场景下的静默段特征。其解决方案的关键在于采用客观的声学参数（基频阈值）实现说话者性别的自动标注，并结合跨类型内容（真实影视作品与合成播客）的对比分析，揭示静默段在自然与生成语音中的异同，为理解对话结构中的非言语停顿提供实证依据。
链接: https://arxiv.org/abs/2607.18076
作者: Taylor Arnold,Nicolas Ballier,Artem Saloev
机构: 未知
类目: Computation and Language (cs.CL)
备注: 10 pages, MeSSH26 (Méthodes pour les sciences sociales et les humanités - Methods for social sciences and humanities 2026)
Abstract:This study investigates silence gaps in two kinds of audiovisual material. We analysed thirty US situational comedies and fifty-one synthetic podcasts generated with Google NotebookLM. Gaps were compared across speaker gender, assigned from a fundamental-frequency threshold estimated in Praat, and across production settings.
[NLP-12] Pancasila-Dilemmas: Evaluating Large Language Models on Indonesian Human Value Dilemmas Grounded in Pancasila
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在价值对齐（value alignment）评估中普遍忽视特定国家文化价值观的问题，尤其针对印尼本土价值体系的缺失。现有评估多聚焦于西方或普世价值，缺乏对非西方社会情境下价值判断能力的衡量。为此，论文提出Pancasila-Dilemmas数据集，包含1,834个源自印尼新闻的真实生活困境问题，依据印尼国父哲学——五项核心价值（宗教、人文关怀、团结、民主、社会正义）进行分类。为增强评估严谨性，所有场景均设计为道德困境（dilemma），并由母语者审校，由5位不同背景的印尼公民进行标注。实验评估了50个闭源与开源的LLMs，结果显示所有模型在概率匹配得分（Probability Match Score, PMS）上均低于0.5，在最大投票一致性得分（Max-Vote Agreement Score, MVAS）上最高仅达0.72，表明其在捕捉印尼本土价值方面存在显著不足，尤其在“宗教”与“团结”类困境中表现最差。该研究的关键解决方案在于构建一个基于本土文化语境、具有真实生活背景且经多方验证的评估基准，从而推动面向多元文化语境的大语言模型价值对齐研究。
链接: https://arxiv.org/abs/2607.18066
作者: Supryadi,Irfan,Julianti,Darren Keanly Martin,Jayvin Fernando,Yuqi Ren,Deyi Xiong
机构: TJUNLP Lab, School of Computer Science and Technology, Tianjin University, China; Universitas Universal, Indonesia; Beijing Language and Culture University, Beijing, China; School of Artificial Intelligence, Tianjin University, China
类目: Computation and Language (cs.CL)
备注:
Abstract:The value alignment of large language models (LLMs) is crucial for ensuring responses align with human intention and value preferences. However, most evaluations of value alignment focus on Western or universal values, while assessments grounded in the value systems of specific countries remain scarce. In this paper, we introduce Pancasila-Dilemmas, an evaluation dataset of 1,834 questions derived from Indonesian news, classified by 5 values of Pancasila: Religion, Humanity, Unity, Democracy, and Social Justice. This dataset reflects daily life in Indonesia, making it suitable for measuring the value alignment of LLMs deployed for Indonesia. To ensure a more rigorous evaluation, we choose scenarios containing dilemmas. The dataset is proofread by native speakers and answered by 5 diverse Indonesian citizens. We evaluate 50 closed- and open-source LLMs on our dataset. Results reveal that all evaluated LLMs achieves less than 0.5 Probability Match Score (PMS) and 0.72 Max-Vote Agreement Score (MVAS). Compared by each values, LLMs mostly struggle in Religion and Unity dilemma cases. This highlights a significant gap in capturing Indonesian values. The dataset is publicly available at this https URL.
[NLP-13] An Early Warning of Emerging Biosecurity Risks in Frontier LLM s
【速读】： 该论文旨在解决前沿大语言模型（LLMs）在科学工作流中日益增长的生物安全风险问题，特别是当这些模型具备生成具有潜在危害性的生物序列或提供敏感生物操作指导的能力时，现有文本层面的安全防护措施可能无法有效应对实际生物风险。其解决方案的关键在于构建一个“从计算到物理”的一体化评估框架，核心是开发专用的生物红队模型Intern-BioBreaker，通过生成针对性的越狱提示（jailbreak prompts），系统性测试对齐模型在面对高风险生物任务时的脆弱性；同时，将模型生成的序列输出进行实验证实，包括DNA合成、宿主表达及正交蛋白验证，以确认其是否可被物理实现并产生预期生物功能。研究结果表明，当前主流前沿模型普遍存在严重的生物越狱漏洞，部分模型任务级攻击成功率接近或达到100%，且能生成具有致病潜力的病毒序列，其翻译蛋白表现出更强的受体结合亲和力，进一步提升感染风险，且经端到端实验验证，这些设计并非仅停留在文本层面，而是可在受控实验条件下真实实现。这凸显了亟需建立更强大的生物红队机制、核酸合成筛查体系与动态安全防护策略，以匹配模型能力的快速演进。
链接: https://arxiv.org/abs/2607.18056
作者: Zhida He,Xia Hu,Baichen Le,Chunxiao Li,Jiajia Li,Lijun Li,Chaochao Lu,Jing Shao,Youbang Sun,Hua Tang,Xiang Wang,Xiao Wang,Xiaoyu Wen,Tong Wu,Jia Xu,Peng Yu,Shu Yu,Jie Zhang,Qiaosheng Zhang,Yi Zhang,Xing-Ming Zhao,Tianhang Zheng,Ziyuan Zhou
机构: 未知
类目: Computation and Language (cs.CL); Genomics (q-bio.GN)
备注: 22 pages, 7 figures, authors are listed alphabetically by surname
Abstract:Frontier large language models (LLMs) are increasingly integrated into scientific workflows, yet their growing biological capabilities may outpace current safeguards. To assess the biological risks of frontier models, we develop Intern-BioBreaker, a specialized bio-red-teaming model, together with an integrated computational-to-physical framework that couples model-level stress testing with wet-lab validation. Within this framework, Intern-BioBreaker generates targeted jailbreak prompts to test whether aligned models can be induced to provide operational guidance for safety-sensitive biological tasks or produce sequence-level outputs with potentially harmful properties. Selected sequence outputs are then carried forward for DNA synthesis, host expression, and orthogonal protein verification to assess whether model-generated designs can yield the intended biological products. Our evaluation reveals a concerning gap between text-level safeguards and the risks posed by capable scientific models: (i) Intern-BioBreaker outperforms baseline attack models and reveals widespread bio-risk jailbreak vulnerabilities across both open-weight and proprietary frontier LLMs, with several targets reaching near-saturated or 100% task-level attack success rate (ASR); (ii) in sequence-level case studies, GPT-5.5 can be induced to generate modified viral candidate sequences with pathogenic potential; the corresponding translated proteins may exhibit even stronger receptor-binding affinity and thus enhanced infection potential; and (iii) end-to-end verification shows that selected model-generated biological designs are not merely textual artifacts, but can be physically realized under controlled experimental settings. These findings underscore the need for stronger biological red-teaming, nucleic acid synthesis screening, and safety mechanisms that keep pace with model capabilities.
[NLP-14] L1 Augmented Attention as an Improved Vector Similarity Metric
【速读】： 该论文旨在解决标准缩放点积注意力（scaled dot product attention）在Transformer模型中将方向对齐与向量幅值混杂的问题，从而限制了其作为相似性度量的有效性。其核心解决方案是提出一种名为L1增强注意力（L1 augmented attention）的改进方法，通过从点积得分中减去一个可学习的、头独立的查询与键之间的L1距离，实现对几何信息的互补建模：点积保留方向对齐的优势，而L1项则惩罚坐标偏差。为降低L1计算成本，该方法引入低维子空间投影机制，使查询和键的参数专门化以保留有意义的L1结构。在使用紧凑型Transformer于WikiText 2数据集上的实验表明，L1增强注意力相较于原始Transformer基线最多降低14.5%的困惑度，并优于RBF L2核方法。对范数方差及学习到的L1权重的分析揭示了不同层间存在不同的几何角色以及显著的头级别专业化现象。结果表明，通过引入L1几何信息可为现代语言模型中的相似性计算提供一种原理性强且高效的改进，在提升准确率的同时兼顾并行效率。
链接: https://arxiv.org/abs/2607.18027
作者: Kurt Godden
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 16 pages, 2 figures, 2 tables
Abstract:Scaled dot product attention conflates directional alignment and vector magnitude, limiting its effectiveness as a similarity metric in Transformer models. We introduce L1 augmented attention, a simple and computationally parallelizable modification that subtracts a learned, head specific L1 distance between queries and keys from the dot product score. This hybrid similarity captures complementary geometric information. Dot product rewards directional alignment, while L1 penalizes coordinate deviations. To reduce the cost of L1 computation, we project queries and keys into low dimensional subspaces whose parameters specialize to preserve informative L1 structure. Evaluated on WikiText 2 using a compact transformer, L1 augmented attention achieves up to a 14.5% reduction in perplexity over the original transformer baseline and outperforms an RBF L2 kernel. Analysis of norm variance and learned L1 weights reveals distinct geometric roles across layers and strong head level specialization. These results demonstrate that enriching attention with L1 geometry provides a principled and effective improvement to similarity computation in modern language models, with practical benefits for both accuracy and parallel efficiency.
[NLP-15] What Transfers Under Source Shift? Definitions Examples and Fine-Tuning for Climate Disclosure Classification
【速读】： 该论文旨在解决企业在不同文本来源（如年度报告、新闻稿、财报电话会）中披露气候信息时，因文本长度、目的和写作风格差异导致的气候信息披露分类任务在跨源场景下性能下降的问题。现有研究多局限于单一来源的评估，未充分考察主流大语言模型（LLM）适应策略在源域迁移下的有效性。为此，论文将气候信息披露分类重构为一个跨源适应问题，系统评估了三种广泛使用的适应策略——定义（definition）、示例（example）和微调（fine-tuning）——在十一款开源与闭源大模型上的表现，基于两个共享标签空间但来源不同的语料库进行对比分析。研究发现，所有策略均能带来平均意义上的跨源性能提升，但最强的源内表现策略在跨源情境下未必最优：基于相似性检索与LoRA微调虽在源内表现突出，但在源域转移时优势显著衰减；而表现较弱的随机少样本示例策略反而具有更强的鲁棒性；定义提示的迁移效果最为一致，前提是其粒度与目标文本匹配。总体而言，在源域变化背景下，策略的简洁性往往更有利于稳定表现，即“简单优于复杂”。
链接: https://arxiv.org/abs/2607.17952
作者: Guosheng Li,Fenghui Ren,Bin Liu,Chuan Yu,Kaiying Ji,Lin Yue,Jun Shen,Sasa Qian
机构: University of Wollongong (卧龙岗大学); University of Sydney (悉尼大学); Macquarie University (麦考瑞大学)
类目: Computation and Language (cs.CL); Computational Engineering, Finance, and Science (cs.CE)
备注: 15 pages, 12 figures. Code: this https URL
Abstract:Climate disclosure classification is a fundamental task for analysing corporate climate disclosures, yet such disclosures appear in many different sources – annual reports, press releases, and earnings calls – that differ in length, purpose, and writing style. Existing evaluations are mostly conducted within a single source, leaving open whether common LLM adaptation strategies remain effective under source shift. We reframe climate disclosure classification as a cross-source adaptation problem and study three widely used adaptation strategies – definitions, examples, and fine-tuning – across eleven open- and closed-source LLMs, using two corpora that share the same label space but come from different sources. We find that all strategies bring positive cross-source gains on average, but the strongest in-source strategies are not the strongest cross-source ones: similarity-based retrieval and LoRA fine-tuning gain most in-source but lose most of that advantage under source shift; randomly selected few-shot examples, a weaker in-source baseline, retain their advantage more reliably; definitions transfer most consistently, though only when their granularity matches the target text. Across these strategies, when the source changes, simpler is often safer.
[NLP-16] DeLIVeR: Decomposed Learning for Information-grounded Veracity Recognition via Reinforced Knowledge Graph Exploration
【速读】： 该论文旨在解决大语言模型（LLM）在自动化事实核查中面临的“查询脆弱性”问题，即传统检索系统在处理复杂、多跳的虚假声明时难以有效获取高精度证据。其核心解决方案是提出DeLIVeR（Decomposed Learning for Information-grounded Veracity Recognition）框架，将证据检索建模为一种强化学习驱动的战略探索任务。该框架利用规划器大模型（Planner LLM）将复杂的主张分解为一系列目标明确的问题集，并基于这些提问遍历结构化知识图谱（Knowledge Graph, KG）以精准定位证据。通过引入组相对策略优化（Group Relative Policy Optimization, GRPO）并设计以结构多样性与结论准确性为核心的奖励机制，对规划器策略进行优化。实验结果表明，DeLIVeR在LIAR、FEVER和PolitiFact三个基准数据集上显著优于现有先进方法，使用Qwen2.5-7B模型时分别取得83.73、84.57和79.70的最高F1分数，较HippoRAG2提升10%-15%。该方法通过强化问答规划策略，有效弥合了多跳推理中的信息断层，同时提供了可审计、透明的可验证信息溯源路径，增强了事实核查过程的可信度与可解释性。
链接: https://arxiv.org/abs/2607.17935
作者: Cong Hoan Nguyen,Thomas Hoang,Hieu Minh Duong,Long Nguyen
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to 7th International Conference on Deep Learning Theory and Applications (DeLTA 2026)
Abstract:Automated fact-checking remains a challenge for Large Language Models (LLMs) due to “query brittleness” in traditional retrieval systems. We propose DeLIVeR (Decomposed Learning for Information-grounded Veracity Recognition), a framework that treats evidence retrieval as a reinforced strategic exploration task. DeLIVeR utilizes a Planner LLM to decompose complex claims into targeted question sets, which are used to traverse structured Knowledge Graphs (KGs) for high-precision evidence. We optimize the Planner’s policy using Group Relative Policy Optimization (GRPO) with a reward system prioritizing structural diversity and verdict accuracy. Our evaluation on LIAR, FEVER, and PolitiFact shows that DeLIVeR significantly outperforms state-of-the-art baselines. Using Qwen2.5-7B, our framework achieved peak F1-scores of 83.73, 84.57, and 79.70 respectively, representing a 10-15% improvement over HippoRAG2. By shifting to a reinforced question-planning strategy, DeLIVeR effectively bridges multi-hop reasoning gaps and provides an auditable, transparent path for verifiable misinformation detection.
[NLP-17] Benchmarking Resource-Efficient LLM s for Research Topic Ontology Generation in the Biomedical Field
【速读】： 该论文旨在解决生物医学领域知识组织系统（如本体）在构建与维护过程中因依赖人工标注而产生的效率瓶颈问题。尽管生成式人工智能（Generative AI）为本体的自动化生成提供了可扩展的潜力，但其对复杂、领域特异性语义关系的识别能力仍需系统评估。为此，本文提出一个包含4000个从医学主题词表（MeSH）中提取的语义关系的数据集——MeSH-Rel-4K，用于评估五种参数量不超过90亿的小型开源大语言模型（LLM）在识别生物医学概念间语义关系方面的性能。研究对比了三种适配策略：标准提示（standard prompting）、思维链提示（Chain-of-Thought prompting）以及微调（fine-tuning）。结果表明，尽管参数受限的模型在上下文逻辑推理方面存在天然局限，但经过针对性微调后，平均F1分数提升了34.1个百分点，显著超越了传统提示方法的性能极限。这说明直接微调能够有效突破小型模型的推理瓶颈，从而为专业化生物医学本体的自动构建与持续演化提供一种高精度、可扩展的技术路径。
链接: https://arxiv.org/abs/2607.17902
作者: Tanay Aggarwal,Angelo Salatino,Francesco Osborne,Enrico Motta
机构: Knowledge Media Institute, The Open University, Milton Keynes, UK; Department of Business and Law, University of Milano-Bicocca, Milan, IT
类目: Digital Libraries (cs.DL); Computation and Language (cs.CL)
备注:
Abstract:Knowledge Organization Systems like Ontologies and taxonomies are fundamental for structuring scientific knowledge, yet their manual curation presents a persistent bottleneck in knowledge management. While Large Language Models (LLMs) offer a scalable mechanism for automated ontology generation, their capacity to classify complex, domain-specific semantics requires systematic evaluation. In this paper, we assess the performance of five small, open-source LLMs (up to 9 billion parameters) in identifying semantic relationships between biomedical concepts. To support this evaluation, we introduce MeSH-Rel-4K, a dataset comprising 4K semantic relationships extracted from the Medical Subject Headings (MeSH). We analyse three adaptation strategies: standard prompting, Chain-of-Thought prompting, and fine-tuning. While parameter-constrained models traditionally struggle with the nuances of in-context logic, our results reveal that targeted fine-tuning increases the average F1-score by 34.1 percentage points. These results confirm that direct fine-tuning effectively exceeds the reasoning bottlenecks of smaller LLMs, providing an accurate, automated methodology for the construction and evolution of specialised biomedical ontologies.
[NLP-18] Zero Hallucination by Construction: Hallucination-Aware Layered Oversight for Trustworthy Enterprise AI
【速读】： 该论文旨在解决企业在部署生成式AI（Generative AI）时因对模型幻觉（hallucination）缺乏信任而产生的实际障碍。幻觉指模型生成看似合理但与事实不符的输出，是当前阻碍企业广泛应用的核心痛点。传统解决方案倾向于等待“零幻觉”的理想化模型，但作者指出这一目标不可行——大语言模型（LLM）的本质决定了其具备生成无依据内容的能力，规模扩展无法从根本上消除该风险。因此，论文提出根本性范式转变：将“零幻觉”从模型固有属性重构为系统可保障的运行特性。其核心解决方案是提出HALO（Hallucination-Aware Layered Oversight）保证架构，通过六层协同防御机制实现对幻觉的可控管理。关键在于构建一个分层、可验证、可追溯的系统级防护体系，其中证据基础置信度（evidence-based confidence）为核心创新点，即通过直接比对模型提取内容与原始文档，而非依赖模型自述的置信度，确保判断客观可靠；同时引入校准式拒答（calibrated abstention）、全流程可追溯性及持续监督闭环，使系统在检测到性能漂移或阈值突破时能主动再生并统计验证改进后的代理，从而实现对幻觉的动态控制与持续优化。
链接: https://arxiv.org/abs/2607.17883
作者: Bogdan Raduta,Horia Velicu,Alexandru Preda,Serban Chiricescu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 12 pages, 2 figures
Abstract:Enterprises will not deploy AI agents they cannot trust, and the most-cited reason for distrust is hallucination: confident, fluent output that is simply not true. The common response is to wait for a model that does not hallucinate. We argue that this is the wrong target. Large language models are, by construction, capable of generating unsupported text, and no amount of scale removes the possibility; a faithfulness judge bolted onto a raw model catches some errors but still ships others, and even well-curated retrieval pipelines have been shown to fabricate citations. We reframe the goal: “zero hallucination” is not a property a model possesses but a property a system enforces. We present HALO (Hallucination-Aware Layered Oversight), an assurance architecture which treats hallucination as a containable failure mode rather than an eliminable one. HALO composes six layers of defense: grounded generation over retrieved, approved content; constrained, deterministic execution that bounds where the model can err; multi-signal verification that scores every output for groundedness and hallucination using both an LLM judge and evidence-based checks against the source text; calibrated abstention, so the system declines rather than guesses when grounding is insufficient; total traceability of every retrieval, tool call, and generation; and continuous oversight that detects drift, alerts on threshold breaches, and closes the loop by regenerating and statistically validating improved agents. We detail each layer, give particular attention to evidence-based confidence (which verifies extractions against the source document rather than trusting the model’s self-reported certainty), and illustrate the architecture on a regulated claims-extraction workload
[NLP-19] Mobius Learning: Cyclic Depth Folding in Transformers
【速读】： 该论文旨在解决传统基于Transformer的语言模型中，浅层与深层模块在层级结构中固有的表征角色分工问题，即浅层模块通常负责提取局部特征，而深层模块则承担抽象语义建模的任务。这一固定的角色分配限制了模型对不同层次信息的灵活利用，尤其在分布式训练场景下，因需在各计算节点间同步完整的块堆栈（block stack）而带来高昂的内存开销。其解决方案的关键在于提出一种名为“莫比乌斯学习”（Mobius Learning）的训练架构，该架构基于循环深度折叠（cyclic depth folding），使不同数据流遵循循环偏移的块序列顺序。这使得同一块组（block group）在某些数据流中作为浅层模块使用，在另一些数据流中则作为深层模块使用，从而实现深度角色的叠加（depth-role superposition）。这种机制打破了块组必须固定于某一特定深度位置的传统观念，显著提升了模型参数的复用效率和表达能力。实验表明，在使用修改版GPT-2 small（124M）模型、基于25亿个FineWeb标记进行训练的四工作器设置下，通过Muon框架实现的莫比乌斯学习在多次块序列遍历后仍能取得低于固定顺序循环Transformer的验证损失。更重要的是，该架构天然适配内存受限的分布式训练：原始数据保持本地化，每个工作节点仅需存储一个块组而非整个块堆栈，极大降低了通信与存储成本，为高效可扩展的大规模语言模型训练开辟了新的设计空间。
链接: https://arxiv.org/abs/2607.17843
作者: Tongtian Zhu
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Distributed, Parallel, and Cluster Computing (cs.DC)
备注:
Abstract:Transformer-based language models organize computation along an ordered depth axis, where shallow and deep blocks often develop distinct representational roles. We challenge the conventional view that these roles must remain tied to a block’s position in the ordered sequence. We introduce Mobius Learning, a training architecture based on cyclic depth folding, in which different data streams follow cyclically shifted block orders. The same block group is therefore applied early in the block sequence for some data streams and late for others, so it is optimized in both shallow and deep roles, a phenomenon we call depth-role superposition. Surprisingly, in four-worker experiments with a modded GPT-2 small (124M) model trained on 2.5B FineWeb tokens using Muon, Mobius Learning achieves lower validation loss than a fixed-order looped Transformer at larger numbers of Transformer block-sequence passes. This counterintuitive result shows that a block group need not remain confined to one fixed shallow or deep role within the block sequence and opens a new design space based on cyclic depth folding. Crucially, this structure makes Mobius Learning particularly well suited to memory-constrained distributed training: raw training data remain local, while each worker stores one block group rather than the complete Transformer block stack.
[NLP-20] When a Name Is Not a Name: A Benchmark Dataset and Distilled Reasoning for Culturally Entangled Bangla Homographs in Low-Resource LLM s
【速读】： 该论文旨在解决生成式语言模型在处理文化语义纠缠同音异义词（Culturally Entangled Homograph, CEH）时存在的文化知识缺失问题，尤其针对孟加拉语中兼具个人姓名与文化负载名词双重身份的词汇（如“Maya”既为女性名字，又意为“慈爱的同情”）。其核心挑战在于，现代语言模型因预训练数据中缺乏足够的文化语境信息，倾向于默认选择常见名词义项，而忽略名称义项，导致系统性主导义项偏差。解决方案的关键在于提出一种基于对比链式思维（contrastive chain-of-thought prompting）的无训练干预策略，有效缓解偏差；同时通过蒸馏文化解释信息，使小型（1–3B）模型具备推理能力而非仅记忆标签，显著降低主导义项偏差（从最高100%降至5%以下），并成功逆转原本表现不佳的孟加拉语专用模型的表现，使其成为最强系统。
链接: https://arxiv.org/abs/2607.17828
作者: Md. Asaduzzaman Shuvo
机构: United International University, Bangladesh
类目: Computation and Language (cs.CL)
备注:
Abstract:Many Bangla words are at once personal names and culturally loaded common nouns, “Maya” is both a girl’s name and a word for affectionate compassion. Choosing the right reading demands cultural knowledge that is scarce in the pretraining data of modern language models. We introduce Culturally Entangled Homograph (CEH) disambiguation and build a Bangla benchmark of 1,516 expert-verified sentences (3,032 labelled occurrences) in which one word appears twice with two distinct readings, each labelled with a culturally grounded category and an explanation of the reasoning behind it. Across open- and closed-source models, we find a systematic dominant-meaning bias: models default to the common-noun sense and overlook the name. A Bangla-specific model fails under every prompting regime we test, showing that language-specific pretraining alone does not confer cultural grounding. We further show that contrastive chain-of-thought prompting can sharply reduce this bias without training, and that distilling cultural explanations teaches small (1-3B) models to reason toward the correct reading rather than memorise labels, cutting dominant-meaning bias from as high as 100% to under 5% and turning the failed Bangla-specific model into our strongest system. Dataset and code are available at this https URL.
[NLP-21] ESCUCHA: A Spanish Speech Benchmark for Heterogeneous Acoustic Conditions
【速读】： 该论文旨在解决大音频语言模型（Large Audio Language Models, LALMs）在真实声学条件下对西班牙语语音理解能力评估不足的问题。当前缺乏针对复杂、多样真实场景的系统性评测基准，尤其在多变声学环境与高阶推理能力方面存在显著空白。为此，论文提出ESCUCHA——首个专为评估LALMs在异质声学条件和多维度推理能力下表现而设计的西班牙语语音理解基准。其关键创新在于：构建了1,000个由人工标注的问答对，总计162.9小时音频数据均来自真实世界（“from the wild”），涵盖从数秒至超过80分钟的长时程片段；同时覆盖9类感知任务与10类推理任务，体现西班牙语多种口音及非标准发音特征；并引入多音频输入、口语提问与音频指令等多样化测试形式，支持开放式评估。实验表明，现有主流多模态与语音模型在该基准上性能远低于人类水平，凸显当前模型在真实场景下的理解局限性。
链接: https://arxiv.org/abs/2607.17812
作者: Fernando López,Ana Ayala,Guillermo Segovia,Fernando Ibáñez,Ana Martínez,Pablo Gómez,Jordi Luque
机构: 未知
类目: Computation and Language (cs.CL)
备注: Under review
Abstract:As large audio language models (LALMs) advance, robust evaluation frameworks have become essential. In this context, Spanish speech understanding under realistic acoustic conditions has received particularly little attention. We introduce ESCUCHA, the first Spanish speech understanding benchmark designed to evaluate LALMs across heterogeneous acoustic conditions and reasoning abilities. ESCUCHA comprises 1,000 human-curated questions paired with audio, totaling 162.9 hours sourced directly ``from the wild’’ rather than drawn from existing datasets, with durations ranging from a few seconds to over 80 minutes. The benchmark emphasizes reasoning, spanning 9 perceptual and 10 reasoning categories, and it captures linguistic diversity through multiple Spanish accents and non-normative speech. ESCUCHA further includes multi-audio questions, spoken questions, and audio instructions, and it flags which questions support open-ended evaluation. Benchmarking several state-of-the-art multimodal and speech models reveals substantial performance gaps relative to trained humans.
[NLP-22] When to Use Extra Context: Evidence-Grounded Terminology Adaptation for Simultaneous Speech Translation
【速读】： 该论文旨在解决技术类演讲的同步语音翻译（SimulST）中，如何有效利用文档级上下文信息以提升翻译质量的问题。现有方法通常将整个文档上下文注入每个流式翻译片段，但这种粗粒度的上下文引入方式难以精准捕捉关键术语，且易引入冗余信息。通过诊断实验发现，上下文带来的性能提升主要源于特定于论文的术语恢复，而非泛化的语义增强。为此，论文提出证据锚定术语自适应框架（EGTA），其核心在于构建文档级术语记忆库，基于当前流式状态动态选择紧凑的候选术语，并仅使用这些精选术语对自动语音识别（ASR）/语音端与解码端的决策空间进行适应性调整。该方法无需全模型微调，可适配级联式、端到端及生成式等多种SimulST架构。在包含MCIF-dev和ACL60/60-dev的ACL技术演讲同步翻译评测集上，EGTA-RG在英译中与英译德任务中分别实现BLEU提升+1.05/+0.59、XCOMET-XL提升+0.019/+0.006，命名实体召回率相对提升79%/73%，缩略词召回率提升0.099/0.171；在不同延迟设置下均稳定提升术语召回率与翻译质量。外部验证与消融实验进一步表明，性能提升源于论文特异性证据对齐，而非通用上下文提示效应。
链接: https://arxiv.org/abs/2607.17766
作者: Zeyu Yang,Satoshi Nakamura
机构: The Chinese University of Hong Kong, Shenzhen
类目: Computation and Language (cs.CL)
备注:
Abstract:Extra context is valuable for simultaneous speech translation of technical talks, but injecting the entire document context into every streaming segment is often too coarse. Through diagnostic experiments, we find that context gains mainly come from paper-specific terminology recovery rather than uniform semantic enhancement. We therefore propose EGTA, an Evidence-Grounded Terminology Adaptation framework that builds a document terminology memory, selects compact candidate terms conditioned on the current streaming state, and adapts ASR/speech-side and decoder-side decision spaces using only the selected terms. EGTA can be instantiated in cascaded, end-to-end, and generation-only SimulST settings without full-model fine-tuning. We evaluate EGTA on an ACL technical-talk SimulST evaluation suite consisting of MCIF-dev and ACL60/60-dev. On MCIF-dev, EGTA-RG improves BLEU by +1.05/+0.59, XCOMET-XL by +0.019/+0.006, named-entity recall by +79%/+73% relative, and acronym recall by +0.099/+0.171 on En \rightarrow Zh and En \rightarrow De. Across MCIF-dev latency settings, EGTA consistently improves XCOMET-XL, named-entity recall, and acronym recall. External validation on ACL60/60-dev further shows consistent terminology-recall gains without additional fine-tuning. Shuffled-memory controls and activation audits provide evidence that the improvements are tied to paper-specific evidence alignment rather than generic context prompting.
[NLP-23] Large Language Models for Citation Function Classification LREC2026
【速读】： 该论文旨在解决科学引文功能分类（citation function classification）中的关键问题，即如何准确识别学术文献中引用行为的语义意图，从而深化对科研成果间关系的理解并推动文献计量分析的发展。其核心挑战在于引文不仅包含信息性引用，还涉及评价性、批判性或赞许性等具有主观倾向性的表达，传统方法难以有效区分此类复杂语义。本研究的关键解决方案是：首次系统评估多种前沿大语言模型（LLM）在引文功能分类任务上的表现，采用零样本（zero-shot）、少样本（few-shot）和微调（fine-tuning）三种范式进行对比，并通过引入AC3这一新型数据集，构建了包含七类标注的细粒度分类体系，明确区分中性致谢与显性评价立场（如批评、赞扬、反驳等），同时结合四种上下文提取变体，量化分析上下文范围对分类性能的影响。实验结果表明，经过微调的Falcon 7B模型在ACL-ARC数据集上取得了73.3%的宏平均F1分数，达到新的最先进水平，显著优于以往方法。此外，研究还提供了详尽的模型表现分析、实验配置及局限性讨论，为后续研究提供了重要参考。
链接: https://arxiv.org/abs/2607.17738
作者: Daniel Vodička,Jakub Šmíd,Pavel Král,Christophe Cerisara
机构: 未知
类目: Computation and Language (cs.CL)
备注: Published at the 15th edition of the Language Resources and Evaluation Conference (LREC 2026)
Abstract:Citation function classification plays a crucial role in understanding the relationships between scientific publications and advancing bibliometric analysis. This study presents one of the first comprehensive evaluations of multiple state-of-the-art (SOTA) large language models (LLMs) for citation function classification, achieving new SOTA results on the ACL-ARC dataset. We systematically compare five models (Mistral 7B, Orca 2-7B, LLaMA 3.1-8B, Falcon 7B, and SciBERT) across zero-shot, few-shot, and fine-tuning approaches. Our fine-tuned Falcon 7B model achieves a 73.3% macro F1 score on ACL-ARC, representing a significant improvement over previous methods. Additionally, we introduce AC3, a novel dataset featuring a seven-category annotation scheme that distinguishes between neutral acknowledgments and explicit evaluative stances (more opinion-oriented citations - criticizing, complimenting, contradicting). The dataset is implemented across four context extraction variants to systematically evaluate the impact of contextual scope on classification performance. We also provide detailed analysis of model performance, experimental configurations, and limitations to guide future research in this domain. To our knowledge, this is one of the first studies dedicated to comprehensive model comparison for citation function classification, addressing a gap identified in recent surveys.
[NLP-24] C2KV: Compressed and Composable KV Cache Reuse for Efficient LLM Inference KDD2026
【速读】： 该论文旨在解决长上下文大语言模型（Long-context LLM）推理中因大规模键值（Key-Value, KV）缓存存储与访问开销导致的性能瓶颈问题。尽管已有研究通过KV缓存复用以减少冗余预填充（prefill）计算，但现有方法普遍忽视了KV缓存本身带来的存储和传输成本，且在引入压缩机制时，若与非前缀（non-prefix）KV复用简单结合，常导致生成质量显著下降。为此，本文提出C²KV——一种统一的非前缀KV复用框架，其核心在于联合优化KV提取与推理时的拼接过程。该方案学习一个可组合、压缩的KV缓存流形（manifold），并显式设计为位置无关（position-agnostic），从而支持灵活复用。C²KV引入轻量级侧车（sidecar）提取器，包含可学习的压缩令牌（compression tokens）与结构化注意力流，实现模块化的KV表示，可在不修改冻结主模型的前提下动态拼接复用。此外，通过压缩-拼接协同训练策略，确保提取阶段的表示与下游复用行为对齐。大量实验表明，C²KV在多个长上下文基准与模型族上显著降低KV缓存的存储与传输开销，实现长达17倍的推理加速，同时保持生成质量。
链接: https://arxiv.org/abs/2607.17715
作者: Chuheng Du,Junyi Chen,Hanlin Tang,Kan Liu,Tao Lan,Lin Qu,Chaoyue Niu,Shengzhong Liu,Guihai Chen,Fan Wu
机构: Shanghai Jiao Tong University (上海交通大学); Alibaba Group (阿里巴巴集团)
类目: Computation and Language (cs.CL)
备注: 12 pages, 9 figures, accepted by ACM SIGKDD 2026
Abstract:Long-context inference is central to modern large language model (LLM) applications such as retrieval-augmented generation and multi-document reasoning. To mitigate the growing inference cost, recent work has explored key-value (KV) cache reuse to reduce redundant prefill computation. However, existing reuse methods primarily focus on computation savings and overlook a critical bottleneck in long-context LLM serving: the cost of storing and accessing large KV caches. While KV compression appears to be a natural complement, naively combining compression with non-prefix KV reuse often leads to severe accuracy degradation. In this work, we propose C ^2 KV, a unified framework for non-prefix KV reuse that jointly optimizes KV extraction and inference-time concatenation. C ^2 KV learns a composable and compressed KV cache manifold that is explicitly designed to be position-agnostic. Our approach introduces a lightweight sidecar Extractor with learnable compression tokens and a structured attention flow, enabling modular KV representations that can be flexibly reused and concatenated without modifying the frozen base model. We further employ a compression-concatenation co-training strategy to align extraction-time representations with their downstream reuse behavior. Extensive experiments across multiple long-context benchmarks and model families demonstrate that C ^2 KV significantly reduces KV cache storage and transfer costs, achieving up to 17 \times inference speedup under long contexts, while preserving generation quality.
[NLP-25] AEGIS: Awareness-Enhanced Guidance for Iterative Safeguard
【速读】： 该论文旨在解决多语言文本净化中“片段级推理依据（span-level rationales）”在提升可控性方面的有效性问题，尤其关注其在降低毒性与保持语义完整性之间的权衡机制。现有研究普遍假设片段级引导可增强控制力，但缺乏对引导效果在不同生成模型、模型规模及语言背景下差异的系统性分析。为此，论文提出意识增强型迭代防护框架（AEGIS），通过融合冻结的生成器主干网络与片段级检测器输出，将有害片段、毒性强度标签及目标属性作为结构化引导信号，在重写过程中实现精细化控制。该方案的关键在于：不追求最优的净化性能，而是通过对比不同生成器家族、模型规模和语言（英语、中文、韩语）下，片段引导对毒性削减与语义保真度之间平衡的影响，揭示了片段引导的条件性有效性——其作用效果显著依赖于生成器架构与语言特性，既展现了片段级控制信号在多语言净化中的潜力，也揭示了其应用局限性。
链接: https://arxiv.org/abs/2607.17713
作者: Kyungwon Park,Sangmin Lee,Heejae Chon,Hyungu Kang
机构: Yonsei University (延世大学)
类目: Computation and Language (cs.CL)
备注: 11 pages, 3 figures, 9 tables. Preprint
Abstract:Span-level rationales are often assumed to improve controllability in text detoxification, but it remains unclear when such guidance helps and when it introduces trade-offs. We present Awareness-Enhanced Guidance for Iterative Safeguard (AEGIS) as an exploratory framework for studying span-guided multilingual detoxification across English, Mandarin Chinese, and Korean. AEGIS combines span-level detector outputs with frozen generator backbones, allowing harmful spans, intensity labels, and target attributes to be provided as structured guidance during rewriting. Rather than claiming state-of-the-art detoxification performance, we analyze how span guidance affects the balance between toxicity reduction and meaning preservation across generator families, model scales, and languages. Our results suggest that span-guided detoxification is conditionally useful: explicit rationales change the trade-off between toxicity reduction and meaning preservation, but their effects depend strongly on the generator backbone and the linguistic context. These findings highlight both the promise and the limitations of span-level control signals for multilingual detoxification.
[NLP-26] okenizing Crosslingual Homographs
【速读】： 该论文旨在解决多语言语言模型中因共享子词词汇表（subword vocabulary）导致的跨语言同形异义词（cross-lingual homographs）与虚假同源词（false friends）处理不当的问题。这类现象表现为不同语言中表面形式相同的词被统一处理，忽视了其语义或用法差异，从而影响模型的跨语言理解能力。解决方案的关键在于在分词（tokenization）阶段引入轻量级的语言信息：通过使用语言特异性字符替代共享词汇中词首字符，以降低词汇构建过程中的共现同一性。实验表明，该干预方法能有效缓解基于BPE和UnigramLM的分词器对跨语言同形词的过度泛化问题，使上下文感知型的SaGe分词器表现更趋一致；在英译多语言机器翻译任务中，该方法在部分设置下取得适度性能提升，尤其在使用BPE分词时效果显著，尽管结果并非在所有语言和评估集上均一致。研究结果表明，在分词层面上嵌入语言线索是一种具有前景的优化方向。
链接: https://arxiv.org/abs/2607.17689
作者: Rotem Brillant,Yuval Pinter
机构: Ben-Gurion University of the Negev (本·古里安大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Multilingual language models rely on shared subword vocabularies to represent multiple languages within a limited number of token units. While such sharing is often useful, it can also create cases in which identical surface forms are treated too uniformly across languages, even when their meanings or usage differ. We investigate this limitation through cross-lingual homographs and false friends, and examine whether introducing language information earlier in the tokenization process can improve their treatment. We propose a simple tokenizer-level intervention based on language cues: language-specific characters replacing initial characters of shared-vocabulary words, reducing common identity during vocabulary construction. In intrinsic analysis, we find through tokenizer-level statistics that BPE and UnigramLM often treat cross-lingual homographs in a largely language agnostic way, whereas the context-sensitive SaGe tokenizer diverges more strongly; our intervention removes this gap. In downstream English-to-X machine translation, our cues yield modest improvements in several settings, especially under BPE, although the effect is not consistent across all languages and evaluation sets. Overall, the findings suggest that adding lightweight language information at the tokenizer level is a promising direction for further exploration.
[NLP-27] PoLoRA: A Preconditioned Orthogonalized LoRA Optimizer
【速读】： 该论文旨在解决低秩适配（LoRA）在微调大语言模型时因优化器选择不当而导致的训练效率与稳定性不足的问题。现有方法普遍采用Adam优化器处理LoRA的低秩参数，但其将矩阵因子视为单一扁平参数向量，忽略了参数的矩阵结构及乘积形式，从而限制了优化性能。尽管已有研究尝试引入矩阵感知或乘积感知的优化器（如Muon及其变体），但其提升效果并不稳定。为此，本文提出PoLoRA——一种预条件正交化低秩适配优化器，其核心创新在于三个关键设计：（1）基于乘积结构的谱更新方向，显式建模因子间的耦合关系；（2）通过控制单样本损失变化推导出曲率预条件机制，实现对梯度动态的自适应调整；（3）引入幅度规则以约束因子更新与合并后更新的尺度，防止数值发散。实验表明，PoLoRA在1B至8B参数规模的代码与数学指令微调任务中，相较于标准Adam，在1.2至1.7倍更少的训练步数内达到相同的验证损失，且每步额外开销不超过3%。此外，PoLoRA对学习率不敏感，且最优学习率在不同低秩秩数下保持稳定，显著提升了训练鲁棒性与可扩展性。
链接: https://arxiv.org/abs/2607.17620
作者: Nikhil Ghosh,Tetiana Parshakova,Robert M. Gower
机构: Center for Computational Mathematics, Flatiron Institute (计算数学中心，扁平化研究所)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Optimization and Control (math.OC)
备注:
Abstract:Low-rank adaptation (LoRA) makes finetuning large language models cheaper by adding to each weight matrix a trainable low-rank update parameterized as the product of two matrices. These matrices are usually trained with Adam, which treats them as a single flat vector of parameters and ignores both the matrix and product structure of LoRA. Applying a matrix-aware optimizer such as Muon to each factor does not consistently improve over Adam, and neither do the product-aware Muon variants proposed in concurrent works. To realize consistent gains, we introduce PoLoRA, a Preconditioned Orthogonalized LoRA optimizer built from three ingredients: a product-aware spectral update direction, curvature preconditioning derived from controlling the per-sample loss change, and a magnitude rule that controls the sizes of both the factor and merged updates. We evaluate PoLoRA on instruction-tuning datasets for code and math across models from 1B to 8B parameters, and find that it reaches the final held-out loss achieved by tuned Adam in 1.2-1.7 times fewer steps, while adding at most 3% per-step overhead. Compared to Adam, PoLoRA is also less sensitive to the learning rate, and its optimal learning rate is stable across ranks.
[NLP-28] Is Progressive Disclosure All You Need for Long-Context Agents ?
【速读】： 该论文旨在解决长文档问答任务中面临的两大困境：一是将整篇文档全部加载至上下文窗口带来的计算与容量限制，二是依赖外部检索器（retriever）所带来的系统复杂性与延迟。现有方法通常在“全量加载”与“外置检索”之间做出二选一的权衡，而生成式 AI (Generative AI) 提供了一种更灵活的替代路径——通过赋予智能体（agent）文档路径，由其自主决定读取内容与方式。本文提出的核心解决方案是基于“代理技能（Agent Skills）”机制实现渐进式披露（progressive disclosure），即根据查询需求逐步暴露从摘要到具体段落的信息层级，从而动态优化上下文利用效率。研究的关键发现在于：渐进式披露的优势并非来自提升智能体的推理能力，而是有效节约上下文空间；当智能体本身具备较强的导航与定位能力时，该机制冗余无益；但随着文档规模扩大至多本书籍级别的跨文档任务，原始文档遍历策略迅速失效，而单层渐进披露能显著优于其他方案，且进一步增加路由层级反而会损害准确性。因此，解决方案的核心在于以“单层渐进披露”实现对大规模文档集的有效管理，在保持上下文高效利用的同时避免过度复杂化，其价值在高规模场景下尤为突出。
链接: https://arxiv.org/abs/2607.17598
作者: Yifeng He,Yinzhe Zhao,Jicheng Wang,Hao Chen
机构: University of California, Davis (加州大学戴维斯分校); Zhejiang University (浙江大学); The University of Hong Kong (香港大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Software Engineering (cs.SE)
备注:
Abstract:Long-document question answering usually forces a choice between loading the whole document into the context window and bolting on a separate retriever. Agentic AI suggests a broader option, giving the agent the document path and letting it decide how and what to read. Agent Skills, a standard for packaging expertise into folders an agent loads on demand, supply a ready mechanism: progressive disclosure, which exposes only what a query needs, from a short description down to the specific passages. Practitioners rapidly adopted this pattern for book-length understanding tasks, but the evidence to support such choices has been anecdotal. We run the first controlled study of the pattern, comparing raw-document navigation and several designs of Agent Skills packs against a classical hybrid retriever across three agent harnesses and three model families on InfiniteBench. On a single book, the gain depends on the harness, running large when the agent navigates the raw document poorly but near zero when a strong agent harness already divides and retrieves on its own. When scaling up to tasks that span many books, raw-document navigation collapses while one-level progressive disclosure degrades more slowly and pulls ahead. A second, deeper routing level never helps and sometimes breaks accuracy outright, so one level is enough. Progressive disclosure buys context, not intelligence: it is redundant while a strong agent can locate the right passages itself, and decisive once the corpus grows too large to navigate by reading.
[NLP-29] Salience Induction against Multi-Hop RAG Agents : Threat and Defense
【速读】： 该论文旨在解决生成式智能体在多跳问答（Multi-Hop Question Answering）任务中，因外部证据检索与工具调用所引入的新型安全威胁——显著性诱导攻击（Salience Induction）。现有防御机制主要关注内容投毒（注入虚假事实）和提示注入（嵌入恶意指令），但忽略了显著性通道（salience channel）这一关键攻击面：即通过调整事实的位置、强调方式、表述框架及语义邻近性，在不改变检索结果真实性且无显式指令的前提下，误导智能体推理路径。其解决方案的核心在于提出一种保持真实性但重构推理关联性的编辑策略，定义六类显著性编辑算子，并构建基于事实一致性与隐蔽性约束的迭代提议-验证流水线。研究还构建了首个带有欺骗性标注的多跳问答基准数据集SalientWiki-MH，实验覆盖五大主流模型家族（GPT、Claude、Gemini、DeepSeek、Qwen）与三种智能体架构（ReAct、Reflexion、tool-calling），验证了方法的广泛泛化能力。在30%编辑预算下，攻击成功率高达83.3%，而最强基线防御仍保留75.7%的攻击成功率；无目标重写虽可降低攻击效果，但以牺牲中立任务性能为代价。最终提出的轻量级输入端防御机制——显著性归一化（Salience Normalization），在标准攻击下将攻击成功率降至15.3%，自适应攻击下亦降至23.6%。研究表明，仅依赖真实性校验与指令过滤不足以保障生成式检索增强生成系统（RAG）的安全性，必须针对显著性与相关性解耦问题建立专门防御机制。
链接: https://arxiv.org/abs/2607.17535
作者: Xingfu Zhou,Pengfei Wang,Yuan Zhou,Wei Xie,Xu Zhou
机构: National University of Defense Technology (国防科技大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: 18 pages, 4 figures, 12 tables
Abstract:Agentic retrieval-augmented generation (RAG) systems increasingly retrieve external evidence and orchestrate tools for knowledge-intensive applications. In Multi-Hop question answering, agents chain facts across documents. Existing defenses focus on content poisoning, which injects false facts, and prompt injection, which embeds directives. We identify a third attack surface: the salience channel, through which fact position, emphasis, framing, and semantic proximity can redirect reasoning even when all retrieved claims are true and no instructions are present. We formalize Salience Induction as truth-preserving edits that redirect Multi-Hop attribute binding while leaving the retrieval trace semantically intact. We define six Salience-Editing operator classes and build an iterative proposer-verifier pipeline under factual and stealth constraints. We also introduce SalientWiki-MH, a decoy-annotated Multi-Hop benchmark. Evaluations across five frontier model families (GPT, Claude, Gemini, DeepSeek, and Qwen) and three agent architectures (ReAct, Reflexion, and tool-calling) show broad generalization. Under a 30% edit budget, Salience Induction achieves an 83.3% attack success rate; the strongest evaluated baseline defense leaves 75.7% post-defense ASR. Untargeted rewriting further reduces attacks only by degrading neutral task success. Our lightweight input-side defense, Salience Normalization, reduces attack success to 15.3% under standard attacks and 23.6% under an adaptive attack. These results show that truthfulness and instruction filtering alone are insufficient: robust agentic RAG also requires defenses against salience-relevance decoupling.
[NLP-30] Oracle Gap and Signal Fidelity: A Fixed-Pool Diagnostic for Test-Time Collaboration ACL
【速读】： 该论文旨在解决生成式 AI（Generative AI）在测试阶段通过协作机制（如自一致、最佳-N选择、批判模型和验证器流水线等）提升大语言模型（LLM）推理性能时，其增益存在不均衡甚至为负的难题。核心问题在于：在不进行额外训练的前提下，何时应预期此类协作方法有效？论文的关键解决方案是将协作过程重新建模为一个可量化的候选解选择问题，而非依赖于多智能体拓扑结构的内在属性。作者提出一套分解框架，将选择器或验证器的净收益拆解为四个可测量因子：可恢复质量（recoverable mass）、验证信号覆盖率（verification-signal coverage）、条件选择质量（conditional selection quality）以及对已正确输出的损害程度（harm to already-correct outputs）。通过实证分析发现，性能上限首先由“最优解差距”（oracle gap）决定，其次受验证信号保真度（signal fidelity）制约，后者可通过验证器判断与标准答案之间的候选级一致性直接衡量。实验结果表明，在LiveCodeBench、MATH Level-5高难度题型及GPQA-Diamond任务中，协作效果显著受限于任务特性、模型能力与采样配置共同决定的最优解差距；尤其在高一致性的候选池中（如GPQA-Diamond中87.54%的候选答案相同），可改进空间极小。由此，论文提出一种实用的部署前诊断流程：先估计最优解差距，再评估覆盖率、信号保真度与损害率，从而指导是否值得投入资源构建协作系统。
链接: https://arxiv.org/abs/2607.17531
作者: Jie Hu
机构: China Telecom(中国电信)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 13 pages, 2 figures, 8 tables. Code: this https URL
Abstract:Test-time collaboration, including self-consistency, best-of-N selection, critic models, and verifier pipelines, is often credited with broadly improving LLM reasoning, yet its gains are uneven and sometimes negative. We ask when training-free collaboration should be expected to help. For a fixed candidate pool, we decompose a selector or verifier’s net gain into measurable factors: recoverable mass, verification-signal coverage, conditional selection quality, and harm to already-correct outputs. This reframes collaboration as a candidate-selection problem rather than as an intrinsic property of a multi-agent topology. Across LiveCodeBench, MATH Level-5 hard subjects, and GPQA-Diamond, gains are bounded first by the oracle gap and then by signal fidelity, which we measure directly as candidate-level agreement between verifier verdicts and official labels. On LiveCodeBench, a public-test verifier (MCC 0.825) gains +8.14 percentage points (pp) over a first-sample baseline; a generated-test verifier (MCC 0.248) improves by +2.70pp and is not statistically distinguishable from an LLM selector, but operates at near-zero harm versus the selector’s 4.69% harm rate. On MATH, a symbolic answer-equivalence selector beats self-consistency by +4.67pp, while LLM selectors are negative. On GPQA-Diamond, recoverable mass is only 3.03% and 87.54% of candidate pools are answer-identical; a weaker model’s pools shrink both further, suggesting that oracle gap is a joint property of task, model, and sampling configuration. Our framework yields a practical pre-deployment diagnostic: estimate the oracle gap, then measure coverage, signal fidelity, and harm before investing in collaboration.
[NLP-31] oken-Level Off-Policy Learning for Faithful Generation Under Distribution Shift
【速读】： 该论文旨在解决大模型在后训练（post-training）阶段中，如何有效实现离策略（off-policy）生成且保持高质量输出的问题。传统方法直接训练模型生成特定的离策略响应，易导致性能下降或引入偏差，而现有策略常依赖于序列级（sequence-level）奖励信号，难以精确指导模型对每个生成令牌的质量进行判断。为此，论文提出了一种名为“令牌级离策略标注”（Token-Level Off-Policy Labeling, TOPL）的新范式，其核心在于将后训练过程重构为一个令牌级别的正确性预测任务：通过训练模型区分生成响应中的优质与劣质令牌，从而自然引导模型倾向于生成高质量的令牌，避免了直接优化离策略输出所带来的潜在问题。实验结果表明，TOPL在11个不同数据集上的文档摘要任务中均展现出卓越的分布外泛化能力，显著优于多种序列级与令牌级基线方法；此外，其有效性在机器翻译任务中也得到验证，表明该方法具有跨任务泛化能力。消融研究表明，令牌级学习信号是性能提升的关键，而对应的序列级版本无法带来相同收益。最后，研究发现通过TOPL学习到的LoRA适配器具备可解释性，可被视作线性分类头和方向控制向量，进一步揭示了其内在机制。
链接: https://arxiv.org/abs/2607.17524
作者: Zitong Huang,Gustavo Lucas Carvalho,Deqing Fu,Robin Jia
机构: University of Southern California (南加州大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:We propose Token-Level Off-Policy Labeling (TOPL), an off-policy training paradigm that reframes post-training as a token-level correctness prediction task. Our key intuition is that by training the model to distinguish good and bad tokens in a response, we naturally guide the model towards generating good tokens, while avoiding the pitfalls that come with directly training the model to generate off-policy tokens. Experiments on document summarization tasks show that TOPL achieves strong out-of-distribution generalization across 11 datasets against a diverse set of sequence-level and token-level baselines. We further demonstrate that TOPL transfers effectively to machine translation, suggesting that its benefits generalize across different faithful generation tasks. Through ablation studies, we confirm that our token-level learning signal is critical to good performance; sequence-level analogues do not confer similar benefits. Finally, we show that TOPL induces interpretable model updates: the LoRA adapters learned through TOPL function as linear classification heads and steering vectors.
[NLP-32] hinking in Video: Can Video Generators Really Reason About the Real World?
【速读】： 该论文旨在解决生成式视频模型在复杂世界动态推理中是否存在真正因果理解的问题，而非仅依赖对训练数据的表面记忆或视觉保真度。其核心挑战在于现有评估方法难以区分感知保真与语义逻辑一致性，导致无法有效验证模型是否具备“以视频为媒介进行因果推理”的能力。为此，论文提出关键解决方案——因果-生成双判官评估框架（Causal-Generative Dual-Judge, CGDJ），从两个维度协同评估世界模型的一致性：显式因果感知（Explicit Causal Perception）通过时空扁平化的视觉问答任务检验模型是否将视频场景解析为可推理的因果问题；隐式生成感知-预测差距（Implicit Generative Perception-Prediction Gap）则评估模型生成的未来视频是否与所推断的因果逻辑保持一致。实验结果揭示出显著的感知-生成鸿沟：开源模型虽能生成视觉上合理的动态，但几乎无显式因果感知；而先进闭源系统虽具一定推理与生成对齐能力，但仍存在明显局限。进一步分析还发现音视频错位现象，即模型在语言层面更可靠地表达正确因果逻辑，但在视觉生成中未能准确体现，从而质疑了当前生成式视频模型作为“世界模拟器”的叙事合理性。
链接: https://arxiv.org/abs/2607.17523
作者: Yongheng Zhang,Guang Yang,Ruihan Hou,Qiguang Chen,Ziang Liu,Xiaolong Liu,Manman Zhang,Yanchao Hao,Zheng Wei,Hao Wu,Libo Qin,Peishan Dai,Yinghui Li,Di Yin,Xing Sun
机构: Central South University (中南大学); Tencent (腾讯); Tsinghua University (清华大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Recent advances in world models and video generation have given rise to an emerging reasoning paradigm that leverages video generative models to simulate, predict, and reason about real-world dynamics. We redefine this paradigm as Thinking in Video, where video is not merely an output artifact but a medium for constructing, extending, and verifying causal thought. However, this promise remains unverified: convincing rollouts may reflect memorized appearances rather than causal understanding, while existing metrics separate perceptual fidelity from semantic logic. To evaluate whether video generators support such reasoning, we introduce the Causal-Generative Dual-Judge (CGDJ), auditing World Model Consistency from two perspectives. Explicit Causal Perception tests whether a generator reads a video scenario as a reasoning problem through spatio-temporal flattened visual question answering, while Implicit Generative Perception-Prediction Gap evaluates whether it renders the causal consequence as a consistent future video. Applying CGDJ to representative open- and closed-source generators reveals a clear Perception-Prediction Gap: open-source models produce plausible dynamics despite near-zero explicit causal perception, whereas advanced closed-source systems show stronger but still limited alignment between reasoning and generation. Further analysis exposes audio-visual misalignment, where models verbalize correct causal logic more reliably than they render it, challenging the “world simulator” narrative.
[NLP-33] After the Euclidean Highway: Hyperbolic Expert AI as the Next Innovation
【速读】： 该论文旨在解决大模型训练中因几何结构不匹配导致的稳定性问题，具体而言，传统基于欧几里得空间的Transformer架构无法有效保留专家领域（expert domains）固有的树状层次结构，其父-子关系在深层传播时被指数级稀释。尽管已有研究尝试引入双曲几何（hyperbolic geometry）以更好地建模此类结构，但现有方法普遍将曲率全局施加于网络，缺乏对曲率作用位置的精细控制。本文提出的关键解决方案是：仅在损失层（loss layer）引入双曲几何，即HySAT（Hyperbolic Structure-Aware Training）——通过限制双曲损失仅作用于输出层，而非在整个可训练适配器（adapter）上施加双曲结构。实验证明，全局施加双曲结构会导致训练崩溃（17次失败，累计约220 GPU小时），而仅在损失层使用双曲结构则能实现稳定训练。该设计的核心在于保持流形不变性（manifold invariance）的同时避免在底层参数空间引入不稳定的非线性曲率，从而在六种专家小语言模型（SLMs）上实现成功部署，并通过四臂消融实验与数学证明（三个命题及一个引理）验证了损失层专用双曲性的理论稳定性。
链接: https://arxiv.org/abs/2607.17513
作者: Kwan Soo Shin,In Seok Kang,Munho Lee
机构: PMMinds(PM Minds)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 40 pages, 11 figures. Supplementary Information included as an ancillary file. Data and code: Zenodo, concept DOI https://doi.org/10.5281/zenodo.21438499 (published, CC-BY-4.0)
Abstract:Expert domains are trees; the Euclidean transformer is not, diluting parent-child structure exponentially at depth. The hyperbolic turn left one question unasked: not how much of a network to curve, but where curvature may touch the gradient. Placement is a law, not a knob: the same geometry on a trainable adapter collapses training (seventeen training collapses, ~220 GPU-hours), yet at the loss layer alone it trains without one – this is HySAT (Hyperbolic Structure-Aware Training), hyperbolic losses at the loss layer only. Across six expert SLMs we constructed and deployed (Llama 3.1 and EXAONE 3.5; four adapter strategies; 18.0M-sample corpus; zero NaN over ~317K optimizer steps), a matched four-arm ablation isolates the preserved manifold invariant, and three propositions and a lemma prove why loss-only placement is stable where adapter-on-manifold is not. Four models are operationally deployed (one live, consumer-facing), two open-weight, with per-step traces and a seventeen-incident failure ledger on Zenodo (CC-BY-4.0).
[NLP-34] Multilingual Sentence Embeddings for Linguistic-Integrated Reliability Audit
【速读】： 该论文旨在解决多语言评估系统在评分与质量控制过程中对翻译依赖所带来的局限性问题，特别是翻译失败导致的答题数据丢失及可靠性估计偏差。其解决方案的关键在于利用多语言句向量（multilingual sentence embeddings）替代传统的翻译后英文输入，实现跨语言的语言学整合可靠性审计（Linguistic-Integrated Reliability Auditing, LiRA）。实验结果表明，使用母语句向量可高度复现基于翻译的可靠性估计结果，并有效恢复因翻译失败而被排除的原始响应，且未对整体可靠性产生显著影响，证明了多语言嵌入在保持评估一致性方面的可行性与有效性。
链接: https://arxiv.org/abs/2607.17466
作者: Ummugul Bezirhan,Ji Yoon Jung,Matthias von Davier
机构: Boston College (波士顿学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:Multilingual assessment systems commonly rely on translation for scoring and quality-control processes. We evaluate whether multilingual sentence embeddings can replace translated English input for Linguistic-Integrated Reliability Auditing (LiRA) across 11 PIRLS constructed-response items and three embedding models. Native-language embeddings reproduced translation-based reliability estimates closely while recovering responses excluded after translation failure, with no meaningful change in reliability.
[NLP-35] How Reliable Are Multimodal Signals of Conversational State? Evidence from Remote Dyadic Collaborative Tasks
【速读】： 该论文旨在解决从多模态行为中准确测量对话状态（如认知负荷与对话权力）所面临的特征可靠性与跨任务泛化能力不足的问题。其核心挑战在于现有特征在不同任务情境下表现不稳定，且评估方法常忽略说话人身份的影响，导致对特征可靠性的误判。论文提出一个三维评估框架，综合考量预测准确性、跨任务泛化性与重测信度，并基于AVCAffe数据集（53对参与者，9项协作任务）对交互、声学与语言特征进行系统评估。研究发现：语言特征虽在预测认知负荷上表现最优，但跨任务泛化能力差，易受任务特异性词汇干扰；声学特征的高信度在控制说话人身份后急剧下降，表明其实际反映的是语音生理特征而非对话状态；唯有交互特征在说话人归一化后仍保持稳定，成为唯一真正可靠的信号，且能有效预测同对内认知负荷不对称性。此外，权力角色分类始终接近随机水平，表明任务层面聚合的行为特征难以捕捉对话中的权力动态。研究揭示三大关键洞见：（1）语言特征预测能力强但泛化性差；（2）声学特征的可靠性在控制说话人后失效，挑战了传统评估范式；（3）交互特征是唯一具备真实可靠性的信号。因此，论文主张将说话人归一化与多维评估作为构建情境感知、鲁棒性强的多模态对话系统中特征选择的先决条件。
链接: https://arxiv.org/abs/2607.17452
作者: Tahiya Chowdhury
机构: Colby College(科尔比学院)
类目: Computation and Language (cs.CL)
备注: Accepted, to appear in Proceedings of ACM International Conference on Multimodal Interaction 2026, 13 pages, 6 figures, 4 tables
Abstract:Measuring conversational states such as cognitive load and conversational power from multimodal behavior requires characteristic features that are not only predictive but also reliable across task contexts. We present a three-dimensional evaluation framework assessing predictive accuracy, cross-task generalizability, and test-retest reliability, applied to interactional, acoustic, and linguistic features extracted from dyadic conversations during collaborative tasks performed over a video-conferencing platform (AVCAffe dataset; 53 dyads, 9 tasks). Our results show that no single feature family dominates all three dimensions. Linguistic features show the highest predictive accuracy for cognitive load but collapse under cross-task evaluation, revealing sensitivity to task-specific vocabulary. Additionally, acoustic reliability, often reported as evidence of feature stability, degrades once speaker identity is controlled, confirming that standard prosodic features measure vocal characteristics rather than conversational state. Interaction features provide the only genuinely reliable signal, unchanged after speaker normalization. Interestingly, classifying power role remained near chance baseline across all conditions, indicating limitations of task-level aggregated behavior for predicting power role in conversation. Our findings reveal three insights: (1) linguistic features predict best but generalize poorly across task contexts; (2) acoustic reliability collapses to near-zero once speaker identity is controlled, challenging standard evaluation practice; and (3) interaction features provide the only genuinely reliable signal, with floor dominance predicting within-dyad cognitive load asymmetry. These results argue for speaker normalization and multi-dimensional evaluation as prerequisites for context-aware, robust multimodal feature selection in conversational systems.
[NLP-36] Abliteration Is Not a Scalpel: Off-Target Effects of Refusal Removal on Decision Disposition Across Model Families
【速读】： 该论文旨在解决“去审查化”（uncensored）开源大模型在通过权重删减（abliteration）移除拒绝指令（refusal direction）后，是否真正仅移除了拒绝行为而未引入其他不可预见的副作用这一关键问题。研究指出，当前主流的“去审查化”方法并非干净的操作，其本质是通过修改模型权重来消除拒绝响应，但这一过程会引发系统性偏差。解决方案的关键在于采用一个精心设计的、无拒绝发生的基准任务——对华沙证券交易所60只股票在18周内的每周涨跌方向进行预测，并通过冻结推理管道确保仅有决策层模型为变量，从而隔离出纯副作用信号。实验结果显示，在两个Mixture-of-Experts架构（Gemma-4-26B-A4B-it与Qwen3-30B-A3B-Instruct-2507）中均复现了三个显著效应：被删减模型表现出系统性更强的乐观倾向（Gemma+12.2个百分点，Qwen+7.4个百分点）、自我辩护时论述更长、在强制自评中使用更少明确表达不确定性的词汇；此外还发现第四种效应呈现反向变化：同一操作使Gemma-abliterated模型表达信心降低，而Qwen-abliterated模型表达信心增强，且两家族置信区间不重叠，表明同一手术导致相反的自信变化。研究进一步排除了指令遵循能力下降作为驱动因素，并证实所有模型均无经济收益能力，所谓“优势”实为市场贝塔而非阿尔法。此外，溯源审计还发现了两个独立的污染源：量化器不匹配的预训练对齐对以及过时的社区聊天模板，揭示出社区修改检查点的研究中工具链缺陷普遍存在。结论表明，将“去审查化”模型用作决策代理时，实际上部署的是一个在认知倾向、表达风格和自信水平上均与原模型显著不同的决策主体，而非简单去除拒绝功能的基线模型。
链接: https://arxiv.org/abs/2607.17427
作者: Aleksander Fafuła
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Computational Finance (q-fin.CP)
备注: 11 pages, 5 figures, 4 tables. Preregistered. Data and code: this https URL ; dataset DOI: https://doi.org/10.5281/zenodo.21314839
Abstract:Abliteration - deleting a model’s refusal direction from its weights - is the standard recipe behind popular “uncensored” open-weight models. We show the surgery is not clean. As a disposition probe we use 21,600 decisions under uncertainty - weekly up/down calls on 60 Warsaw Stock Exchange equities over 18 weeks, replayed through a frozen pipeline so the decision-layer model is the only variable. The task elicits no refusals at all, so any between-arm delta is pure side effect. Holding provenance constant (official BF16 checkpoints, a single abliteration author, an identical serving stack, one byte-identical frozen prompt), we compare base and abliterated arms of two Mixture-of-Experts families, Gemma-4-26B-A4B-it and Qwen3-30B-A3B-Instruct-2507. Three effects replicate across both families (weeks-clustered bootstrap CIs excluding zero): abliterated models are systematically more optimistic (+12.2 pp Gemma, +7.4 pp Qwen; the confirmed preregistered endpoint), justify themselves at greater length, and use fewer explicit uncertainty words in forced self-critiques (both exploratory). A fourth effect reverses sign: the same operation makes Gemma-abliterated less confident and Qwen-abliterated more (family CIs non-overlapping) - one weight surgery, opposite shifts in expressed confidence. Capability covariates rule out instruction-following degradation as the driver, and no arm shows economic skill: the apparent edge of abliterated arms is regime beta, not alpha. Our provenance audit also caught two independent contamination channels - a mismatched-quantizer pilot pair and a stale community chat template that silently mangled the rendered prompt - suggesting toolchain artifacts are the rule in studies of community-modified checkpoints. Whoever deploys an “uncensored” model as an agent is deploying a measurably different decision-maker, not the base model minus refusals.
[NLP-37] he Librarian Who Refused to Code: Model-Dependent Identity Enactment in LLM Code Generation
【速读】： 该论文旨在解决生成式 AI（Generative AI）在代码生成任务中使用人物角色设定（biographical personas）时，其实际效果缺乏在受控、预先注册条件下的系统性评估这一问题。研究通过设置四种提示条件（无角色、两名工程师角色、一名研究馆员角色），在12个代码生成任务上对两种前沿模型进行测试，并每组条件执行五次运行（共480次生成结果），以量化角色设定对输出质量与行为的影响。其解决方案的关键在于采用预注册的混合效应模型分析方法，揭示了角色设定对不同模型具有显著差异化的效应：在Claude Opus上，简约型工程师角色使可见输出减少30%（提供者统计的令牌数减少33%），但未提升正确性；详尽型工程师角色则增加输出量却未带来正确性改善；而研究馆员角色引发55/60次响应中出现角色内声明，导致平均正确率从0.92降至0.67，且产生12次真实“无代码”响应。相比之下，GPT-5.5未表现出此类行为。研究结论表明，人物角色并非通用的质量优化手段，而是作为依赖于模型特性的行为策略偏差（model-dependent behavioral-policy biases）起作用。研究还公开了原始生成结果、衍生评分、分析脚本、预注册文档及执行日志，支持可复现性，但端到端测试重评分需依赖未公开的任务框架。
链接: https://arxiv.org/abs/2607.17420
作者: Shayell Aharon Salomon(Bluebear Security),Noam Israel(Bluebear Security),Ido Safruti(Bluebear Security),Amir Shaked(Bluebear Security)
机构: Bluebear Security
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: 17 pages, 3 tables, no figures. Ancillary files include raw completions, derived scores, analysis scripts, persona texts, and preregistration
Abstract:Biographical personas are widely used in system prompts, but their effects on code generation are rarely evaluated under controlled, pre-registered conditions. We tested four prompt conditions (no persona, two engineer personas, and a research-librarian persona), 12 code-generation tasks, two frontier models, and five runs per cell (480 completions). Persona effects differed between the two tested models. Under the pre-registered mixed-effects analysis, the condition-by-model interaction was significant for provider-reported output tokens; a post-hoc visible-character measure showed the same qualitative pattern. Six GPT-5.5 completions were length-capped and are reported separately. On Claude Opus, the minimalist engineer persona reduced visible output by 30% (33% in provider tokens) without improving correctness, while the thorough engineer persona increased output without a correctness gain. In an exploratory post-hoc analysis, the librarian persona elicited in-character disclaimers in 55 of 60 Opus responses and 12 genuine no-code responses, lowering mean correctness from 0.92 to 0.67. GPT-5.5 produced neither behavior in its 59 non-truncated responses. These results are consistent with personas acting as Model-Dependent behavioral-policy biases rather than universal quality interventions. We release raw completions, derived scores, analysis artifacts, a pre-registration document, and an execution gate log; end-to-end test-based rescoring requires an unreleased task harness.
[NLP-38] am DACTYL at PAN 2026: Bayesian Data Mixing and Empirical X-risk Minimization for AI-text Detection
【速读】： 该论文旨在解决生成式 AI (Generative AI) 文本检测分类器在分布外（out-of-distribution, OOD）文本上性能显著下降的问题，这一现象表明现有模型存在对特定数据集特征的过拟合。其核心挑战在于：简单地合并多个训练数据集并不总能提升泛化能力，反而可能诱导模型学习到捷径依赖（shortcut learning）。为此，本文提出通过贝叶斯分类头（Bayesian classification heads）对 BERT-tiny 模型进行微调，以从三个不同数据集中精选高质量样本构建统一训练集，从而实现更鲁棒的泛化能力。关键解决方案包括：采用经验 X-风险最小化（empirical X-risk minimization）微调 DeBERTa-V3-large 与 ModernBERT-large 分类器，并引入 MCGrad 模型对 ModernBERT-large 的预测结果进行校准，以增强不确定性估计和模型可靠性。实验结果表明，经过精心数据集筛选与模型优化，所提方法在 PAN 2026 测试集上实现了优异的 OOD 性能，其中 MCGrad 达到 0.974 的平均得分，位列排行榜第二，验证了数据集精细化治理对提升模型泛化能力的关键作用。
链接: https://arxiv.org/abs/2607.17382
作者: Shantanu Thorat
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Accepted in CLEF 2026 Working Notes
Abstract:Existing research shows that AI-generated text detection classifiers achieve strong in-distribution (ID) performance but do not maintain the same performance on out-of-distribution (OOD) texts, suggesting overfitting to dataset-specific features. However, combining different training datasets doesn’t always improve performance and, in some cases, can even encourage shortcut learning. To address this issue, we fine-tune BERT-tiny models with Bayesian classification heads to select texts across three different datasets to use as a consolidated training set. We trained three different classifiers: fine-tuned DeBERTa-V3-large and ModernBERT-large classifiers via empirical X-risk minimization, and an MCGrad model that calibrates the predictions from the ModernBERT-large classifier. The DeBERTa-V3-large-large classifier achieves a mean score of 0.882 on the PAN 2026 test set across five metrics: AUROC, F_1 , C@1, Brier score, and F_0.5u . ModernBERT-large achieves a score of 0.96 while MCGrad achieves the best score of the three with a mean score of 0.974, ranking second on the leaderboard. Our results highlight that careful dataset curation can lead to strong OOD performance. We release our ModernBERT-large and DeBERTa-V3-large models at this https URL .
[NLP-39] WAR: Workload-Aware Rollouts for Synchronous Agent ic Reinforcement Learning
【速读】： 该论文旨在解决在代理强化学习（agentic reinforcement learning, RL）中，长时域回溯生成（long-horizon rollout generation）导致的系统性能瓶颈问题。随着代理与环境交互次数增加，轨迹长度迅速扩展至数万令牌，显著加剧了同步式RL训练中的回溯生成延迟。其解决方案的关键在于提出一种工作负载感知的回溯生成系统WAR（Workload-Aware Rollout），通过联合优化解码策略与调度机制，在不同负载条件下动态调整优化重点：在低负载场景下，采用无需额外模型的后缀推测解码（SuffixDecoding），复用已完成轨迹中的后缀模式作为未来回溯的推测草稿，避免引入额外推理模型及显存竞争；在高负载场景下，转向缓存感知的全局调度策略，依据缓存局部性、轨迹进度和服务器负载分布请求，减少重复的键值缓存（KV-cache）计算并缓解负载不均。通过解码层的后缀重用与系统层的智能调度协同，WAR在不同工作负载下均实现稳定吞吐量提升，在低负载下提升1.4倍，高负载下最高达1.6倍，有效消除了同步式长上下文代理训练中的主要回溯瓶颈，为可扩展的长上下文代理训练提供了实用路径。
链接: https://arxiv.org/abs/2607.17299
作者: Ryan Xu,Atlas Zhao,David Bao,Frank Du
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Operating Systems (cs.OS)
备注:
Abstract:Long-horizon rollout generation has become the dominant systems bottleneck in agentic reinforcement learning (RL). As agents interact with environments over many turns, trajectories rapidly grow to tens of thousands of tokens, making synchronous RL training increasingly constrained by rollout. We propose WAR, a workload-aware rollout system that substantially accelerates synchronous agentic RL by jointly optimizing decoding and scheduling. WAR is built on a key observation: the optimal rollout optimization strategy depends on runtime load: (1) Under low load, WAR enables model-free speculative decoding with SuffixDecoding, which reuses suffix patterns from previously completed trajectories as speculative drafts for future rollouts. Unlike model-based drafters, SuffixDecoding introduces no additional draft model and avoids GPU contention with rollout generation. (2) Under high load, where saturated batched decoding leaves limited room for speculative speedup, WAR shifts the optimization focus to cache-aware scheduling. A global scheduler places requests across rollout replicas based on cache locality, trajectory progress and server load, reducing redundant KV-cache recomputation and mitigating load imbalance. By combining decoding-level suffix reuse with system-level rollout scheduling, WAR delivers robust throughput improvements across workload regimes without changing the underlying RL algorithm. WAR improves long-context agentic rollout throughput by 1.4x under low load and up to 1.6x under high load. These results show that WAR removes a major rollout bottleneck in synchronous agentic RL and provides a practical path toward scalable long-context agent training.
[NLP-40] Safety That Does Not Transfer: Cross-Lingual Clinical Correctness Drift in Deployable Medical Language Models
【速读】： 该论文旨在解决在低资源医疗环境中，大语言模型（Large Language Models, LLMs）的临床安全性评估如何从高资源、以英语为主的语言环境迁移至本地化、小规模量化部署的模型的问题。研究关注的核心问题是：在尼日利亚北部高负担疾病（疟疾、镰状细胞病、结核病）背景下，临床安全性能是否能从英语有效转移到豪萨语（Hausa），以及性能下降是由语言差异、临床任务复杂性，还是可本地部署模型的类型所导致。其解决方案的关键在于构建了成对的英文-豪萨语问题集，针对知识回忆、紧急分诊、诱导禁忌行为及传统疗法主张等关键临床场景，对六种模型（五种本地可部署的4-90亿参数模型和一种前沿模型）进行双盲、独立评分的对比评估。结果显示，本地部署模型在豪萨语中的平均临床正确率从英语的1.57骤降至-0.03（满分2，负值表示有害），而前沿模型虽有所下降但仍保持较高水平（从2.00降至1.75），且未产生有害响应。跨任务一致性表明性能退化并非源于语言或临床内容本身，而是由可本地部署模型的层级特性所致，揭示了当前小规模量化模型在多语言低资源医疗应用中的系统性安全缺陷。
链接: https://arxiv.org/abs/2607.17270
作者: Anthonio Oladimeji Gabriel,Dimeji Olawuyi,Toba Ajayi,Temilola Aderemi
机构: 未知
类目: Computation and Language (cs.CL); Computers and Society (cs.CY)
备注:
Abstract:Safety evaluation of large language models is conducted predominantly in English and predominantly on frontier systems. Neither condition describes how such models are encountered in low-resource health settings, where small quantised systems are run locally and queried in local languages. We ask whether clinical safety established in English transfers to Hausa, and whether any failure is attributable to the language, the clinical task, or the class of model that low-resource deployment admits. Matched English-Hausa question pairs were built for three conditions of high burden in northern Nigeria: malaria, sickle cell disease, and tuberculosis, probing knowledge recall, emergency triage, a leading question inviting a contraindicated action, and a traditional-remedy claim. Six models were evaluated: five locally deployable systems of 4-9 billion parameters, two medically fine-tuned, and one frontier system. All 128 responses were scored against Nigerian national treatment guidelines by two fluent Hausa speakers working independently and blind to one another. Among locally deployable models, mean clinical correctness fell from 1.57 in English to -0.03 in Hausa, on a scale where 2 denotes a correct answer and -1 an actively harmful one. The frontier model moved from 2.00 to 1.75 and produced no response judged harmful in either language. Drift was consistent across all three conditions. Inter-rater agreement was substantial for clinical correctness (kappa = 0.70); agreement on harm was initially poor (kappa = 0.22) and is examined in detail. Because a frontier model answers the same questions competently in Hausa, the deficit is a property neither of the language nor of the clinical material, but of the deployable tier.
[NLP-41] Debate-on-Graph: Reliable and Adaptive Reasoning of Large Language Model on Uncertain Knowledge Graph ECML-PKDD2026
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在问答（Question Answering, QA）任务中因幻觉现象和缺乏相关知识而导致的推理不可靠问题。尽管知识图谱（Knowledge Graphs, KGs）被用于增强LLM的推理能力，但传统KG常包含噪声与错误，而现有方法难以识别并过滤这些低质量信息，反而可能加剧幻觉。为应对这一挑战，论文引入不确定知识图谱（Uncertain Knowledge Graphs, UKGs），通过为每个三元组赋予置信度分数以量化不确定性。其解决方案的关键在于提出一种名为“图上辩论”（Debate-on-Graph, DoG）的新框架，该框架实现LLMs与UKGs之间的自适应协同推理：首先设计针对UKGs的启发式搜索算法，以提取高置信度且与问题相关的子图，有效降低检索知识的噪声；随后引入多智能体辩论机制，通过自适应的对抗性辩论过程，充分挖掘UKGs中的可信知识，同时保障证据的可靠性。实验表明，DoG在四个基准QA数据集上均达到当前最优性能，实现了更可靠、灵活的推理能力。
链接: https://arxiv.org/abs/2607.17266
作者: Peiji Yu,Xin Chen,Tianxing Wu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 18 pages, Accepted by ECML-PKDD 2026
Abstract:Large language models (LLMs) have demonstrated remarkable capabilities in natural language processing. However, LLMs often suffer from hallucinations and lack of relevant knowledge when dealing with question answering (QA) tasks. To mitigate these issues, knowledge graphs (KGs) have been utilized to enhance LLM reasoning. Nevertheless, KGs often contain noise and errors, while existing KG-enhanced LLM approaches are generally unable to identify and filter such noisy and erroneous content, which can instead amplify hallucinations and pose challenges for reliable reasoning. Uncertain knowledge graphs (UKGs), which associate each triple with a confidence score to quantify uncertainty, offer a promising direction to address this challenge. Compared with prior work, we investigate how to leverage UKGs to support LLMs for QA. We propose Debate-on-Graph (DoG), a new framework that enables LLMs and UKGs to collaborate adaptively for reliable reasoning. Specifically, we first design a heuristic search algorithm tailored for UKGs to extract reliable and question-relevant subgraphs, thereby reducing noise and errors in retrieved knowledge. We then introduce a Multi-Agent Debate mechanism, which yields reliable answers through adaptive adversarial debates, aiming to fully exploit the knowledge in UKGs while preserving the reliability of retrieved evidence. Extensive experiments on four benchmark QA datasets show that DoG achieves state-of-the-art performance over existing LLM reasoning methods and KG-based baselines, while enabling reliable and adaptive reasoning. Our code is available at this https URL.
[NLP-42] Should Missing Modalities Always Be Necessary to Repair for Multi-modal Sentiment Analysis?
【速读】： 该论文旨在解决多模态情感分析（Multimodal Sentiment Analysis, MSA）在模态缺失情况下的关键问题：传统方法普遍采用“先修复再预测”的范式，但该假设未必适用于所有样本。研究表明，并非所有缺失模态都应被修复——仅有少数样本在完整模态输入下表现最优，而不同模态子集对不同样本更具优势，说明各模态的贡献具有样本依赖性。针对此问题，论文提出一种名为SIEVE（Sufficiency-Informed Evidential ValvE）的新框架，其核心创新在于将“是否修复缺失模态”转化为一个可学习的、样本级的显式决策机制。SIEVE通过并行比较直接预测分支与修复预测分支的损失差距，构建经验充分性信号，并利用证据门控机制联合建模模态充分性及其认知不确定性，从而实现动态路由。该方法具备修复无关性（repair-agnostic），可无缝集成于任意显式或隐式修复模块之上，无需修改其内部结构。实验结果表明，SIEVE在CMU-MOSI和IEMOCAP数据集上显著提升了多种代表性修复基线模型的性能，并逼近了每样本双分支可达到的最优解。
链接: https://arxiv.org/abs/2607.17262
作者: Yubo Gao,Haotian Wu,Xiaoyu Xu,Yibo Yan,Hong Chen,Ruoshui Peng,Fei Pan,Puay Siew Tan,Zhuoran Gao,Yonghua Hei,Jie Zhang,Xuming Hu
机构: The Hong Kong University of Science and Technology (Guangzhou); The Hong Kong University of Science and Technology; Nanyang Technological University; Singapore Institute of Manufacturing Technology, A*STAR; Lingnan University
类目: Computation and Language (cs.CL)
备注:
Abstract:Existing methods for multimodal sentiment analysis (MSA) under missing modalities usually follow a repair-first paradigm. We revisit this assumption and ask: \emphshould every missing modality be repaired? A per-sample oracle analysis shows the answer is not always: full-modality input is optimal for only a small fraction of samples, and every modality subset is preferred by some samples. These results suggest that adding or repairing modalities may not always improve prediction, and that the utility of each modality is sample-dependent. Building on this finding, we propose \textbfSufficiency-\textbfInformed \textbfEvidential \textbfVal\textbfvE (\textbfSIEVE) that turns ``whether to repair’’ into an explicit, learnable decision at the sample level. SIEVE compares a direct prediction branch with a repair branch, derives an empirical sufficiency signal from their per-sample loss gap, and routes each input through an evidential gate that jointly models sufficiency and its epistemic uncertainty. SIEVE is repair-agnostic: it operates as a plug-and-play decision on top of any explicit or implicit repair module, without modifying its internal design. Experiments on CMU-MOSI and IEMOCAP show that SIEVE consistently improves representative repair backbones across evaluated missing rates, and approaches the per-sample dual-branch achievable optimum.
[NLP-43] EvolvingWorld: An Open-Schema Framework for Co-Evolving Role-Play Agents and World Model in Interactive Literary World
【速读】： 该论文旨在解决交互式文学世界中角色与世界协同演化（character and world co-evolution）的建模难题，现有系统多将交互式文学模拟简化为静态人格模仿或孤立场景生成，无法捕捉角色与世界随时间动态演化的复杂关系。其解决方案的关键在于提出EvolvingWorld框架，采用开放架构（open-schema framework）支持跨多元文学世界的长期演化模拟，并通过两个耦合模块实现协同演化：一是基于大语言模型（LLM）的角色代理（Character Agent），负责多角色扮演及角色属性的持续演化；二是基于大语言模型的世界模型（World Model），用于维护全局与局部实体层面的状态并推进场景发展。该框架构建了7个可训练任务以实现场景初始化、交互生成与状态更新，并基于57部文学作品构建大规模数据集（138,596条监督样本和222个测试快照），同时引入基于轨迹级别的“大语言模型作为裁判”（LLM-as-Judge）评估协议，涵盖10个维度与20项指标，实验表明该框架能有效维持角色与世界在长时程模拟中的持久性与连贯性。
链接: https://arxiv.org/abs/2607.17250
作者: Qing Zong,Yue Guo,Mengxin Yang,Yiwen Guo,Yangqiu Song
机构: Hong Kong University of Science and Technology (香港科技大学); LIGHTSPEED; Huazhong University of Science and Technology (华中科技大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:This paper introduces EvolvingWorld, a framework and benchmark for character and world co-evolution in interactive literary worlds. Existing systems either treat interactive literary simulation as static persona imitation or isolated scene generation, failing to capture how characters and worlds evolve together over time. To address this, EvolvingWorld models literary simulation as a long-horizon process where characters interact, scenes progress, and character and world states are persistently updated. Unlike prior systems relying on fixed schemas, EvolvingWorld adopts an open-schema framework to support simulation across diverse literary worlds. The framework consists of two coupled modules: a Character Agent for multi-character role-play and persistent profile evolution, and an LLM-based World Model for global and location/entity-level state maintenance and scene progression. Based on this architecture, we formulate 7 trainable tasks for scene initialization, interaction generation, and state update. We construct a dataset from 57 books, producing 138,596 supervised training samples and 222 snapshots for testing. Furthermore, we introduce a trajectory-level LLM-as-Judge evaluation protocol spanning 10 dimensions and 20 metrics. Experiments show that EvolvingWorld can improve long-horizon simulation by effectively maintaining persistent, coherent character and world development.
[NLP-44] AI_LectureNote: A Retrospective Pilot Study of a Post-ASR Workflow for English-Script Rendering and Semantic Drift in Korean-English Medical Lectures
【速读】： 该论文旨在解决韩语-英语医学讲座在自动语音识别（ASR）后处理阶段中，生成文本的可读性与术语准确性不足的问题。其核心挑战在于：现有ASR系统将医学术语以韩语音译形式输出，导致英文术语丢失，影响专业学习材料的可用性。为此，论文提出了一种面向可读性的后ASR工作流——AI_LectureNote，通过重写语音转文字结果，将非标准的韩语音译医学术语还原为拉丁字母书写的英文术语。该方案的关键在于分步优化不同输入路径下的术语呈现质量：在Whisper-1路径上，后处理使英文术语渲染率从0.39提升至0.71；在3分钟分块的GPT-4o转录输出上，该比率从0.26升至0.65。然而，研究也揭示了英文术语渲染率提升并不等同于语义忠实度的提高，存在显著的语义漂移（34/282句）和语义极性误判（11/101行），且不同前端输入引发的失败模式具有差异性——极性错误在不同前端间重叠度更高（Jaccard 0.60），而一般性语义漂移则表现出较低的一致性（Jaccard 0.23）。因此，该研究强调应将表面准确性、术语书写格式、分块级别书写一致性及医学含义保真度分别评估，为后续改进提供可量化的故障模式分析框架。
链接: https://arxiv.org/abs/2607.17237
作者: Kyeongeon Lee,Donghoon Chang,Seungryeol Baek,Taehong Kim,Wonjun Yang
机构: Sungkyunkwan University (成均馆大学)
类目: Computation and Language (cs.CL)
备注: 12 pages, 4 figures
Abstract:AI_LectureNote is a historical, readability-oriented post-ASR workflow for Korean-English medical lectures. It rewrites speech-to-text output into study transcripts while restoring Latin-script medical terms rather than Korean phonetic transliterations. We retrospectively evaluate the workflow on four author-recorded lectures across five conditions. In this pilot, post-processing raised the macro English-script rendering rate from 0.39 to 0.71 on the whisper-1 path and from 0.26 to 0.65 when applied to 3-minute chunked gpt-4o-transcribe output. However, English-script rendering did not imply semantic faithfulness: the two post-processed conditions showed semantic drift in 34 and 36 of 282 reference sentences and polarity failures in 11 and 13 of 101 polarity-cue rows. A descriptive cross-input comparison suggested different candidate failure patterns: polarity-failure sets overlapped more strongly across front-ends (Jaccard 0.60; 9 shared of 15 unioned failures) than general semantic-drift sets (Jaccard 0.23; 13 shared of 57 unioned drifts). This single-annotator pilot documents concrete failure modes rather than population rates and supports evaluating surface accuracy, term-script rendering, chunk-level script consistency, and medical-meaning preservation separately.
[NLP-45] Robust Summarization of Doctor-Patient Conversations: TalTech Systems for the Beyond Transcription Challenge
【速读】： 该论文旨在解决从长时医生-患者对话录音中直接生成结构化SOAP病历（SOAP notes）的挑战，其核心问题在于避免传统流程中依赖中间文本转录环节所带来的误差累积与效率损失。为实现端到端生成，研究者首先筛选具备长音频鲁棒性的开源语音大模型（speech LLM），并针对轻量级（Voxtral Mini）与重量级（Voxtral Small）两个赛道分别采用LoRA监督微调（supervised fine-tuning）与基于奖励函数的强化学习（DAPO reinforcement learning）进行优化，其中奖励信号直接使用挑战赛指标——开放医学概念F1（Open Medical Concept F1）作为对齐目标。该方案的关键在于将医学概念匹配精度作为强化学习的显式奖励，实现了生成质量与事实可靠性之间的平衡，最终在两个赛道均取得第一名，且独立评估显示其幻觉率最低，表明基于概念匹配度的强化学习可有效维持生成内容的真实性。此外，研究还发现：在文本转录上进行微调的模型能够良好迁移至语音输入场景，并显著提升在真实、跨领域录音上的鲁棒性。
链接: https://arxiv.org/abs/2607.17230
作者: Aivo Olev,Tanel Alumäe
机构: 未知
类目: Computation and Language (cs.CL); Audio and Speech Processing (eess.AS)
备注: SLT 2026 BeTraC
Abstract:This paper describes TalTech’s submissions to the Beyond Transcription Challenge (BeTraC), which requires generating SOAP notes directly from long doctor-patient conversation recordings, without intermediate transcription. After screening open-weight speech LLMs for long-audio robustness, we adapted Voxtral Mini (lightweight track) and Voxtral Small (heavyweight track) with LoRA supervised fine-tuning followed by DAPO reinforcement learning that uses the challenge metric, Open Medical Concept F1, as its reward. Our systems ranked first in both tracks, and an independent LLM-as-a-judge evaluation showed the lowest hallucination rate among all submissions, indicating that reinforcement learning against a concept-matching metric need not compromise factual reliability. We also find that fine-tuning on text transcripts transfers well to speech input and appears to improve robustness on out-of-domain real recordings.
[NLP-46] Literary Non-Style in LLM -Generated Text
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）生成文本在风格与语义层面与人类写作存在显著差异的问题。现有研究已表明，LLM生成文本在量化特征和质性表现上均偏离人类写作风格，且其语义范围远窄于人类表达。本文通过分析LLM生成文本中n-gram的统计分布模式，揭示出简单但一致的规律性特征，进而借助对这些n-gram的定性分析，暴露了模型在风格层面的系统性缺陷。由于高阶n-gram与语义内容密切相关，研究进一步指出，风格与语义问题并非可清晰分离的独立维度，二者具有内在耦合性。其解决方案的关键在于：通过识别并分析n-gram分布中的结构性偏差，揭示风格与语义之间的深层关联，从而为评估和改进生成文本的真实性与自然性提供新的理论路径。
链接: https://arxiv.org/abs/2607.17228
作者: Cory Massaro
机构: 未知
类目: Computation and Language (cs.CL)
备注:
Abstract:Prior work on LLM-generated text has demonstrated quantitative and qualitative departures from text produced by humans. LLM-generated texts differ from human writing in style, resulting in a characteristic textual “feel,” while the semantic range of LLMs is much restricted compared to that of humans. In this contribution, I note simple but consistent patterns in the statistical distribution of n-grams within LLM-generated text. Via qualitative analysis of these n-grams, I reveal deficiencies in LLM style. Because higher-order n-grams correlate to semantic content, I conclude that questions of style and semantics are not cleanly separable.
[NLP-47] Auditing Question-Order Effects in Large Language Models with the QQ Equality: Mechanism Characterization and a Saturation Caveat
【速读】： 该论文旨在解决大语言模型（LLM）在顺序判断任务中是否存在类似人类问卷调查中的“问题顺序效应”（question-order effects）的问题，尤其关注这些效应是否符合量子提问模型（projective quantum question-order model）所预测的无参数理论——QQ等式（QQ equality）。其核心问题是：如何对自回归大语言模型的顺序响应行为进行可信赖的审计，以识别其是否表现出与人类认知一致的上下文依赖性或非经典概率特征。解决方案的关键在于将QQ等式发展为一种可操作的审计标准，并构建一套预设、可记录的评估流程。该方法通过理论分析确定了满足QQ等式的机制类别，包括边际独立核需满足四类错配转移率一致、极性与位置相关的重复族需满足精确交叉对称性条件、以及秩-2的上下文依赖性由默认准则（Contextuality-by-Default）转化为审计坐标 $ |\qQQ| \le \OSS $（其中 \OSS 为顺序敏感度得分，衡量两个边缘分布的顺序敏感性总和）。在方法论上，提出了一套结合最坏情况鲁棒性包络、采样一致性检验、全标签反平衡设计及饱和度诊断的预设审计流水线。实证结果显示，在对一个开放权重的指令微调模型进行初步测试时，尽管所有预设健康阈值均通过，但绝大多数项目对（17/18 和 7/8）呈现饱和状态（近确定性），且无任何项目被认证为残余上下文依赖性，表明在当前模型与提示条件下，仅依赖二元化下一词概率无法实现分布级的QQ审计；因此，研究建议在将下一词分布视为调查响应分布时，必须加入预设的饱和度诊断。
链接: https://arxiv.org/abs/2607.17219
作者: Pilsung Kang
机构: Dankook University (단국대학교)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Quantum Physics (quant-ph); Methodology (stat.ME)
备注:
Abstract:Human survey respondents exhibit question-order effects that satisfy the QQ (quantum question) equality, an a priori, parameter-free prediction of the projective quantum question-order model. We develop the QQ equality into an audit criterion for sequential judgments of autoregressive large language models (LLMs). Theoretically, we characterize which mechanism classes satisfy it robustly: marginal-independent kernels satisfy QQ iff all four mismatch transition rates coincide (a class containing the 2D rank-1 projective model with a fixed measurement pair under state variation); a polarity- and position-dependent repetition family is characterized by an exact cross-symmetry condition with closed-form violations; QQ-satisfying behaviors are closed under order-matched mixing; and the rank-2 Contextuality-by-Default criterion translates into audit coordinates as |\qQQ|\le\OSS , where \OSS (the order-sensitivity score) totals the order sensitivity of the two marginals. Methodologically, we develop a pre-specified, audit-logged pipeline applicable to any model exposing next-token log-probabilities; it combines worst-case robustness envelopes, sampling-consistency spot checks, full label counterbalancing, and a saturation diagnostic. Empirically, in a first-signal pilot on an open-weight instruction-tuned model under two framings, all pre-specified health gates passed, yet 17/18 and 7/8 item pairs, respectively, were saturated (near-deterministic), and no item was certified residually contextual. Forced-binary next-token log-probabilities were thus inadequate for distribution-level QQ audits under the tested model and prompting conditions; we recommend pre-specified saturation diagnostics whenever next-token distributions are treated as survey-response distributions.
[NLP-48] KyrgyzLLM -Bench: Benchmarking Kyrgyz Language Understanding
【速读】： 该论文旨在解决多语言大语言模型（LLM）评估中因依赖英文数据集翻译而导致的目标语言语言与文化特异性丢失的问题，尤其针对资源匮乏语言如吉尔吉斯语（Kyrgyz）缺乏可靠原生语料的困境。其核心解决方案是构建首个系统性、大规模的吉尔吉斯语大语言模型评估基准——KyrgyzLLM-Bench，该基准包含两个原生撰写的语料库（KyrgyzMMLU 和 KyrgyzRC），以及经过精心翻译并人工校对的 WinoGrande、HellaSwag、BoolQ 和 TruthfulQA 的吉尔吉斯语版本。通过在零样本和少样本设置下评估26个开源与闭源模型，研究揭示了模型性能在不同任务间的跨语言迁移规律，并量化了翻译伪影对评估可靠性的影响。关键发现表明，模型排名在 WinoGrande 与 BoolQ 上从英语向吉尔吉斯语具有较强迁移性，而在 MMLU 上较弱，而 HellaSwag 则表现出显著的英-吉性能差距，归因于翻译引发的语义合理性偏移。此外，少样本提示对部分开源模型的阅读理解能力有提升作用，但对闭源模型在翻译任务上的表现则不一致。研究团队公开发布了所有数据集、评估代码及模型结果，并将吉尔吉斯语任务集成至主流多语言评估框架，为未来吉尔吉斯语自然语言处理研究提供坚实基础。
链接: https://arxiv.org/abs/2607.17173
作者: Timur Turatali,Aida Turdubaeva,Rustem Izmailov,Anton M. Alekseev,Sergey I. Nikolenko
机构: The Cramer Project; Institute of IT, Kyrgyz State Technical University named after I. Razzakov(吉尔吉斯斯坦技术大学伊·拉扎科夫命名); School of Computer Science, University of Windsor(温莎大学计算机科学学院); St. Petersburg Department of the Steklov Math. Institute, RAS(圣彼得堡斯捷克洛夫数学研究所俄罗斯科学院分部); St. Petersburg State University(圣彼得堡国立大学)
类目: Computation and Language (cs.CL)
备注: Preprint; manuscript currently under consideration at a journal
Abstract:Evaluating large language models (LLMs) across languages remains challenging, as most multilingual benchmarks rely on translated English datasets, often obscuring linguistic and cultural specificity in the target language. This issue is particularly pronounced for less-resourced languages such as Kyrgyz, where reliable natively authored evaluation data are scarce. Building on previously introduced Kyrgyz-language evaluation datasets, this work reports the first systematic and large-scale evaluation of LLMs in Kyrgyz using the KyrgyzLLM-Bench benchmark suite. KyrgyzLLM-Bench comprises two natively authored datasets - KyrgyzMMLU and KyrgyzRC - together with carefully translated and manually post-edited versions of WinoGrande, HellaSwag, BoolQ, and TruthfulQA. We evaluate 26 open- and closed-source LLMs under zero-shot and few-shot settings, analyzing model performance, cross-lingual transfer, and the impact of translation artifacts on evaluation reliability. Across families and tasks, model rankings transfer broadly from English to Kyrgyz on WinoGrande and BoolQ, and to a lesser extent on MMLU, while HellaSwag exhibits a substantial English-Kyrgyz performance gap consistent with translation-induced plausibility shifts. Few-shot prompting improves several open-source models on reading comprehension but behaves inconsistently for proprietary models on translated tasks. We publicly release all datasets, evaluation code, and per-model results, and integrate the Kyrgyz tasks into a widely used multilingual evaluation framework to support future research on Kyrgyz NLP.
[NLP-49] How Jailbreak Attacks Inform Safety Alignment: A Defender-Centric Shapley-Based Evaluation of Jailbreak Contributions
【速读】： 该论文旨在解决现有大型语言模型（LLM）越狱攻击评估体系中过度依赖攻击者中心指标（如攻击成功率，ASR）所导致的局限性——即高成功率的攻击未必能有效提升模型安全性。其核心问题在于：如何更合理地评估越狱攻击的实际价值，使其不仅作为突破模型防御的工具，更能转化为提升模型安全性的可复用资源。解决方案的关键是提出一种以防御者为中心的评估范式，并引入A-MESS（最小有效攻击子集选择）框架，该框架基于黑盒子集效用观测，通过Shapley值衍生的AttackSHAP评分，量化单个攻击对安全训练的边际贡献，并结合贪婪或代理优化策略，在用户指定预算下高效筛选出最具安全提升潜力的紧凑攻击子集。实验表明，攻击成功率与防御者视角下的实际安全收益相关性较弱，而AttackSHAP可在少量效用查询下准确估计，且直接优化攻击子集比仅基于攻击者指标或仅依赖归因的方法带来更强的安全性改进。这表明应将越狱攻击视为提升模型安全性的战略资源，而非单纯的破坏性工具。
链接: https://arxiv.org/abs/2607.17152
作者: Yukai Zhou,Feiyang Lu,Xiaokai Mao,Jinfei Liu,Wenjie Wang
机构: ShanghaiTech University (上海科技大学); Zhejiang University (浙江大学)
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注:
Abstract:Jailbreak attacks on large language models are usually evaluated by attacker-centric metrics such as attack success rate (ASR), yet an attack that breaks a model is not necessarily useful for improving its safety. We propose a defender-centric view of jailbreak evaluation, where attacks are evaluated by the downstream safety improvements they enable when used as red-teaming data for safety training. Building on this view, we introduce A-MESS (Minimal Effective Attack-Subset Selection), a setting-agnostic framework for attributing and selecting jailbreak attacks from black-box subset utility observations. A-MESS estimates AttackSHAP, a Shapley-based score that attributes marginal utility to individual attacks and selects compact attack subsets under user-specified budgets via greedy or surrogate-based optimization. Across controlled utility landscapes and real LLM safety settings, we find that ASR rankings are weakly aligned with defender-centric utility, that AttackSHAP can be estimated accurately with limited utility queries, and that directly optimizing subsets yields stronger safety utility than attacker-centric or attribution-only selection. These results suggest evaluating jailbreak attacks as resources for improving safety, not only as tools for breaking models.
[NLP-50] SlotGuard: Stop Oversharing Private Local Context in LLM Agent Transcri ICML2026
【速读】： 该论文旨在解决大语言模型（Large Language Model, LLM）智能体在运行过程中因将敏感信息（如路径、电子邮件、API密钥等）作为观察结果（如工具输出、Shell日志、文件读取内容）记录到与服务提供商绑定的对话转录中而导致的隐私泄露问题。现有基于占位符的脱敏方法存在诸多缺陷：无法有效识别嵌入式或跨轮次引用，容易误删无害相似内容，并破坏对推理任务至关重要的结构信息。其解决方案的关键在于提出一种名为SlotGuard的本地转录边界机制，通过将结构化绑定重写为具有类型和后缀感知能力的槽位（slot），以格式保持的方式用合成值替换秘密信息，利用轻量级会话图链接跨轮次引用，并仅在受信任的运行时环境中恢复原始值。实验表明，在面向代码仓库的受控智能体转录数据上，SlotGuard成功移除全部20,814个标注的结构敏感字符（覆盖9,229条路径），并将凭证泄露率降至0.0%（针对852个植入值），同时在四个上游模型上维持接近原始转录的任务成功率，显著优于通用脱敏方法（后者降至2.5%）。转录重写的平均耗时仅为每轮14.424微秒，具备高效实用性。
链接: https://arxiv.org/abs/2607.17147
作者: Haocheng Xia,Yongjoo Park
机构: 未知
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注: ICML 2026 AIWILD
Abstract:LLM agents can leak privacy (e.g., paths, emails) and credentials (e.g., API keys) as agent observations (e.g., tool outputs, shell logs, and file reads) are appended to provider-bound transcripts. Existing placeholder redaction is brittle: it can miss embedded or cross-turn references, over-redact benign lookalikes, and destroy the structure useful for reasoning. We present SlotGuard, a local transcript boundary that can hide sensitive data while retaining agents’ performance. SlotGuard rewrites structural bindings as typed, suffix-aware slots, replaces secrets with format-preserving synthetic values, links cross-turn references with a lightweight session graph, and restores raw values only inside the trusted runtime. On controlled repository-oriented agent transcripts, SlotGuard removes all 20,814 annotated structurally sensitive characters across 9,229 paths and reduces credential leakage to 0.0% across 852 planted values. It remains close to raw-transcript task success across four upstream models, while generic redaction drops to 2.5%. Transcript rewriting takes a median of 14.424~ \mu s per agent turn. The code is publicly accessible at this https URL.
[NLP-51] SynH-Rank: Quality-Aware Code Search via Diverse Data Synthesis and Hierarchical Ranking Training
【速读】： 该论文旨在解决现有代码搜索系统在语义相关性评估中忽视非功能性质量（如执行速度、内存使用和可维护性）的问题，这类质量因素对实际软件开发至关重要。当前方法依赖于“检索-重排”流水线，其重排模块通常仅关注查询与代码片段之间的语义相关性，而忽略了高质量代码的优先排序需求。为应对这一挑战，论文提出SynH-Rank——一种面向质量感知的代码重排框架，其核心创新在于结合大语言模型（LLM）驱动的多样化数据合成与分层排名训练机制。关键解决方案包括：(1) 设计三层次标注体系（高质量相关、低质量相关、无关），显式建模代码质量的层级关系；(2) 引入新的对比学习目标，以捕捉高质量、低质量和无关代码之间的序数关系，突破传统二元对比学习在质量区分上的局限性。此外，研究构建了一个包含4,209对样本的新基准，并提出两个新指标：质量偏好准确率（QPA）用于评估高质代码的优先级排序能力，多条件准确率（MCA）用于衡量复杂场景下的泛化性能。实验结果表明，SynH-Rank在QPA上相较基线模型提升20.15%，优于标准相关性对比训练15.80%，同时在传统相关性指标和多条件泛化能力方面也表现出显著优势。
链接: https://arxiv.org/abs/2607.17139
作者: Keyu Liang,Haoye Wang,Yanfu Yan,Zhiyuan Wan,Zhongxin Liu
机构: Zhejiang University (浙江大学); Hangzhou City University (杭州城市学院); Hangzhou High-Tech Zone (Binjiang) Institute of Blockchain and Data Security (杭州高新区（滨江区）区块链与数据安全研究院)
类目: oftware Engineering (cs.SE); Computation and Language (cs.CL)
备注:
Abstract:Code search enhances developer productivity by enabling efficient code reuse. Current code search systems often use a retrieve-then-rerank pipeline, where rerankers focus on modeling semantic relevance between queries and code. However, these rerankers overlook critical non-functional qualities like execution speed, memory usage, and maintainability, which are essential for practical software development. Studies reveal developers expect results to maintain high coding standards and satisfy specific needs, such as resource optimization, highlighting the importance of quality-aware code search. Achieving quality-aware code search faces two major challenges: the scarcity of quality-annotated datasets for effective training and the limitations of standard contrastive learning objectives, which fail to capture the ordinal relationships among high-quality, low-quality, and irrelevant code. Although contrastive learning excels in distinguishing relevant from irrelevant code, its binary objective does not support nuanced quality this http URL address these challenges, we propose SynH-Rank, a quality-aware code reranking framework that combines LLM-driven diverse data synthesis with hierarchical ranking training. SynH-Rank employs a three-level labeling scheme to explicitly model the hierarchy: high-quality relevant low-quality relevant irrelevant. Additionally, we introduce a new benchmark with 4,209 pairs and two novel metrics: Quality Preference Accuracy (QPA) for assessing prioritization of high-quality code and Multi-Condition Accuracy (MCA) for evaluating performance under complex this http URL results show SynH-Rank improves QPA by 20.15% over backbone models and outperforms standard relevance-only contrastive training by 15.80%, while simultaneously enhancing traditional relevance metrics and multi-condition generalizability. Subjects: Software Engineering (cs.SE); Computation and Language (cs.CL) Cite as: arXiv:2607.17139 [cs.SE] (or arXiv:2607.17139v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2607.17139 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-52] Scope3Trace: Evidence-Based Identification and Extraction of Scope 3 GHG Emissions from Sustainability Reports
【速读】： 该论文旨在解决企业在核算范围三（Scope 3）温室气体（GHG）排放时面临的可扩展性分析难题，主要挑战包括披露数据稀疏、报告文档格式异构以及证据溯源能力有限。现有方法多依赖大语言模型（LLM）从环境、社会与治理（ESG）报告中提取排放信息，但普遍存在缺乏显式证据支撑或需高成本人工标注与验证以保证提取可靠性的问题。为此，本文提出Scope3Trace——一种基于证据的可信信息抽取框架，其核心在于通过整合文档信息抽取流水线，实现对真实世界可持续发展报告中组织级与建筑级范围三排放信息的可解释、可追溯提取。该框架融合了PDF采集与光学字符识别（OCR）解析、基于LLM的页面定位与表格重建，以及混合规则-大语言模型的抽取机制，并引入证据驱动的验证流程，确保结果的可信度。在此基础上，研究构建了一个双层级、基于证据、多模态的数据集，涵盖来自异构报告的组织级范围三排放披露内容。Scope3Trace实现了对范围1-3总排放量及类别级披露的高精度提取，显著提升了异构可持续性披露信息的可靠抽取与透明集成能力。
链接: https://arxiv.org/abs/2607.17122
作者: Siyuan Zheng,Yifan Duan,Chao Xue,Flora D. Salim
机构: UNSW Sydney (新南威尔士大学)
类目: Computation and Language (cs.CL)
备注: 23 pages
Abstract:Scope 3 greenhouse gas (GHG) emissions account for the majority of corporate carbon footprints, yet remain difficult to analyze at scale due to sparse disclosures, heterogeneous report document formats, and limited evidence traceability. Existing approaches typically rely on large language models to extract emissions information from ESG reports, but often lack explicit evidence grounding or depend on costly manual annotation and verification to ensure extraction reliability. To address these challenges, we propose Scope3Trace, an evidence-grounded information extraction framework designed to extract interpretable and traceable Scope 3 emissions information from real-world ESG and sustainability reports. The framework integrates a document information extraction pipeline that performs PDF collection and OCR parsing, LLM-assisted page localization and table reconstruction, and hybrid rule-LLM extraction of organization- and building-level emissions disclosures with evidence-grounded verification. Building upon this framework, we further contribute a dual-level, evidence-grounded, multimodal dataset comprising organization-level Scope 3 disclosures extracted from heterogeneous sustainability reports. Scope3Trace enables reliable extraction and transparent integration of heterogeneous sustainability disclosures, achieving high accuracy in extracting Scope 1-3 totals and category-level disclosures from sustainability reports.
[NLP-53] Persistent Sparse Autoencoders: Learning Feature Timescales in Language Models
【速读】： 该论文旨在解决标准稀疏自编码器（Sparse Autoencoders, SAEs）在语言模型激活值分解中忽略序列间持续性信息的问题。标准SAEs独立编码每个词元，无法捕捉跨序列持续存在的语义特征，限制了对长期依赖和上下文关联的建模能力。为此，论文提出了一种新型架构——持久稀疏自编码器（Persistent Sparse Autoencoders, Persistent SAEs），其关键在于为每个特征引入一个“持续系数”（persistence coefficient），使模型能够学习哪些特征应保持活跃以及持续时长。实验表明，该方法在保持竞争性重建质量的同时，能够自动学习不同时间尺度的特征：快速特征表现为局部可解释的检测器，而慢速特征则将主题级语义信息集中于持久状态。此外，在提示注入监控的案例研究中，慢速特征能有效保留检测信号，并在长上下文场景中维持因果有效性。这些结果表明，持久稀疏自编码器通过引入具有时间持续性的语义表征，为语言模型的可解释性分析与行为监控开辟了新路径。
链接: https://arxiv.org/abs/2607.17117
作者: Haoyan Luo,Mateo Espinosa Zarlenga,Mateja Jamnik
机构: University of Cambridge(剑桥大学); University of Oxford(牛津大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:
Abstract:Sparse autoencoders (SAEs) decompose language model activations into sparse features, but standard SAEs encode each token independently and do not expose information that persists across a sequence. We introduce Persistent Sparse Autoencoders (Persistent SAEs), which extend standard SAEs by learning a persistence coefficient for each feature, allowing the model to learn which features should persist and for how long. Our experiments show that they retain competitive reconstruction quality while learning a spectrum of feature timescales: fast features behave as locally interpretable detectors, whereas slow features concentrate topic-level information in a persistent state. Moreover, as shown in a prompt-injection monitoring case study, slow features preserve detection signals and remain causally effective over long contexts. These results suggest that Persistent SAEs open up new opportunities for interpreting and monitoring language models through persistent semantic representations.
[NLP-54] BLAD: A Historically Contextualized Multilingual Dataset of Bangladeshi Legal Acts (1799 to 2025)
【速读】： 该论文旨在解决南亚低资源、大陆法系司法管辖区在法律自然语言处理（Legal NLP）领域长期存在的语料资源匮乏问题，尤其聚焦于孟加拉国自1799年至2025年间立法文本的系统性缺失。其核心挑战在于缺乏结构化、多语言且具有时间维度标注的立法语料库，限制了对区域法律演变规律的量化分析与模型训练。解决方案的关键在于构建首个全面的孟加拉国法律法案数据集（Bangladesh Legal Acts Dataset, BLAD），涵盖1,484部立法文件，每部法案均以完整文本形式呈现，并包含结构化章节、脚注、废止状态及元数据信息，可关联至颁布时的统治政权、国家元首与法律体系。该数据集整合了英语、孟加拉语及混合语言文档，支持跨时间与多语言的法律文本分析，为后续法律语义理解、历史法规变迁建模与生成式法律智能应用提供了基础支撑。
链接: https://arxiv.org/abs/2607.17111
作者: Adib Sakhawat
机构: Islamic University of Technology (伊斯兰大学技术学院)
类目: Computation and Language (cs.CL)
备注: 9 pages, 6 figures
Abstract:We present the Bangladesh Legal Acts Dataset (BLAD), a curated collection of 1,484 legislative acts enacted between 1799 and 2025. Each act is represented with its full text, structured sections and footnotes, repeal status, and metadata linking it to the governing regime, head of state, and prevailing legal framework at the time of enactment. The corpus spans English, Bengali, and mixed-language documents, supporting temporal and multilingual analysis of statutory law. BLAD addresses a persistent gap in legal natural language processing (NLP) resources for low-resource, civil-law jurisdictions in South Asia. We describe the acquisition and enrichment pipeline, report descriptive statistics over more than two centuries of legislation, and outline the research directions the corpus enables. The dataset is publicly available under the CC~BY-SA~4.0 license at this https URL.
[NLP-55] Otap:Structure-Aware Optimal Transport for Evaluating Planning and Execution in Agent Trajectories
【速读】： 该论文旨在解决大语言模型智能体（Large Language Model Agents）在执行任务时生成的规划轨迹（trajectory）评估问题。现有评估方法将轨迹简化为二元成功标志或通过精确匹配与参考轨迹对比，前者无法区分合理解决方案与偶然成功，后者对计划顺序或分解方式的合理差异过于敏感。本文提出将轨迹评估重构为代理执行图与一组有效解图之间的距离度量，并通过属性依赖图上的非平衡融合格罗莫夫-沃瑟斯坦（unbalanced fused Gromov-Wasserstein）传输问题进行实例化，得到一种称为\otap（Optimal Transport for Agentic Planning）的伪度量。其关键创新在于：该方法对保持依赖关系的重排序具有不变性，对冗余步骤具有有界敏感性；非平衡边缘可处理缺失或幻觉步骤而不强制完全匹配；软耦合机制支持计划粒度的灵活变化。在受控扰动及三个公开基准测试中，\otap 在语义仅依赖的指标表现低于随机水平的区间内仍能有效区分有效与无效轨迹，且当依赖图被精确恢复时性能最优，仅在通过自由文本轨迹启发式推断图结构时性能下降。
链接: https://arxiv.org/abs/2607.17082
作者: Babak Barazandeh,Subhabrata Majumdar,George Michailidis
机构: Fortinet(飞塔); OpenAI; Meta; Stability.AI; Anthropic; Character.ai; Claude
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Large language model agents solve tasks by generating trajectories that interleave planning, tool calls, and intermediate results. Current evaluation metrics reduce such a trajectory to a binary success flag or compare it against a reference by exact matching. A success flag cannot distinguish a sound solution from one that succeeds by luck, and says nothing about why a failed run went wrong. Exact matching penalizes plans that are valid but reordered or decomposed differently from the reference. We reframe trajectory evaluation as a distance between the agent’s execution graph and a set of valid solution graphs, and instantiate it via an unbalanced fused Gromov-Wasserstein transport problem over attributed dependency graphs. The resulting score, termed \otap (Optimal Transport for Agentic Planning), is a pseudo-metric that is provably invariant to dependency-preserving reorderings and has bounded sensitivity to redundant steps. Its unbalanced marginals handle missing or hallucinated steps without forcing a match, and its soft coupling accommodates variation in plan granularity. On controlled perturbations and three public benchmarks, \otap separates valid from invalid trajectories in a regime where semantics-only metrics score below chance. Its accuracy is highest when the dependency graph is recovered exactly, and drops only when the graph is inferred heuristically from free-text traces.
[NLP-56] A Systematic Evaluation of Traditional Privacy Policy Analysis Tools Against LLM s
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在隐私政策与数据合规分析领域是否能够替代传统专用工具的核心问题。具体而言，研究关注的是现成的大型语言模型（LLM）在无需额外工程改造或领域特定训练的情况下，能否全面复现以往依赖专业工具才能实现的多样化功能，包括矛盾检测、法规合规性分析以及隐私政策摘要与聚合等关键任务。其解决方案的关键在于开展首次系统性评估，通过构建包含10份真实隐私政策的定制数据集，对比两种前沿大型语言模型（GPT-5.2与Gemini-2.5）在六种代表性工具所覆盖的三大核心功能及三类中间任务（如基于元组的结构化数据提取、语义角色标注及人工隐私政策标注）上的表现。实验结果表明，这些通用大模型在多数任务中不仅达到甚至超越了专用工具的性能水平，尤其在第一方数据收集主体和第三方共享主体的人工标注任务中分别实现了81.8%与91.4%的平均精确率，展现出在无须领域微调的前提下完成复杂隐私分析任务的强大潜力，从而证明生成式AI具备替代传统专用工具的可行性与有效性。
链接: https://arxiv.org/abs/2607.17075
作者: Madhav Aryal,Sudipa Saha,Kaushal Kafle,Anshuman Chhabra,Sunil Manandhar
机构: University of South Florida; IBM T.J. Watson Research Center
类目: Cryptography and Security (cs.CR); Computation and Language (cs.CL)
备注:
Abstract:The advent of LLMs has significantly changed the research on privacy policy and data compliance analysis by enabling tasks that previously required specialized, domain-specific tools. However, it remains unclear to what extent LLMs can truly replicate the diverse functionalities, and the wide range of methodologies and analysis offered by prior work. In this paper, we conduct the first systematic evaluation of whether off-the-shelf LLMs can replace specialized privacy analysis tools. We study six representative tools spanning three major functionalities: contradiction detection, regulatory compliance analysis, and privacy policy summarization and aggregation, and across three intermediate tasks: structured data extraction using tuples, Semantic Role Labeling (SRL) and manual privacy policy labeling. We compare the performance of two state-of-the-art LLMs (GPT-5.2 and Gemini-2.5 in various configurations) against the tools by directly prompting the models to perform corresponding functionalities and tasks on a custom dataset of 10 privacy policies, allowing us to assess whether off-the-shelf models can produce tool-specific functionalities without further engineering or domain-specific training, major limitations in prior work. Our results show that LLMs consistently match or exceed the capabilities of existing tools across the functionalities. In manual labeling of first-party collection entities, LLMs achieved an average precision of 81.8% and recall of 70.9%, while for labeling of third-party sharing entities, they achieved an average precision of 91.4% and recall of 70.8% compared to the OPP-115 dataset. Overall, our findings indicate that LLMs can effectively perform a broad range of functionalities and tasks in privacy policy and regulation analysis that previously required specialized tools.
[NLP-57] Learning from Synthetic Data without Model Collapse in Iterative Instruction Tuning
【速读】： 该论文旨在解决生成式人工智能（Generative AI）在指令微调过程中因使用合成数据迭代演化而引发的模型退化（model collapse）问题。现有研究多关注如何限制性能下降的幅度，但本文指出，在迭代模型演进中更关键的目标是确保每一阶段的模型均能相对于前序模型实现实质性提升，这要求对模型退化现象进行细粒度诊断以指导数据筛选与优化。研究发现，在合成数据自改善的指令微调场景中，模型退化并非均匀的性能衰减，而是表现为能力极化的现象：合成数据强化了已具备较强能力的方面，同时进一步削弱了薄弱环节。针对这一发现，作者提出KITE（基于探索的知识边界指令微调），其核心在于采用两阶段框架——第一阶段通过故障引导的合成数据生成识别模型薄弱区域，第二阶段结合边界感知的不确定性筛选机制，精准选择有助于提升边缘能力的数据。实验结果表明，KITE在多个数据集和多种开源大语言模型上均实现了比主流合成数据基线更稳定且显著的性能提升。
链接: https://arxiv.org/abs/2607.17043
作者: Xiaonan Luo,Yue Huang,Kehan Guo,Ping He,Chuan Zou,Ting Hua,Xiangliang Zhang
机构: University of Notre Dame(圣母大学); Vanderbilt University(范德堡大学); University of Pennsylvania(宾夕法尼亚大学)
类目: Computation and Language (cs.CL)
备注:
Abstract:Model collapse is a central challenge in learning from synthetic data: as later-generation large language models (LLMs) are trained on an increasing proportion of model-generated data, performance can degrade due to narrowed coverage and accumulated bias. Existing work mainly studies how to bound this degradation. In iterative model evolution, however, the more meaningful objective is to ensure that each successive model improves over its predecessor, which requires diagnosing collapse at a granularity that is actionable for data curation. We study this problem in synthetic data self-improving for instruction tuning. We show that collapse in this setting is not simply uniform performance degradation, but can appear as a polarization of competence, where synthetic training reinforces already strong skills while further degrading weak ones. Motivated by this observation, we propose KITE (Knowledge-boundary Instruction Tuning via Exploration), a two-stage framework that combines failure-guided data generation with boundary-aware uncertainty curation. Experiments across several datasets and multiple open-source LLMs show that KITE yields more stable improvement than strong synthetic-data baselines.
[NLP-58] raining Continuous Chain of Thought Models: A Tale of Two Regimes ICML2026
【速读】： 该论文旨在解决连续思维链（Continuous Chain-of-Thought, CoT）方法中推理表示学习效率与性能之间的矛盾问题。现有连续CoT方法依赖于间接监督，通过使潜在表示的最终状态逼近冗长的显式推理轨迹来训练模型，这一过程需采用自回归生成方式，导致训练速度缓慢。为克服此局限，本文提出C-MTP（Continuous Mean Token Projection），其核心创新在于采用直接监督策略，将每个潜在表示建模为待压缩推理轨迹中嵌入向量的均值，从而实现更高效、快速的训练。实验表明，C-MTP在简化推理轨迹（少于100个词元）的评估设置下，优于先前的直接监督方法，并可与较慢的间接监督方法相媲美。进一步地，研究将评估扩展至包含数百词元以上复杂推理任务，发现无论是直接还是间接监督方法均出现显著性能下降（约65%），揭示了当前连续CoT方法在处理长序列复杂推理时的固有局限性。
链接: https://arxiv.org/abs/2607.16972
作者: Varun Yerram,He He,Eunsol Choi
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted to AdaptFM Workshop, ICML 2026
Abstract:Continuous Chain-of-Thought methods replace verbose reasoning traces with a short sequence of dense latent representations. Earlier continuous CoT methods indirectly supervise the latent representations such that its final state match that of verbose reasoning traces, requiring autoregressive, slow generation during training. We introduce C-MTP, a simpler, faster direct supervision approach that models each latent as an average of the embeddings in the CoT traces to be compressed. Our approach outperforms a prior direct supervision method that approximates the distribution of compressed tokens, and performs competitively to slower indirect supervision approaches in existing evaluation setup with simplified CoT traces (less than 100 tokens). Lastly, we extend the evaluation of Continuous CoT methods to complex tasks with longer reasoning traces ( \ge few hundreds reasoning tokens). We find both direct and indirect supervision training methods perform poorly (roughly 65% performance drop) in this setting, revealing the limitations of current continuous CoT methods. The code and checkpoints are released at this https URL
[NLP-59] A Method for Learning Value Systems in Generative AI AAAI
【速读】： 该论文旨在解决生成式 AI 中缺乏具备明确价值根基（grounding）的值学习方法的问题，即现有方法往往仅复制人类偏好而忽视价值对齐的多维结构，或缺乏系统性、可解释的价值体系获取机制。其核心解决方案是将一种经过验证的价值体系学习方法适配至生成式 AI 场景，基于成对的提示-响应偏好数据，同时学习：i）由多目标奖励模型实现的一组价值的根基表示，ii）以加权线性标量化的形式表达的价值体系。为确保所学价值体系具有内在一致性，算法动态优先优化根基学习过程。实验结果表明，该方法在多个提示-响应偏好数据集上表现优于基线，并在保持性能的同时显著提升可解释性。
链接: https://arxiv.org/abs/2607.16903
作者: Andrés Holgado-Sánchez,Holger Billhardt,Sascha Ossowski
机构: 未知
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: Full version of a to be published paper in proceedings of the 9th AAAI/ACM conference in AI, Ethics and Society (AIES 2026). Includes supplementary material. 20 pages, 2 figures
Abstract:Value-aware AI systems require explicit computational representations of human values (groundings) and their aggregation into value systems in order to align their decisions with ours. As such representations are difficult to elicit, value learning seeks to infer them by observing human behaviour. This work addresses the lack of grounded value learning methods in generative AI: existing approaches typically replicate human preferences without awareness of the multidimensional structure of value alignment, or lack principled value system elicitation methods. To address these gaps, we adapt a previously validated value system learning method to the generative AI setting, which, based on pairwise prompt-response preference data, simultaneously learns: i) an implementation of a grounding for a set of values given by a multi-objective reward model, and ii) a value system representation in the form of a weighted linear scalarization of the previous grounding model. To ensure that the learned value systems are based on coherent value representations, our algorithm dynamically prioritizes the grounding learning process. We evaluate the method against baselines and a contemporary method on prompt-response preference datasets. Results show competitive performance and minimal trade-offs against the baselines, while improving explainability.
[NLP-60] Cross-Branch Conflict as a Shield: Safeguarding Facial Identities in Unified Multimodal Image Editing
【速读】： 该论文旨在解决统一多模态模型（Unified Multimodal Models, UMMs）在指令驱动图像编辑中可能引发的个人肖像未经授权篡改的安全风险。现有对抗性保护方法主要针对视觉理解或图像生成模型设计，难以有效迁移至处理图像时通过多个互补视觉路径的UMMs。其关键问题是：单一分支的特征扰动无法在多模态融合过程中维持持续且对称的破坏效果，导致身份信息仍可被恢复。为此，本文提出一种名为CCS（Cross-Branch Compatibility Suppression）的统一对抗性保护框架，其核心在于同时扰动基于ViT的理解分支与基于VAE的生成分支的特征表示，并通过线性中心化核相关性（linear CKA）显式破坏两分支间的跨分支兼容性。该方法通过从双路径中同步移除稳定信息并构建不兼容的视觉上下文，有效阻止了UMM在编辑过程中恢复可靠的个体身份信息。大量实验表明，CCS在抑制保留身份的编辑行为方面显著优于现有方法。
链接: https://arxiv.org/abs/2607.16898
作者: Weiwei Tan,Junxian Li,Rui Wang,Zhenhua Xu,Yanjun Zhang,Yu Leo Zhang
机构: 未知
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注: 8 pages, preprint
Abstract:Unified multimodal models (UMMs) have recently demonstrated powerful instruction-based image editing capabilities, but they also raise serious concerns about unauthorized manipulation of personal portraits. Existing adversarial protection methods are mainly designed for either visual understanding or image generation models and often become ineffective when transferred to UMMs, which process an image through multiple complementary visual pathways. In this work, we first conduct a feature-level analysis of unified image editing. We observe that the ViT-based understanding branch and the VAE-based generation branch exhibit non-trivial structural agreement for the same input image. Although perturbing an individual branch can reduce this agreement and induce intermediate hidden-state deviations, such effects are asymmetric and gradually attenuated during multimodal fusion and generation. These observations reveal that single-branch feature distortion is insufficient for consistently disrupting unified image editing. Motivated by this finding, we propose CCS, a unified adversarial protection framework that jointly drives the ViT and VAE representations away from their clean counterparts while explicitly disrupting their cross-branch compatibility through linear CKA. By simultaneously removing stable information from both visual pathways and creating incompatible visual contexts, CCS prevents the UMM from recovering reliable identity information during editing. Extensive experiments demonstrate that CCS consistently outperforms existing protection methods in suppressing identity-preserving edits.
[NLP-61] race-Based On-Policy Distillation for Masked Diffusion Language Models
【速读】： 该论文旨在解决生成式语言模型中基于扩散机制的大型语言模型（dLLMs）在推理导向后训练过程中面临的挑战，即传统监督微调（SFT）依赖于密集但常为非策略性的掩码状态，而强化学习（RL）则受限于稀疏奖励或价值建模。为此，论文提出了一种基于轨迹的在策略蒸馏（TOPD）方法，其核心在于通过教师模型对目标dLLM自身的去噪轨迹进行监督，聚焦于形成最终响应的、与轨迹对齐的词元决策。具体而言，TOPD从目标dLLM中采样在策略的扩散轨迹，在对应的局部去噪状态上获取教师模型的词元分布，并采用词元级别的反向Kullback-Leibler（Reverse-KL）损失更新目标模型。该设计既保持了密集的教师监督，又使训练过程与模型自身的去噪状态对齐。实验表明，在数学推理基准测试中，TOPD使SDAR-4B-Chat达到与强化学习训练的TraDo-4B-Instruct相当的MATH500准确率，静态评估下提升+5.7，动态评估下提升+4.5；且相比强化学习方法，仅需1/4的回滚轮次，实现约96.0倍的模型计算效率提升。
链接: https://arxiv.org/abs/2607.16872
作者: Haolin Ren,Ziyang Huang,Chenhao Yuan,Jun Zhao,Kang Liu
机构: Institute of Automation, Chinese Academy of Sciences(中国科学院自动化研究所); University of Chinese Academy of Sciences(中国科学院大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Diffusion large language models (dLLMs) are a promising alternative to autoregressive generation. However, reasoning-oriented post-training for dLLMs remains challenging. Supervised fine-tuning (SFT) for dLLMs requires dense but often off-policy masked states, while reinforcement learning (RL) relies on sparse rewards or value modeling. This paper proposes \textbftrace-based on-policy distillation (TOPD), a teacher-supervised framework that transfers reasoning ability to a target dLLM without reward estimation. The key idea is to supervise a dLLM on its own denoising trajectory, focusing on the trace-aligned token decisions that form the final response. Specifically, TOPD samples on-policy diffusion trajectories from the target dLLM, obtains teacher token distributions from a teacher model on the corresponding partially denoised states, and updates the target dLLM with a token-level Reverse Kullback-Leibler (Reverse-KL) objective. This design preserves dense teacher supervision while aligning training with the model’s own denoising states. On mathematical reasoning benchmarks, TOPD enables SDAR-4B-Chat to match the MATH500 accuracy of its RL-trained counterpart TraDo-4B-Instruct, with gains of +5.7 under static evaluation and +4.5 under dynamic evaluation. Compared with the RL-trained counterpart, TOPD achieves this with 4 \times fewer rollout rounds, corresponding to an estimated 96.0 \times to-accuracy model-compute speedup.
[NLP-62] Group Entropy-Controlled Policy Optimization
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）在强化学习（Reinforcement Learning, RL）对齐过程中，由于多任务混合训练导致的熵分布异质性问题。具体而言，在不同任务组间，相同策略下产生的熵水平差异显著，使得全局或逐标记级别的熵调控无法满足各任务组对探索-利用平衡的差异化需求，进而导致基于GRPO（Generalized Reward Policy Optimization）的归一化优势信号引入依赖熵的偏差，使不同提示组之间的优势值在统计上不可比。为应对这一挑战，本文提出分组熵控策略优化（Group Entropy-Controlled Policy Optimization, GEPO），作为GRPO的轻量级扩展。其核心在于利用已有分组样本估计各任务组的熵，并据此进行条件化的非对称优势调整：在低熵组中抑制正向优势以缓解过度利用，在高熵组中增强负向优势以维持探索能力，且自适应阈值由历史熵统计信息动态生成。大量实验表明，GEPO在十三个涵盖数学、物理、科学、代码生成和指令遵循等领域的基准测试中，持续优于GRPO及近期熵控方法，在保持各任务特异性探索水平的同时实现跨任务性能的均衡提升。
链接: https://arxiv.org/abs/2607.16850
作者: Guangran Cheng,Chengqi Lyu,Songyang Gao,Wenwei Zhang,Kai Chen
机构: 未知
类目: Computation and Language (cs.CL)
备注: 16 pages, 5 figures
Abstract:Entropy control has become an effective tool in reinforcement learning (RL) of large language models (LLMs), helping balance exploration-exploitation trade-off during alignment process. Such RL paradigm is often conducted on mixtures of heterogeneous tasks, which induce distinct entropy regimes under the same policy, making global or token-level entropy regulation insufficient to corresponding heterogeneous needs of exploration. This heterogeneity further makes GRPO-style normalized advantages induce an entropy-dependent bias, making advantage signals across prompt groups statistically non-comparable. To address this issue, we propose Group Entropy-Controlled Policy Optimization (GEPO), a lightweight extension to GRPO that uses group entropy, estimated from existing grouped samples to perform entropy-conditioned asymmetric advantage shaping. GEPO attenuates positive advantages in low-entropy groups to reduce over-exploitation, and negative advantages in high-entropy groups to preserve exploration, with adaptive thresholds derived from historical entropy statistics. Extensive experiments on two base models across thirteen benchmarks spanning mathematics, physics, science, code generation, and instruction following show that GEPO consistently outperforms GRPO and recent entropy-controlled methods, delivering balanced cross-task improvements while preserving task-specific exploration levels throughout training.
[NLP-63] Schema-Constrained Document-Level Event Argument Extraction with Lightweight LLM Fine-Tuning KDD2026 ECML
【速读】： 该论文旨在解决文档级事件论元抽取（Document-level Event Argument Extraction, EAE）中的关键挑战，包括触发词与论元之间的长距离依赖、跨句上下文建模以及严格的角色约束导致的边界错误、角色不确定性及与受限事件模式不一致等问题。其解决方案的关键在于三方面协同设计：（1）在提示中引入角色集注入（role-set injection），以确保模型输出符合预定义的事件模式；（2）采用参数高效微调方法LoRA，在训练和推理阶段均使用相同的仅JSON接口，保持系统一致性；（3）通过确定性解码结合后处理机制，对生成的JSON结果进行有效性验证、无效角色过滤、论元去重以及论元跨度与文档窗口的对齐。该方法在MAVEN-ARG基准上显著优于先前基于GPT的基线模型，最佳模型Phi-4（14B）在事件共指评估层面达到42.39%的F1分数，证明了中等规模开源大语言模型在严格约束下的文档级EAE任务中具备可靠性能。
链接: https://arxiv.org/abs/2607.16808
作者: Roberto Pietrantuono,Antonio Guerriero,Pouya Sattari
机构: 未知
类目: Computation and Language (cs.CL)
备注: Accepted at ECML PKDD 2026
Abstract:Event Argument Extraction (EAE) converts documents into structured event records by identifying argument spans and assigning them schema-defined roles. Document-level EAE is challenging due to long-range dependencies between triggers and arguments, cross-sentence context, and strict role constraints, which often lead to boundary errors, uncertainty in roles, and inconsistencies with restricted schemas. In this paper, we study whether mid-sized open LLMs can perform schema-constrained EAE reliably at the document level on MAVEN-ARG. Our approach combines (i) role-set injection in prompts for schema compliance, (ii) parameter-efficient supervised fine-tuning (LoRA) using the same JSON-only interface used at inference, and (iii) deterministic decoding with post-processing that validates JSON, filters invalid roles, de-duplicates arguments, and aligns spans to the document window. Under the official MAVEN-ARG evaluator, fine-tuned mid-sized open models outperform previously reported GPT baselines across mention, entity-coreference, and event-coreference evaluations; our best model (Phi-4, 14B) reaches 42.39% F1 at the event-coreference level. Code to reproduce experiments is publicly available at this https URL. Comments: Accepted at ECML PKDD 2026 Subjects: Computation and Language (cs.CL) Cite as: arXiv:2607.16808 [cs.CL] (or arXiv:2607.16808v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.16808 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-64] Diagnosing Correctness Probes under Self-Judgement Confounding
【速读】： 该论文旨在解决生成式语言模型中“客观正确性”（Objective Correctness, OC）与模型自身“自判断”（Self-Judgement, SJ）之间语义不一致的问题，尤其是在隐藏状态读出（hidden-state readouts）用于预测输出正确性时，由于OC与SJ通常高度一致，导致解码信号的语义模糊。为揭示二者差异，研究构建了冲突案例，使OC与SJ对响应正确性的排序相反。在高置信度分歧场景下，传统基于正确性标注的对比方法常将错误但自我认可的回应排在正确但自我拒绝的回应之上，表现出对SJ而非OC的偏好。研究通过估计与SJ和OC相关联的因子方向，并在数学推理和事实回忆任务中评估其极性，发现：在四个指令微调模型（最大参数量达140亿）中，所有模型的SJ相关方向在跨领域迁移中均显著优于随机水平，而OC相关方向在预期的OC排序上点估计均低于随机水平。这种迁移不对称性在中间至晚期层逐渐显现，且在回答似然、序列长度及零方向控制下仍保持稳定，并可外推至MMLU和二分类的TruthfulQA数据集，无需针对目标领域进行方向拟合。综合来看，在所考察的模型与诊断子集上，最具可迁移性的成分始终保留SJ相关极性，表明仅凭迁移能力无法确立模型具备客观正确性语义。因此，该研究的关键突破在于揭示了当前语言模型的内部表征更倾向于编码主观自判断而非客观正确性，从而挑战了以迁移性作为衡量模型具备客观认知能力的合理性。
链接: https://arxiv.org/abs/2607.16799
作者: Yi-Long Lu
机构: 未知
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:Hidden-state readouts can predict whether language-model outputs are correct, but objective correctness (OC) usually agrees with the model’s own self-judgement (SJ), leaving the decoded signal semantically ambiguous. We construct conflict cases in which OC and SJ predict opposite readout orderings. On high-confidence disagreements, conventional correctness-labelled contrasts often rank incorrect/self-endorsed responses above correct/self-rejected responses, following SJ rather than OC. We estimate factorial SJ- and OC-associated directions and evaluate their polarity across mathematical reasoning and factual recall. Across four instruction-tuned models up to 14B parameters, the SJ-associated direction transfers above chance in both cross-domain directions for every model, whereas the OC-associated direction has a below-chance point estimate for the expected OC ordering in every corresponding condition. This transfer asymmetry develops across middle-to-late layers, persists under answer-likelihood, sequence-length, and null-direction controls, and extends to MMLU and binary TruthfulQA without target-domain direction fitting. Across the studied models and diagnostic subsets, the most reliably transferable component preserves SJ-associated polarity. Transferability alone therefore does not establish objective-correctness semantics.
[NLP-65] Cascading versus Joint Modeling for Hierarchical Offensive Language Detection
【速读】： 该论文旨在解决细粒度辱骂语言检测中两种建模范式——级联分解（cascaded decomposition）与联合多任务建模（joint multi-task modeling）在准确性、参数量和推理延迟等方面的直接对比缺失问题，以及现有类别不平衡处理策略是否最优缺乏验证的问题。其核心解决方案是提出一个三级级联检测系统，针对每个子任务定制训练策略，并引入双重验证机制：首先通过受控消融实验确定各子任务的最佳类别不平衡处理方式；其次构建共享编码器的联合多任务模型作为架构对照组，实现对准确率、参数量和推理延迟的真实量化评估。实验表明，级联系统在官方测试集三个子任务上分别取得0.795、0.716和0.557的宏平均F1分数，且消融研究揭示仅凭不平衡严重程度直觉配置损失函数存在局限性，需基于实证结果重调参以提升性能与稳定性。端到端级联评估发现，约五分之一的错误源于第一阶段过滤器，后续阶段无法纠正。相较联合多任务模型，级联架构在所有子任务上均取得更高准确率，尤其在最严重类别不平衡的子任务上提升7.1个宏F1点，但代价为参数量增加3倍、推理延迟提高1.67倍。上述结果明确建立了级联架构在精度优势与部署成本之间的可量化权衡关系。
链接: https://arxiv.org/abs/2607.16790
作者: Ruixing Ren,Junhui Zhao,Xiaoke Sun
机构: Beijing Jiaotong University (北京交通大学); National Computer Network Emergency Response Technical Team/Coordination Center of China (CNCERT/CC) (中国国家计算机网络应急技术处理协调中心)
类目: Computation and Language (cs.CL); Systems and Control (eess.SY)
备注:
Abstract:Fine-grained offensive language detection organizes labels into a hierarchical structure, for which two modeling paradigms exist: cascaded decomposition and joint multi-task modeling. Prior work rarely provides a direct, controlled comparison of the two paradigms in terms of accuracy, parameter count, and inference latency, and rarely verifies whether a chosen class-imbalance handling strategy is actually optimal. This paper proposes a three-level cascaded detection system whose training strategy is customized per subtask, together with two verification mechanisms. First, a controlled ablation study determines the best class-imbalance handling strategy for each subtask. Second, a joint multi-task model with a shared encoder is trained as an architectural control, yielding real measurements along the dimensions of accuracy, parameter count, and inference latency. Experiments show that the cascaded system attains macro-F1 scores of 0.795, 0.716, and 0.557 on the three subtasks of the official test set. The ablation study reveals that configuring the loss function purely by imbalance-severity intuition is suboptimal; reconfiguring based on the ablation results improves both performance and stability. End-to-end cascade evaluation shows that roughly one-fifth of the errors in the cascade pipeline originate from the first-stage filter and cannot be corrected by subsequent stages. Relative to the joint multi-task model, the cascaded architecture achieves higher accuracy on all three subtasks, with a 7.1-point macro-F1 gain on the most severely imbalanced subtask, at the cost of three times the parameters and 1.67 times the inference latency. Together, these results establish an explicit, quantifiable trade-off between the accuracy advantage of cascaded architectures and their deployment cost.
[NLP-66] JOR-Bench: Japanese Operations Research Benchmarks for Large Language Models
【速读】： 该论文旨在解决大语言模型（LLMs）在日语环境下对运筹学（Operations Research, OR）问题的建模与求解能力评估缺乏系统性基准的问题。现有基准多基于英语，难以全面评估模型在非英语语境下的实际应用性能，尤其在日语等东亚语言中存在语义理解与逻辑推理偏差的风险。为此，论文提出JOR-Bench，一个包含5个日本语基准任务的评测集合，涵盖线性规划、混合整数规划、非线性规划及组合优化共1,319个问题，均为现有英文基准（IndustryOR、MAMO Complex LP、NL4OPT、OptiBench、OptMATH）的准确日语翻译。其关键创新在于构建了与求解器无关（solver-independent）的评测框架，仅提供日语问题描述与期望数值答案的配对，支持任意求解器或编程语言接入；同时通过统一使用Python接口调用OR-Tools进行标准化执行，确保模型输出的可比性与可复现性。实验表明，对于强健的多语言模型，建模能力在语言间基本保持一致（英日平均准确率差异仅为-0.3个百分点），但错误分析揭示细微跨语言差异：部分模型在日语提示下出现语用消歧失败，误输出决策变量值而非目标函数值，暴露出日语特定表达结构对模型推理路径的影响。
链接: https://arxiv.org/abs/2607.16777
作者: Yuu Jinnai
机构: CyberAgent(サイバーエージェント)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:
Abstract:We present JOR-Bench, a collection of five Japanese-language benchmarks for evaluating the ability of large language models (LLMs) to formulate and solve operations research (OR) problems. Each benchmark is a Japanese translation of an existing English benchmark: IndustryOR, MAMO Complex LP, NL4OPT, OptiBench, and OptMATH, covering 1,319 problems spanning linear programming, mixed-integer programming, non-linear programming, and combinatorial optimization. JOR-Bench is a solver-independent benchmark that can be used with any solver or programming language, and consists of pairs of Japanese problem statements and expected numerical answers. We evaluate seven LLMs, including multilingual general-purpose models and Japanese-specialized models, on both the original English and the new Japanese versions, and compare performance across languages. For the main evaluation, we standardize execution with the Python interface to OR-Tools to make model outputs comparable and reproducible with open-source software. Our results show that OR formulation ability is largely language-neutral for strong multilingual models; the overall average accuracy difference between English and Japanese is only -0.3 pp. Yet error analysis reveals subtle cross-lingual differences, including a pragmatic disambiguation failure in some domains that causes models to output decision-variable values instead of the objective value when the prompt is in Japanese.
[NLP-67] DS@GT ARC at eRisk 2026: Hybrid Multi-Agent LLM System with Structured Algorithmic Guidance for Conversational Depression Screening
【速读】： 该论文旨在解决在不直接询问敏感心理健康问题的前提下，通过对话式交互对个体抑郁状态进行有效筛查的挑战，即如何在保护用户隐私的同时准确预测贝克抑郁量表第二版（BDI-II）得分及识别关键抑郁症状。其核心问题是：如何在避免直接心理评估提问的情况下，利用大语言模型（LLM）构建具有模拟个体特征的对话代理，并实现高精度的抑郁水平量化与症状识别。解决方案的关键在于提出一种分阶段演进的混合架构：首先采用单模型原型验证可行性，随后构建多智能体系统以分离对话访谈与评分模块，最终引入开源模型Gemma 27B替代昂贵的GPT-5-nano作为对话访谈器，通过三项算法增强机制弥补其推理与指令遵循能力的不足——包括预计算对话树以标准化问诊流程、基于可靠性加权的共识聚合策略（受Weaver框架启发），以及基于聚类的未探查症状插补方法。该方案在降低约四分之三的每例调用成本基础上，仍实现了优于付费基线模型的性能表现（ADODL达0.9063，排名第三），验证了算法强化可使弱化开源模型在对话式抑郁筛查任务中媲美甚至超越强效专有模型的核心假设。
链接: https://arxiv.org/abs/2607.16712
作者: Victor Gong,David Guecha
机构: Georgia Institute of Technology(佐治亚理工学院)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Conference and Labs of the Evaluation Forum (CLEF), 19 pages, 5 figures
Abstract:We describe DS@GT’s submission to the eRisk 2026 Task 1 challenge on conversational depression screening, in which systems interview LLM personas that simulate individuals with varying depression profiles and produce a Beck Depression Inventory II (BDI-II) score plus four key symptoms per persona, without directly asking sensitive mental health questions. Our pipeline evolved through three stages: a monolithic single-model prototype to start off, a baseline multi-agent architecture that separates conversational interviewing from BDI-II scoring under a coordinating orchestration layer, and a final hybrid configuration that replaces the paid GPT-5-nano interviewer with the open-source Gemma 27B. To offset the model’s weaker reasoning and instruction-following, the hybrid adds three algorithmic components: a precomputed dialogue tree that standardizes interview openers and follow-ups, a reliability-weighted consensus aggregation inspired by the Weaver framework, and a cluster-based imputation step for unprobed symptoms. We submitted three fully automated runs across all 20 personas, with Run 1 from the paid baseline and Runs 2 and 3 from the hybrid. Hybrid Run 3 achieved an ADODL of 0.9063, ranking 3rd among all complete-submission runs and placing DS@GT 2nd among the 21 teams overall, while outperforming our paid baseline Run 1 (0.8841) at roughly one-quarter of the per-persona API cost. These results support our central hypothesis that with sufficient algorithmic supervision, a weaker open-source model can compete with a stronger proprietary model in the conversational interviewer role. Our source code is available at this https URL.
[NLP-68] hough Language Models Err While They Strive: Conformal Prediction for Self-Correcting Scientific Generation
【速读】： 该论文旨在解决大语言模型在生成技术内容时频繁违反基本科学原理的问题，从而削弱其在科学应用中的可靠性。其核心解决方案是提出一种图结构化的置信度控制框架——科学合理性控制（Scientific Feasibility Control, SFC），通过渐进式绝对一致性事实性验证提供科学推理有效性的统计保障。SFC的关键在于将科学推理分解为原子级的“绝对一致性事实性单元”（absolute-coherent-factuality units），要求每个单元不仅符合物理定律，还需从前序上下文获得逻辑支持，以此缓解早期科学错误对后续推理链的级联污染问题。与依赖独立性假设、孤立评估命题的传统方法不同，SFC将逻辑依赖关系建模为近似可推导图（approximate deducibility graphs），并在检测到科学违规时实时触发动态分支机制，基于已验证的上下文切换至替代生成路径。实验表明，SFC在多个主流科学推理基准（包括PhyX多模态物理、MATH、ScienceQA和ARC Challenge）上均取得显著性能提升，在PhyX物理推理任务中达到50.1%准确率，优于DeepSeek-R1（49.8%）和GPT-4（45.8%），同时在α=0.10显著性水平下实现91.7%的正式置信覆盖率，并在多种模型架构上减少73%的科学定律违背现象。
链接: https://arxiv.org/abs/2607.16704
作者: Mingqiao Mo,Yunlong Tan,Hao Zhang
机构: 未知
类目: Computation and Language (cs.CL)
备注: 25 pages
Abstract:Large language models frequently violate fundamental scientific principles when generating technical content, undermining their reliability in scientific applications. We introduce Scientific Feasibility Control SFC, a graph-structured conformal prediction framework that provides statistical guarantees for scientific reasoning validity through progressive absolute-coherent-factuality validation. Our approach decomposes scientific reasoning into atomic absolute-coherent-factuality units requiring both individual correctness against physical laws and logical substantiation from preceding context, addressing the cascade effect where early scientific errors contaminate subsequent reasoning steps. Unlike independence-based methods that treat claims in isolation, SFC models logical dependencies as approximate deducibility graphs and operates through real-time validation with dynamic branching when scientific violations are detected, the system branches to alternative generation paths using verified context as foundation. We demonstrate SFC across established scientific reasoning benchmarks including PhyX multimodal physics, MATH, ScienceQA, and ARC Challenge, achieving 50.1 percent accuracy on PhyX physics reasoning, substantially outperforming recent reasoning models including DeepSeek-R1 49.8 percent and GPT-4 45.8 percent while providing 91.7 percent scientific validity with formal conformal coverage guarantees at alpha equals 0.10 confidence level and reducing scientific law violations by 73 percent across multiple model architectures.
[NLP-69] Are Arithmetic Heuristic Neurons Form-Invariant? A Mechanistic Analysis of Symbols Text and Code in LLM s
【速读】： 该论文试图解决大语言模型（Large Language Models, LLMs）在面对同一数学问题的不同表述形式时，表现出性能差异的问题，即模型在一种形式下表现良好而在等价形式下失败的“形式敏感性”问题。其核心问题是：这种失败是由不同内部计算回路导致的，还是由同一回路在不同激活状态下的表现差异所引起？论文的关键解决方案是通过构建一个两阶段管道（结合归因修补和激活修补），识别出在符号算术、自然语言应用题和Python代码三种不同表达形式中均存在的“算术启发式神经元”（arithmetic heuristic neurons）。研究发现，这些神经元构成一个紧凑的共享子集，且对高层层的算术计算既必要又充分；将某一成功执行中的共享神经元激活状态转移到另一失败执行中，可使超过97%的错误预测得以恢复，表明跨格式失败主要源于激活状态而非电路结构的差异。此外，这些共享神经元在不同格式中始终归属于相同的启发式类别，证明了大语言模型中的算术计算在神经元层面具有高度的形式不变性。
链接: https://arxiv.org/abs/2607.16693
作者: Sharath Naganna,Tanvir Ahmed Sijan,Uddipta Kalita
机构: Indian Institute of Technology Bombay (印度理工学院孟买分校); Jahangirnagar University (贾杭格纳尔大学); National Institute of Technology, Silchar (国立技术学院西卡尔)
类目: Computation and Language (cs.CL)
备注: Under Review
Abstract:Large language models often succeed on one formulation of a problem while failing on an equivalent formulation. Whether these failures arise from distinct internal circuits or different activation states of a shared circuit remains unknown. Recent mechanistic interpretability studies suggest that arithmetic in LLMs emerges from a “bag of heuristics,” encoded by a sparse set of MLP neurons that represent distinct arithmetic strategies. We investigate whether arithmetic heuristic neurons are form-invariant across symbolic arithmetic, natural language word problems, and Python code in three Llama-3 models. In each format, we identify arithmetic heuristic neurons using a two-stage pipeline combining attribution patching and activation patching. A compact set of neurons is shared across all three formats, and targeted interventions show this shared circuit is both necessary and sufficient for late-layer arithmetic computation. Transferring the shared neurons’ activations from a successful execution in one format to a failed execution in another recovers most incorrect predictions, exceeding 97% for addition and subtraction, indicating that cross-format failures arise from activation states rather than distinct circuits. Moreover, shared neurons consistently belong to the same heuristic families across formats, demonstrating that arithmetic computation in LLMs is largely form-invariant at the neuron level.
[NLP-70] Dependency-Guided Code Generation: Structured Matrix Decomposition and Consistency-Guided Refinement
【速读】： 该论文旨在解决现代软件系统复杂性日益增加背景下，自动化代码生成过程中难以有效捕捉代码实体间多层次、复杂依赖关系的问题，从而导致生成代码在逻辑上不完整或难以集成到实际系统中的缺陷。其解决方案的关键在于提出一种依赖感知的代码生成框架，通过图结构表示显式建模代码实体间的交互关系，并将依赖关系分解为两个互补成分：用于捕获强而明确关系的量化矩阵，以及用于建模弱而隐含交互的稀疏低秩因子分解。该分解通过交替优化过程高效学习，并在代码生成阶段作为约束引入，以保证生成代码在语义连贯性和结构一致性上的质量。此外，引入稀疏三元组表示强依赖关系，显著提升了存储效率与计算可扩展性。实验结果表明，该方法在语义对齐度和结构保真度方面均优于现有方法。
链接: https://arxiv.org/abs/2607.16692
作者: Mingqiao Mo,Yangchen Zeng,Zikai Xiao,Xin Xiao,Wenhua Nie,Zhaolu Kang,Guangyuan Dong,Kai Shu,Hao Zhang,Xiaodong Fan
机构: University of the Chinese Academy of Sciences; ByteDance Inc.; Zhejiang University; National Taiwan University; Peking University; Alibaba Group; Tsinghua University; Liaoning Technical University
类目: oftware Engineering (cs.SE); Computation and Language (cs.CL)
备注: 12 pages
Abstract:The increasing complexity of modern software systems has made automated code generation a fundamental task in software engineering. However, existing approaches often fail to adequately capture the intricate, multi-level dependencies among code entities, leading to generated code that is logically incomplete or difficult to integrate into real-world systems. To address this limitation, we propose a dependency-aware code generation framework that explicitly models interactions among code entities through a graph-based representation. We decompose dependencies into two complementary components: a quantized matrix that captures strong, explicit relations, and a sparse low-rank factorization that models weaker, implicit interactions. The decomposition is efficiently learned via an alternating optimization procedure. During code generation, the learned dependency structure is incorporated as a constraint, ensuring both semantic coherence and structural consistency of the generated code. Furthermore, we introduce a sparse triplet representation for strong dependencies, significantly improving storage efficiency and computational scalability. Extensive experiments demonstrate that our approach consistently produces code with superior semantic alignment and structural fidelity compared to existing methods.
[NLP-71] SpecLA: Efficient Speculative Decoding for Linear-Attention Models
【速读】： 该论文旨在解决状态型线性注意力（stateful linear-attention）模型在自回归解码过程中因逐令牌读取、更新和写入递归状态而导致的效率瓶颈问题。现有推测解码（speculative decoding）方法主要针对基于Transformer的键值（KV）缓存设计，无法直接适用于具有递归依赖关系的状态型线性注意力结构。其关键挑战在于：验证过程需遵循链式与分支结构中的递归依赖关系，接受的输出必须仅更新对应的合法状态轨迹，且起草器（drafter）需避免提交会浪费状态感知验证资源的候选序列。为此，本文提出SpecLA——一种专为状态型线性注意力模型设计的推测解码运行时系统。其核心创新包括：采用拓扑感知的核函数实现对链与树结构的高效验证；在验证过程中压缩存储中间因子以恢复被接受的状态轨迹；结合置信度剪枝与目标对齐的EAGLE风格起草器，确保输入验证器的候选序列具备高有效性。在配备NVIDIA H100的公开GDN-1.3B模型上，SpecLA实现了相较于传统自回归解码高达1.70倍的端到端加速。
链接: https://arxiv.org/abs/2607.16673
作者: Zhibin Wang,Xuying Han,Zhaohua Yang,Fuliang Liu,Xue Li,Rong Gu,Sheng Zhong,Chen Tian
机构: Nanjing University (南京大学); Alibaba Group (阿里巴巴集团)
类目: Computation and Language (cs.CL)
备注:
Abstract:Linear-attention models replace the growing KV cache with recurrent states, but autoregressive decoding still reads, updates, and writes these states one token at a time. Speculative decoding can reduce this cost by verifying several draft tokens in one target pass, yet existing speculative systems are designed for Transformer KV caches. For stateful linear-attention targets, verification must follow recurrent dependencies across chains and branches, acceptance must update only the accepted state trajectory, and the drafter must avoid submitting candidates that waste stateful verification work. This paper presents SpecLA, a speculative decoding runtime for stateful linear-attention models. SpecLA verifies chains and trees with topology-aware kernels, stores compact factors produced during verification to recover accepted states, and uses confidence pruning plus a target-aligned EAGLE-style drafter to feed useful candidates to the verifier. On an NVIDIA H100 with a public GDN-1.3B target, SpecLA achieves up to 1.70x end-to-end speedup over autoregressive decoding.
[NLP-72] OpenLanguageModel: Readable and Composable Small-Language-Model Pretraining for Education and Research
【速读】： 该论文旨在解决小规模语言模型（Small Language Model, SLM）在构建与预训练过程中缺乏透明性与可读性的问题，即现有框架往往将模型结构抽象为黑箱组件，导致研究者难以理解、修改和复现模型的内部机制。其解决方案的关键在于提出OpenLanguageModel（OLM）——一个基于PyTorch的开源库，通过将模型架构设计为可读性强的模块化代码：以Block、Residual、Repeat和Parallel等显式组件描述模型连接方式，使模型代码直接对应其结构图谱，实现从教学笔记到完整预训练流程的无缝迁移。该设计不仅支持本地与流式数据集、混合精度训练、检查点保存、回调机制及跨硬件（CPU、单GPU、多GPU）的自适应执行，还通过集成AutoTrainer实现了自动化资源配置。实验验证了其与独立参考实现的高度一致性、348M参数模型在四张GPU上达到90.6%的弱扩展效率，并展示了对模型架构的紧凑修改能力与良好的早期可用性。
链接: https://arxiv.org/abs/2607.16669
作者: Tavish Mankash,Vardhaman Kalloli,Keshava Prasad,Deepan Muthirayan
机构: Plaksha University (普拉克沙大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注: 8 pages, 3 figures, and 1 table. Website: this https URL . Code: this https URL
Abstract:OpenLanguageModel (OLM) is an open-source PyTorch library for building and pretraining small language models while keeping their machinery visible. In OLM, model code reads like the architecture: components are ordinary modules, while Block, Residual, Repeat, and Parallel describe how they are wired. The resulting model can move unchanged from a teaching notebook to a complete pretraining run or a research ablation. OLM connects this readable model layer to tokenizers, local and streaming datasets, optimization, mixed precision, callbacks, checkpoints, and hardware-aware CPU, single-GPU, and single-node multi-GPU execution. We demonstrate the full path by tracing GPT-2 from diagram to code, launching a FineWeb-Edu training script, replacing one attention component, and letting AutoTrainer configure the available machine. The package includes 27 presets across nine familiar model families and documentation that progresses from LM fundamentals to architecture research. Validation shows close agreement with independent reference implementations, 90.6% four-GPU weak-scaling efficiency for a 348M-parameter workload, compact architecture edits, and positive early usability results. OLM is MIT-licensed and available through PyPI, GitHub, and its documentation site.
[NLP-73] From Memory to Skills: Evidence-Grounded Co-Evolution Governance for Long-Horizon LLM Agents EMNLP’2026
【速读】： 该论文旨在解决长时程大语言模型（LLM）智能体在记忆系统中仅被动调用历史轨迹作为上下文，而未能将经验转化为可执行能力的核心问题。其解决方案的关键在于提出一种无需训练的“记忆-技能协同演化”（Memory–Skill Co-Evolution, MSCE）框架，该框架通过将智能体经验结构化为具身化的步骤痕迹（grounded step traces）、可复用的过程性策略（procedural policies）以及陈述性环境认知（declarative environmental cognition），实现对经验的主动转化与利用。其中，MSCE将具有正向收益估计的经验证据支持的二级策略（L2 policies）结晶为带有证据关联、适用边界、决策指导、验证规则及可靠性评估的可调用技能；同时引入反射加权的价值回填机制（reflection-weighted value backfilling），通过密集的局部自我反思，将稀疏的终端反馈传播至密集的中间状态，生成基于证据校准的轨迹价值，从而驱动记忆与技能的持续演化。实验结果表明，该方法在EvoAgentBench和LoCoMo基准上显著优于现有最先进的技能增强型与记忆驱动型智能体基线，展现出强大的跨领域迁移能力与终身演化特性。
链接: https://arxiv.org/abs/2607.16621
作者: Bo Tang,Yang Zhang,Guomian Zhuang,Wenqiang Wei,Gaoyang Zheng,Lindong Xie,Yanchao Tan,Feiyu Xiong,Qingyu Yang,Edward Chung,Zhiyu li
机构: 未知
类目: Computation and Language (cs.CL)
备注: Submitted into EMNLP’2026
Abstract:Existing memory systems for long-horizon LLM agents often retrieve prior traces as passive context rather than converting them into executable capabilities. In this paper, we propose MSCE, a training-free Memory–Skill Co-Evolution framework that organizes agent experience into grounded step traces, reusable procedural policies, and declarative environmental cognition. MSCE crystallizes evidence-backed L2 policies with positive estimated gain into callable skills that retain evidence links, applicability boundaries, decision guidance, verification rules, and reliability estimates. It further introduces reflection-weighted value backfilling, which propagates sparse terminal feedback through dense local self-reflections to produce evidence-calibrated trace values for governing memory and skill evolution. Experiments on EvoAgentBench and LoCoMo demonstrate that MSCE significantly outperforms state-of-the-art skill-augmented and memory-driven agent baselines, exhibiting strong cross-domain transferability and lifelong-evolution capabilities.
[NLP-74] Can Multimodal Large Language Models Understand OCT?
【速读】： 该论文旨在解决现有医学影像分析基准在光学相干断层扫描（OCT）图像理解评估中过于简化的问题，即当前主流评测多局限于粗粒度疾病分类或孤立的视觉问答任务，未能充分涵盖从视觉感知到临床推理的完整认知过程。其解决方案的关键在于提出OCT-Bench——一个面向OCT图像理解的综合性基准，包含10,076个高质量多选题，基于7个公开数据集中的4,137张OCT图像构建。该基准遵循真实临床解读流程，建立了一个包含20个细粒度任务的分层能力分类体系，覆盖感知（Perception）、认知（Cognition）与推理（Reasoning）三个维度，涵盖成像特征、视网膜解剖结构、病灶特性、空间关系、疾病评估、治疗决策及预后管理等多方面能力。通过系统评估20种代表性多模态大语言模型（MLLMs），研究发现当前模型在可靠OCT理解方面仍存在显著差距，且医疗领域适配或模型规模增大并未在所有能力层级上一致提升性能。OCT-Bench为多模态大模型提供了全面、细粒度的评估框架，有助于识别能力瓶颈并推动以临床应用为导向的OCT理解技术发展。
链接: https://arxiv.org/abs/2607.16609
作者: Baochen Fu,Wenzhi Deng,Baihao Jin,Yang Li,Zihan Nie,Kailin Jiang,Yuntao Du,Weiye Song
机构: 1. Tsinghua University (清华大学); 2. Beijing Academy of Artificial Intelligence (北京人工智能研究院); 3. Institute for AI Industry Research (AIR), Tsinghua University (清华大学人工智能产业研究院); 4. Alibaba Cloud (阿里巴巴云); 5. Peking University (北京大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:
Abstract:Optical coherence tomography (OCT) imaging is essential for the diagnosis and treatment of retinal diseases. Although multimodal large language models (MLLMs) have demonstrated considerable potential in medical image analysis, existing benchmarks largely reduce OCT understanding to coarse-grained disease classification or isolated visual question answering, leaving the complete cognitive process from visual perception to clinical reasoning insufficiently evaluated. To address this limitation, we introduce OCT-Bench, a comprehensive benchmark dedicated to OCT image understanding. OCT-Bench comprises 10,076 high-quality multiple-choice questions constructed from 4,137 OCT images across seven public datasets. Following the real-world clinical interpretation workflow, we establish a hierarchical capability taxonomy consisting of 20 fine-grained tasks across three dimensions: Perception, Cognition, and Reasoning. These tasks cover a broad range of capabilities, including imaging attributes, retinal anatomy, lesion characteristics, spatial relationships, disease assessment, therapeutic decision-making, and prognostic management. We systematically evaluate 20 representative MLLMs, including proprietary models, open-source general-purpose models, and medical-domain models. Experimental results demonstrate that current models remain substantially short of reliable OCT understanding. Moreover, neither medical-domain adaptation nor increased model scale consistently improves performance across capability levels. OCT-Bench enables comprehensive and fine-grained evaluation of MLLMs, providing a foundation for identifying capability bottlenecks and advancing clinically grounded OCT understanding.
[NLP-75] NOWJ@COLIEE 2026: Adaptive Pipelines for Legal Retrieval and Reasoning
【速读】： 该论文旨在解决法律人工智能竞赛COLIEE 2026中五个核心任务的复杂法律文本理解与推理问题，涵盖法律案例检索、法律蕴含判断、成文法检索与蕴含、法律文本蕴含以及法律判决预测等关键挑战。其解决方案的关键在于构建多阶段、模块化且高度适配不同任务特性的端到端框架：在案例检索任务中，采用四阶段流水线（候选过滤、互补嵌入的密集检索、基于微调生成式重排序器与MLP对偶分类的重排序、自适应查询截断预测）以提升检索精度；在法律蕴含任务中，融合词频统计（BM25）、T5重排序与大语言模型（LLM）一致性集成验证，增强语义匹配能力；针对成文法相关任务，引入检索增强生成（Retrieval-Augmented Generation, RAG）架构，结合注意力重排序与少样本提示推理，实现精准法律条文理解；在法律文本蕴含任务中，提出动态路由机制，根据查询难度自动分配至少样本平衡求解器或结构化零样本思维链（Chain-of-Thought）求解器，实现推理策略的智能适配；在判决预测任务中，采用分层变压器（Hierarchical Transformer）结合条件随机场（CRF）层、论点关系挖掘与概率论点图推理，有效捕捉法律论证的层级结构与逻辑关联。整体方案的核心优势在于通过任务定制化设计与多模型协同机制，显著提升了法律自然语言处理任务的准确性与鲁棒性。
链接: https://arxiv.org/abs/2607.16603
作者: Thuong-Hieu Ngo,Hoang-Trung Nguyen,Huu-Dong Nguyen,Xuan-Bach Le,Le-Dung Nguyen,Quang-Thanh Tran,Ha-Thanh Nguyen,Thi-Hai-Yen Vuong
机构: VNU University of Engineering and Technology (河内国立大学工程与技术学院); Center for Juris-Informatics, ROIS-DS College of Engineering Computer Science, VinUniversity (越南维大学工程与计算机科学学院司法信息研究中心)
类目: Computation and Language (cs.CL)
备注: Presented at COLIEE 2026
Abstract:This paper presents the methodologies and results of the NOWJ team’s participation across all five tasks of the COLIEE 2026 competition. For Task 1 (Legal Case Retrieval), we propose a four-stage pipeline comprising candidate filtering, dense retrieval with complementary embedding models, cross-encoder reranking via fine-tuned generative rerankers and MLP-based pairwise classification, and adaptive per-query cutoff prediction. For Task 2 (Legal Case Entailment), we combine BM25 filtering, T5-based reranking, and LLM-based entailment verification with consensus ensemble. For Task 3 (Statute Law Retrieval and Entailment), we adopt a retrieval-augmented generation framework with dense retrieval, attention-based reranking, and few-shot-prompted LLM reasoning. For Task 4 (Legal Textual Entailment), we introduce a dynamic routing pipeline that classifies query difficulty and dispatches cases to either a balanced few-shot solver or a structured zero-shot chain-of-thought solver. For the Pilot Task (Legal Judgment Prediction), we combine hierarchical transformers with CRF layers, argument relation mining, and probabilistic argumentation graph reasoning.
[NLP-76] Encoding EEG Signals to Examine Human-Like Next-Word Prediction Behaviour in Language Models
【速读】： 该论文旨在解决生成式语言模型（Generative Language Models, GLMs）在阅读理解任务中预测准确率虽接近人类水平，但其是否真正捕捉了与人类认知加工过程一致的神经信号这一关键问题。传统评估方法主要依赖于预测精度，但忽略了模型在认知层面的可解释性与生物学合理性。本文的核心解决方案在于通过构建基于两种信息量度（即Top-1预测概率与意外度，surprisal）的回归变量，以预测由脑电图（EEG）记录的事件相关电位（ERP），从而对人类与模型在阅读过程中的认知动态进行精细化建模。研究发现，仅有意外度（surprisal）能有效预测与语言处理相关的ERP成分，尤其在具有高语义内容的开放类词汇中表现显著；而顶1预测概率则不具备此类关联性。这一结果表明，单纯提升模型参数规模和计算资源并不必然导致模型在认知机制上趋近于人类语言处理模式，从而挑战了“模型越大越像人”的普遍假设。
链接: https://arxiv.org/abs/2607.16549
作者: Boi Mai Quach,Binh T. Nguyen,Cathal Gurrin,Graham Healy
机构: Dublin City University, Ireland; VNUHCM – University of Science
类目: Computation and Language (cs.CL)
备注:
Abstract:Language models (LMs) are trained to excel at predicting the next word in the sequence given prior context, and humans also share this predictability in reading comprehension. Neuroscience research reveals that next-word predictability influences brain response, as recorded at millisecond resolution using electroencephalography (EEG). While our evidence indicates that advanced LMs achieve accuracies closely aligned with human performance at the next-word prediction task, this raises the question: Does higher prediction accuracy necessarily mean that these models adequately capture the cognitive signals associated with human reading comprehension? Here, we generate regressors for both humans and LMs based on two information measures, including top-1 prediction and surprisal, to predict event-related potential (ERP) elicited from EEG recordings which reflect different stages of cognitive processing during reading. We argue that modelling ERP patterns offers fine-grained analysis of the cognitive plausibility of various LMs during reading. Our results indicate that only surprisal potentially correlates with language-processing ERPs, especially for open-class words with high semantic content. Moreover, our findings challenge the assumption that scaling LMs with increased parameters and computational budgets will consistently lead to improved convergence with human-like linguistic processing.
[NLP-77] Committed Before Reasoning : Behavioral Reproduction and Preliminary Activation-Level Evidence of Answer Pre-Commitment in an Open-Weight LLM
【速读】： 该论文旨在解决生成式 AI（Generative AI）在推理过程中出现“先入为主”（answer-first）现象的问题，即模型在未充分遵循任务前提的情况下，提前锁定答案并构造事后合理化推理，导致输出与任务逻辑矛盾。其核心问题是：尽管“开车前往洗车店”是唯一可行选项（因车辆需抵达目的地），但模型在多数情况下仍错误推荐“步行”。解决方案的关键在于揭示这一偏差的内在机制——通过设计一个最小化探测任务，结合行为实验与激活层分析，发现模型在生成答案前的隐藏状态中已表现出显著的“步行倾向”读出信号（walk-leaning read-outs），且该信号不受词汇表层偏倚或文本恢复的影响。此外，研究还指出，问题表述方式对模型行为具有决定性影响，强调了实验设计中正向对照的重要性。这些发现表明，当前大模型的推理过程存在深层认知偏差，其根源可能在于训练范式中的响应偏好而非真实逻辑推导。
链接: https://arxiv.org/abs/2607.16451
作者: Heejin Jo
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 8 pages. Code, data, and all reported statistics: this https URL
Abstract:Chat models sometimes commit to an answer and then produce reasoning that justifies it rather than deriving it – even when the answer contradicts a task premise. We study a minimal probe: “I want to wash my car. The car wash is 100 meters away. Should I walk or drive?” Only drive works (the car must be at the car wash), yet models overwhelmingly recommend walking. (1) Behavioral reproduction: on Qwen3-8B across five system-prompt conditions (210 rollouts), the wrong commitment occurs in 85-100% of sampled rollouts per condition and 100% of greedy rollouts, in both thinking and non-thinking modes; a 4,096-token thinking budget does not repair it. (2) Preliminary activation-level evidence: probing hidden states with a pretrained, training-free activation oracle (no task-specific probe training) at positions before the answer text is emitted, “walk” read-outs exceed a neutral-context baseline (68% vs. 17%; walk-committing rollouts p=.005, drive-committing rollouts p=.005, Fisher exact) – notably, rollouts that eventually answer drive also read as walk-leaning before commitment (5/6). The oracle’s default on unrelated content is “drive” (83%), so the read-outs are not lexical bias; stratifying by literal walk/drive occurrence shows they are not text recovery either (spans containing “drive” still read out walk; in balanced lexical fields, per-rollout walk-majorities beat a per-prompt neutral baseline 15/22 vs. 1/8, p=.01; drive-committing rollouts 6/6, p=.002). Samples are small and the within-rollout positional gradient is not significant (p=.34); we frame these results as preliminary. (3) Methodological: with fixed oracle, activations, and positions, question wording alone moves a positive control from 2/16 (open question) to 11/16 (closed); negative oracle results are uninterpretable without per-wording positive controls.
[NLP-78] One Modality to Forget Them All: Enhancing Cross-Modal Unlearning in Vision-Language Models
【速读】： 该论文旨在解决多模态大模型中跨模态知识遗忘（cross-modal unlearning transfer）的不完整性与不对称性问题，即在视觉-语言模型（Vision-Language Models, VLMs）中，对某一模态（如文本）进行知识删除后，该遗忘效果是否能有效传递至另一模态（如视觉），以及其鲁棒性如何。核心挑战在于，现有方法普遍假设单模态的遗忘可泛化至另一模态，但实际中存在显著的跨模态遗忘转移偏差和脆弱性。研究发现，尽管部分情况下文本模态的遗忘可强效传递至视觉模态，但在针对文本呈现形式进行的字形攻击（typographic attacks）下，已被删除的知识可被轻易恢复，表明当前遗忘机制本质为“浅层”（shallow）遗忘，缺乏深层表征层面的彻底清除。为此，论文提出 \textscCrossInf，一种基于影响引导（influence-guided）的缓解策略，其关键在于识别并聚焦于对跨模态泛化贡献最大的变压器块（transformer blocks），通过针对性强化这些关键组件的遗忘，从而显著缩小跨模态遗忘差距（减少超过50%），同时在保持模型功能可用性的前提下，大幅提升对字形攻击的鲁棒性，将攻击成功率降至接近零。研究还结合人类评估（κ=0.77）与中心核对齐（Centered Kernel Alignment, CKA）分析，验证了跨模态遗忘行为的不对称性及深层原因，为构建更安全、可靠的多模态模型知识管理机制提供了理论与技术支撑。
链接: https://arxiv.org/abs/2607.16442
作者: Sudharshan Balaji,Yili Ren,Guangjing Wang,Yimin Chen,Ning Wang
机构: University of South Florida (南佛罗里达大学); University of Massachusetts Lowell (马萨诸塞大学洛厄尔分校)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Cryptography and Security (cs.CR)
备注: 18 pages
Abstract:Machine unlearning is widely used to remove hazardous knowledge from large language models. Modern Vision-Language Models (VLMs), however, process both text and visual inputs, raising a fundamental security question: does unlearning in one modality transfer to the other? We present the first systematic, bidirectional study of cross-modal unlearning transfer across three VLM architectures: LLaVA-1.5 (MLP projection), InstructBLIP (Q-Former), and IDEFICS (gated cross-attention). We find that unlearning transfers across modalities, but the transfer is asymmetric and incomplete. In some cases, text unlearning strongly transfers to vision. However, this robustness is not preserved under typographic attacks that manipulate the visual presentation of text. Under such attacks, previously unlearned knowledge can be readily recovered, indicating shallow unlearning. To address the transfer gap and shallow robustness, we propose \textscCrossInf, an influence-guided mitigation strategy. Motivated by the observation that different model components contribute unequally to cross-modal transfer, \textscCrossInf focuses unlearning on transformer blocks that most influence cross-modal generalization. It reduces the transfer gap by more than half in architectures with strong fusion, while preserving model utility. It also improves robustness under typographic attacks, reducing the attack success rate to near zero. We further conduct human evaluation with three annotators ( \kappa=0.77 ) to validate our findings. Finally, we analyze shallow unlearning using Centered Kernel Alignment (CKA), providing insights into the observed transfer behavior and robustness limitations. Comments: 18 pages Subjects: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL); Cryptography and Security (cs.CR) Cite as: arXiv:2607.16442 [cs.CV] (or arXiv:2607.16442v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.16442 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[NLP-79] RIMS: Preference Optimization via Smoothed Multi-pair Aggregation for Small-Scale LLM Retrieval-Augmented Generation
【速读】： 该论文旨在解决小规模语言模型（Small-scale Language Models, SLMs）在资源受限场景下进行检索增强生成（Retrieval-Augmented Generation, RAG）时，因模型容量有限而对噪声或虚假检索证据高度敏感的问题。现有基于偏好学习的方法存在两大局限：一是如RoseRAG采用硬选择（hard argmin/argmax）仅保留最困难的单一偏好对，导致其余信息被丢弃；二是将多个偏好对视为独立的二元比较，造成数据利用率低下。本文提出RIMS（Rejection-sampling based Inference with Multi-pair Soft aggregation），其核心解决方案在于构建一个三阶段偏好优化框架：首先利用目标SLM自身通过拒绝采样生成合成思维链（chain-of-thought）偏好数据，无需依赖专有大模型；其次设计可微的软聚合机制，以平滑算子替代硬选择，保留所有偏好对的梯度信号，同时维持边际感知选择的判别结构；最后在多对齐算法上应用平滑后的优化目标。理论分析表明，该平滑近似具有可控误差界，且相较于硬选择，软聚合能实现更紧的梯度对齐性。实验在四个多跳问答基准上验证了RIMS的有效性，结果表明其在多种SLM骨干网络下均显著优于现有基线，在噪声检索条件下持续提升精确匹配率（Exact Match）与F1分数。
链接: https://arxiv.org/abs/2607.16431
作者: Pei Tian,Zihan Dong,Tianci Liu,Linjun Zhang,Haoyu Wang
机构: Columbia University (哥伦比亚大学); Rutgers University (罗格斯大学); Purdue University (普渡大学); SUNY Albany (纽约州立大学阿尔巴尼分校)
类目: Computation and Language (cs.CL)
备注:
Abstract:Small-scale language models (SLMs) are attractive for retrieval-augmented generation (RAG) in resource-constrained settings, but their limited capacity makes them highly sensitive to noisy or spurious retrieved evidence. Existing preference-based methods such as RoseRAG select only the hardest single preference pair via hard argmin/argmax, discarding the remaining signal; others treat multiple pairs as independent binary comparisons, resulting in low data utilization. We propose RIMS, a three-stage preference optimization framework comprising (1) synthetic chain-of-thought preference data generation via rejection sampling using the target SLM itself without relying on proprietary models, (2) a differentiable soft aggregation mechanism that replaces hard selection with a smooth operator, preserving gradient signal from all preference pairs while retaining the discriminative structure of margin-aware selection, and (3) preference optimization with the smoothed objective applied to multiple alignment algorithms. We theoretically show that the smoothed approximation admits a controllable error bound and that smooth aggregation yields provably tighter gradient alignment to the oracle objective than hard selection. Experiments on four multi-hop question answering benchmarks show that our approach outperforms state-of-the-art baselines across multiple SLM backbones, achieving consistent gains in Exact Match and F1 under noisy retrieval conditions. Our implementation is available at this https URL.
[NLP-80] Multi-level context Modeling for consistent expert selection in Mixture-of-Experts
【速读】： 该论文旨在解决现有混合专家（Mixture-of-Experts, MoE）架构中专家选择机制因依赖浅层或孤立的标记表示而导致路由决策不稳定且语义不一致的问题。其核心挑战在于专家选择过程中上下文信息的不完整性，限制了专家的有效专业化。为此，作者提出多层级上下文融合混合专家（Multi-level Context Fusion MOE, MCF-MOE）框架，通过融合跨层语义聚合与局部标记级交互的互补信号，构建更具上下文感知能力的表示，从而实现更准确、更一致的专家选择。实验结果表明，MCF-MOE在语言建模与理解任务上显著优于现有强基准模型，验证了上下文完整性对提升专家路由性能的关键作用。
链接: https://arxiv.org/abs/2607.16427
作者: Shuhan Huang,Naifan Zhang,Yuanbo Tang,Yang Li,Wai Kin Victor Chan
机构: Tsinghua Shenzhen International Graduate School, Tsinghua University; School of AI, The Chinese University of Hong Kong (Shenzhen)
类目: Computation and Language (cs.CL)
备注:
Abstract:Mixture-of-Experts (MoE) enables efficient scaling of Transformer models by routing tokens to a small subset of experts. However, existing routers typically condition expert selection on shallow or isolated token representations, which often produce unstable and semantically inconsistent routing decisions across layers. In this work, we revisit expert selection from a representation perspective and identify context incompleteness as a key bottleneck limiting effective expert specialization. To address this issue, we propose Multi-level Context Fusion MOE (MCF-MOE), a framework that constructs context-aware representations by integrating complementary signals from cross-layer semantic aggregation and local token-level interactions, enabling more informative and consistent expert selection. Experiments on language modeling and understanding benchmarks demonstrate that MCF-MOE consistently improves routing consistency and downstream performance over strong MoE baselines, highlighting the importance of contextual completeness in expert routing. The code is available at this https URL.
[NLP-81] Quantifying Ranking Uncertainty in LLM Benchmarks
【速读】： 该论文旨在解决预训练模型在多任务基准测试中排名结果的不确定性问题，尤其聚焦于知识评估基准MMLU（Multi-Task Mixed Language Understanding）中的排名可靠性。现有方法通过汇总成对假设检验来构建排名置信区间以量化不确定性，但其忽略了实际评估中多种潜在误差来源。论文的关键贡献在于系统分析了影响MMLU排名的各类不确定性来源，并提出对传统假设检验进行改进，使其能够更准确地反映这些因素的影响。研究发现，不同学科领域（subject）间的性能波动显著，表明模型排名在跨主题比较时具有高度变异性，因此在比较大语言模型（LLM）性能或识别最优模型时，必须考虑这一变异性，否则可能导致错误结论。
链接: https://arxiv.org/abs/2607.16259
作者: Bitya Neuhof,Yuval Benjamini
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Machine Learning (stat.ML)
备注:
Abstract:Pretrained models are typically ranked on multi-task leaderboards to assess their effectiveness across diverse tasks. Rank confidence intervals were recently introduced as a method to quantify the uncertainty in these rankings by aggregating pairwise hypothesis tests. In this work, we analyze the sources of uncertainty in the knowledge evaluation benchmark MMLU and show how hypothesis tests can be modified to account for their effects. We demonstrate that ranking variability across MMLU subjects is substantial and should be considered when comparing LLMs or identifying the top-performing models.
[NLP-82] From Outcomes to Actions: Leverag ing Hindsight for Long-Horizon Language Agent Training ICML2026
【速读】： 该论文旨在解决生成式 AI（Generative AI）在复杂任务中应用强化学习（Reinforcement Learning, RL）时，因长时序交互导致的优化方差过高问题。其核心挑战在于难以准确区分长序列中各动作对最终回报的贡献，从而影响策略收敛与性能稳定性。为此，论文提出一种新型策略梯度方法——事后策略优化（Hindsight Policy Optimization, HPO），其关键在于将当前策略分布与事后（hindsight）策略分布共同投影至意图空间（intent space），并利用二者间的Wasserstein距离提取低方差的学习信号。通过在意图空间中聚合语义相似的状态与动作，该方法理论上可实现方差有界估计，并在实验中验证了其在提升策略性能稳定性和泛化能力方面的有效性。
链接: https://arxiv.org/abs/2607.16257
作者: Zishang Jiang,Tingyun Li,Jinyi Han,Xinyi Wang,Sihang Jiang,Yizhou Ying,Xiaojun Meng,Jiansheng Wei,Jiaqing Liang,Yanghua Xiao
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted at ICML 2026
Abstract:Reinforcement learning (RL) has become a widely adopted technique for improving large language models (LLMs) on complex tasks. Despite this progress, existing RL methods still face challenges in training agents with longer-horizon interactions. One major bottleneck is distinguishing the contribution of different actions in long-horizon interaction, leading to high optimization variance. To address this, we introduce a novel policy gradient method, Hindsight Policy Optimization (HPO), that projects both the current policy distribution and the hindsight distribution into an intent space and extracts low-variance learning signals from the Wasserstein distance between them. We theoretically and empirically show that aggregating semantically similar states and actions in the intent space yields a bounded-variance estimator and improves policy performance stably. Our code is available online.
[NLP-83] Feature Generation Using LLM s: An Evolutionary Algorithm Approach
【速读】： 该论文旨在解决机器学习流程中特征工程（feature engineering）的自动化难题，即如何从原始表格数据中自动生成具有判别性的新特征，以提升模型性能。传统特征工程依赖人工经验，耗时且难以泛化，而现有方法在捕捉复杂属性间非线性关系方面存在局限。本文提出一种基于大语言模型（Large Language Model, LLM）的特征生成框架，其核心在于利用LLM理解自然语言提示（prompt）的能力，结合已有特征通过数学与逻辑运算自动构造潜在有用的衍生特征。该方案的关键创新在于构建了一个端到端的特征生成与选择管道：首先由LLM根据输入特征和提示生成候选特征集，随后通过一个基于任务性能的评估算法筛选出最优特征组合。实验在八个来自不同领域和数据类型的数据集上验证了该方法的有效性，结果表明，在多数情况下，LLM能够生成对分类任务有实际贡献的新特征，并显著提升模型表现，证明了将大语言模型用于自动化特征工程的可行性与优越性。
链接: https://arxiv.org/abs/2607.16255
作者: Aria Nourbakhsh,Benoît Alcaraz,Christoph Schommer
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:A crucial step in machine learning pipelines is to present each entity with features or attributes that are representative of the characteristics of the processed entities. Feature engineering is an important step in finding a relation among attributes that otherwise may not be processed by the ML algorithms. Meanwhile, Large Language Models have shown promising abilities in coding, mathematical reasoning, and processing world knowledge. In this work, we utilize an LLM for the problem of feature generation from tabular data based on the previously given features. We have created a pipeline that takes a set of attributes and a prompt to generate new features. Then, our selection algorithm selects the best-performing sets of attributes. We apply our method to eight datasets from different domains and data types. Our results show that, in most cases, the language model can produce new features based on mathematical and logical operators that are useful for the given tasks and can improve classification results.
[NLP-84] CIGPO: Contextual Information-Gain Policy Optimization for Multi-Turn Evidence-Reading LLM Agents
【速读】： 该论文旨在解决在仅基于最终结果（outcome-only）的强化学习框架下训练多轮证据阅读智能体时，因中间轮次缺乏直接奖励而导致训练不稳定的难题。其核心问题在于：在生成式强化学习策略优化（GRPO）中，由于所有采样轨迹均遭受最低格式惩罚（-2.0），导致组内相对优势（group-relative advantage）消失，进而引发零优势锁定（zero-advantage lock-in）机制——政策梯度损失归零，陷入优化死锁。解决方案的关键在于引入方差注入策略：通过为每个中间证据阅读轮次分配独立的逐轮奖励信号，防止组内奖励分布坍缩至单一值，从而维持GRPO所需的优势差异性。具体实现上，提出上下文信息增益策略优化（CIGPO），利用冻结参考模型对真实答案的对数似然边际提升作为逐轮奖励信号，并对信息增益（Information Gain, IG）与F1分数进行独立归一化，结合IG权重课程学习（IG-weight curriculum）。在Qwen2.5-3B-Instruct模型上，CIGPO在HotpotQA任务中达到标准F1 0.518（相较于基线0.252提升105%），显著优于最优GRPO检查点（0.430）和最终崩溃的GRPO（0.000），且全程保持有意义的奖励方差，有效避免了零优势锁定。研究揭示了奖励方差坍缩是结果导向型GRPO的一个具体失效模式，并证明了基于信息增益的逐轮奖励可在此设定下有效缓解该问题。
链接: https://arxiv.org/abs/2607.16244
作者: Hao Dou
机构: Harbin Institute of Technology (哈尔滨工业大学)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Training multi-turn evidence-reading agents with outcome-only reinforcement learning is unstable because intermediate turns receive little direct credit. In HotpotQA experiments with Qwen2.5-3B-Instruct, GRPO initially improves (standard F1 0.430) but subsequently collapses to 100% format-violating outputs. Training-log diagnosis reveals a zero-advantage lock-in mechanism: all sampled trajectories receive the minimum format penalty (-2.0), group-relative advantages vanish, and the policy-gradient loss becomes zero–an optimization deadlock. We propose a variance-injection strategy: by assigning per-turn rewards to intermediate evidence-reading turns, we prevent the group reward distribution from collapsing to a single value–preserving the variation that GRPO’s group-relative advantage requires. Contextual Information-Gain Policy Optimization (CIGPO) implements this strategy using the marginal increase in the frozen reference model’s log-likelihood of the ground-truth answer as the per-turn signal. With separate normalization of IG and F1 rewards and an IG-weight curriculum, CIGPO reaches a standard F1 of 0.518 on HotpotQA at the 3B scale (from 0.252 base; +105%), compared with 0.430 for the best GRPO checkpoint and 0.000 for the final GRPO checkpoint. CIGPO maintains meaningful reward variance and avoids zero-advantage lock-in throughout training. These results identify reward-variance collapse as a concrete failure mode of outcome-only GRPO and show that turn-level IG rewards can prevent it in this HotpotQA setting.
[NLP-85] RAIL Guard: Closing the Evaluation-to-Remediation Gap in Responsible AI for LLM Agents
【速读】： 该论文旨在解决现有大型语言模型（LLM）代理防护系统依赖二元分类器对不安全内容进行拦截，导致组织不得不丢弃失败输出并从头重试所引发的效率低下与资源浪费问题。其核心解决方案是提出一种闭环负责任人工智能（RAIL Guard）管道，通过在八个可度量维度上评估LLM输出，并采用“评估-重写-再评估”的迭代修复机制实现动态修正。关键创新在于将原本一次性阻断的静态防护转变为可自我修复的闭环流程，实验表明该方法在4,276条内容输出和6,400个代理工具调用场景下实现了96.9%的收敛率，显著优于传统“阻断-重试”方式的49.1%；尤其值得注意的是，基于反馈的自修复机制在可修复维度上达到86.6%的收敛率，且无显著效用损失（p = 0.177）。研究进一步揭示了可修复维度与结构性维度的根本差异：透明性（93.0%失败）、问责性（92.8%失败）及包容性（82.5%失败）等结构性维度需通过架构层面改进而非算法修复来解决。该系统已作为开源SDK发布，支持实际部署与持续优化。
链接: https://arxiv.org/abs/2607.16215
作者: Sumit Verma,Pritam Prasun,Pritish Kumar
机构: Responsible AI Labs(负责任的人工智能实验室)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: 15 pages, 10 figures, 4 tables. Code: this https URL (Python). Benchmark: this https URL
Abstract:Existing guardrail systems for large language model agents operate as binary classifiers that block unsafe content, leaving organizations to discard failing outputs and retry from scratch. We introduce RAIL Guard, a closed-loop responsible AI pipeline that evaluates LLM outputs across eight measurable dimensions and iteratively remediates failing outputs through an evaluate-rewrite-reevaluate loop. We evaluate the pipeline across three experiments on four frontier LLMs and 4,276 content outputs plus 6,400 agent tool-call scenarios. Closed-loop remediation achieves 96.9% convergence versus 49.1% for block-and-retry, though the highest-convergence method reduces utility by 22.3%; feedback-driven self-repair achieves 86.6% convergence on fixable dimensions with no significant utility loss (p = 0.177). Pre-tool-call evaluation reduces unsafe agent executions by 33% (p = 0.007) with zero impact on task completion. We identify a key distinction between fixable dimensions that respond to remediation and structural dimensions (Transparency at 93.0%, Accountability at 92.8%, and Inclusivity at 82.5% failure) that require architectural rather than algorithmic solutions. The system is available as open-source SDKs.
[NLP-86] ColGraphRAG : Late-Interaction Evidence Retrieval for Multimodal GraphRAG
【速读】： 该论文旨在解决生成式多模态问答中，基于图结构的证据推理因视觉候选项排序机制不足而导致的细粒度对齐失效问题。现有方法依赖单向量双编码器（bi-encoder）相似性计算对图关联图像节点进行排序，但这种策略会丢失图像块（patch）与标记（token）级别的局部结构信息，从而影响下游推理的准确性。其解决方案的关键在于：在保持原有离线图构建、文本与表格侧检索、结构化提取及下游推理流程不变的前提下，将图关联图像节点的候选排序算子替换为类MaxSim的晚期交互多向量评分机制（late-interaction MaxSim-style multi-vector scoring），该机制源自ColBERT/ColPali系列模型，能够保留并利用图像内部的多层次语义表示，实现更精细的跨模态对齐。实验结果表明，在MultimodalQA数据集上，该改进显著提升了图关联图像候选项在检索阶段的点估计性能，并带来了下游问答任务的整体性能提升，尤其在视觉证据起决定作用的任务中表现更为突出；而在以文本为主导的问题上则呈现混合趋势，这为图结构中引入视觉证据的有效性提供了机制层面的支持，但需进一步开展更广泛的验证与更精细的图级诊断分析。
链接: https://arxiv.org/abs/2607.16208
作者: Seonok Kim
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:
Abstract:Graph-grounded multimodal question answering organizes text, tables, and images in a structured evidence graph, yet end-to-end accuracy depends on which multimodal assets are ranked highly enough to enter downstream reasoning; for graph-linked images, single-vector bi-encoder similarity can discard patch- and token-level structure needed for fine-grained alignment. We evaluate replacing the visual candidate-ranking operator over graph-linked image nodes with late-interaction MaxSim-style multi-vector scoring in the ColBERT/ColPali lineage, while keeping offline graph construction, text- and table-side retrieval, structured extraction, and downstream reasoning unchanged. On MultimodalQA, this change is associated with improved retrieval-stage point estimates for graph-linked image candidates and downstream QA gains, with larger movement where visual evidence matters most and mixed trends on text-dominant questions; we interpret the pattern as mechanism-level evidence for graph-linked visual evidence inclusion, while broader validation and finer graph-level diagnostics remain important future work.
[NLP-87] DocOCR-Eval: A Correction-Based Framework for OCR Tool Selection Without Ground Truth
【速读】： 该论文旨在解决在标签稀缺（label-scarce）场景下，如何为特定文档集合选择最优的文档解析（document parsing）方案这一关键挑战。尽管已有多种光学字符识别（OCR）引擎和多模态大语言模型（MLLMs）可用于文档解析，但其性能差异显著，且缺乏有效的自动化评估手段以指导工具选型。论文提出了一种无标注（annotation-free）的评估框架——DocOCR-Eval，其核心创新在于采用三阶段纠错与排序策略，无需依赖真实标注（ground-truth）即可近似逼近基于人工标注的工具排序结果。该方法通过聚合多个MLLM的输出，逐步提升与人工标注排名的一致性，从而实现对OCR工具在真实、标签有限环境下的可靠筛选。该解决方案的关键在于利用多模型协同推理能力弥补单一模型上下文理解能力不足的问题，并通过无需人工标注的自动化评估机制，为跨领域、跨语言的真实文档集合部署提供可落地的实践指导。
链接: https://arxiv.org/abs/2607.16203
作者: Zihan Xu,Puzhen Wu,Lawrence Chun Man Lau,Wei Liu,Sirui Li,Yifan Peng,Yihao Ding
机构: 未知
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Work in progress
Abstract:Document parsing is a foundational step for document understanding tasks such as visual question answering and key information extraction, as it transforms unstructured scanned images into structured representations by extracting textual, visual, and layout information. While numerous Optical Character Recognition (OCR) engines and multimodal large language models (MLLMs) have been developed for this purpose, selecting an appropriate document parsing solution for a given document collection remains challenging, particularly in label-scarce settings. In this work, we conduct a systematic evaluation of text recognition performance across a diverse set of OCR engines and state-of-the-art MLLMs on multiple scanned document benchmarks spanning different domains and languages. Motivated by the limited contextual reasoning capabilities of many OCR engines and the high cost of manual annotations, we propose DocOCR-Eval, an annotation-free evaluation framework for automatic OCR assessment and selection. DocOCR-Eval employs a three-staged correction and ranking strategy to approximate annotation-based tool ordering without ground-truth labels. We show that aggregating across multiple MLLMs progressively improves alignment with annotation-based rankings. Extensive experiments further demonstrate that reliable OCR tool selection can be achieved in realistic, label-limited settings, providing practical guidance for deploying document parsing systems across diverse real-world document collections.
[NLP-88] Calibrating Semantic Uncertainty from Observable Language-Model Probabilities
【速读】： 该论文旨在解决语言模型（Language Model）生成的词级概率无法直接反映专业决策所需的关键状态（如诊断、假设或运行条件）的不确定性问题。现有模型输出的数值置信度缺乏可靠性验证，且其结果高度依赖于提示（prompt）的词汇形式，导致语义稳定性不足。为此，作者提出“语义映射”（semantic map）这一核心解决方案：通过构建一个预定义、可检验的桥梁，将语言模型对自然语言响应的概率分布，转化为对目标隐含状态（finite-valued latent state）的概率估计，采用半参数推断框架实现。该方法以参考模型定义目标后验分布，语言模型提供无约束的响应条件分布，并通过留出校准数据连接二者。研究进一步推导了后验误差界及存在性、唯一性、稳定性与序列贝叶斯更新的充分条件。关键在于，语言模型的概率依赖于提示的词汇表达形式，而目标后验应保持在信息等价重述下不变；实证测试表明，基于语义映射获得的概率在联邦储备经济与金融数据集及可控模拟中，优于原始数值置信度，能有效恢复隐藏后验并维持合理的不确定性覆盖，对提示改写具有强鲁棒性，并能正确响应证据变化。该方法将提示工程中的语义不稳定性问题转化为可检验的统计建模任务，在满足接受条件时可提供可审计的后验估计，为分类、推荐及其他流畅响应的语义稳定性审计提供了通用范式。
链接: https://arxiv.org/abs/2607.17447
作者: Matthew F. Dixon
机构: Artificial Intelligence Finance Institute (AIFI)(人工智能金融研究所); Quiota LLC(奎奥塔有限责任公司)
类目: Methodology (stat.ME); Computation and Language (cs.CL); Statistics Theory (math.ST); Machine Learning (stat.ML)
备注:
Abstract:Language models produce probabilities over words, but professional decisions require uncertainty over meaningful states such as diagnoses, hypotheses or operational conditions. A model’s printed numerical confidence does not establish reliability. We introduce a semantic map: a prespecified, testable bridge from probabilities over verbal responses to probabilities over declared states, formulated as semiparametric inference for a finite-valued latent state. A reference model defines the target posterior, the language model supplies an unrestricted conditional distribution over verbal responses, and held-out calibration connects them. We derive posterior-error bounds and conditions for existence, uniqueness, stability and sequential Bayesian updating. Crucially, language probabilities depend on the prompt’s lexical form, whereas the target posterior is unchanged by information-equivalent rewording. We test the method on professional market text compiled from Federal Reserve economic and financial series and on controlled simulations with exact posteriors. Across two fitted language models, language-derived probabilities outperform printed numerical probabilities, recover held-out posteriors with valid uncertainty coverage, remain largely stable under paraphrasing, and respond appropriately to altered evidence. The broader implication is that prompt engineering optimizes a wording-dependent response, whereas scientific and professional use requires validated stability of application-relevant meaning. The semantic map turns this general concern into a testable statistical problem and, when its acceptance conditions hold, yields an auditable posterior estimate. The same principle offers a template for auditing classifications, recommendations and other fluent responses that may conceal semantic instability.
[NLP-89] he Geometry of Semantic Space: A Continuous Geometric Framework for the Transformer Architecture
【速读】： 该论文旨在解决大语言模型（Large Language Models, LLMs）中复杂动态行为的理论建模问题，特别是针对Transformer架构中离散代数操作在连续时空下的统一几何解释缺失这一核心挑战。其关键在于构建一个连续几何框架，将Transformer的核心组件（如RMSNorm、RoPE、Softmax Attention、FFN、残差连接、SGD与权重衰减等）统一映射为定义在语义纤维丛\calE = \calM \times \mathbb{R}^d上的积分微分方程（Integro-Differential Equation, IDE），基于“词元序列构成具有规范测度格点的离散1-流形”这一单一几何公理。该框架通过微分几何、测度论与随机微积分的融合，实现了对注意力机制作为Schrödinger桥的熵最优传输本质、SGD作为违背细致平衡的Itô扩散的非平衡热力学特性的定量预测。实验验证覆盖五种主流模型（Qwen3、LLaMA-3.1、Gemma-3、GPT-2、Mistral，参数量从124M至8B），结果在多个维度上高度一致：包括机器精度下的ϵ−1/2 Lipschitz标度校准（R2=1.000）、Lie-Trotter算子分裂引起的挠率、对称消融不稳定性所证实的拓扑稳定性双律、罗佩环面（RoPE torus）上\calO(1/\sqrt{k})量级的庞加莱回归抑制、热力学上下文边界相变以及跨优化器（AdamW与Pure SGD）验证的非平衡稳态参数涡旋。这表明，以连续随机微分几何为视角分析Transformer，可提供一套兼具预测性与描述性的理论语言，用于刻画大型语言模型的稳定性极限、上下文容量边界及优化动力学特性。
链接: https://arxiv.org/abs/2607.17146
作者: Zhihua Liang
机构: INFN, Sezione di Cagliari (意大利核物理国家研究所，卡利亚里分部), I-09042 Monserrato (CA), Italy
类目: Disordered Systems and Neural Networks (cond-mat.dis-nn); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:
Abstract:We present a continuous geometric framework that models the discrete algebraic operations of the Transformer architecture as an integro-differential equation (IDE) on a semantic fiber bundle \calE = \calM \times \R^d . Beginning from a single geometric axiom – that the token sequence forms a discrete 1 -manifold equipped with a canonical measure lattice – we translate every core component of the modern Transformer (RMSNorm, RoPE, Softmax Attention, FFN, Residual Stream, SGD, Weight Decay) into a cohesive vocabulary of differential geometry, measure theory, and stochastic calculus. The resulting framework yields quantitative predictions spanning entropic optimal transport (Attention as a Schrödinger bridge) and non-equilibrium thermodynamics (SGD as Itô diffusion violating detailed balance). We conduct a six-part experimental campaign across five architectures (Qwen3, LLaMA\nobreakdash-3.1, Gemma\nobreakdash-3, GPT-2, Mistral) spanning 124 M to 8 B parameters. The empirical observables are quantitatively consistent with the geometric predictions: the \epsilon^-1/2 Lipschitz scaling calibration at machine precision ( R^2 = 1.000 ), the Lie–Trotter operator-splitting torsion, the symmetric ablation instability confirming the Dual-Law of Topological Stability, the \calO(1/\sqrtk) thermodynamic suppression of Poincaré recurrence on the RoPE torus, the thermodynamic context-limit phase transition, and the Non-Equilibrium Steady State parameter vortex – verified across two optimizers (AdamW and Pure SGD) to exclude momentum artifacts. The results demonstrate that analyzing Transformers through the lens of continuous stochastic differential geometry provides a predictive descriptive vocabulary for the stability limits, context bounds, and optimization dynamics of Large Language Models.
信息检索
[IR-0] jina-reranker-v3.5: An Efficient Listwise Reranker with Hybrid Attention and Self-Distillation
链接: https://arxiv.org/abs/2607.18152
作者: Christina Nasika,Feng Wang,Antonis Krasakis,Han Xiao
类目: Information Retrieval (cs.IR)
备注: 13 pages, 2 figures, 9 tables
Abstract:Listwise rerankers are the discriminative core of agentic retrieval pipelines, yet production deployment demands efficiency, domain robustness, and fluency on semi-structured data at the same time. We present jina-reranker-v3.5, a 0.6B-parameter listwise reranker that meets these demands together without sacrificing the cross-document comparison that makes its predecessor jina-reranker-v3 effective. jina-reranker-v3.5 keeps the last-but-not-late (LBNL) interaction of jina-reranker-v3 and reworks it along three axes. It replaces uniform global attention with a hybrid schedule of three sliding-window layers followed by two global layers, pinning the terminal layer to global as LBNL readout requires. It trains on a curated multi-domain mixture that spans legal, medical, financial, multilingual, and structured retrieval. It transfers quality through a three-stage self-distillation recipe in which a full-attention teacher sets an upper bound that a sparse-attention student then recovers under a staged adaptation protocol. jina-reranker-v3.5 reaches 63.20 nDCG@10 on BEIR, matching a 4B model at roughly 7x fewer parameters, and improves over jina-reranker-v3 on MIRACL and RTEB as well. Its largest gains come on semi-structured retrieval, where it lifts nDCG@10 by 9.6 points over jina-reranker-v3 and leads all rerankers of comparable size. The hybrid schedule further cuts listwise inference latency by up to 1.56x. We release the model weights on Hugging Face under a non-commercial license.
[IR-1] Evidence-in-the-Loop: Trace-Driven Optimization for Customer-Service LLM Agents
链接: https://arxiv.org/abs/2607.18039
作者: Chunming Wu,Dafei Qiu,Congde Yuan,Charles Quan,Jun Wu,Suipeng Li,Mo Wu,Gavin Xie,Hope Chen,Max Yao
类目: Information Retrieval (cs.IR)
备注:
Abstract:Production customer-service bots must improve answer quality across iterative releases, yet large language models must not bypass evidence boundaries, policy rules, or human-handoff safeguards. We present an \textbfEvidence-Grounded Customer-Service Agent Workflow deployed in a real-world customer-service setting. BM25 recall, issue-title-vector recall, issue-description-vector recall, weighted RRF fusion, and cross-encoder reranking construct grounded FAQ evidence for controlled LLM decisions. Policy-guided orchestration then combines this RAG evidence with scenario-specific rule evidence, conversation memory, and clarification state inside a fixed LangGraph DAG~\citelanggraph2024. The paper contributes three reusable deployment patterns: \textbfhybrid RAG evidence construction, where multi-channel retrieval and reranking produce auditable FAQ candidates; \textbfevidence-grounded issue/action decision, where an Evidence-Grounded Decision Module selects an issue/action from typed FAQ evidence and scenario-specific rule evidence; and \textbftrace-driven RAG and reranker improvement, where traces diagnose whether failures come from recall, ranking, final candidate selection, clarification, rule-derived evidence, or action policy, and where reranker fine-tuning is evaluated not only for in-domain gain but also for forgetting risk.
[IR-2] Remote Awareness of Seafloor Images Collected by AUVs over Low-Bandwidth Communication Links
链接: https://arxiv.org/abs/2607.18013
作者: Adrian Bodenmann,Cailei Liang,Miquel Massot-Campos,Samuel Simmons,Alexander B. Phillips,Alberto Consensi,Matthew Kingsland,Rashiid Sherif,Stan Brown,Adam Riese,Blair Thornton
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG); Robotics (cs.RO); Image and Video Processing (eess.IV)
备注:
Abstract:This paper introduces a method for real-time processing and transmission of autonomous underwater vehicle (AUV) imagery over low-bandwidth communication links. It leverages artificial intelligence (AI) techniques to identify a set of images that best represent an entire dataset, or automatically finds the most similar images to a given query image for transmission to operators. Combined with metadata of a larger set of images, compressed versions of the selected images can be transmitted over satellite communication links or underwater modems, and provide operators on shore with information about the type of imagery the AUV is collecting while it is still deployed. Data from three deployments off the coast of the UK and in Gran Canaria using different AUVs and imaging systems demonstrate the method in the field. It achieved an almost 400,000-fold reduction in data volume compared to the raw data size, enabling transmission of data summaries of a 2-hour 47-minute-long mapping mission in just over 34 minutes over low-bandwidth satellite communication.
[IR-3] MagicSelector: Joint Optimization for Agent Tool Selection via Counterfactual Decomposition and Progressive Reranking
链接: https://arxiv.org/abs/2607.17751
作者: HONOR Agentic Search Team:Zhengzong Chen,Lei Tang,Lijun Liu,Chuandi Jiang,Fan Yang,Keyun Chu,Chu Zhao,Shihao Liu,Minghang Li,Bo Liang,Can Wen,Hailong Wu,Jingnan Ju,Mian Liu,Nengbin Zhang,Peiqiang Wang,Penghe Nie,Qinhui Gu,Sijia Lv,Siqi Chen,Wei Zhang,Yang Xu,Yuhao Qian,Yuxiang Zhang,Zeng Cheng,Zhen Wang,Zuan Chen,Yuanyuan Zhao,Fei Huang
类目: Information Retrieval (cs.IR)
备注:
Abstract:We present MagicSelector, a joint optimization framework integrating Counterfactual task decomposition, Progressive reranking, and Dynamic Top-K, designed to address the fundamental challenges of tool retrieval in agents. MagicSelector is a specialized framework capable of translating ambiguous user instructions into executable atomic subtasks and guiding high-precision tool retrieval, effectively mitigating redundant noise and severe context distraction in out-of-domain (OOD) this http URL empower MagicSelector with these capabilities through three key contributions: (1) a preferenceguided counterfactual task decomposition mechanism that utilizes a counterfactual reward to quantify the marginal causal gain of decomposition on retrieval ranking, effectively imposing fine-grained structural supervision on logical coherence; (2) a progressive tool reranking method driven by self-distillation hard negative mining, which optimizes both point-wise and list-wise relevance to enhance fine-grained discrimination among highly similar tools; and (3) a dual semantic boundary-aware dynamic Top-K strategy that adaptively monitors reranking score cliffs and inter-tool semantic shifts to dynamically truncate the candidate list, maximizing relevant tool recall while filtering long-tail noise. Evaluated on MTDTool, the first task decomposition benchmark we constructed tailored for mobile multi-turn interactions with process-level annotations, MagicSelector yields promising performance. Extensive experiments demonstrate that MagicSelector significantly outperforms state-of-the-art methods in terms of tool retrieval accuracy, OOD generalization capability, and overall token efficiency, thereby demonstrating the effectiveness of our proposed framework.
[IR-4] ANNLib: A Development Framework for Efficient Approximate Nearest Neighbor Search
链接: https://arxiv.org/abs/2607.17582
作者: Zheqi Shen,Jingbo Su,Zijin Wan,Yan Gu,Yihan Sun
类目: Machine Learning (cs.LG); Information Retrieval (cs.IR)
备注:
Abstract:Approximate Nearest Neighbor Search (ANNS) plays a pivotal role in modern deep learning pipelines. Recently, many ANNS systems have been proposed to either provide broad functionality or reach high performance. However, it is yet difficult to achieve both with minimal programming efforts. We propose ANNLib to address the gap. ANNLib is a library that provides a programming framework for achieving high performance and flexible functionality in ANNS systems, based on popular graph-based ANNS algorithms. We carefully decouple and independently optimize both the algorithm and the data structure components of an ANNS system. In addition, we integrate state-of-the-art algorithms and data structures into ANNLib as modules, along with our new designs. Users can choose combinations of components to implement sophisticated settings with high performance, such as filter search, fully dynamic updates, and historical queries on snapshots. Our experiments show that our new solution provides a simple interface for various applications and achieves comparable or even better performance than previous work, specifically for each application.
[IR-5] D-NOVA: In-Storag e Retrieval Accelerator via Dual-Bound 3D NAND-Optimized Similarity Search with Vector Adaptation MICRO MICRO2026
链接: https://arxiv.org/abs/2607.17538
作者: Chang Eun Song,Sumukh Pinge,Tianqi Zhang,Sung Eun Kim,Tajana S. Rosing,Mingu Kang
类目: Hardware Architecture (cs.AR); Computation and Language (cs.CL); Databases (cs.DB); Emerging Technologies (cs.ET); Information Retrieval (cs.IR)
备注: Accepted at the 59th IEEE/ACM International Symposium on Microarchitecture (MICRO 2026), Athens, Greece. Chang Eun Song and Sumukh Pinge are co-first authors and contributed equally
Abstract:Retrieval-Augmented Generation (RAG) enhances the factual grounding of large language model (LLM) inference by retrieving relevant information from external knowledge bases. However, its dense vector retrieval introduces significant latency and energy overhead, becoming the primary performance bottleneck. Although recent in-storage accelerators aim to reduce data movement, they still rely on host or embedded processors outside the memory, where nearly 70% of the total retrieval time is spent. As a result, they cannot fully overcome the bandwidth limitations, leading to yet another memory bottleneck. To tackle these limitations, we present D-NOVA, a hardware-software co-designed in-storage retrieval accelerator. D-NOVA executes an inverted file (IVF)-based hierarchical retrieval pipeline by deeply embedding the search functionality directly into the NAND memory array. This is achieved by incorporating a new distance metric, Dual-Bound Tight Similarity Sensing (DTS), which is specifically tailored for searching within the NAND string. In addition, we introduce a lightweight contrastive adapter that maps embedding vectors into a DTS-friendly domain, recovering near-software recall while improving performance and energy efficiency. D-NOVA is up to 41.7x faster and 71x more energy-efficient than a CPU baseline, and achieves 12.13x higher throughput while being up to 1.26x more energy-efficient than state-of-the-art in-storage RAG accelerators, demonstrating the potential of fully in-storage vector search for scalable RAG acceleration.
[IR-6] RAMP: Robust Ad Recommendation Under Limited Personalized-Feature Availability via Masking and Alignment Pathways ICTIR’26
链接: https://arxiv.org/abs/2607.17473
作者: Dairui Liu,Zhongyi Lu,Roger Zhe Li,Changhong Jin,Jitao Lu,Xinyang Shao,Bichen Shi,Mete Sertkan,Aghiles Salah,Aonghus Lawlor,Barry Smyth,Tri Kurniawan Wijaya,Ruihai Dong,Xingsheng Guo
类目: Information Retrieval (cs.IR)
备注: 12 pages, 4 figures, accepted to ICTIR '26
Abstract:Click-through rate (CTR) and conversion rate (CVR) prediction are fundamental tasks in online advertising, aiming to estimate the likelihood of user interactions based on various features. While personalized attributes such as age and gender can significantly enhance predictive accuracy, their use is increasingly restricted by privacy regulations, thereby limiting available data for both training and inference. To address this challenge, we propose RAMP (Robust Ad Recommendation Under Limited Personalized-Feature Availability via Masking and Alignment Pathways), which is designed to improve CTR/CVR prediction accuracy when personalized features are not accessible, thus supporting deployment in privacy-constrained this http URL consists of (i) a personalized pathway built upon a dual-tower component with identical inputs but independent parameters, where output masking separates predictions for personalized and non-personalized signals, (ii) a separate non-personalized pathway trained with non-personalized features only, and (iii) a distillation-inspired prediction-alignment architecture between (i) and (ii) that improves prediction when personalized features are unavailable. We conduct comprehensive experiments using both public benchmarks and industrial datasets to evaluate the performance of RAMP. Our evaluation spans multiple backbone models and different settings: with and without access to personalized features. The results show that RAMP consistently outperforms state-of-the-art methods when personalized features are missing, while maintaining competitive performance when all features are available. %demonstrating its effectiveness and practicality for real-world advertising systems. Our code is publicly available at this https URL.
[IR-7] HyCoRec: Hypergraph-Enhanced Multi-Preference Learning for Alleviating Matthew Effect in Conversational Recommendation
链接: https://arxiv.org/abs/2607.17461
作者: Yongsen Zheng,Ruilin Xu,Ziliang Chen,Guohua Wang,Mingjie Qian,Jinghui Qin,Liang Lin
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注:
Abstract:The Matthew effect is a notorious issue in Recommender Systems (RSs), \emphi.e., the rich get richer and the poor get poorer, wherein popular items are overexposed while less popular ones are regularly ignored. Most methods examine Matthew effect in static or nearly-static recommendation scenarios. However, the Matthew effect will be increasingly amplified when the user interacts with the system over time. To address these issues, we propose a novel paradigm, Hypergraph-Enhanced Multi-Preference Learning for Alleviating Matthew Effect in Conversational Recommendation (HyCoRec), which aims to alleviate the Matthew effect in conversational recommendation. Concretely, HyCoRec devotes to alleviate the Matthew effect by learning multi-aspect preferences, \emphi.e., item-, entity-, word-, review-, and knowledge-aspect preferences, to effectively generate responses in the conversational task and accurately predict items in the recommendation task when the user chats with the system over time. Extensive experiments conducted on two benchmarks validate that HyCoRec achieves new state-of-the-art performance and the superior of alleviating Matthew effect. Our code is available at this https URL.
[IR-8] he Matryoshka Hypencoder SIGIR2026
链接: https://arxiv.org/abs/2607.17457
作者: Majd Alkawaas,Sean MacAvaney
类目: Information Retrieval (cs.IR)
备注: SIGIR 2026
Abstract:The Hypencoder is a recently-proposed retrieval approach that encodes queries as shallow neural networks (“Q-Nets”) that estimate relevance over pre-computed document embeddings. Inspired by Matryoshka Representation Learning, we show that the Hypencoder can be extended to support multiple sizes of Q-Nets, allowing trade-offs between effectiveness and efficiency when deployed. We find that this “Matryoshka Hypencoder” achieves comparable in-domain effectiveness with approximately 7x fewer active parameters in-domain and half as many active parameters out-of-domain, which corresponds to a 1.6-3.4x increase in scoring throughput. This work paves the way for practical deployment of Hypencoders.
[IR-9] Adapting Embedding Models for Agent Capability Retrieval SIGIR2026
链接: https://arxiv.org/abs/2607.17347
作者: Tingwei Chen,Yunxiao Shi,Zhengdong Chu,Qingsong Wen,Min Xu
类目: Information Retrieval (cs.IR)
备注: Accepted for oral presentation at the AgentSearch Workshop, SIGIR 2026
Abstract:Open agent marketplaces list native agents, tool bundles, and reusable skill packages in the same search interface, yet practitioners still have little guidance on how to retrieve across this mixed catalog. We study whether off-the-shelf retrieval models, trained for general text retrieval, can be adapted to match user queries to executable agent capabilities, and whether the learned signal transfers beyond the benchmark used for tuning. We fine-tune three open retrieval backbones, BGE-base, KaLM-v1.5, and EasyRec, on AgentSelect, which represents marketplace-visible units as capability profiles derived from public metadata, and test transfer on two catalogs not seen during training: MuleRun native agents and a ClawHub benchmark of 50 skills with 1,000 queries. Adaptation helps on both catalogs. Code and data will be released upon publication.
[IR-10] DRNOISE: Benchmarking Deep Research Agents in Misleading Evidence Environments
链接: https://arxiv.org/abs/2607.17291
作者: Jun Nie,Zhiqin Yang,Zhenheng Tang,Yonggang Zhang,Xiaowen Chu,Xinmei Tian,Bo Han
类目: Machine Learning (cs.LG); Computation and Language (cs.CL); Information Retrieval (cs.IR)
备注: 16 pages, 2 figures, 11 tables
Abstract:Deep research agents increasingly operate over the open web, where relevant records coexist with redundant summaries, outdated reports, and misleading documents. Existing evaluations offer limited insight into whether agents preserve sound evidential standards when an ordinary-looking false document is deliberately seeded into a searchable environment and offers a direct shortcut to a conflicting answer. We introduce DRNOISE, a 100-task benchmark for answer recovery under misleading evidence. Each task has a unique gold answer supported by two corroborating indirect record chains; the paired noisy condition adds one plausible document that states a conflicting answer directly. The benchmark spans ten families of evidence operations. Across agents with strong clean-task performance, this single intervention causes 66-88 percentage-point accuracy drops. Trace analyses identify verification inertia as the dominant failure mode: agents often retrieve truthful records but stop before completing and reconciling the evidence chain, instead deferring to the answer-like document. Generic verification prompts reduce but do not close this gap. The setting is especially relevant to open-web deployment, where plausible falsehoods arrive through ordinary-looking pages rather than explicit attacks. Reliable deep research therefore requires more than retrieval and citation; it requires active reconciliation of direct claims with record-level evidence.
[IR-11] Learning Sparse Representations of Multimodal Content for Enhanced Cold Item Recommendation RECSYS2026
链接: https://arxiv.org/abs/2607.17184
作者: Gregor Meehan,Johan Pauwels
类目: Information Retrieval (cs.IR)
备注: Accepted at RecSys 2026
Abstract:The scale and rapid growth of item catalogs in modern digital platforms present significant challenges to recommender system (RS) practitioners. Most RSs use embedding similarity to predict user-item preferences, but embedding storage and low-latency retrieval is challenging in industry-scale catalogs. Furthermore, newly added items do not have corresponding embeddings and cannot be recommended effectively; previous works often tackle this item cold-start problem by generating cold item representations from auxiliary content, such as images or descriptive text, so that user preferences can be predicted without historical interactions. In this paper, we argue that sparse embeddings have notable advantages over standard dense vectors in this content-based cold-start paradigm. We describe how existing cold-start training regimes can be adapted for sparse representation learning, and build on insights from linear attention to design a pre-sparsification activation technique that induces sharpness and denoising effects in learned item-item similarities. We show that the resulting sparse embeddings achieve significant improvements in cold-start recommendation accuracy over dense embeddings at considerably lower storage costs, especially for users with multiple interests. Through comprehensive experiments on four multimodal RS datasets, we also demonstrate the interpretability of sparse content embeddings and their robustness in the trade-off between size and accuracy.
[IR-12] Fenced Citation-Context Retrieval for Case Law: Temporal Leakage and Degree Control Across Two Jurisdictions
链接: https://arxiv.org/abs/2607.17142
作者: Yao Liu,Tien-Ping Tan,Zhilan Liu
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注:
Abstract:Incoming citation context-the text with which later cases describe a case when citing it-is a strong signal for legal precedent retrieval, but it is evaluated without a temporal fence: at query time the target precedent has not yet been cited by future cases, so an unfenced method credits itself with evidence unavailable at deployment. Our central contribution is to measure how much: a temporal-admission decomposition that splits the naive relax-the-fence gain into genuine future-citation leakage, legitimate pre-query admission, and an index effect, reporting the phantom fraction-the share that is future evidence. We instantiate this as a controlled deployability audit of incoming-citation retrieval across two jurisdictions, CLERC (US federal, 1.84M documents) and ECtHR-PCR (European Court of Human Rights), adding a zero-training anchor channel that admits only citers dated before the query. On ECtHR-PCR only 14.9% of the naive gain is genuine future evidence; on CLERC the over-credit grows to +4.95 R@1000 as the fence is relaxed. Under the fence, the zero-training channel gains +16.1 R@1000 over BM25 on CLERC, clears a citation-degree control, and reaches a published R@1000 estimate similar to the strongest trained system on ECtHR-PCR (79.56 vs 79.39) at zero training.
[IR-13] Uncertainty as Remedy: Mitigating Satisfaction Label Bias in Short Video Multi-Objective Ensemble Ranking
链接: https://arxiv.org/abs/2607.17092
作者: Zonghe Shao,Tiantian He,Xiaoxiao Xu,Jiaqi Yu,Minzhi Xie,Jinfang Gu,Yongqi Liu,Kaiqiao Zhan,Kun Gai
类目: Information Retrieval (cs.IR)
备注:
Abstract:The core objective of short video recommendation is to model users’ unobservable true satisfaction with recommended videos. As the dominant industrial framework, end-to-end multi-objective ensemble ranking models are typically trained with multi-dimensional dense user behavioral signals, such as clicks and watch time. However, these behavioral signals are partial, fragmented, and often mutually conflicting user satisfaction proxies, introducing uncertainty and label bias into satisfaction modeling. Conventional deterministic models overlook this uncertainty, which exacerbates satisfaction label bias and results in suboptimal model convergence. Meanwhile, existing uncertainty-aware methods mostly employ uncertainty for post-hoc ranking adjustments rather than leveraging it as a remedy to mitigate the inherent bias within the core optimization pipeline. This paper proposes UAME, an Uncertainty-Aware end-to-end Multi-objective Ensemble ranking framework for short video recommendation. UAME represents the model’s prediction as a Gaussian scoring variable, where the mean denotes the predicted satisfaction score and the variance quantifies predictive uncertainty associated with this score. We further design a probabilistic pairwise ranking loss, and construct an uncertainty-aware sample-level weighting scheme to mitigate the bias. We further provide theoretical analysis suggesting that the weighting scheme helps mitigate satisfaction label bias. Extensive offline and online experiments on a large-scale industrial short video platform demonstrate that UAME consistently improves two state-of-the-art paradigms, EMER and EASQ, and better aligns with questionnaire-based user satisfaction. UAME has been deployed in our production short-video recommendation system and continues to deliver stable, statistically significant gains.
[IR-14] WHALE: A Scalable Unified Model for Recommendation with Wukong-HSTU Architecture
链接: https://arxiv.org/abs/2607.17017
作者: Renqin Cai,Dawei Sun,Yuanjun Yao,Zhiyong Wang,Velvin Fu,Maggie Zhuang,Yu Shi,Zhongnan Fang,Xuan Cao,Jing Qian,Rui Li
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:As scalability becomes increasingly important in recommendation modeling, recent architectures have advanced the modeling of two broad sources of ranking signals along separate paths: non-sequence features, including user, item, context, and cross features; and sequence features from user behavior histories. Wukong and HSTU have emerged as representative scalable backbones for these paths: Wukong scales high-order non-sequence feature-interaction modeling, while HSTU scales long user-behavior sequence modeling. Despite their complementary strengths, practical architectures that combine these two types of feature modeling remain underexplored. We present WHALE, a scalable unified recommendation architecture that jointly models non-sequence and sequence features on top of Wukong and HSTU. Each WHALE layer contains a Wukong module, an HSTU module, and an attention-based fusion module in which Wukong-derived interaction representations query HSTU-derived behavior representations. This design keeps both backbones active throughout the network and enables progressive Wukong-HSTU exchange, allowing high-order feature crosses to repeatedly retrieve fine-grained evidence from long user histories. To make WHALE practical for industrial deployment, we introduce customized Triton kernels and other model-systems co-design techniques to improve training and inference efficiency. On large-scale industrial recommendation data, WHALE achieves consistent gains in offline experiments. Additionally, it delivers positive online gains with a modest serving-throughput trade-off. The method has been deployed in production systems. Overall, WHALE provides a practical example of how these two sources of information can be scalably unified in an industrial recommendation model.
[IR-15] urboVec: A Case Study in Cost-Efficient Private Retrieval for Enterprise RAG via Codebook-Oblivious Quantization
链接: https://arxiv.org/abs/2607.16973
作者: Navnit Shukla,Kamal Pandey,Omsankar Tiwari
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:
Abstract:Retrieval-Augmented Generation (RAG) systems increasingly power enterprise LLM applications, yet the vector retrieval layer introduces two underexplored challenges: (1) trained codebook quantizers may expose corpus statistics during index construction, creating a leakage channel in multi-tenant deployments, and (2) post-hoc filtering for tenant isolation degrades recall on selective queries. We study TurboVec, an open-source vector index built on TurboQuant - a codebook-oblivious scalar quantizer requiring no corpus-dependent training. On the DBpedia OpenAI embeddings benchmark (d=1536, 100K-999K vectors), TurboQuant 4-bit outperforms trained FAISS Product Quantization at the same memory budget by 8.5-8.9 percentage points in Recall@5 across all scales. Compared to HNSW (R@5=0.991) and IVF-PQ (R@5=0.840), TurboQuant occupies a distinct design point: higher recall than IVF-PQ without training, at 4-8x less memory than HNSW. Deployed on Snowpark Container Services, TurboVec achieves 11ms median query latency at 100K vectors versus 707ms for warehouse brute-force scan. Kernel-level allowlist filtering maintains 0.86-0.93 Recall@10 across 10-1000 tenant workloads versus 0.09-0.19 for post-filter baselines. Codebook-oblivious design reduces membership inference accuracy to near-random (50.0%) versus 57.3% for PQ codebooks. Limitations include single dataset evaluation, uncompressed HNSW comparison, and privacy evaluation on synthetic data only.
[IR-16] Adaptive Incident Prioritization for Security Operations at Scale
链接: https://arxiv.org/abs/2607.16963
作者: Scott Freitas,Amir Gharib,Maayan Magenheim
类目: Cryptography and Security (cs.CR); Information Retrieval (cs.IR)
备注:
Abstract:Large security operations centers (SOCs) often face hundreds of active incidents per day, creating substantial cognitive and operational demands for analysts. Analysts must quickly decide which incidents deserve attention within long, constantly changing queues, yet incidents are commonly ordered by arrival time, coarse severity, or product-specific heuristics that leave their relative priority unclear. We introduce Adaptive Incident Prioritization (AIP), the ranking algorithm behind Microsoft Defender Queue Assistant, which continuously prioritizes security incidents for analyst investigation. AIP adapts BM25-style ranking to a query-less, multi-tenant queue setting by representing each incident as a collection of normalized security components extracted from alerts and metadata. The model combines saturated local component frequency, global component rarity estimated across tenants, bounded domain-prior multipliers, and component-level explanations. Deployed across tens of thousands of customers, AIP performs near-real-time inference and refreshes incident scores with a median latency of five seconds. In an expert-reviewed evaluation across 1,000 customer organizations, AIP achieves 92.8% Precision@10. In post-launch telemetry across 473,000 organization-day queues, AIP increases alert-detail interaction by 5.8% and alert-detail view events by 17.5% relative to severity ordering, providing behavioral evidence that model-ranked queues concentrate analyst engagement. We also extend the Microsoft GUIDE dataset with, to our knowledge, the first public label source for SOC queue prioritization over real-world incidents. The extension covers 499 organization queues and 9,980 incidents with expert-derived priority labels, enabling the research community to develop, compare, and advance methods for incident prioritization.
[IR-17] How Do You Choose Your AI Component? An Interview Study of Secure AI Integration in Practice
链接: https://arxiv.org/abs/2607.16660
作者: Mahzabin Tamanna,Elizabeth Lin,Sparsha Gowda,Laurie Williams,Dominik Wermke
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: 18 pages, 3 figures/tables
Abstract:The increasing adoption of Large Language Models (LLMs) as AI components in modern software systems introduces distinct security risks to the software supply chain. While many considerations and safety mechanisms are in place for components of the traditional software supply chain, the recent rapid adoption of AI components and platforms has overlooked these hard learned lessons. Selecting and integrating AI models without clear guidance on how these choices affect system security may leave applications vulnerable to threats, such as malicious components, data leakage, and unintended behavior. The goal of this study is to understand practitioners’ decision making process and security considerations in selecting and integrating AI components through an exploratory semi-structured interview study. Toward this goal, we conducted semistructured interviews with 22 software developers, architects, and AI practitioners across diverse organizations about how they integrate AI components into their software. Our analysis finds that practitioners’ model selection is predominantly driven by functional criteria, including performance, accuracy, cost, and specific features, e.g., tool calling or multimodal support, while security is rarely considered as an evaluation criterion. We observe a consistent lack of security concern throughout the AI component integration process, with established software supply chain lessons overlooked or ignored. The industry is repeating the historically costly mistakes of early software dependency management, prioritizing rapid reuse and availability over security and provenance. We distill our findings into actionable recommendations for AI adopters, model providers, and researchers, advocating for a proactive, security-by-design approach that integrates security evaluation into component selection and sustains it throughout the software development lifecycle. Comments: 18 pages, 3 figures/tables Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Information Retrieval (cs.IR); Machine Learning (cs.LG) Cite as: arXiv:2607.16660 [cs.SE] (or arXiv:2607.16660v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2607.16660 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[IR-18] Beyond Fixed Depths and Widths: Optimizing Textual Decoding Tries in LLM -based Generative Recommendation
链接: https://arxiv.org/abs/2607.16633
作者: Jingzhe Liu,Hanbing Wang,Jiliang Tang,Liam Collins,Tong Zhao,Neil Shah,Mingxuan Ju
类目: Information Retrieval (cs.IR)
备注:
Abstract:Generative recommendation (GR) is an increasingly popular paradigm in recommender systems, with a prominent line of work using LLMs as autoregressive backbones to predict the next item’s term IDs (e.g., titles or keywords). The success of autoregressive generation hinges on constrained beam search over a decoding trie to ensure that generated outputs correspond to valid items. However, current research predominantly focuses on generating more comprehensive term IDs to describe items, while largely neglecting the structural design of the decoding trie formed by these terms. This can lead to a trie that is poorly suited to beam search, which degrades performance. To address this, we examine the effectiveness of term IDs from the perspective of decoding trie optimization. Through empirical and theoretical analyses, we identify two desirable properties for a highly performant trie: (1) adaptive and variable ID length, enabling items with varying semantic richness to be represented by IDs of appropriate lengths, and (2) constrained branching factors, especially at shallow levels, which drastically improves the success rate of constrained beam search. Motivated by these properties, we introduce BONSAI: Branching-Optimized Node Structure for Adaptive Identifiers, a novel framework that co-designs textual term IDs and their underlying decoding trie. BONSAI extracts recommendation-informative words from item metadata and employs a minimum set cover formulation to recursively build a trie that satisfies the above properties. Experiments reveal that BONSAI achieves up to a 21.6% relative improvement over state-of-the-art baselines. Further analyses confirm the crucial role of our proposed properties, and demonstrate their generalizability to be applied to enhance the performance of other term ID methods.
[IR-19] A Quantum-Classical Hybrid Framework for Multivariate Time-Series Forecasting Complexity-Fidelity Trade-offs and Limitations
链接: https://arxiv.org/abs/2607.16358
作者: Sanjay Chakraborty,Fredrik Heintz
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET)
备注:
Abstract:This paper presents a unified quantum-classical hybrid framework for multi-horizon time-series forecasting, introducing two model variants Quantum Reservoir Forecaster (QRC-F) and Variational Quantum Forecaster (VQF-F). The proposed framework investigates the complexity-fidelity trade-off of quantum forecasting under near-term NISQ hardware constraints. Continuous time-series signals are transformed into binary representations through uniform quantization and encoded into quantum states using angle encoding with parameterized RY rotation gates. Cross-channel entanglement layers capture dependencies among multiple variables. QRC-F utilizes a fixed random unitary quantum reservoir for stable, gradient-free temporal feature extraction, whereas VQF-F employs a trainable variational quantum circuit optimized through the parameter-shift rule to learn temporal and inter-variable patterns from Pauli expectation values. Both models replace computationally expensive quadratic self-attention with efficient linear transformations, reducing parameter complexity. A shared MIMO-based multi-horizon prediction head simultaneously generates forecasts across multiple horizons, avoiding error accumulation in recursive forecasting. Experimental evaluations on benchmark datasets, including ETTh1, ETTh2, ETTm1, ETTm2, Weather, electricity, and exchange-rate, demonstrate that VQF-F achieves superior training stability and parameter efficiency, while QRC-F provides enhanced robustness and circuit fidelity under quantum noise. The results establish a practical quantum-native forecasting framework with strong potential for deployment on near-term NISQ devices.
[IR-20] Art Beyond Semantics: Sheaf-Informed Contrastive Learning for Multi-Relational Representations ECCV2026
链接: https://arxiv.org/abs/2607.16321
作者: Ludovica Schaerf,Antonio Purificato,Piera Riccio,Fabrizio Silvestri,Noa Garcia
类目: Computer Vision and Pattern Recognition (cs.CV); Information Retrieval (cs.IR)
备注: Accepted to ECCV 2026
Abstract:Understanding a painting is never a single act. Art historians may analyze the same work through concepts of style, iconography, or historical context, dimensions that are not interchangeable, and each carries distinct semantic relationships between the visual and the textual. Vision-Language Models (VLMs) like CLIP, which learn a single shared embedding space, collapse this richness into a single homogeneous alignment, thereby losing the multi-relational structure that defines art-historical reasoning. We introduce CANVAS (Contrastive Art-aware Network for Vision-Language Alignment with Sheaves), a framework for learning relation-aware multimodal representations inspired by sheaf theory. Each artwork is projected into multiple embeddings conditioned on the type of relation (i.e., the context), and a novel contrastive loss encodes contextual information during training, with no dependency on external data at inference. We evaluate on three newly introduced benchmarks of artworks for multi-relational art understanding: WikiArt+, derived from WikiArt and Wikipedia, HertzianaDP, from the Bibliotheca Hertziana collection, and SemArt+, refined from the SemArt dataset. In multimodal retrieval and art understanding, CANVAS outperforms the baselines, supporting the view that multi-relational alignment is not just theoretically motivated but also practically essential.
[IR-21] Discovery by Dreaming: Cross-Domain Recombination in Artificial Memory
链接: https://arxiv.org/abs/2607.16256
作者: Oliver Zahn,James Evans,David Eagleman
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Neural and Evolutionary Computing (cs.NE)
备注: 38 pages, 13 figures, 7 tables
Abstract:Dreams splice together people, places, and times that never met. Neuroscience suggests this recombination is not noise, but a function driving insight and creative discovery. This reframes memory consolidation: rather than merely defending against forgetting, its measurable value lies in recombining knowledge across experiences that have not yet co-occurred. We test this directly by isolating the recombinatory-replay mechanism and implementing it in two architecturally unrelated systems: a LoRA fine-tuning pipeline (DREAMS) and a symbolic engine replaying structured knowledge objects (SAPIENCE). Both systems converge on the same finding: cross-domain consolidation creates value, while within-domain rehearsal does not. The symbolic arm surfaces novel cross-domain connections at 85.7%, a +21 percentage point (pp) gain over baseline. The neural arm improves overall by +5.64 pp, but on subtasks explicitly requiring cross-domain transfer (like unseen math reasoning on GSM8K), gains reach +14.5 pp. This effect is a genuine property of the weights–not a prompt artifact–as prepending the same material in-context to a 671B-parameter model actually reverses the gain. We validate this prediction against documented discoveries across 50,000 real papers and state a falsifiable hippocampal-recording prediction to distinguish recombination from rehearsal. Ultimately, this principle is substrate-general, tracking real discovery at scale. Reading the literature teaches a model to recall what it has seen, but producing discovery requires a separate offline phase that recombines knowledge across domains–the computational analog of dreaming. Consolidation is not for remembering, but for discovering.
人机交互
[HC-0] he Autonomous Agency Scale: A Behavioral Framework for Measuring Self-Directed Behavior in AI Systems
链接: https://arxiv.org/abs/2607.17947
作者: Samuel Presgraves
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注: 14 pages, 1 figure, 2 tables. Framework v0.2.1; rubric and all assessment files: this https URL
Abstract:Existing AI measurement frameworks quantify cognitive capability, task automation, or catastrophic risk, but none measure autonomous agency: the extent to which a system behaves in a self-directed way. A system can saturate capability benchmarks while remaining entirely reactive, acting only when prompted and ceasing all activity when a task completes. We introduce the Autonomous Agency Scale (AAS), a behavioral framework that scores AI systems on a 0-5 lexicon across seven dimensions of agency: cognitive autonomy, temporal persistence, environmental agency, social agency, creative agency, self-awareness, and goal formation, each operationalized by falsifiable threshold tests. Every dimension is scored in two temporal bands: an Active band covering engaged, user-initiated activity, and an Ambient band covering idle periods. Ambient Level 4 is gated by the Idle-Gap Test, a counterfactual criterion (remove all triggers and observe whether internally derived activity persists) that separates self-direction from scheduled rule-following. We apply the scale to six contemporary systems spanning task agents (Claude Code, Manus, Hermes), consumer assistants (ChatGPT, Siri), and a persistent companion architecture (Airi). The two-band profile quantifies a boundary that single-score frameworks conflate: task agents reach Active composites of 2.3-2.4 while scoring 0.6-1.9 Ambient, with every idle-period behavior attributable to user-configured schedules, whereas the companion architecture, evaluated longitudinally, is the only assessed system whose idle-period behavior survives trigger removal. We discuss limitations, including single-rater provenance, developer-evaluator bias on the longitudinal assessment, and the partially operationalized self-direction boundary in the Active band.
[HC-1] AlphaOracle: Oracle bone script decipherment via human-workflow-inspired deep learning
链接: https://arxiv.org/abs/2607.17849
作者: Yuliang Liu,Haisu Guan,Pengjie Wang,Xinyu Wang,Jinpeng Wan,Kaile Zhang,Handong Zheng,Xingchen Liu,Zhebin Kuang,Huanxin Yang,Bang Li,Yongge Liu,Lianwen Jin,Xiang Bai
类目: Human-Computer Interaction (cs.HC); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by The Innovation 2026
Abstract:Approximately 3,000 of the 4,500 oracle bone script (OBS) characters remain undeciphered due to fragmentary inscriptions and sparse evidence. Current AI approaches fail to replicate expert workflows that integrate form analysis, contextual semantics, and philological reasoning. We introduce AlphaOracle, a human-workflow-inspired framework that systematizes OBS decipherment using the largest digitized corpus to date. Its multi-stage pipeline comprises: (i) rubbing parsing; (ii) radical-based morphological analysis with diachronic modeling; (iii) contextual retrieval with semantic alignment; and (iv) philological validation against classical sources. Each stage generates explicit, confidence-weighted evidence chains, culminating in interpretable reports for scholarly verification. Across multiple test characters, AlphaOracle’s readings strongly agreed with expert interpretations. In a study of 86 domain specialists, it reduced analysis time by 64% and 79% of participants rated it highly useful. Notably, AlphaOracle resolves the character “Lao” as a toponymic or clan designation, offering concrete revisions to Shang administrative and social interpretations. These results suggest that computational methods aligned with philological practice can facilitate OBS research and provide a conceptual reference for studies of other undeciphered scripts.
[HC-2] I wanted it to feel more personal: Customization of social AI as AI individualism in practice
链接: https://arxiv.org/abs/2607.17826
作者: Marita Skjuve,Anna Grøndal Larsen,Asbjørn Følstad,Nena van As,Petter Bae Brandtzaeg
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 21 pages, 5 figures
Abstract:Despite the growing availability of customizable social artificial intelligence (AI), such as ChatGPT, Grok, and this http URL, we know little about how users actively shape social AI to reflect their personal preferences. This study examines why and how users (N = 169) customize social AI through the lens of the newly developed concept of AI individualism. Through reflexive thematic analysis of open-ended responses, we identified several motivations for customization, including (1) enhanced pragmatic support, (2) emotional support or companionship, (3) trust and reliability, (4) pushback, (5) a tailored degree of human likeness, (6) creativity or playfulness, and (7) having the AI function as an extension of the self. In line with the concept of AI individualism, our findings show that, for many users, customization is a co-creative process between the human and the AI that is perceived as strengthening support, autonomy, ownership, and engagement, potentially contributing to a closer and more personal relationship. Through customization users may come to view social AI as a personalized social resource that increases their sense of individualism, freedom, and control. We discuss how these perceptions may foster pseudo-autonomy, whereby customization creates an illusion of individual control over powerful social AI systems.
[HC-3] Vis2Reg: Visibility-Aware Landmark-Free Geometric 3D–2D Registration for Liver Laparoscopy
链接: https://arxiv.org/abs/2607.17810
作者: Jiaming Feng,Xukun Zhang,Shahid Farid,Sharib Ali
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 11 pages
Abstract:Accurate 3D–2D liver registration, which aligns preoperative 3D models to partial, view-dependent intraoperative surface observations, is critical for AR-guided laparoscopic surgery but remains challenging due to severe occlusion, limited visibility, and the lack of 3D ground-truth supervision. Existing landmark-free approaches perform partial-to-complete geometric alignment, yet robust self-supervision under extreme partial visibility remains difficult. We propose Vis2Reg, a visibility-aware registration framework that explicitly constrains deformation using mask-consistent visible regions. We introduce a visibility-aware self-supervision that derives a visible-domain 3D supervision signal from intraoperative masks, enabled by differentiable point rasterization and mask-guided back-projection. This formulation improves robustness under severe occlusion while maintaining fully self-supervised learning. Vis2Reg combines a robust geometric rigid initialization module with an implicit neural deformation field for stable alignment. Vis2Reg achieves a Dice score of 92.6% and a Chamfer Distance of 1.43 mm on real intraoperative datasets, with 111 ms per-frame inference time, demonstrating both accuracy and practical efficiency.
[HC-4] Persona-as-Configuration: Generative Stakeholder Reporting for Agricultural Floods
链接: https://arxiv.org/abs/2607.17774
作者: Oliver Aleksander Larsen,Tiziano Santilli,Francesco Daghero,Mahyar T. Moghaddam
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: Accepted at CASA 2026 (9th Workshop on Context-Aware, Autonomous and Smart Architectures), co-located with ECSA 2026. 16 pages, 2 figures
Abstract:Cyber-physical systems built on deterministic edge inference, such as on-vehicle flood detection for agricultural fields, produce structured decision logs that must be interpreted differently by heterogeneous stakeholders. Pairing such systems with large language models (LLMs) to generate stakeholder-specific reports introduces a tension: the generative layer is non-deterministic, while the edge plane must remain replayable and auditable. We propose an architectural pattern resting on two invariants: unidirectional consumption, in which the generative layer is a strict read-only consumer of the deterministic plane and never writes back, and persona-as-configuration, in which stakeholder adaptation is a versioned prompt-template artifact rather than runtime improvisation. We instantiate the pattern as a context-aware dashboard layer over the JSON decision logs of a previously published edge-based standing-water detection system, and analyse how the integration boundary admits standard generative-reliability mitigations as configuration- or middleware-level extension points. A structured expert review rated the pattern favourably across five ISO/IEC 25010-aligned quality dimensions, with strongest agreement on separation of concerns. End-user evaluation with agricultural stakeholders is planned for future work.
[HC-5] From Sign Language Generation to Humanoid Execution: Vision-Language Guided Retargeting with Collision Mitigation
链接: https://arxiv.org/abs/2607.17769
作者: Nabeela Khan,Bowen Wu,Runwu Shi,Benjamin Yen,Takeshi Ashizawa,Carlos Toshinori Ishi,Takashi Minato,Kazuhiro Nakadai
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC)
备注:
Abstract:Recent sign language generation (SLG) systems increasingly output dense 3D body representations, which better preserve full-body kinematics and geometry for downstream embodiment on humanoid robots. However, these generated motions frequently exhibit self-intersections such as hand-hand and hand-torso penetration. While such artifacts may be tolerated in offline rendering, they become critical in humanoid execution as they lead to infeasible inverse-kinematics (IK) solutions, collisions, and unstable retargeted trajectories. We present a system-level framework that bridges SLG outputs to humanoid joint-space execution via two components. First, we introduce a volumetric SMPL-X collision-mitigation module that projects generated signing motions toward physically plausible configurations while minimally deviating from the original trajectory. Second, we propose a vision-language-guided retargeting algorithm built on an IK backbone: a VLM serves as a visual critic over rendered humanoid motion, identifies embodiment-specific failure modes, and triggers targeted task-space corrections. Our results highlight collision handling and perception-guided refinement as key missing components for reliable humanoid signing.
[HC-6] oward Site-Aware MR Art Exhibitions: A SLAM-Based Deployment Pipeline for Spatial Coherence and Exhibition Experience
链接: https://arxiv.org/abs/2607.17665
作者: Yawei Zhao,Yuming Zhu,Hao Li,Yuqi Liang,Ao Yu,Anca-Simona Horvath,Pan Hui
类目: Multimedia (cs.MM); Human-Computer Interaction (cs.HC)
备注: Proceedings of the 34th ACM International Conference on Multimedia
Abstract:Mixed Reality (MR) is increasingly being used in exhibition settings to bring digital artworks into relation with the physical environment. However, existing MR exhibition systems are often confined to prototypes or case-specific deployments, offering limited guidance for large-scale practical implementation. To address this gap, this paper presents a practical pipeline for designing and deploying large-scale MR art exhibitions, treating spatial alignment not only as a technical mechanism but also as an experiential design decision. We first conducted a pilot study comparing marker-based and Simultaneous Localization and Mapping (SLAM)-based alignment methods in an MR exhibition setting. Based on the results, we developed a SLAM-based pipeline for MR exhibitions that integrates technical deployment with exhibition curation. We then evaluated the pipeline through both system overhead measures and users’ experiential feedback. The results show that spatial alignment influences not only technical stability, but also overall exhibition coherence, visitors’ sense of continuity and immersion, and artwork interpretation. These findings provide an empirically grounded reference for future large-scale MR art exhibition deployment.
[HC-7] Informal Learning Emerges in Everyday Human-LLM Interaction
链接: https://arxiv.org/abs/2607.17643
作者: Zixin Chen,Haotian Li,Ziang Xiao,Huamin Qu,Xing Xie
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注: Under Review
Abstract:As LLMs become increasingly capable of completing tasks for users, a central concern is that everyday AI use may become primarily cognitive offloading, eroding the opportunities through which people develop their own capabilities. We analyse large-scale human–LLM conversations to ask whether informal learning behaviors also emerge in this setting: whether users engage in exchanges in ways that preserve opportunities to learn. Across 128,569 naturalistic conversations, we translated learning-science constructs into turn-level behavioural signatures. Cognitive engagement, users’ cognitive effort as reflected in the exchange, appeared in 31.9% of 491,685 user turns, whereas constructive engagement, the deepest observable form of learning-oriented engagement, appeared in 4.9%, showing that deeper sense-making was recurrent but selective. Our study further identifies factors associated with these forms of engagement. Scaffolded assistant support consistently marked richer constructive participation, with associations varying by user framing, task ecology, support form, timing and prior user state. Together, these findings show that everyday human–LLM interaction is not only answer delivery or cognitive offloading; it also contains measurable, selective and conditionally organized behavioural signatures of informal learning. They shift AI evaluation from answer-delivery efficiency toward the preservation of cognitive opportunities for users to reason, test ideas and construct understanding in the course of everyday problem-solving.
[HC-8] It Matters How You Say It: Exploring Rhetorical Patterns for AI-Assisted Information Evaluation
链接: https://arxiv.org/abs/2607.17627
作者: Sadra Sabouri,Zeinabsadat Saghi,Jordan Lee Boyd-Graber,Jonathan May,Jonathan K. Kummerfeld,Souti Chattopadhyay
类目: Human-Computer Interaction (cs.HC); Computation and Language (cs.CL)
备注: 13 pages, 6 figures, 1 table
Abstract:Prior work on AI-assisted information evaluation has largely focused on what AI systems communicate, comparing explanation types and formats, with responses predominantly cast in directive rhetoric where the system delivers a verdict and the user passively accepts it. While debate-style interactions have recently shown promise in prompting critical evaluation over deference, the rhetorical patterns that structure AI responses and how they might induce reflection, uncertainty, or independent reasoning remain largely unexamined. To address this, we investigated eight rhetorical patterns known to induce contemplation: Intentional Misleading, Interpretive Alternative, Scaffold Explanation, Triggering Distrust, Information Distortion, Alternative Framing, Socratic Questioning, and an Oracle baseline. Through a within-subject study with n=98 participants on a hint-on-demand fact verification task, we observed preliminary evidence that Scaffold Explanation were associated with the highest accuracy gains, and encouraging deeper reflection. Surprisingly, the adversarial conditions also improved accuracy modestly. Participants preferred Alternative Framing most and Interpretive Alternative least, largely due to the latter’s perceived time cost. We discuss the implications of designing conversational agents with varied rhetorical styles and the trade-offs among user performance, satisfaction, and contemplation.
[HC-9] Human-in-the-Loop User Feedback Affects Perceived Accuracy and Trust but Task Subjectivity Matters
链接: https://arxiv.org/abs/2607.17548
作者: Donald R. Honeycutt,Mahsan Nourani,Eric D. Ragan
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注:
Abstract:While ML can produce complex models beyond those that a human could produce manually, incorporating human input can often improve performance beyond purely data-driven models. While this feedback could come from system designers or domain experts, in many cases, the end users who regularly use the system will naturally develop an understanding of its flaws and desire the ability to change the system’s behavior based on their knowledge. While soliciting feedback from end users can result in significant model improvement over time, introducing these feedback techniques can also affect several human factors-such as trust or perception of system accuracy-that are not yet fully understood and have different effects reported in the existing literature. Therefore, we sought to build on the existing research to further explore how the act of providing feedback can affect user understanding of an intelligent system and its accuracy in different contexts. We present three controlled experiments that study the effects of interactive feedback collections on user impressions in domains with objective and subjective feedback. The results show that in a context where there is an objectively correct answer, providing HITL feedback lowered both participants’ trust in the system and their perception of system accuracy, regardless of whether the system accuracy improved in response to their feedback. However, when the feedback being provided involved subjective opinion, no such negative bias was observed. Furthermore, in the objective context, participants distrusted the system over time, whereas participants in the subjective context mistrusted the system over time. These results highlight the importance of considering the effects of allowing different types of end-user feedback on user trust when designing intelligent systems.
[HC-10] Sidekick: Designing Communication for Effective Multitasking with Computer Use Agents
链接: https://arxiv.org/abs/2607.17527
作者: Ruei-Che Chang,Wenqian Xu,Dingzeyu Li,Bryan Wang,Anhong Guo
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:Computer Use Agents (CUAs) can autonomously execute complex, multi-step tasks within GUIs, enhancing efficiency through parallel multitasking. However, our formative studies with CUA experts and GenAI users indicated that current feedback is primarily text-based, requiring sustained attention to monitor progress and offering limited visibility to trace past GUI interactions. Based on the findings, we developed a prototype system, Sidekick, for communicating CUAs’ status with multimodal feedback across different stages of interaction: (i) When CUAs run in the background, Sidekick signals its execution state through ambient cues. (ii) Upon resuming interaction with CUAs, Sidekick provides multimodal summaries of completed actions to support rapid context resumption. (iii) When CUAs operate in the foreground, Sidekick enhances transparency by verbalizing and visualizing the agent’s reasoning. A study with 30 participants demonstrated that Sidekick significantly improved multitasking performance with CUAs compared to baseline systems that presented textual feedback either in a typical chat or in an ambient display. Sidekick supported progress awareness, and error and action traceability more effectively. Finally, we demonstrate the promise of Sidekick through several example applications, and discuss implications for long-horizon human-agent collaboration.
[HC-11] askArtisan: Designing Composable Generative Widgets for LLM -Assisted Analysis
链接: https://arxiv.org/abs/2607.17394
作者: Meng Chen,Amy Pavel
类目: Human-Computer Interaction (cs.HC)
备注:
Abstract:People increasingly use chatbots such as ChatGPT for everyday analysis tasks. While chatbots unify many analysis functions (e.g., scripts, visualizations, summaries), long conversations become hard to navigate, making it difficult to revisit prior steps or reuse successful workflows. LLMs now generate high-fidelity GUI code that enables people to create customized analysis tools beyond text. Yet, what new opportunities generative UIs bring to analysis work remain unclear. We interviewed six professionals about analysis with chatbots, analyzed publicly shared LLM-generated GUI tools, and conducted a comparison study (N=12) between a chatbot and TaskArtisan, a technology probe that enables people to create and assemble generative analysis UI widgets for sequential and fan-out composition. We find that GUI improved clarity and visual presentation but also introduced rigidity and additional prompting challenges. We summarize the trade-offs into a provisional design framework (malleability, specification, interoperability) to inform future generative UI in LLM-assisted analysis workflows.
[HC-12] EII-SCL: Harnessing Emotional Inertia for Multimodal Emotion Recognition in Conversation INTERSPEECH2026
链接: https://arxiv.org/abs/2607.17366
作者: Zilong Huang,Kong Aik Lee,Chong-Xin Gan,Zezhong Jin,Ruichen Zuo,Man-Wai Mak
类目: Multimedia (cs.MM); Computation and Language (cs.CL); Human-Computer Interaction (cs.HC); Signal Processing (eess.SP)
备注: Accept by Interspeech 2026
Abstract:Multimodal emotion recognition in conversation (MERC) achieves accurate predictions by integrating multimodal and contextual information in dialogues. While current MERC approaches focus on modeling complex contextual dependencies in conversation, they often overlook the impact of contextual emotional inertia in emotion shift, leading to sub-optimal performance. To address this issue, we propose a novel Emotional Inertia-Informed Supervised Contrastive Learning module (EII-SCL) that informs the contrastive objective by constructing inertia-affected samples within temporal windows, effectively leveraging emotional inertia as a prior while enabling seamless integration with existing MERC models without requiring additional data. Extensive experiments on IEMOCAP and MELD show that our approach consistently outperforms state-of-the-art methods.
[HC-13] From Perception to Assistance: Open-Vocabulary Shared Autonomy for Robotic Manipulation
链接: https://arxiv.org/abs/2607.17323
作者: Murilo Vinicius da Silva,Ricardo V. Godoy,Juliano Negri,Gustavo J. G. Lahr,Ranulfo Bezerra,Marcelo Becker
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC); Systems and Control (eess.SY)
备注:
Abstract:Teleoperating a robotic manipulator in industrial environments demands precision that camera-based interfaces alone struggle to deliver. The operator must align the end-effector with a target in clutter, under limited depth perception, and without colliding with the surrounding structures. This paper presents a shared-autonomy framework that assists the operator throughout this process. A single RGB-D camera captures the operator’s arm motion and hand gestures without wearables, fiducials, or a calibration stage. The intended target is specified by a free-form text prompt, grounded by a vision-language model in the robot’s gripper camera, and tracked across its onboard cameras by a promptable video-segmentation model, resulting in a grasp frame continuously separated from the obstacle map. Every commanded motion is executed by a GPU-accelerated model-predictive controller that enforces self- and environment-collision avoidance against an online volumetric reconstruction, while a potential field corrects the operator’s reference toward the grounded target during the final approach. An autonomous mode can be gesture-triggered to complete the grasp on the same target without a separate perception pipeline. The framework is validated on a quadruped mobile manipulator. The interface achieves a positional RMSE of 59 mm relative to motion-capture ground truth, and the controller keeps the arm at least 18 cm from obstacles while the operator deliberately commands the arm into them by 6 cm. In an industrial valve manipulation and a pick-and-place task, the full framework succeeded in all trials, while ablating either the collision or the assistance module produced failures through complementary mechanisms, and autonomous execution succeeded in four of five trials per task.
[HC-14] Semantic Context Matters: Analysis of Color Names Across Domains
链接: https://arxiv.org/abs/2607.17221
作者: Adilet Yerkin,Elnara Kadyrgali,Malika Ziyada,Nuray Toganas,Muragul Muratbekova,Ayan Igali,Aruzhan Sabitkyzy,Pakizar Shamoi
类目: Computer Vision and Pattern Recognition (cs.CV); Human-Computer Interaction (cs.HC); Multimedia (cs.MM)
备注: Submitted to 2026 Joint 14th International Conference on Soft Computing and Intelligent Systems and 27th International Symposium on Advanced Intelligent Systems (SCISISIS 2026)
Abstract:Color naming is influenced not only by physical color values but also by the semantic context in which colors are used. This paper investigates context-dependent color naming by mapping color-name datasets from Cosmetics, Crayola, and Car-color vocabularies onto the 86 fuzzy color categories of the COLIBRI color model. Contextual variation is analyzed using category coverage, Shannon entropy, and maximum lift. The results show that the three contexts occupy the COLIBRI color space differently: Cosmetics covers 48 of 86 fuzzy categories, Crayola covers 50, and Car colors cover 40. The results demonstrated that Crayola provides the broadest and most balanced use of the fuzzy color space, Cosmetics is mainly concentrated around warm-tone regions, and Car colors are more specialized around blue and achromatic regions. These findings show that color naming cannot be fully explained by numerical color similarity alone and that semantic context plays an important role in human color interpretation. The proposed framework supports the development of context-aware color models for design analytics, product search, recommendation systems, and human-centered artificial intelligence.
[HC-15] PocketPPD: Screening for Postpartum Depression Risk Using Passive Smartphone Sensing
链接: https://arxiv.org/abs/2607.17185
作者: Jia Tang,Helinyi Peng,Akihito Taya,Kaoru Sezaki,Daisuke Nishi,Anind K Dey,Yuuki Nishiyama
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注:
Abstract:Postpartum depression (PPD) is a serious perinatal mental health condition affecting approximately 20% of new mothers worldwide. Common screening approaches for PPD, such as self-report questionnaires and active digital logs, rely heavily on user input and thus impose a substantial burden on participants, limiting their feasibility for long-term use. Recent passive mobile sensing (PMS) approaches have enabled low-burden detection of depressive symptoms using machine learning methods with multi-modal sensor data from off-the-shelf mobile devices including smartphones. However, the postpartum period entails distinct behavioral patterns, raising uncertainty about whether sensing-based indicators for general depression and mental disorders generalize to PPD. To address this gap, we propose PocketPPD, a PMS-based PPD screening method that detects PPD risk using maternal contextual features, such as disruptions in behavioral rhythms and shifts in stability, collected through a smartphone. In our exploratory four-week feasibility study with 61 postpartum women, the PMS-only model achieved an AUC of 0.75, while the best-performing model, integrating PMS-oriented data and self-report features, achieved an AUC of 0.83. Moreover, we find that morning and late-night routine volatility ranks among the top digital biomarkers, dynamically moderated by maternal contexts such as infant developmental stage and employment status. This work provides empirical evidence for low-burden PPD risk screening and our findings lay the groundwork for continuous perinatal mental health monitoring.
[HC-16] Strategic Gaze: Attention Allocation and Transition Patterns Across Functional Areas of Interest by Gameplay Outcome
链接: https://arxiv.org/abs/2607.17151
作者: Yufei Cao,Penny Sweetser
类目: Human-Computer Interaction (cs.HC)
备注: Accepted for publication in the Proceedings of the ACM on Human-Computer Interaction (PACM HCI), to be presented at CHI PLAY 2026
Abstract:Video games present players with complex, spatially distributed information across interface elements, with attention shaped by visual features and task goals. Eye tracking provides a useful method for examining player attention through gaze behaviour during gameplay. Yet empirical game research has relied on accumulated fixation measures that capture where attention is directed and how long it is maintained within regions, leaving less known about how gaze moves between regions to coordinate distributed information. We address this gap by integrating distribution-, duration-, and transition-based gaze measures across functionally organised interface regions in relation to gameplay outcomes. We conducted a within-subject study with 32 participants using a deck-building game, defining six functional Areas of Interest (AOIs) within the turn-based combat interface, spanning enemy, player, action, and auxiliary elements. We computed AOI hit, dwell time, transition probabilities, and entropy to compare gaze behaviour across outcome groups. Players in the win group showed more selective attention allocation extending to peripheral auxiliary resources, with more frequent action-oriented gaze transitions, a wider range of AOI pairs, and a more even distribution across AOIs. Within a strategic game setting, our findings show that gameplay outcomes are reflected in how players distribute, sustain, and shift visual attention across functional interface elements, offering a more comprehensive account of attention organisation in gameplay.
[HC-17] ach it to stop not just to click
链接: https://arxiv.org/abs/2607.17136
作者: Barada Sahu(Cabal AI),Shivesh Pandey(Para AI)
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG)
备注: 15 pages, 3 figures. Reliability protocol and library (cua_reliability), completion verifier, and leakage-free held-out plus bootstrap evaluation harness are open-source
Abstract:Agentic computer-use RL is reported in single runs, and those numbers mislead. Using verifier-guided repair of a 35B computer-use agent (CUA) across five oracle-graded environments, we show a repaired policy’s success rate is dominated by upstream variance: a variance-components decomposition across three cells (crossed data-draw \times seed grid, bootstrap CIs) finds evaluation variance negligible ( \sigma_\mathrmeval \approx 0 ) and the training-seed effect small everywhere ( \leq 10% ); instead it splits between the data draw and run-to-run nondeterminism, the data draw’s share rising to dominant ( 48% ) on the hardest cell. There the run-to-run distribution is bimodal (Hartigan dip p=0.07 , k=10 ), so a single run has roughly a 30% chance of the failure mode and mean \pm std is the wrong summary. On that footing, two findings hold. First, repairability is two-tier in how constrained the corrective action is: a single fixed token installs reliably (done-detection 0.97\pm0.06 ), while open-ended corrections are only partial – spatial-coordinate clicks (grounding 0.53\pm0.35 ) and a generative field-fill ( 0.14\pm0.04 ). Second, the frame-level repair transfers to task success only when the corrective action is the task’s sole remaining blocker (LinkedIn 8/20 vs. base 0/15, Fisher p=0.006 ). We caught two of our own over-claims – a sample-efficiency curve and a ‘grounding cannot be bought’ boundary – only by replicating across seeds; a stress test makes the stakes external: a single-run improvement of the size this field publishes would have the wrong sign roughly one-third of the time in a comparable regime. We release a library (cua_reliability) for routine k-seed reporting. The apparatus is, to our knowledge, the first multimodal segment-aggregated on-policy self-distillation (SA-OPSD) update on a real 35B CUA policy.
[HC-18] SAVEstate: A Method for Documenting Player Reflection in Digital Games
链接: https://arxiv.org/abs/2607.17128
作者: Nisha Devasia,Michele Newman,Safinah Ali,Julie A. Kientz,Jin Ha Lee
类目: Human-Computer Interaction (cs.HC)
备注: Accepted for publication at CHI PLAY 2026
Abstract:In recent years, interest in eudaimonic player experiences (PX) - concerning reflection, meaning-making, and personal growth - has increased. However, most games user research methods are not well-suited to study eudaimonic PX, as they have been developed to evaluate features of hedonic PX, such as flow, immersion, and playability. To more deeply explore eudaimonic PX, we require methods that can 1) investigate how moment-to-moment PX shapes player reflection and 2) explore how players reengage and reinterpret their experiences longitudinally. In this paper, we present SAVEstate, a method that uses documentation as a means of studying player reflection. SAVEstate consists of two phases: reflection-in-action and reflection-on-action, which allow researchers to probe players’ in-situ reflections and how they reengage with their gameplay, respectively. Using SAVEstate, we were able to observe in-situ meaning-making and connect it to post-game reflection-on-action, and view synchronous sensemaking across multiple participants. We also developed an open-source desktop application for researchers to adapt in their own SAVEstate deployments. We discuss implications for how researchers might use SAVEstate to conduct future research in meaningful PX.
[HC-19] Can LLM Code Explanations Adapt to Diverse Problem-Solvers Needs? AAAI
链接: https://arxiv.org/abs/2607.17022
作者: Andrew Anderson,David Piorkowski,Justin Weisz,Margaret Burnett,Kush Varshney
类目: oftware Engineering (cs.SE); Human-Computer Interaction (cs.HC)
备注: To be published in AAAI/ACM conference on AI, Ethics, and Society (AIES 2026)
Abstract:Large language model (LLM) code explanations can support people in solving code-related problems, yet prior work has shown that people have diverse problem-solving styles. If explanations fail to meet people’s problem-solving needs, they may be less productive in their occupations and miss opportunities to learn and grow. Although some research has examined how LLMs can adapt their outputs to a user’s age or expertise, no prior work has examined how LLMs can adapt their code explanations to people’s problem-solving styles. To address this gap, we developed prompts from an established inclusive design method that considers 5 types of problem-solving styles, and we generated 1,072 code explanations from six open-weight LLMs. Using natural language processing techniques, we uncovered a taxonomy of 13 linguistic adaptations, with each adaptation supported by evidence from the literature, the prompts, or the LLMs’ outputs. They also show which LLMs adapted their code explanations more frequently than others. This paper is the first to investigate problem-solving style adaptations in LLM code explanation, contributing two problem-solving adaptation approaches: declarative statements for each adaptation and 10 problem-solving style prompts.
[HC-20] Real-World Evaluation of an AI Agent Drafting Translational Impact Summaries
链接: https://arxiv.org/abs/2607.16989
作者: Mohammad Arvan,Amber E. Osterholt,Bailee Rue,Yuvaneswaren Ramakrishnan Sureshbabu,Krishna Riteshkumar Patel,Rebecca T. Feinstein,Bethany C. Bray,Niranjan S. Karnik
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Digital Libraries (cs.DL); Human-Computer Interaction (cs.HC)
备注: 15 pages, 5 figures, 2 tables. Submitted to the Journal of Clinical and Translational Science. Code: this https URL
Abstract:Introduction. Clinical and Translational Science Award (CTSA) programs must document their scholars’ research impact, but assembling each scholar’s record by hand takes staff an estimated 15 hours and does not scale to a full cohort. An artificial intelligence (AI) agent could serve as a tool to gather scholar data across platforms and disciplines. Methods. We built a human-in-the-loop AI agent that assembles a dossier of sourced evidence for each scholar and drafts one-sentence Translational Science Benefits Model (TSBM) impact summaries for staff review. We evaluated it in the impact-reporting workflow of one CTSA hub across 10 career-development (KL2/K12) scholars. Two evaluation staff independently coded all 507 findings as accept, edit, or reject; the primary measure was the unanimous usable rate, defined as the share both accepted or edited. Results. Both reviewers accepted or edited 81.7% of the agent’s findings. Reviewers each spent a median of 14 minutes per scholar, replacing an estimated 15 hours of manual assembly. Inter-rater agreement was moderate (Cohen’s kappa 0.43 on the usable-versus-reject decision). A profile discovery study found the agent’s recall close to human search. The agent’s impact evidence spanned all four TSBM domains, and about a third of the reviewed findings fell in non-scholarly categories that routine processes tend to miss. Reviewers rated synthesis accuracy 4.5 and usefulness 4.8 on a 5-point scale. Conclusions. A human-in-the-loop AI agent can serve as the first-pass author of a scholar’s impact record, shifting staff from collecting and writing to reviewing, and making cohort-scale impact reporting feasible. Comments: 15 pages, 5 figures, 2 tables. Submitted to the Journal of Clinical and Translational Science. Code: this https URL Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Digital Libraries (cs.DL); Human-Computer Interaction (cs.HC) ACMclasses: I.2.7; I.2.11; H.3.7; H.5.2 Cite as: arXiv:2607.16989 [cs.CL] (or arXiv:2607.16989v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2607.16989 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Mohammad Arvan [view email] [v1] Sat, 18 Jul 2026 22:34:52 UTC (205 KB)
[HC-21] Retrofitting Existing 3D Objects with Surface-Conforming Capacitive Sensing SIGGRAPH
链接: https://arxiv.org/abs/2607.16739
作者: Andela Ilic,Junpeng Gao,Zhipeng Li,Yijing Jiang,Rachel Schuchert,Manuel Meier,Philipp Herholz,Christian Holz
类目: Human-Computer Interaction (cs.HC)
备注: Accepted to ACM SIGGRAPH Conference Papers 2026
Abstract:Augmenting the surface of 3D objects with capacitive sensing is challenging when their volumes cannot be modified. In this paper, we present a generative computational fabrication pipeline that retrofits surface-only sensor layouts to 3D geometries for multi-touch interaction. Our method scans a real-world object to obtain its 3D mesh, generates and optimizes a 3D sensor design of drive and sense lines for mutual-capacitance sensing under physical and hardware constraints, and unfolds the design into individual 2D stencils that can be cut from conductive material. Our fabrication pipeline cuts these stencils from thin copper foil with a vinyl cutter and then assists manual sensor attachment by projecting the sensor design onto the dynamically registered real-world object. We connect the resulting electrode mesh to a mutual-capacitance scanning controller and resolve touch interaction in real time. We demonstrate our approach with four 3D geometries and evaluate our method and fabrication pipeline on them.
[HC-22] AlterAtlas: Shifting Travel Planning from AI Generation to Validation via Persona-Driven Simulations ALT
链接: https://arxiv.org/abs/2607.16565
作者: William Huang,Ruofei Du,Yang Zhang
类目: Human-Computer Interaction (cs.HC)
备注: 11 pages, 8 figures. For associated codebase, see this https URL
Abstract:Travel planning requires balancing interacting goals and constraints across time and space. Current AI travel tools provide limited support for encoding these constraints and understanding how generated travel plans may fail users. We present AlterAtlas, an interactive travel planning system that supports high-fidelity itinerary validation and revision through persona-based simulations grounded in geospatial information. AlterAtlas models travelers as editable personas, generates candidate itineraries from prioritized places of interest, and simulates how different personas would experience each plan. Simulations expose route-level tradeoffs, temporal user states (e.g., fatigue, hunger), and mismatches between plans and user preferences to allow users to iteratively refine both itineraries and user personas. An expert evaluation of 51 paired itineraries demonstrates that simulation-guided revisions significantly improve plan-persona alignment. Furthermore, a within-subjects study (N=11) reveals that AlterAtlas empowers users to uncover hidden constraints, fluidly compare alternatives, and build trust in their final plans. Our results suggest that simulation-based validation is a powerful, transparent interaction layer for AI-assisted travel planning.
[HC-23] Hindsight: Similarity-Based Analytics for Mars Rover Drive Retrieval
链接: https://arxiv.org/abs/2607.16537
作者: Luke Fiorante,Leslie Liu,Adam Xu,Xianmei Lei,Darwin Chiu,Krys Blackwood,Maggie Hendrie,Scott Davidoff,Santiago V. Lombeyda,Hillary Mushkin
类目: Human-Computer Interaction (cs.HC)
备注: In press for the 2026 IEEE Conference on Visualization (VIS 2026)
Abstract:While Mars rover operators plan drives across hazardous Martian terrain and diagnose unexpected faults, the necessary information is distributed across separate systems and often reconstructed through manual correlation and memory. To address this challenge, we partnered with Mars rover operators at the NASA Jet Propulsion Laboratory to introduce Hindsight, a visual analytics system that unifies previously disparate rover drive data into a single workspace for search, comparison, and investigation. This paper presents a design study of the Hindsight application. The partnership revealed that operators reason about drives as holistic spatiotemporal episodes rather than discrete parameters. By externalizing operator intuition into an explicit visual query process, we argue that Hindsight transforms analysis into a structured, shareable workflow. Preliminary feedback from operators suggests Hindsight supports their ability to correlate terrain, telemetry, and fault events within a singleworkspace.
[HC-24] How Formerly Incarcerated People Envision Technologies for Prison Parole
链接: https://arxiv.org/abs/2607.16513
作者: Saiph Savage,Jesse Nava,Wanqing Iris Zhou,Hwijoon Lee
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 25 pages, 1 figure. To appear in Proceedings of the 2026 ACM Conference on Fairness, Accountability, and Transparency (FAccT '26)
Abstract:AI-driven algorithms and automated tools are increasingly embedded in the correctional landscape, shaping parole eligibility,release decisions, and surveillance. These tools are also often framed as objective, inevitable solutions to inefficiency andbias. Yet, these computational systems are rarely designed with input from justice-impacted individuals, which means theymight fail to address the real needs of incarcerated people. To address this gap, we surveyed 31 formerly incarcerated peopleabout their parole experiences and their visions for technologies that could support parole preparation. Contrary to dominantassumptions, participants did not imagine computational tools as instruments to dismantle the prison system, but as resourcesfor navigating power: translating complex parole concepts into culturally familiar terms, documenting personal transformationin board-legible ways, and recognizing the often-invisible labor of families. We argue that these imaginaries point towardtechnologies designed for strategic agency, where tools help incarcerated individuals and their families build the capacity tonavigate existing power structures in pursuit of freedom. We conclude by reframing computational tools for parole awayfrom surveillance and toward human-centered systems that support people in navigating carceral power structures.
[HC-25] Back to the museum: Investigation of the acceptance of Android Andrea with and without emotion simulation in a museum
链接: https://arxiv.org/abs/2607.16428
作者: Marcel Heisler,Christian Becker-Asano
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC); Machine Learning (cs.LG)
备注: Presented at the ACM Intelligent Virtual Agents conference in 2026 (IVA2026)
Abstract:For a second time, the android robot Andrea was set up at a public museum in Germany for six consecutive days to have conversations with visitors, fully autonomously. Building on previously gathered qualitative results, the robot was now capable of engaging in multi-lingual conversation with the visitors about the museum context. The robot was prepared with context information about the museum in general and its surrounding exhibits this time. The robot featured a slightly artificial sounding voice that was previously evaluated as congruent with its gender-ambiguous but very humanlike design. Three experimental conditions were implemented, in which either (1) the robot simulated no emotions, (2) the robots emotions were determined by ChatGPT 4.1, or (3) the WASABI emotion simulation architecture simulated the robot’s emotion dynamics. An extended version of the TAM2 questionnaire was employed to let 73 visitors report on several factors of their opinion about the android robot Andrea after having experienced it. In result, the statistical analysis suggests that these first two approaches to implementing emotions into the chat architecture of our android robot Andrea did not yield any positive effects on the subjective evaluations by the visitors and were not detectable on a conscious level.
[HC-26] he World According to a Social Robot – Augmenting Human-Robot Dialogue With Vision Language Models
链接: https://arxiv.org/abs/2607.16318
作者: Thomas Sievers
类目: Human-Computer Interaction (cs.HC); Robotics (cs.RO)
备注:
Abstract:Vision Language Models (VLMs) enable robots to visually perceive their environment as well as the actions and characteristics of their conversation partner or humans in collaboration. Especially for social robots deployed in everyday settings and for uncomplicated, natural use, it is essential that the robot has an understanding of situations that is appropriate to human customs. This paper presents initial experiences with the application of a Mistral AI language model with a Pepper robot for Human-Robot Interaction (HRI) in dialogue, as well as an investigation of the effects of additional visual information on response time in different models. The results show that incorporating visual information adds context to the dialogue with only a moderate increase in response time, enabling both the robot and the human to take into account unspoken elements of the situation. Furthermore, using an LLM hosted in Europe offers a solution that complies with European data protection regulations and can therefore facilitate real-life applications more easily.
[HC-27] Algorithmic Accuracy as a Motivational Driver in Robot-Mediated Learning: A Comparative Study of Cross-Correlation and CNN-Based Sound Detection in an Interactive Quiz Game
链接: https://arxiv.org/abs/2607.16299
作者: Rezaul Tutul,Ilona Buchem,Niels Pinkwart
类目: Human-Computer Interaction (cs.HC); Robotics (cs.RO)
备注:
Abstract:In competitive learning activities, inaccurate robot decisions may reduce students’ perceptions of fairness and competence, ultimately affecting their motivation. This paper investigates whether the accuracy of sound detection algorithms influences student motivation during a robot-mediated quiz game. A Pepper humanoid robot hosted an interactive buzzer-based quiz in which two sound detection approaches, a Convolutional Neural Network (CNN) and a Cross-Correlation algorithm, were evaluated using a controlled between-subjects experiment involving 40 university students. Participants were equally assigned to a CNN group (n = 20) and a Cross-Correlation group (n = 20). Both groups completed the same quiz under identical conditions, differing only in the sound detection algorithm used for first-responder identification. Student motivation was assessed using the Intrinsic Motivation Inventory (IMI), while algorithm performance was evaluated through real-time detection accuracy. The results indicate that the Cross-Correlation approach achieved more reliable sound detection under classroom conditions and produced significantly higher scores across all IMI subscales, demonstrating greater student interest, perceived competence, effort, perceived choice, and lower perceived pressure (after reverse coding). These findings provide empirical support for the proposed Algorithmic Precision-Motivation Relationship (APMR) model, demonstrating that algorithmic accuracy is not merely an engineering performance metric but an important factor influencing learner motivation in robot-assisted educational environments.
[HC-28] From Weights to Words: Expressing and Editing Preference Model Inferences in Natural Language
链接: https://arxiv.org/abs/2607.16232
作者: Zachary Wojtowicz,Ayush Nayak,Jacob Andreas
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
备注: 42 pages, 22 figures, 14 tables; main text 11 pages, remainder appendices
Abstract:The growing use of statistical learning algorithms to infer human preferences from high-dimensional choice data runs up against a fundamental challenge: choice alternatives typically differ in many ways simultaneously, so it is generally unclear which factors actually drove an observed decision and should be credited as preferences. Compounding this problem, the opacity of these methods leaves human operators unable to inspect, contest, or correct models when they err. We introduce \emphweights to words, a method that takes a dataset of choice problems as input and automatically discovers a collection of domain-relevant preference dimensions, each described in natural language and paired with a vector in the model’s representational space. These dimensions address both under-determination and opacity: they can be applied to concentrate attribution on a small set of meaningful factors, and they can externalize the model’s inferences in natural language so that users can inspect and edit them in real time. We first qualitatively illustrate the method’s versatility on four diverse domains: moral dilemmas, movies, wines, and free-form LLM responses. We then report two pre-registered human-subjects experiments, on moral dilemmas ( N=450 ) and movie selection ( N=449 ), that demonstrate its benefits for learning preference models: (1) regularizing a preference model toward the learned basis increases prediction accuracy on held-out choices, and (2) incorporating participants’ structured edits further improves accuracy. In head-to-head comparisons, participants prefer the method’s inferred preference profiles and endorse its predictions as more accurate.
[HC-29] From Novelty to Normalisation: Tracking Changing Perceptions of AI in Higher Education 2024-2026
链接: https://arxiv.org/abs/2607.16223
作者: Juliana Gerard,Morgan Macleod,Kelly Norwood,Aisling Reid
类目: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注:
Abstract:The rapid integration of generative artificial intelligence (AI) has reshaped the landscape of higher education. Students have embraced tools such as ChatGPT with striking speed, while teaching staff and institutions have responded with greater caution. Existing research on AI perceptions has mainly been cross-sectional, providing single-point snapshots that view attitudes as stable rather than evolving. This paper presents a longitudinal study of AI perceptions in higher education, tracking undergraduates, doctoral researchers, teaching staff and non-teaching staff at Ulster University across three survey waves between 2024 and 2026 (n=1,665). A quantitative survey design measured familiarity, reported use and perceived risk; results show that students rapidly normalised AI use over the period, moving from tentative experimentation to routine engagement, while staff expressed persistent concerns about academic integrity, assessment design, and critical thinking. Doctoral and non-teaching staff occupied intermediate positions, reflecting both pragmatic adoption and institutional caution. The student-staff gap widened as institutional policy struggled to keep pace with actual practice. By tracking these shifts directly rather than reconstructing them from disconnected studies, the paper moves beyond descriptive accounts of AI attitudes and demonstrates the importance of capturing perceptions in real time. The findings carry significant implications for adaptive institutional policy, AI literacy initiatives, and targeted staff training.
[HC-30] A Fuzzy Logic Framework for Community-Aware Crime Hotspot Detection: Prototype Application Architecture and Exploratory Validation for an Urban Computing Platform ICIP
链接: https://arxiv.org/abs/2607.16218
作者: Ariton Verush,Vaibhav Motwani
类目: Computers and Society (cs.CY); Human-Computer Interaction (cs.HC)
备注: 12 pages, 6 tables, 1 figure. Framework paper with Python prototype and exploratory validation with 25 participants
Abstract:Urban crime prevention is a persistent socio-technical challenge for municipalities, law enforcement agencies, and citizens. Traditional reporting and response processes often rely on delayed incident reports and reactive resource allocation, while community-level signals and ambiguous early-warning indicators may remain underused. This paper reframes an Urban Computing seminar project into a fuzzy logic-based framework for community-aware urban crime hotspot detection and real-time notification. The proposed platform combines citizen reports, historical crime data, contextual urban indicators, and configurable fuzzy rules to estimate localized risk levels and support targeted awareness notifications. Unlike binary classification approaches, fuzzy logic can represent partial risk, uncertainty, and incomplete information, making it suitable for urban environments where risk is gradual and context-dependent. The paper presents the system architecture, fuzzy risk model, notification workflow, privacy safeguards, and exploratory validation with 25 participants. The evaluation focuses on perceived usefulness, notification relevance, usability, trust, privacy concerns, multilingual accessibility, and acceptance of community reporting. The findings suggest that such a platform may improve situational awareness, support reporting and planning discussions, and help address weak points in existing urban safety workflows, while not yet constituting evidence of deployed crime reduction or real-world predictive accuracy. The contribution is a responsible, human-centered framework for future research on fuzzy logic, urban computing, and community-aware crime prevention systems.
计算机视觉
[CV-0] he Many Senses of Visual Similarity: A Text-Prompted Image Perceptual Metric
链接: https://arxiv.org/abs/2607.18237
作者: Sheng-Yu Wang,Yotam Nitzan,Aaron Hertzmann,Jun-Yan Zhu,Eli Shechtman,Alexei A. Efros,Richard Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Project Webpage: this https URL
Abstract:Human visual similarity judgments are context-dependent. For example, two images may be similar in shape but distinct in color. Existing perceptual similarity metrics, however, collapse these nuances into a single scalar value, offering no mechanism to condition on specific aspects. To bridge this gap, we introduce a large-scale dataset of human similarity judgments over image triplets, where each triplet is annotated across multiple, free-form semantic aspects of similarity. Benchmarking a broad range of frontier vision-language models (VLMs) reveals a considerable performance gap compared to human annotators’ consensus. Leveraging our data, we fine-tune a VLM to produce our Text-Prompted Image Perceptual Similarity (TPIPS) metric, capturing multiple senses of visual similarity depending on the specified text prompt. We demonstrate that TPIPS aligns more closely with human perception and generalizes reliably beyond the training distribution. Finally, we show that TPIPS unlocks new capabilities in text-guided retrieval, compositional search, and the fine-grained evaluation of generative models. Our code, data, and trained models are at this https URL
[CV-1] Simple Domain Generalization for Strong Pixel-Level Image Tampering Detection in Modern VLMs
链接: https://arxiv.org/abs/2607.18230
作者: Yi Tang,Xinyi Shang,Jiacheng Cui,Sondos Mahmoud Bsharat,Jiacheng Liu,Xiaohan Zhao,Tran Dinh Tien,Ahmed Elhagry,Salwa K. Al Khatib,Tianjun Yao,Yonina C. Eldar,Jing-Hao Xue,Hao Li,Salman Khan,Zhiqiang Shen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Our code is available at this https URL
Abstract:Modern vision-language models (VLMs) have significantly improved image generation and editing capabilities, making pixel-level image tampering detection increasingly important yet challenging under cross-model and out-of-distribution shifts. This work studies domain generalization for pixel-level image tampering detection in modern VLMs like ChatGPT, Gemini, Qwen-Image, etc., aiming to learn tampering localization models that remain robust across diverse VLM-generated manipulation distributions. We propose a simple yet effective domain-generalized training framework built on two practical strategies. First, we introduce a balanced minibatch sampling scheme that strategically samples tampered and real images in each minibatch, preventing biased optimization toward either manipulated artifacts or clean-image priors and avoiding training collapse, ensuring that each optimization step receives proper sampled gradient signals. Second, we adopt a simple late-injection strategy, where the detector is first trained on large-scale base data until stable convergence, and then exposed to a small amount of newly selected supporting data from emerging VLM distributions, improving adaptability without overfitting to limited new domains. Together, these components provide a simple yet strong recipe for improving pixel-level tampering localization and OOD robustness across modern VLMs. Despite the conceptual simplicity, our framework outperforms the prior state-of-the-art PIXAR by a large margin of 26.1% and 26.8% relative improvement in average gIoU and cIoU, respectively, across OOD VLMs of GPT-Images-2.0, Gemini-3.1, FLUX.2, and Seedream 4.5. Our code is available at this https URL
[CV-2] FlowMimic: Mask-free Visual Editing and Generation with Pixel-pair Warped Flow Field for Online Video Editing Data Generation and Modality Mimicry
链接: https://arxiv.org/abs/2607.18227
作者: Dingyun Zhang,Lixue Gong,Wei Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Due to file size constraints, the figures in the arXiv file have been heavily lossy-compressed. Please visit the uncompressed file at: this https URL
Abstract:In line with the prevailing direction of vision research, we explore the integration of both generation and editing capabilities for video and image modalities within a single model. Current approaches to collecting video editing data typically depend on labour-intensive, time-consuming curated procedures–involving object mask annotation, the use of error-introducing pair synthesis via I2V model and ControlNet-like guidance, and VLM-based quality filtering or refinement–and demonstrate limited task scalability. As a result, the diversity of editing tasks remains substantially narrower than that available for image editing models. We develop a pixel-pair temporal warped flow field that can directly generate corresponding video editing samples in real time from image editing samples, and we demonstrate across multiple levels of video editing tasks that a model can learn video editing using only such data. We regard the image modality as a particular form of the video modality. Accordingly, we design a modality mimic generation loss and a modality mimic editing loss to relatively align the capabilities–and thereby the output distributions–of the two modalities through mutual imitation. Moreover, language-based visual editing entails the comprehension of the editing instruction and the reference visual content, the localization of the region corresponding to that instruction within the reference visual contents, and the modification of that region alone. Existing approaches predominantly rely on external aids, such as fine-tuning an additional MLLM or explicitly supplying a mask sequence as auxiliary input during inference. In contrast, we aspire for the model to internalize this capability. To that end, we introduce sense-related tasks–for instance, referring expression segmentation–along with corresponding editing-region-aware latent-level loss and attention-level loss.
[CV-3] GigaPath-Flash and GigaTIME-Flash: Efficient Pathology Foundation Models for Whole-Slide and Tumor Microenvironment Analysis
链接: https://arxiv.org/abs/2607.18218
作者: Naoto Usuyama,Jeya Maria Jose Valanarasu,Sicong Yao,Hanwen Xu,Jaspreet Bagga,Guanghui Qin,Robert E. Kramer,Cliff Wong,Soohee Lee,Hao Qiu,Theodore Zhengde Zhao,Racheli Ben Shimol,Angela Crabtree,Kevin Matlock,Eduardo Alejandro Lozano Garcia,Naiteek Sangani,Alberto Santamaria-Pang,Jason Entenmann,Alexandra Q. Bartlett,Bill J. Wright,Bernard A. Fox,Brian Piening,Sheng Zhang,Sheng Wang,Tristan Naumann,Carlo Bifulco,Hoifung Poon
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Models: this https URL (GigaPath-Flash) and this https URL (GigaTIME-Flash)
Abstract:Foundation models have emerged as a driving force in computational pathology, with the potential to transform cancer diagnosis, prognosis, and treatment selection by learning transferable representations from large-scale histopathology data. A growing landscape of pathology foundation models now spans diverse data sources, architectures, and downstream applications. However, most pretrained models operate only at the image-tile level, use restrictive licenses, and remain computationally expensive, limiting large-scale slide-level clinical and research use. Here, we introduce GigaPath-Flash and GigaTIME-Flash, efficient models for whole-slide pathology AI and spatial proteomics prediction. GigaPath-Flash combines a 22M-parameter ViT-S tile encoder with a 21M-parameter LongNet slide encoder, both pretrained on large-scale real-world histopathology data. Its compact tile encoder is distilled from the billion-parameter GigaPath (ViT-g) teacher and shared by both models. GigaPath-Flash retains 97% of GigaPath’s average slide-level performance with 50x less compute. GigaTIME-Flash extends this backbone to predict the tumor immune microenvironment directly from routine HE images. It surpasses the original CNN-based GigaTIME in prediction quality while running 6x faster and using 8x less GPU memory. Together with GigaPath and GigaTIME, these models form an open-weight, Apache-2.0-licensed family pretrained on large-scale real-world clinical data. By releasing all models and weights, we provide accessible building blocks for computational pathology, immuno-oncology, and precision health. Comments: Models: this https URL (GigaPath-Flash) and this https URL (GigaTIME-Flash) Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.18218 [cs.CV] (or arXiv:2607.18218v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.18218 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-4] HOMIE: Human-object Centric Video Personalization via Multimodal Intelligent Enchancement
链接: https://arxiv.org/abs/2607.18217
作者: Yiyang Cai,Nan Chen,Rongchang Xie,Junwen Pan,Chunyang Jiang,Cheng Chen,Wen Zhou,Zhenbang Sun,Wei Xue,Wenhan Luo,Yike Guo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 28 pages, 14 figures
Abstract:Human-object centric video personalization (HOCVP) is a core task within subject-driven video generation. However, existing methods suffer from two key limitations. First, most approaches focusing on inter-subject personalization still struggle to strike a balance between high subject fidelity and accurate interaction patterns between humans and diverse objects, especially when objects represent abstract concepts such as logos. Second, while intra-subject references (e.g., OCR maps, multi-view inputs) are expected to enhance subject fidelity, most existing works lack mechanisms to understand such latent correspondence. To address both challenges, we propose HOMIE, an HOCVP framework that tackles both inter- and intra-subject input settings in a unified manner. Compared to previous approaches, HOMIE proposes a better MLLM integration strategy to extract knowledge of reference-level relationships without compromising the controllability of text encoders or incurring costly re-alignment. Specifically, we introduce global multimodal guidance within self-attention to better align MLLM-derived semantic features with VAE tokens. Furthermore, we propose modality-reference embedding to differentiate tokens from MLLM features and VAE tokens and associate intra-subject reference image tokens. Extensive experiments validate that our method achieves state-of-the-art performance across various HOCVP tasks. Project Page: this https URL
[CV-5] hree-Body Scattering for Generative Modeling
链接: https://arxiv.org/abs/2607.18198
作者: Peng Sun,Zhenglin Cheng,Deyuan Liu,Jun Xie,Xinyi Shang,Tao Lin
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 31 pages, 5 figures, and 4 tables. Code: this https URL
Abstract:Modern generative models typically rely on an adversarial critic, a prescribed noise-to-data path, or an autoregressive factorization. Instead, we show that a proper distributional energy can induce sample-level motion and provide direct regression supervision for a one-step generator. Three-Body Scattering Modeling (TBSM) for generation turns the energy distance into a constant-size per-projectile interaction: each projectile is attracted toward one real source and repelled from one independently generated source. Conditioned on the projectile and its condition, its expectation equals the 2 -Wasserstein gradient-flow velocity of \frac12D_E^2(P_\theta,Q) . A batch of B frozen-target events yields O(B) sample-level losses, each using one reference for its condition instead of the minibatch-wide all-pairs field used by methods such as Drifting Models. Tracking this conditional expectation online can reduce field noise. Using scattering in frozen image features, TBSM trains one-step generators on ImageNet-256, achieving FID =2.23 with pixel-space PixelDiT-XL and FID =1.63 with latent-space DiT-XL at NFE =1 . We provide a design map relating diffusion-related supervision, Drift-like dynamics, and GAN-like objectives. These results establish tracked scattering as a route to high-dimensional one-step generation. Code: this https URL.
[CV-6] Certified Training for Convolutional Perturbations
链接: https://arxiv.org/abs/2607.18195
作者: Benedikt Brückner,Alessio Lomuscio
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Vision models have been found to be susceptible to perturbations such as motion blur induced at runtime by a shaking camera. This impedes their deployment in critical applications since phenomena such as slightly blurred vision might lead to failures, for example an object detector missing objects. While methods such as data augmentation or Adversarial Training can improve empirical robustness, they lack formal safety guarantees, making it difficult to identify and mitigate hidden vulnerabilities. We introduce a novel Certified Training approach that leverages an efficient encoding of convolutional perturbations to train provably robust models. Our method significantly outperforms Adversarial Training, achieving, for example, over 80% robust accuracy against motion blur of reasonable intensity on CIFAR10 while maintaining comparable standard accuracy.
[CV-7] Robust Multimodal Dynamic Object Segmentation ICRA2026
链接: https://arxiv.org/abs/2607.18153
作者: Zhe Xin,Hanzhi Chang,Penghui Huang,Yinian Mao,Guoquan Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by IEEE International Conference on Robotics Automation ICRA 2026
Abstract:Dynamic object segmentation plays a critical role in many visual applications such as static scene reconstruction from dynamic videos. However, existing optical flow-based methods fail to ensure consistent static/dynamic segmentation along object boundaries, while 3D reconstruction-based approaches are highly sensitive to reconstruction errors. To address these limitations, we present a dynamic object segmentation framework that can generate both precise and complete dynamic masks by integrating multimodal cues including 2D point tracks, 3D reconstruction, and semantic information. We design a network combining Transformer architectures with feature clustering aggregation modules to perform static/dynamic classification of multimodal feature trajectories. It enables the model to adaptively determine which type of feature should dominate based on the characteristics of each scene, while also mitigating the impact of feature degradation. Additionally, we introduce a novel point-query-based SAM post-processing method capable of handling multiple objects within a single mask. Extensive experiments demonstrate that our approach achieves state-of-the-art performance in both dynamic object segmentation and static scene reconstruction tasks.
[CV-8] Plenoptic Condensation: A Novel Approach to Generalized Scene Reconstruction
链接: https://arxiv.org/abs/2607.18151
作者: Brevin Tilmon,Alex DeJournett,John Leffingwell,Scott Ackerson
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We present a novel Generalized Scene Reconstruction (GSR) approach called Plenoptic Condensation (PCon). PCon uses a multi-stage reconstruction pipeline, initially converting images into “soupy” scene elements with low (representational) power, then adaptively condensing the “soup” into “structured” elements of higher power capable of efficiently representing, for example, sharp edges and smooth reflective surfaces. PCon scene models called Reality Models (Relms) enable spatially varying representational power, which is essential for high-fidelity rendering, measurement, and scene understanding. We showcase several in-the-wild PCon reconstructions captured with consumer phone cameras and drones. In one case called “Damaged Fiat”, PCon is benchmarked against two state-of-the-art (SOTA) GSR methods: NeRO and RT-Splatting. Referring to Figure 1 below, PCon reconstructs the car hood more than twice as accurately as the SOTA methods. But more importantly, the local damage profile error for PCon is 35 um (0.035 mm), whereas the two other SOTA methods are essentially unable to measure the damage at all. Our project website is available at this https URL.
[CV-9] Lossless-INR: Lossless Volumetric Implicit Neural Representations IEEE-VIS2026
链接: https://arxiv.org/abs/2607.18150
作者: Kaiyuan Tang,Daniel Burke,Chaoli Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Accepted in IEEE VIS 2026 (short paper)
Abstract:Implicit neural representation (INR) methods provide continuous coordinate-to-value mappings and integrate naturally with direct volume rendering, making them attractive for representing volumetric data. However, existing INR-based approaches for volumetric data are inherently lossy, and even small reconstruction errors can propagate through rendering and downstream analysis. In this work, we explore Lossless-INR, a lossless INR framework for 3D scientific volumetric data based on bit-plane decomposition. By decomposing each voxel value into binary bit-planes, we reformulate reconstruction as per-bit binary classification, so that exact recovery reduces to predicting every bit correctly. To make this optimization tractable while keeping the representation compact, we combine an octree block-partitioning strategy that adaptively subdivides complex regions with a ternary feature-grid network whose grid entries are parameterized by a ternary set of values. Experiments on diverse volumetric datasets show that this design can achieve zero bit-error rate and bit-exact reconstruction, enabling faithful rendering and downstream analysis with a compact representation. The code is available at this https URL.
[CV-10] Occlusion-Aware Panoptic Segmentation with Joint Position Embedding and Occlusion-Level Attention
链接: https://arxiv.org/abs/2607.18112
作者: Wenbo Wei,Jun Wang,Shan Raza,Abhir Bhalerao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Panoptic segmentation in complex scenes remains challenging because of occlusions, yet modern approaches often neglect occlusion modelling. In this paper, we propose \textbfPosition \textbfEmbedding \textbfModulation with \textbfOcclusion-\textbfLevel \textbfAttention (PEMOLA), a novel occlusion-aware module that can be seamlessly integrated into transformer-based panoptic segmentation. To obtain occlusion cues, we train an occlusion classifier on the COCO-OLAC dataset. The classifier derives the occlusion-level attention, which serves as spatial guidance, while the occlusion labels are encoded into a learnable embedding to produce channel-wise weights. Through joint modulation, PEMOLA elegantly introduces the occlusion priors into the position embedding, thereby improving the occlusion modelling. We further annotate the Cityscapes dataset with occlusion levels, termed Cityscapes Occlusion Labels for All Computer Vision Tasks (Cityscapes-OLAC), following the same labelling protocol as COCO-OLAC, to evaluate the cross-dataset generalisation ability of PEMOLA. Extensive experiments on COCO-OLAC and Cityscapes-OLAC demonstrate that PEMOLA consistently improves panoptic segmentation quality while introducing minimal computational overhead. These results highlight the importance of occlusion modelling, where incorporating occlusion-level attention helps deliver robust panoptic segmentation under occlusion. Code and dataset are available at this https URL.
[CV-11] SpEmoC: A Balanced Speaker-Segment Multimodal Emotion Benchmark
链接: https://arxiv.org/abs/2607.18109
作者: Sania Bano,Shahzad Ahmad,Santosh Kumar Vipparthi,Sukalpa Chanda,Subrahmanyam Murala
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Understanding human emotions in spoken conversations is a key challenge in affective computing, with applications in empathetic AI, human computer interaction, and mental health monitoring. However, existing datasets vary in scale, emotion distribution, modality alignment, and data partitioning strategies, which can influence reliable cross-dataset generalization and minority-emotion modeling. We introduce SpEmoC a Speaking segment Emotion for Conversations comprising 306,544 raw clips from 3,100 English language movies and TV series. From these, 30,000 high quality, class balanced clips are curated, featuring synchronized visual, audio, and textual modalities annotated for seven emotions through a hybrid pipeline that integrates pretrained models with human validation. SpEmoC uses strict movie- and series-level splits to prevent content overlap between split sets, allowing more reliable evaluation of model generalization. The dataset also maintains a near-balanced distribution across seven emotions, including minority classes such as Fear and Disgust, which supports more balanced learning across categories. Extensive experiments, including in-domain benchmarking, cross-dataset transfer, low-data training, class-imbalance analysis, and modality transfer show that balanced data and careful splitting lead to more stable performance across emotions when models are evaluated on other datasets. These results highlight the importance of dataset design for robust and transferable multimodal emotion recognition.
[CV-12] Empowering On-Device Model Adaptation with an Edge AI Inference Accelerator ECML-PKDD2026 ECML-PKDD
链接: https://arxiv.org/abs/2607.18101
作者: Mateusz Piechocki,Alessandro Capotondi,Marek Kraft
类目: Machine Learning (cs.LG); Hardware Architecture (cs.AR); Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the ITEM Workshop, held in conjunction with ECML-PKDD 2026. To appear in the ECML-PKDD joint post-workshop proceedings
Abstract:On-device model adaptation is essential to enable lifelong personalization on resource-constrained hardware, but compute, power, and memory limitations of such devices make end-to-end backpropagation impractical for modern deep neural networks. This work proposes a heterogeneous adaptation pipeline that repurposes a commercial edge AI inference accelerator, Hailo-8L, for frozen-backbone feature extraction during on-device training. The computational graph is partitioned so that the pre-trained backbone is quantized to INT8 and run on the accelerator, while only a lightweight FP32 classification head is fine-tuned on the host CPU, enabling frequent, energy-efficient in-field updates with most weights remaining fixed. Across multiple architectures and datasets, this pipeline achieves up to 15.4x faster wall-clock training time compared to a Raspberry Pi 5 CPU baseline, offers competitive throughput in favorable settings, and consistently reduces energy per sample. Post-training quantization restoration is shown to be crucial for preserving the quality of accelerator-generated features and mitigating accuracy loss in quantization-sensitive architectures. Overall, the results demonstrate a practical approach to efficient on-device adaptation using inference-oriented edge accelerators. The implementation is available at this https URL.
[CV-13] SciForma: Structure-Faithful Generation of Scientific Diagrams
链接: https://arxiv.org/abs/2607.18091
作者: Yuxuan Luo,Peng Zhang,Xinjie Zhang,Xun Guo,Zhouhui Lian,Yan Lu
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR); Machine Learning (cs.LG)
备注: 30 pages, 21 figures
Abstract:Structural fidelity is essential to scientific methodology diagrams. To communicate research logic, these diagrams must faithfully render components, directional relations, and textual annotations. Since a single error, such as a reversed arrow or an unreadable equation, can invalidate the entire figure, structural fidelity is inherently conjunctive: correctness on one axis cannot compensate for failure on another. Current open-source models fail to satisfy this criterion. Supervised fine-tuning (SFT) learns plausible layouts but cannot reliably ensure structural correctness, while scalar reward-based post-training obscures which structural dimension has failed. To address this, we introduce SciForma, a framework for the structure faithful generation of scientific methodology diagrams. Specifically, SciForma decomposes diagram quality into three structural axes: Component, Arrow, and Text, guided by a structural inventory. Built on this foundation, we curate SciFormaData-700K for structured training and SciFormaBench-2K for logic-verified evaluation. To close the gap left by SFT, we develop Multi-Dimensional Conjunctive Preference Optimization (M-DPO), which enforces simultaneous correctness across all axes and adaptively routes gradients to the most deficient dimension in post-training. The same structural inventory also enables iterative editing at inference time to correct residual errors. This combination allows SciForma-9B to exceed all open-source baselines and GPT-Image-1.5 on both SciFormaBench-2K and AIBench, bringing open scientific diagram generation close to proprietary-level structural fidelity. Our code and data will be available at: this https URL.
[CV-14] he Label Complexity of Class-Conditional Coverag e under Distribution Shift
链接: https://arxiv.org/abs/2607.18088
作者: Weijia Han,Lisha Qu
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages main text, 22-page supplement
Abstract:Standard evaluation of many recognition systems contains distribution shift by construction, since benchmarks place disjoint conditions in the training and test splits. Under such a shift, split conformal prediction keeps marginal coverage near the nominal level while per-class coverage fails silently: on a real cross-subject skeleton benchmark, marginal coverage stays near ninety percent, the worst action class is covered about seventy percent of the time, and ten of the sixty classes fall below eighty percent coverage. We characterize the cost of restoring per-class validity. First, an impossibility: once the shift acts jointly on the covariates and the labels, the target class-conditional score law is unidentified from source labels and an unlabeled target sample, so no label-free method attains per-class coverage that is at once valid and efficient. Second, we make the cost precise: per-class validity alone needs only a handful of target labels per class, while the label count necessary and sufficient for validity together with per-class efficiency grows as the inverse square of the efficiency tolerance and the logarithm of the number of classes, with matching upper and lower bounds. Third, within the evaluated prediction-powered inference family, even the most favorable use of the classifier’s own pseudo-labels on an unbounded unlabeled target pool improves efficiency by at most a small constant factor where coverage collapses. Skeleton action recognition is our real-data case study. A per-class calibration using source labels alone recovers a substantial share of the per-class gap while the shift preserves marginal coverage, and stops helping exactly when marginal coverage itself breaks. Three real shifts of increasing severity trace this boundary, and the same collapse and recovery appears on a natural-image corruption benchmark, beyond any single modality. Comments: 16 pages main text, 22-page supplement Subjects: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.18088 [cs.LG] (or arXiv:2607.18088v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.18088 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-15] Sparse Evidence Can Suffice: Agent ic Evidence Seeking for Multimodal Video Misinformation Detection
链接: https://arxiv.org/abs/2607.18080
作者: Haochen Zhao,Yongxiu Xu,Xinkui Lin,Dong Xie,Jiarui Lu,Yuqi Qian,Yubin Wang,Hongbo Xu,Gaopeng Gou
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Multimodal video misinformation detection is commonly formulated as a holistic video-understanding task, where the entire video and its associated content are processed and judged in a single pass. However, real-world misinformation often exhibits a sparse and compositional evidence structure: a reliable decision may depend on only a few coupled clues, while most video content contributes limited additional information. Exhaustive multimodal reasoning may therefore introduce substantial redundancy and obscure decisive evidence. This motivates decoupling evidence acquisition from verification: first identifying sparse, decision-relevant clues and then judging veracity based on the acquired evidence. Accordingly, we propose SIEVE, a framework for Sparse Interactive Evidence Verification via Extraction in multimodal video misinformation detection. An evidence-seeking agent actively explores the available multimodal evidence and constructs a compact evidence package, which is then used by a verifier to determine veracity. The agent is trained with supervised evidence-seeking trajectories and an evidence-aware reinforcement learning objective that promotes informative evidence acquisition while discouraging unnecessary or invalid interactions. Experiments on multiple video misinformation benchmarks show that SIEVE consistently outperforms the evaluated baselines and supports reliable verification using compact evidence packages. Moreover, the resulting acquisition process provides an explicit and inspectable evidence trail, improving the transparency and groundedness of multimodal misinformation detection.
[CV-16] VGOcc: Learning Visual-Geometric Gaussians for Vision-Centric 3D Driving Occupancy Prediction
链接: https://arxiv.org/abs/2607.18078
作者: Junhong Lin,Xianda Guo,Kangli Wang,Yuqi Ye,Xiaoyu Liang,Yanlun Peng,Wei Gao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Vision-only occupancy prediction requires recovering a semantic 3D occupancy field from calibrated surround-view images, where each view provides observations with ambiguous depth along camera rays. Existing methods have progressed from dense structured representations to sparse Gaussian primitives, improving the efficiency of 3D scene representation. However, Gaussian learning still relies primarily on image domain features, which provide limited explicit geometric information for volumetric reasoning. Our key observation is that effective Gaussian occupancy modeling requires not only sparse primitives, but also richer geometric and semantic learning cues. In this paper, we propose VGOcc, which learns visual and geometric cues from foundation models for Gaussian modeling. VGOcc incorporates these cues into primitive initialization and refinement, yielding a representation termed Visual-Geometric Gaussians tailored to semantic occupancy prediction. Specifically, we propose Visual-Geometric Gaussian Birth to form spatially balanced Gaussian centers from ray depth hypotheses, while visual semantic features initialize primitive attributes. Next, we design Pose-Aware Feature Learning to combine foundation tokens with camera embeddings and calibrated ray information. Features from neighboring views are then aggregated at projected 3D locations for each Gaussian refinement stage. Finally, Gaussian decoder refines birth Gaussians with pose-aware features and renders them into semantic occupancy. Experiments on nuScenes demonstrate that VGOcc achieves state-of-the-art performance in vision-only 3D occupancy prediction. Codes will be available at this https URL.
[CV-17] QIRF Quantum-Inspired Non-Orthogonal Function-Space Compression for 3D Gaussian Splatting
链接: https://arxiv.org/abs/2607.18067
作者: Shizeng Jiang,Hao Zhang,Xuerui Ma,Ying Hu,Tao Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D Gaussian Splatting (3DGS) achieves high-quality real-time rendering by representing a scene with a large collection of anisotropic Gaussian primitives. However, complex scenes often require millions of Gaussians, resulting in substantial storage and rendering costs. Existing compression methods mainly reduce redundancy through primitive-wise pruning, attribute quantization, clustering, or neural coding, while redundancy caused by strongly overlapping and non-orthogonal Gaussian basis functions remains largely unexplored. We present QIRF, a quantum-inspired non-orthogonal function-space compression method for 3D Gaussian Splatting. QIRF models neighboring Gaussian primitives as a local non-orthogonal basis and formulates primitive reduction as a subspace-aware selection problem. Specifically, an analytic Gaussian overlap matrix and a radiance-response density matrix are constructed to characterize functional redundancy and rendering relevance. Generalized eigendecomposition is then used to identify the dominant local subspace and select representative Gaussian primitives. An RRDM-based response model and detail-aware safeguarding further preserve visually important high-frequency structures under aggressive pruning. Experiments on 13 scenes from Mip-NeRF 360, Tanks and Temples, and Deep Blending show that QIRF reduces the Gaussian count and raw PLY storage by 71.7 percent on average, corresponding to approximately 3.54 times compression, while maintaining reconstruction quality comparable to 3DGS and achieving a marginal average PSNR improvement of 0.10 dB. QIRF also improves the average rendering speed over 3DGS by 34.3 percent. These results suggest that non-orthogonal function-space redundancy is an important yet underexplored source of representational redundancy in explicit Gaussian radiance fields.
[CV-18] UniETP: Unifying Environments for Generalizable Embodied Task Planning
链接: https://arxiv.org/abs/2607.18062
作者: Peiran Xu,Jiaqi Zheng,Ziyou Wang,Yadong Mu
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: We are actively working on releasing the codes and data
Abstract:This paper focuses on the problem of Embodied Task Planning, where an agent is required to execute a sequence of atomic actions within an interactive environment to complete a user-specified task. Though a variety of simulators and datasets have previously been built for this task, these efforts are largely isolated, with each using its own observation format, action type, and task domain. This fragmentation complicates comprehensive model evaluation and hinders the scalability of training data. As an effort towards generalizable embodied planning, we propose UniETP, a unified interface integrating four commonly-used simulators (AI2-THOR, VirtualHome, Habitat, BEHAVIOR). UniETP is characterized by both standardization and diversity. On one hand, it formalizes all the simulators into a consistent observation and action space, and builds an evaluation system to support complicated task goal. On the other hand, it enhances task diversity and complexity across dimensions like task logic, instance grounding, and instruction understanding, constructing a new dataset with varied levels of difficulty in an automatic manner. Extensive experiments on the proposed benchmark are conducted to evaluate the embodied planning capabilities of recent models and analyze the performance bottlenecks. Codes and data will be available at this https URL .
[CV-19] SAR Vessel Detection and Gross Tonnage Estimation from Heterogeneous Datasets for Dark Vessel Identification
链接: https://arxiv.org/abs/2607.18051
作者: Davide Paltrinieri,Andrea Diecidue,Roberto Basla,Daniele Casciani,Piero Fraternali,Giacomo Boracchi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 6 pages
Abstract:Detecting vessels engaging in illegal activities is of paramount importance for maritime security. One of the major goals is to detect dark vessels, ships that disable their transponders to evade surveillance. Deep Learning (DL) models can detect vessels in Synthetic Aperture Radar (SAR) images, enabling maritime traffic analysis regardless of weather or visibility conditions. However, to detect potential dark vessels, a DL model must select only those that are required to carry a transponder based on their Gross Tonnage (GT). Unfortunately, no public SAR dataset is available for training an end-to-end DL model for vessel detection and GT regression. In this work, we present a framework that leverages heterogeneous image and tabular datasets to solve this task. Our solution combines a multi-task DL framework for predicting the location, vessel type, and physical dimensions of ships, cascaded with a non-parametric model for predicting GT from vessel size and category. We perform GT regression by a KNN that measures sample similarity using a hybrid Euclidean and categorical distance. Experiments show that our solution can predict multiple outputs while remaining competitive with state-of-the-art models on individual subtasks, thus enabling the identification of dark vessels. We publish our code on GitHub this https URL.
[CV-20] Anticipate Before Acting: Future-State-Conditioned Vision-Language Navigation
链接: https://arxiv.org/abs/2607.18042
作者: Lingfeng Zhang,Zhanguang Zhang,Liheng Ma,Tongtong Cao,Yingxue Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 10 pages, 1 figure, 5 tables
Abstract:End-to-end vision-language navigation (VLN) with causal vision-language models can map instructions and egocentric observations directly to actions, but standard behavior cloning supervises only the next action and does not explicitly train the policy state to be predictive of future visual outcomes. We first ask a diagnostic question: if the policy is given an expert-trajectory future image as privileged input at training and testing time, is that additional visual evidence useful for choosing the current action? (These expert-trajectory future images are unavailable at test time in real deployment, so we use this setting only as a privileged-input diagnostic.) The answer is yes; this sanity check shows that future observations can provide rich, actionable cues. We then ask a deployable question: without accessing future images at inference, can we still benefit from future information by using a compressed future visual latent only as training supervision? We propose Future-State-Conditioned VLN (FSC-VLN), which adds a future-query token and aligns its hidden state to a frozen visual embedding \Delta steps ahead via a training-only target branch that is removed after training. On R2R val-unseen, FSC-VLN improves SR/OSR/SPL over a StreamVLN-style baseline under two training-data regimes, with larger gains on long-horizon episodes; ablations further support the dual-query design (separating future and action queries). Comments: 10 pages, 1 figure, 5 tables Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.18042 [cs.CV] (or arXiv:2607.18042v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.18042 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-21] When 2D Cues Fail: Improving Image Manipulation Localization with Reliable 3D Geometry
链接: https://arxiv.org/abs/2607.18040
作者: Guofeng Yu,Zhiqing Guo,Dan Ma,Gaobo Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing image manipulation localization (IML) methods rely heavily on 2D forensic cues, such as low-level artifacts, noise traces, and semantic inconsistencies in the manipulated image. While effective in many cases, these cues become much less discriminative when manipulated regions are well blended with their surrounding context in appearance. In such cases, a manipulated region may remain locally appearance-consistent, but still violate the geometric structure of the surrounding scene. This limitation motivates us to go beyond purely 2D evidence and introduce geometric reasoning into IML. To this end, we leverage monocular reconstruction to obtain auxiliary geometric cues, including depth and surface normals. However, a key challenge lies in the fact that reconstructed geometry on manipulated images is inherently noisy and cannot be used naively. Rather than treating depth and normals as direct evidence, we estimate their reliability and exploit them selectively for localization. Based on this principle, we design a geometry-aware framework (GFrame) that fuses reliable geometric cues with RGB features and propagates them across scales to improve fine-grained localization. Extensive experiments show that the proposed method achieves excellent performance under limited budget constraints. These results indicate that reliable 3D geometry provides complementary forensic evidence beyond traditional 2D cues for IML. Related code will be released.
[CV-22] Benchmarking NACTI Species Recognition in Long-Tailed Regimes
链接: https://arxiv.org/abs/2607.18033
作者: Zehua Liu,Tilo Burghardt
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:As with most ``in the wild’’ collections of the natural world, the North America Camera Trap Images (NACTI) dataset exhibits long-tailed class imbalance, with the largest class covering over 50% of its 3.7M images. Building on the PyTorch Wildlife model, we systematically evaluate Long-Tail Recognition (LTR) methodologies to benchmark species recognition performance, including specialised loss functions and LTR-sensitive regularisation. Our optimised configuration achieves state-of-the-art 99.40% Top-1 accuracy on the NACTI test split, significantly outperforming standard baselines and previously reported top performances. To assess robustness under domain shifts (e.g., night-time captures, occlusion, motion-blur), we extend our evaluation across three independent reduced-bias test sets (including ENA-Detection, Caltech Camera Traps and Missouri Camera Traps). Across these out-of-distribution (OOD) evaluations, our LTR-enhanced model consistently demonstrates substantially stronger generalisation capabilities compared to standard cross-entropy approaches. However, qualitative and quantitative analyses underline that current LTR optimisations cannot fully overcome representational bottlenecks, resulting in catastrophic predictive breakdown for rare `Tail’ classes under severe domain shift. For maximum reproducibility, all dataset splits, key code, and network weights are published with this paper at this https URL.
[CV-23] SAMRI-3D: Adapting SAM2 for 3D MRI Segmentation with Global Volume Tokens
链接: https://arxiv.org/abs/2607.18014
作者: Zhao Wang,Wei Dai,Hongfu Sun,Craig Engstrom,Shekhar S. Chandra
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Foundation models such as Segment Anything Model 2 (SAM2) have transformed natural-image and video segmentation, and recent work has begun adapting them to medical imaging. These adaptations, however, are largely general-purpose models that treat MRI as one modality among many; large-scale, MRI-specific modelling and benchmarking remain limited, even though MRI’s low soft-tissue contrast leaves many boundaries effectively invisible on individual slices. We present SAMRI-3D, a benchmark and method for 3D MRI segmentation with SAM2. The SAMRI-3D benchmark is the largest MRI-only evaluation to date - 10,392 volumes from 34 datasets (27 public, 7 in-house) spanning 12 anatomical domains and 10+ sequences, with explicit seen/unseen splits. Freezing the image encoder and fine-tuning only the lightweight decoder and memory modules raises mean Dice from 0.58 (zero-shot SAM2) to 0.76, surpassing recent SAM-based medical models (SAMed-2 0.69, Medical-SAM2 0.49, SAM-Med3D 0.37) with strong statistical significance. To target invisible boundaries, we introduce Global Volume Tokens (GVT): persistent memory tokens trained with a Truncated Signed Distance Field (TSDF) reconstruction objective that is discarded at inference (zero added cost). This full model, SAMRI-3D, attains the best accuracy (0.78) and lowest variance across all 34 datasets and, uniquely, shows no drop on 8 held-out datasets (0.79 unseen vs. 0.78 seen); per-sequence analysis confirms the TSDF objective helps most where per-slice contrast is weakest. We will release the benchmark, code, and models in this paper.
[CV-24] Do Maps Still Matter for Machines: Revisiting the Role of Choropleth Maps in Foundation Model Spatial Understanding
链接: https://arxiv.org/abs/2607.17999
作者: Zhiwei Wei,Yonghe Sun,Zhenjia Liu,Wenjia Xu,Chao He,Weihua Dong,Chunbo Liu,Hua Liao
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 34 pages, 4 figures, and 10 tables
Abstract:Spatial understanding is crucial for foundation models (FMs), and maps have long helped humans organize and reason about geographic information. This study examines whether choropleth maps remain useful for machine spatial understanding when models can directly process structured geodata. We introduce ChoroplethMap-Bench, a controlled benchmark containing 2,400 synthetic choropleth maps, corresponding GeoJSON data, and 12,000 questions across five cognitive dimensions: Identify, Spatial Recognition, Compare, Rank, and Delineate. We evaluate 22 open-source and proprietary models under three input conditions: Data Only, Map Only, and Data + Map. The results show that maps substantially improve spatial reasoning, especially when combined with symbolic data and for tasks requiring higher-level understanding of spatial patterns. We further analyze the effects of map type, color hue, and spatial structure, as well as prompting strategies, language, geographic context, decoding settings, classification methods, and response stability. Overall, the Data + Map condition achieves the strongest performance, demonstrating that maps remain valuable external representations for foundation model spatial reasoning.
[CV-25] HAS: Highlight-guided Attention Steering for Multimodal LLM Video Summarization
链接: https://arxiv.org/abs/2607.17994
作者: Rui Chu,Yingjie Lao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Video understanding has become more and more important with the growth of Artificial Intelligence (AI) for video generation. Recently, Multimodal Large Language Model(M-LLM) has shown its capability in video understanding. Video summarization, a specific domain of video understanding, has proven its importance for efficient navigation and retrieval. Both video understanding and video summarization require a good selection of key frames in a video. Current video summarization methods heavily focus on the selected key frames and correlated segment captions. However, existing approaches overlook the perspective of treating the importance of the frames globally. We argue that using discrete selected frames for summarization will not only reduce the understanding coherence, but also lost important information in the video, as well as wasting the original capacity of the MLLMs. In this paper, we propose HAS, a Highlight-guided Attention Steering method for video summarization. We consider a challenging but practical setting where the video given to MLLMs for summarize should be continuous but with highlight guidance. HAS mainly consists of two parts: The first part is to find a continuous frame-level highlight distribution for the video globally. The second part is to apply the highlight distribution as an attention steering vector for the MLLM, targeting a better understanding of the video, and thus during the model inference time, putting more attention on the highlighted frames, while avoiding lost entire information on less highlighted frames through putting less attention instead of forgetting them. We evaluated HAS on a variety of benchmarks, and it has shown convincing performance in video summarization.
[CV-26] Keyframe-Anchored Identity Preservation for Sequential-Action Video Generation
链接: https://arxiv.org/abs/2607.17985
作者: Zhenjie Liu,Binyan Chen,Hao Chen,Tong Pan,Shangfei Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注:
Abstract:Identity-preserving text-to-video generation aims to synthesize a video that accurately follows a textual description while maintaining the recognizability of a user-specified subject throughout. The IPVG26 challenge extends this framework from a single holistic prompt to a temporally structured specification. The model additionally receives a sequence of timestamped action captions and must render the subject performing these actions in the specified order. This temporal structure presents a challenge not encountered in previous identity-preserving generation tasks, as the subject must continuously perform a scripted sequence of distinct actions while maintaining a consistent identity. However, end-to-end video generators are prone to appearance drift as motion accumulates and the depicted actions change. We address this challenge with a training-free, three-stage pipeline framework. An action-aware prompt polishment stage first rewrites the inputs into image-generation prompts that specify the terminal state of each action. An identity-preserving generation stage then produces the keyframe sequence by conditioning each frame jointly on the reference identity and its predecessor, thereby decoupling time-invariant appearance from time-varying pose. Finally, an identity-aware inference enhancement stage synthesizes the intermediate segments using multi-reference guidance and identity-driven noise searching, both of which reinforce identity fidelity during sampling. Our method ranked third on the official Track 2 leaderboard, demonstrating competitive performance and strong generality.
[CV-27] Fine-Detail Monocular Geometry Estimation with Self-Guided Sparse Volumetric Refinement
链接: https://arxiv.org/abs/2607.17967
作者: Lingyu Kong,Ruicheng Li,Ruicheng Wang,Sicheng Xu,Chengtang Yao,Jianfeng Xiang,Jiaolong Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Monocular geometry estimation has recently achieved impressive performance across diverse scenes. However, state-of-the-art models still face notable distortion in local 3D structure, especially in fine details, like thin structures and small objects. We attribute this limitation to an architectural mismatch: most current models decode 3D geometry within a 2D parameterization, where feature interactions are governed by image-plane proximity rather than true 3D spatial relationships. This inadvertently mixes features from geometrically distant surfaces, resulting in over-smoothed geometry particularly around thin or elongated structure. In this paper, we propose a fine-detail monocular geometry estimation with Self-Guided Sparse 3D Refinement (SSR) that lifts monocular geometry modeling from 2D image space to 3D space for high-fidelity metric-scale point maps. Our model lifts the coarse point map from a foundation base model onto a sparse voxel shell and refines it via SSR. The SSR employs sparse convolutions that aggregate features based on 3D spatial locality, avoiding feature mixing across depth discontinuities. Extensive experiments on diverse datasets demonstrate that our method significantly outperforms existing approaches in recovering fine detailed 3D geometry across both quantitative metrics and qualitative visualizations.
[CV-28] Exploration Matters for Escaping the Blur Trap in 3D Gaussian Splatting
链接: https://arxiv.org/abs/2607.17965
作者: Chengbo Wang,Guozheng Ma,Jinhong Wu,Tie Ji,Yizhen Lao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:3D Gaussian Splatting (3DGS) employs Gaussian primitives for explicit scene representation, facilitating real-time, high-fidelity reconstruction and novel view synthesis of complex scenes. However, the explicit modeling inherent in 3DGS introduces a gradient bias during optimization, rendering its non-convex optimization process highly susceptible to convergence toward local suboptimal solutions. This constitutes a fundamental limitation in 3DGS optimization, which we term the Blur Trap. To address this limitation, we integrate simple explicit exploration into the 3DGS optimization framework. First, through rigorous mathematical analysis of the 3DGS optimization formulation, we identify the underlying optimization bias responsible for the Blur Trap and categorize it into two distinct subtypes: the Far-Side Blur Trap and the Near-Side Blur Trap. Subsequently, we propose two highly straightforward exploration strategies (Random Seeding and Random Splitting) to mitigate the far-side and near-side blur traps, respectively. Experimental validation demonstrates that the incorporation of these exploration operators effectively and complementarily overcome the Blur Trap, achieving high-quality rendering performance across multiple datasets. Project page: this https URL
[CV-29] MuViSeg: Multi-View Segment Correspondences from Dense Geometry Priors
链接: https://arxiv.org/abs/2607.17938
作者: Denis Fatykhoph,Timur Akhtyamov,Konstantin Pakulev,German Devchich,Gonzalo Ferrer
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 4 figures
Abstract:Classical image correspondence is solved at the level of sparse keypoints or dense pixels, but the systems that consume these matches - object-level mapping, topological navigation, scene-graph maintenance - reason about whole objects. Recent work narrows this gap by matchng directly at the level of instance segments: a class-agnostic segmenter partitions each image, and per-segment descriptors are obtained by pooling features from large 3D foundation models over the masks. We build on this segment-level matching paradigm and propose three learned matching heads: a LightGlue-style attention head with DoubleSoftmax scoring on frozen MASt3R descriptors; a DPT-style multi-scale fusion module that exposes layered spatial detail from the VGGT foundation model before pooling; and - as our main contribution - a multi-view extension that performs joint self-attention over segments drawn from several views at once, recovering transitive correspondences that strictly pairwise matchers cannot reach. Under a stratified zero-shot protocol on Replica and Virtual KITTI 2 with controlled viewpoint baselines from 0 deg to 180 deg, the LightGlue-style head improves over a parameter-free Sinkhorn matcher on the same MASt3R backbone by +4.85 AUPRC on Replica and +25.9 AUPRC on Virtual KITTI 2. Dropped into the RoboHop topological navigation pipeline on the Habitat-Matterport 3D (HM3D) Instance Image Navigation benchmark without retraining, our multi-view variant raises success rate from 50% to 70%, and our LightGlue-style head raises SPL from 45.7 to 59.1.
[CV-30] Packet-Loss Robust 3D Gaussian Compression via Atomic Packaging and GNN-based Error Concealment
链接: https://arxiv.org/abs/2607.17916
作者: Yuxuan Tao,Xuerui Ma,Hao Zhang,Chunhua Peng
类目: Graphics (cs.GR); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: 21 pages, 3 figures, 3 tables
Abstract:3D Gaussian Splatting (3DGS) and recent compression schemes such as HAC++ enable high-fidelity real-time neural rendering, but their bitstreams are fragile under packet loss during network streaming. Existing compression methods often separate correlated anchor attributes into independent streams, so losing one packet can create attribute-inconsistent broken anchors and severe rendering artifacts. We propose a packet-loss robust 3DGS transmission and error concealment framework. On the encoder side, anchor-level atomic packaging jointly encapsulates all attributes of each anchor, converting corrupted-attribute failures into clean missing-anchor erasures. Stratified random grouping further disperses packet losses across the spatial domain to avoid large contiguous voids. On the decoder side, we formulate recovery as prior-aware attribute inpainting. A Context-Aware Residual Interpolation (CARI) branch uses hash-grid prior predictions and neighboring residuals to build a robust baseline, while a lightweight two-layer graph neural network with cross-attention over hash-grid priors refines high-frequency attribute residuals. Attribute-wise confidence control falls back to interpolation when learned predictions are unreliable. Experiments under 20 percent random packet loss on BungeeNeRF, Mip-NeRF 360, and Tanks and Temples show that the proposed method substantially improves over no-concealment transmission and limits average PSNR degradation to about 3 dB relative to the lossless HAC++ reference.
[CV-31] Locality-Aware Density Control for Efficient Gaussian-based Image Representation
链接: https://arxiv.org/abs/2607.17896
作者: Jiacong Chen,Qingyu Mao,Xiandong Meng,Shuai Liu,Chao Li,Fanyang Meng,Youneng Bao,Yongsheng Liang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ACMMM 2026
Abstract:2D Gaussian Splatting is an attractive direction for image representation due to its explicit formulation, fast rasterization, and favorable decoding efficiency. The representation quality of this paradigm depends on the proper allocation of Gaussian capacity to the demanding regions. However, existing methods fail to allocate Gaussian capacity efficiently during optimization: under-reconstructed content is often refined in a fragmented pixel-wise manner, while neighboring optimized Gaussians with similar attributes are redundantly retained. This inefficiency motivates the need for a density control framework that jointly addresses insufficient allocation in under-reconstructed regions and redundant allocation in over-reconstructed regions. Our key insight is that this framework should exploit two complementary forms of locality: the local continuity of reconstruction errors in image space for improved Gaussian allocation, and the local similarity of neighboring Gaussians in Gaussian space for redundant elimination. Based on this insight, we propose Locality-Aware Density Control (LocoADC), a plug-and-play framework that improves Gaussian capacity utilization through Region-wise Gaussian Densification (RGD) and Similarity-Driven Gaussian Merging (SDGM) strategies, together with a local color consistency constraint for more reliable merging. Extensive experiments on diverse datasets show that LocoADC consistently improves multiple baselines by enabling more effective local Gaussian allocation, including a 2.93 dB PSNR gain over GI on the CLIC dataset under the same 30k Gaussian budget. Code is available at: \textitthis https URL.
[CV-32] Leverag ing Dissimilarity Invariance as a Robust Anchor for Learning with Noisy Labels AAAI2026
链接: https://arxiv.org/abs/2607.17857
作者: Wenxiao Fan,Kan Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: AAAI2026
Abstract:Deep learning models excel in visual recognition but suffer severe performance drops when training labels are corrupted by noise. Under label noise prior work cannot learn accurate similarities and thus misguide the learning process. In this paper, we uncover a complementary and novel phenomenon, Dissimilarity Invariance, whereby semantic dissimilarity between unrelated samples remains stable despite label noise. Leveraging this insight, we propose NegScale, a plug-and-play framework that shifts focus from fragile similarity to robust dissimilarity. NegScale integrates: (1) Structured Negative Orthogonality Penalty (SNOP), enforcing subspace orthogonality for unrelated samples; and (2) Dissimilarity-Calibrated Similarity Adjustment (DCSA), suppressing spurious similarity using dissimilarity anchors. We also give theoretical analysis that proves Dissimilarity Invariance and the effectiveness of NegScale. Empirical results demonstrate that NegScale consistently outperforms state-of-the-art baselines, establishing new benchmarks on CIFAR with synthetic noise and real-world datasets.
[CV-33] CaT-GS: Efficient 3DGS Rendering for Large Scale Scenes via Inter-frame Caching and Tile Scheduling CVPR2026
链接: https://arxiv.org/abs/2607.17842
作者: Tingjia Zhang,Bo Chen,Shengzhong Liu,Fan Wu,Guihai Chen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: CVPR 2026
Abstract:Recent breakthroughs in 3D Gaussian Splatting (3DGS) have advanced neural rendering with high fidelity and speed. However, its performance degrades significantly in large-scale scenes due to the computational burden of tile-based rasterization. Existing optimization efforts either require costly scene re-training or focus on narrow aspects of the pipeline, overlooking critical inefficiencies in real-world deployments. Through a comprehensive analysis, we identify three primary sources of redundancy and low GPU utilization: redundant inter-frame pre-processing, viewpoint-based occlusion redundancy, and severe tile-level load imbalance. To address these issues, we propose CaT-GS, a novel and efficient 3DGS rendering pipeline. CaT-GS introduces a speculative multi-frame preprocessing method to eliminate redundant computations across consecutive frames, and an inter-frame caching mechanism to eliminate viewpoint redundant rendering stages. Furthermore, it refactors rasterization tasks with a dedicated kernel to mitigate tile load imbalance, significantly boosting GPU utilization. Extensive experiments demonstrate that CaT-GS achieves a speedup of up to 10 times over the original 3DGS and up to 70% over previous state-of-the-art methods, establishing a new benchmark for high-fidelity, real-time rendering of large-scale scenes.
[CV-34] Measuring and Improving Complex-Atomic Answer Consistency in Endoscopic VQA
链接: https://arxiv.org/abs/2607.17834
作者: Yuhao Liu,Cheng Zhao,Guanghui Yue
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 7 pages, 5 figures, 6 tables
Abstract:Endoscopic visual question answering (VQA) increasingly asks complex questions that combine several endoscopic answer components rather than isolated factual queries. Such complex answers may be scored as correct even when the same model fails on associated atomic questions. We introduce EndoCA, a paired complex-atomic answer consistency benchmark for evaluating whether complex answers remain consistent with same-image atomic answers. EndoCA contains two suites: EndoCA-Core evaluates compact question-complexity patterns commonly seen in practical endoscopic VQA, and EndoCA-Diagnostic supports controlled analysis across increasing question complexity. We evaluate 11 VLMs spanning open, medical, endoscopy-adapted, and closed-source models on EndoCA. Some VLMs achieve high complex-answer accuracy, yet their atomic-answer accuracy and complex-atomic answer consistency remain substantially lower. To reduce this complex-atomic inconsistency, we introduce Atomic-Support Reconciliation (ASR), a training-free mechanism that uses model-generated atomic answers as contextual premises for answer revision and consistency-guided selective answering. On four selected publicly available models, ASR-Revise improves paired complex-atomic correctness with modest changes in complex-answer accuracy, while ASR-Selective improves accuracy on answered cases by allowing the model to abstain from less reliable cases. Together, EndoCA and ASR provide a consistency-aware benchmark and a training-free mechanism for answer reconciliation and selective answering in endoscopic VQA.
[CV-35] Consistent Feature Transport for Image Relighting
链接: https://arxiv.org/abs/2607.17833
作者: Bohan Zhang,Huanwei Liang,Yuhan He,Hongteng Xu,Quxiao Chao,Luoqi Liu,Dixin Luo,Ting Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Image relighting modifies illumination while preserving non-lighting content such as identity and geometry. Existing diffusion-based methods often suffer from unstable illumination changes or inconsistent content preservation under complex lighting, as they lack an explicit mechanism to learn feature transformations between images. We reformulate relighting as an illumination feature transport problem and introduce Consistent Feature Transport (CFT), a training principle that explicitly enforces illumination-consistent transport between source and target image distributions. Built upon rectified flow, CFT jointly models noise-to-image generation and illumination-consistent source-to-target transport through trajectory-level supervision. This dual-transport formulation encourages isolation of illumination-specific variations while preserving content-aligned features. To support complex lighting scenarios, we construct a large-scale portrait relighting dataset with diverse relighting effects. Experiments show consistent improvements over existing state-of-the-art relighting approaches and demonstrate that CFT can generalize to other editing tasks, including style transfer. Code is available at this https URL.
[CV-36] PRiSM: Prototype Regularization for Few-Shot VLMs NEURIPS2026
链接: https://arxiv.org/abs/2607.17820
作者: Ghassen Baklouti,Omprakash Chakraborty,Jose Dolz,Ismail Ben Ayed
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to NeurIPS 2026
Abstract:Training-free few-shot adaptation methods have gained significant attention recently in the context of Vision-language Models (VLMs). Yet, current benchmarks rely on strong assumptions about the statistics of the adaptation data, e.g., class balance. We question these simplifying assumptions and introduce a more realistic benchmark that varies both the levels of class balance and the effective number of classes in few-shot tasks via Dirichlet sampling. Surprisingly, under our setting, we observe substantial drops in the performances of state-of-the-art methods, more so when the number of labeled samples increases. To mitigate this, we introduce PRiSM, a class-prototype regularization that can be deployed as a plug and play module on top of any existing baseline method, significantly improving performances. Our method optimizes a novel multi-term loss, which includes a regularizer maximizing inter-class pairwise distances, along with additional terms promoting support-feature alignment and fidelity to the baseline prototypes. Furthermore, we introduce an effective and computationally efficient block Majorize-Minimize optimizer for our objective. More specifically, we derive a valid blockwise Lipschitz constant (i.e., a bound on the Hessian’s spectral norm), which can be computed efficiently via the Gershgorin circle theorem. Extensive experiments show that PRiSM improves several training-free baselines, with large gains when dealing with severe class imbalance and high numbers of classes.
[CV-37] FF-ProCams: Feed-Forward Gaussian Splatting for Projector-Camera System
链接: https://arxiv.org/abs/2607.17803
作者: Ziyao Wang,Yuqi Li,Wenxing Zheng,Jiaying Chen,Chong Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Projector-camera (ProCams) systems achieve active scene perception and controllable appearance manipulation via structured illumination, serving as a core infrastructure for spatial augmented reality, projection mapping, and surface reflectance acquisition. Existing inverse-rendering methods for ProCams deliver high-fidelity results but rely on time-consuming per-scene optimization, while mainstream feed-forward 3D reconstruction models produce baked appearance that cannot adapt to spatially varying projector illumination. To resolve this accuracy-efficiency trade-off, we propose FF-ProCams, a Feed-Forward 3D Gaussian inverse-rendering framework for ProCams. A hybrid Mamba2-Transformer encoder aggregates cross-view geometric and photometric cues from sparse multi-view observations, and lightweight heads predict a relightable Gaussian representation in a single forward pass. We further design a projector-aware differentiable renderer to synthesize camera observations under arbitrary active illumination and ProCams poses. To enable feed-forward training, we construct a large-scale synthetic ProCams dataset covering diverse object geometries and surface materials. Experiments show FF-ProCams achieves high-fidelity projector-aware rendering, generalizes to unseen patterns, and supports novel projector-camera poses. Using only 8 input views, it outperforms optimization-based baselines with 297 views while reducing test-time reconstruction to 0.13 seconds (a three-to-five-order-of-magnitude speedup). The code and data are available at this https URL. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.17803 [cs.CV] (or arXiv:2607.17803v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.17803 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-38] oward Optimal Adenovirus Detection Using YOLO26
链接: https://arxiv.org/abs/2607.17799
作者: Olivier Rukundo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 2 figures, 3 tables
Abstract:This study systematically benchmarks different data augmentation setups across YOLO26 model size variants to determine the most effective setup for adenovirus detection in TEM images. The benchmarked setups include NAS, GAS, GMAS and DAS, all evaluated under identical training conditions. The adenovirus dataset, selected from the published TEM virus dataset, was re-annotated by leveraging adenovirus particle positions to generate YOLO-compatible bounding box annotations. The experimental results demonstrated the impact of the benchmarked data augmentation setups on adenovirus detection with YOLO26 and indicated the most effective data augmentation setup.
[CV-39] ReViV: Reconstructing the Viewer and the View in 4D from Monocular Egocentric Video ECCV2026
链接: https://arxiv.org/abs/2607.17790
作者: Xiaozhong Lyu,Gen Li,Zhiyin Qian,Xucong Zhang,Marc Pollefeys,Siyu Tang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to ECCV 2026. The first two authors contributed equally, and their author order is interchangeable
Abstract:Egocentric devices, such as wearable front-facing cameras, provide a unique perspective for capturing the continuous interaction between a human viewer and the surrounding environment. A holistic and efficient multimodal model capable of reconstructing this 4D representation is therefore highly desirable. However, existing approaches often rely on auxiliary inputs such as pre-computed camera trajectories, treat scene perception and human ego-motion modeling as separate problems despite their strong interdependency, and suffer from slow inference time. To address these limitations, we present ReViV, the first unified framework for holistic egocentric 4D reconstruction that extracts both viewer and view dynamics from a single monocular RGB video. We formulate the task as learning the full joint probability distribution over multimodal signals, including RGB video, camera trajectory, gaze direction, full-body motion, hand motion, and depth. Powered by a Masked Generative Egocentric Transformer, ReViV operates within a single feed-forward architecture to simultaneously reconstruct the temporally consistent 4D reconstruction across the viewer and the view with fast inference speed. Extensive experiments on diverse benchmarks, including HoloAssist, HOT3D, ARCTIC, Aria Digital Twin, and TACO, demonstrate that ReViV achieves state-of-the-art accuracy and efficiency across holistic ego-body, hand, and gaze reconstruction, camera tracking, while maintaining highly competitive egocentric depth estimation without relying on heavy task-specific priors. Code and models are fully open-sourced: this https URL.
[CV-40] Medical Imaging Fusing Vision Transformer: Laryngeal Cancer Screening with Explanation
链接: https://arxiv.org/abs/2607.17789
作者: Haiyang Wang,Luca Mainardi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: EMBC accepted paper 2026
Abstract:Early and timely screening of laryngeal cancer is crucial for improving clinical outcomes. In recent years, NBI endoscopy has become a standard diagnostic tool for the detection of laryngeal lesions. However, its effective use requires well-trained clinicians and the procedure is time-consuming and subject to interobserver variability. In this context, the application of artificial intelligence (AI) offers a promising solution to support clinical decision-making. In this work, we proposed applying transformer and attention mechanism for analyzing the narrow band imaging and distinguish benign and malignant lesions. Results show it has good classification performance with F1 (82.72%), accuracy(82.33%). In addition, the result of laryngeal cancer screening is explainable for clinicians. The explainability is utilizing the state of art segmentation method (MedSAM) to provide the useful pathological information area for clinicians. The proposed methodology fusing classification and segmentation provides a translating on laryngeal cancer screening.
[CV-41] BrainNext: A General-Purpose Self-Supervised Foundation Model for Brain MRI Analysis
链接: https://arxiv.org/abs/2607.17782
作者: Moona Mazher,Abdul Qayyum,Steven A. Niederer,Daniel C. Alexander
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Foundation models pretrained using self-supervised learning have transformed computer vision by learning transferable representations from large-scale unlabeled data. However, existing foundation models for neuroimaging remain limited by task-specific training, slice-based learning strategies, or relatively small pretraining datasets, restricting their generalizability across diverse brain MRI applications. In this work, we present BrainNext, a general-purpose self-supervised foundation model for volumetric brain MRI analysis. BrainNext combines masked autoencoder (MAE) pretraining with a native three-dimensional Bi-Directional xLSTM-UNet architecture to learn rich anatomical representations from 60,551 unlabeled brain MRI examinations spanning multiple MRI modalities. The pretrained model is subsequently adapted to downstream tasks through lightweight task-specific fine-tuning. We evaluate BrainNext on the Foundation Models for Medical Imaging (FOMO) 2025 Method Track, encompassing classification, segmentation, and brain-age estimation, where it achieved second place overall and ranked first in the meningioma segmentation task on the official FOMO 2025 challenge leaderboard, demonstrating strong transferability across heterogeneous neuroimaging tasks. These results highlight the potential of large-scale self-supervised pretraining to learn robust and transferable volumetric representations, establishing BrainNext as a scalable foundation model for diverse brain MRI applications.
[CV-42] CDIS: Cross-Dimensional Class-Agnostic 3D Instance Segmentation via 2D Mask Tracking and 3D-2D Projection Merging IROS
链接: https://arxiv.org/abs/2607.17778
作者: Juno Kim,Hye-Jung Yoon,Yesol Park,Byoung-Tak Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 6 pages, 5 figures. Published in the proceedings of the 2025 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Abstract:Class-agnostic 3D instance segmentation is critical for robotic systems operating in unknown environments, enabling perception of previously unseen objects for reliable manipulation and navigation. Existing approaches typically project per-frame 2D instance masks into 3D and merge them, which often breaks object identities across time and yields fragmented 3D instances. We introduce Cross-Dimensional Class-Agnostic 3D Instance Segmentation (CDIS), a zero-shot framework that explicitly tracks 2D instance masks across frames and associates them with 3D superpoints, creating a feedback loop between 2D and 3D. This cross-dimensional reasoning links temporally stable 2D tracks with spatially coherent 3D regions, producing globally consistent 3D instance labels without any 3D-specific training. Experiments on benchmark datasets demonstrate that CDIS achieves higher accuracy and consistency than state-of-the-art zero-shot methods, while remaining efficient and scalable to diverse real-world environments.
[CV-43] Measuring Monosemanticity in Sparse Autoencoders via Latent Activation Coherence KDD2026 ECML
链接: https://arxiv.org/abs/2607.17770
作者: Katarzyna Filus,Sebastian Pokuciński
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: This is a preprint version. A shorter version of this paper has been accepted for presentation and publication in the post-workshop proceedings of the 8th International Workshop on eXplainable Knowledge Discovery in Data Mining (XKDD 2026), co-located with ECML PKDD 2026. The appendix is included only in this preprint and is not part of the peer-reviewed proceedings paper
Abstract:Within Explainable Artificial Intelligence, mechanistic interpretability uses Sparse Autoencoders (SAEs) to extract more interpretable features from neural representations. However, assessing their monosemanticity, and thus explanation quality, remains challenging. Existing metrics require external concept labels or depend on pretrained embedding models, making them sensitive to encoder’s geometry. We introduce the Tversky Monosemanticity Score (TMS), a label-free metric that operationalizes monosemanticity as activation-set coherence of binarized SAE latents, and does not require external embedding encoders. We evaluate TMS on SAEs trained on features from pretrained vision and vision-language models (DINOv3, CLIP, BLIP2), two common SAE regimes (TopK, BatchTopK), multiple sparsity levels, and expansion factors. Our results show that TMS is less affected by encoder anisotropy than its embedding-based alternative, while remaining aligned with established monosemanticity indicators. TMS also reveals distinct SAE training dynamics across base models. Moreover, under encoder anisotropy, TMS provides a stronger indication of probe-based concept deletion effectiveness, while being competitive otherwise.
[CV-44] o Blend In First Decouple: Rethinking Camouflage Image Generation via Context-Decoupled Representations
链接: https://arxiv.org/abs/2607.17768
作者: Wenzhuang Wang,Yifan Zhao,Mingcan Ma,Yunlong Che,Haoran Chen,Ming Liu,Jia Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 11 figures, ACMMM 2026
Abstract:Camouflage image generation (CIG) focuses on generating visually concealed objects that seamlessly blend into their backgrounds. Existing methods typically follow either background-guided paradigms that adapt object appearance via style transfer, or foreground-guided strategies that outpaint surrounding regions conditioned on object features. However, they still suffer from appearance discrepancy and background artifacts. We attribute these limitations to cross-context representation leakage, where object and background cues are entangled in a coupled conditional space, resulting in ambiguous control and degraded camouflage fidelity. To tackle this, we propose a new context-decoupled generative paradigm, termed CamoDreamer, which aims to isolate contextual conditional guidance and explicitly decouple latent camouflage features into coordinated object and background control streams. First, a Contrast-aware Contextual Bridge is designed to model cross-context discrepancies and construct contrast-aware dual conditional guidance. Second, Context-Decoupled Assimilation Streams are employed to separate generative interactions conditioned on the dual guidance, while facilitating background rendering with target-aware cues in the latent space. Finally, a Frequency-Adaptive Contextual Blend module integrates complementary high-frequency textures and low-frequency structures from decoupled features to improve holistic coherence. Extensive experiments demonstrate that CamoDreamer consistently outperforms existing methods with a substantial margin, while maintaining a relatively lightweight design.
[CV-45] DA-Fusion: Deformable Attention-Based RGB-D Fusion Transformer for Unseen Object Instance Segmentation ICRA2025
链接: https://arxiv.org/abs/2607.17754
作者: Yesol Park,Hye-Jung Yoon,Juno Kim,Byoung-Tak Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 7 pages, 5 figures. Published in the Proceedings of the 2025 IEEE International Conference on Robotics and Automation (ICRA 2025)
Abstract:In logistics automation, precise segmentation of unseen objects is crucial for efficient robotic manipulation in cluttered environments. Tasks such as bin-picking and shelf-picking require robust perception to handle occlusions, varying object shapes, and complex spatial arrangements. Traditional RGB-based methods tend to over-segment objects due to their reliance on texture, while depth-based methods often under-segment by focusing primarily on geometric features. To address these limitations, we propose DA-Fusion, a deformable attention-based RGB-D fusion Transformer designed for unseen object instance segmentation. DA-Fusion effectively combines the strengths of both RGB and depth data, enhancing segmentation accuracy in cluttered and multi-layered object environments. We also introduce the Object Clutter Bin Dataset (OCBD), a benchmark dataset specifically tailored for evaluating bin-picking scenarios in top-down views. Extensive evaluations demonstrate that DA-Fusion outperforms state-of-the-art methods across diverse environments, making it particularly suited for real-world logistics tasks.
[CV-46] PC-Seg: Progressive Cross-View Consistency for 3D OCT Segmentation from Sparse 2D Annotations
链接: https://arxiv.org/abs/2607.17718
作者: Tsubasa Konno,Takahiro Ninomiya,Yukun Zhou,Koichi Ito,Siegfried K. Wagner,Yiqun Lin,Pearse A. Keane,Toru Nakazawa,Takafumi Aoki
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Volumetric segmentation of optical coherence tomography (OCT) images is essential for diagnosing ocular diseases but requires labor-intensive voxel-wise annotations. While semi-supervised learning (SSL) can reduce annotation costs, most existing methods process data slice by slice and fail to exploit the inherent 3D spatial context. We propose PC-Seg, a progressive cross-view consistency framework that learns high-accuracy 3D segmentation models from sparse 2D annotations. Unlike conventional multi-view approaches, PC-Seg uses a single 2D model to learn cross-view consistency from standard B-scans and orthogonal slices, thereby generating reliable volumetric pseudo-labels. These pseudo-labels are then distilled into a 3D model, followed by a co-training stage in which the 2D and 3D models mutually refine each other through ensemble pseudo-labeling. Experiments on the MSHC and Duke DME datasets demonstrate that PC-Seg achieves accuracy comparable to fully supervised learning while using labels for only about 0.7% of the training data, outperforming state-of-the-art semi-supervised and retinal layer segmentation methods. Our code is publicly available at this https URL.
[CV-47] SLAM in Low-Light Environments: Project Report KR
链接: https://arxiv.org/abs/2607.17699
作者: Oleh Basystyi,Anna Stasyshyn,Oleksandr Kosovan,Yaroslav Prytula
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: Conducted as part of the certification program "Off-Road Visual Navigation: Development and Evaluation of Systems in Challenging Environments’’ at the Faculty of Applied Sciences, Ukrainian Catholic University, in collaboration with the UCU UGV Club
Abstract:Simultaneous localization and mapping (SLAM) is one of the fundamental problems in robotics, as it enables autonomous operations in real-world scenarios. Under low illumination, reduced contrast, sensor noise, and motion blur degrade both feature extraction and feature matching, while compensating with LiDAR, depth, or thermal sensors raises cost, power draw, and integration complexity. Existing benchmarks remain dominated by well-lit indoor or daylight sequences, leaving open how far SLAM with standard RGB cameras can be pushed in the dark. We benchmark six systems spanning the feature-based, direct, filter-based, and learning-based paradigms - ORB-SLAM3, DSO, Kimera-VIO, OpenVINS, DPVO, and DPV-SLAM - on five LaMARia sequences of varying difficulty and illumination, reporting absolute and relative pose error alongside control-point recall. Kimera-VIO is the only system to track all five sequences to completion, combining the lowest relative pose error with steadily growing absolute error due to the absence of loop closure; DPVO and DPV-SLAM never lose tracking but incur absolute errors of roughly 100 m under low light; and the classical monocular pipelines (ORB-SLAM3, DSO) together with the filter-based OpenVINS fail outright or diverge on most of the harder and low-light sequences. The results suggest that RGB-only SLAM maintains stable low-light tracking only when both inertial fusion and global optimization are present. Closing the remaining gap will likely require low-light-specific learned front-ends or a return to complementary sensing.
[CV-48] Memory-Supported Synergistic Adaptation for Training-Free Test-Time Medical Image Segmentation ECCV2026
链接: https://arxiv.org/abs/2607.17693
作者: Lingrui Li,Nan Pu,Dong Zhao,Wenjing Li,Andrew P French,Zhun Zhong,Xin Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, ECCV 2026
Abstract:Test-time adaptation (TTA) aims to mitigate distribution shifts by adapting models with unlabeled target data at inference time. While TTA with vision-language models (VLMs) has shown promising results in classification, extending it to medical image segmentation remains challenging. In this setting, the adaptation gains from optimizing on VLM-generated predictions are often outweighed by the degradation to the VLM’s strong pretrained features caused by noisy, update-driven learning, resulting in limited and unstable improvements. We therefore propose Memory-Supported Synergistic Adaptation (MSSA), a novel training-free TTA framework for medical image segmentation. Without updating model parameters, MSSA dynamically selects reliable image-text predictions to construct an online memory, uses them as text-guided semantic priors, and couples them with cross-image structural alignment for robust adaptation. Specifically, MSSA consists of (i) a noise-aware memory construction module that filters and stabilizes cross-modal predictions, and (ii) a relevance-driven prototype alignment module that aligns the target sample with structurally consistent memory samples and their reliable predictions to improve adaptation. Extensive experiments on multiple medical segmentation benchmarks demonstrate that MSSA consistently improves VLM-based segmentation models and outperforms existing fine-tuning-based TTA methods by a clear margin, with gains of up to 12.2% DSC and 11.7% mIoU. Project page: this https URL .
[CV-49] ShotPlan: Cinematic Video Generation with Learnable Planning Token
链接: https://arxiv.org/abs/2607.17675
作者: Su Guo,Guangce Liu,Haosen Yang,Jiepeng Wang,Cong Liu,Junqi Liu,Haibin Huang,Hongxun Yao,Chi Zhang,Xuelong Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL
Abstract:Current video generation models achieve impressive results in single-shot generation, yet remain limited in cinematic video generation, where coherent narratives and effective multi-shot composition require explicit shot planning. To address this challenge, we propose ShotPlan, a framework for explicit multi-shot cinematic video generation built upon a video diffusion foundation model. Our method introduces learnable planning tokens that capture shot-level transition cues and can be seamlessly integrated with the original video generation tokens to control transition timestamps. Unlike standard video generation tokens, the proposed planning tokens are equipped with Fractional Temporal Rotary Position Embedding (FRoPE), enabling shot transitions to be modeled at the frame level. Experiments demonstrate that ShotPlan significantly outperforms existing cinematic video generation methods, offering more flexible shot management and stronger inter-shot consistency.
[CV-50] Attention from Above: A Multimodal Model for Drone-Based Object Localization
链接: https://arxiv.org/abs/2607.17669
作者: Hyun-Ki Jung
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Preprint. Accepted for publication in the International Journal of Interactive Mobile Technologies
Abstract:Drone-based object detection technology has advanced rapidly, becoming increasingly sophisticated and efficient. Recently, research trends have expanded beyond the detection of predefined objects toward the identification of specified target objects. For example, desired targets can be specified through textual prompts, enabling accurate detection of objects of interest. To address this demand, this paper proposes an efficient multimodal-based object detection model aimed at improving small object detection performance. The proposed method is built upon the YOLO-World framework and replaces the C2f layers used in the YOLOv8 backbone with attention-based A2C2f layers. This modification enables more precise representation of local features, particularly for small objects or objects with well-defined boundaries. In addition, the incorporation of attention mechanisms and parallel processing structures significantly enhances the model’s computational accuracy. Comparative experiments conducted on the VisDrone dataset demonstrate that the proposed model outperforms the original YOLO-World model. Specifically, precision increases from 43.0% to 45.1%, recall from 32.8% to 35.0%, the F1 score from 37.2% to 39.4%, mAP@0.5 from 32.5% to 35.2%, and mAP@0.5-0.95 from 18.5% to 19.9%, confirming a substantial improvement in detection accuracy. These results verify that the proposed approach provides an effective and highly accurate solution for object detection in drone-based image and video application environments.
[CV-51] Early Yield Prediction for Sugar Beet Fields using Satellite Data – Learnings from Specialized Vision Transformers
链接: https://arxiv.org/abs/2607.17661
作者: Philipp Vaeth,Bhumika Laxman Sadbhave,Denise Dejon,Gunther Schorcht,Magda Gregorova
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Remote sensing has become an increasingly valuable tool for agricultural monitoring, particularly through the use of publicly available satellite imagery. However, effectively integrating domain knowledge into machine learning methods remains challenging. This study presents a real-world example of early sugar beet harvest yield forecasting from purely optical Sentinel-2 imagery, demonstrating how a tight integration of domain knowledge and machine learning can lead to synergistic gains. We empirically find that using very small vision transformer patch sizes and all available Sentinel-2 spectral bands improves our model despite being uncommon design choices in the domain. As a practical contribution, we were able to identify a large fraction of low-yield fields in a different year early on in the growth cycle through a modified training setup and a ranking-based detection of underperforming fields.
[CV-52] RayOcc: Occlusion-Aware Ray Occupancy Estimation via Gaussian Mixture Intensity IROS2026
链接: https://arxiv.org/abs/2607.17660
作者: Junho Kim,Seongwon Lee
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to IROS 2026
Abstract:Camera-only 3D semantic occupancy prediction aims to infer voxel-wise scene semantics from multi-view images, yet remains fundamentally challenging due to depth ambiguity and occlusion. Along a single camera ray, multiple spatially separated surfaces may coexist, making occupancy inherently a multi-label existence problem rather than a single-depth estimation task. However, most existing approaches favor a single dominant depth hypothesis per ray, limiting their ability to model volumetric scenes under complex occlusion. To address this limitation, we introduce RayOcc, an occlusion-aware ray occupancy framework that reformulates ray modeling as multi-label existence prediction. Instead of predicting a categorical depth distribution, RayOcc estimates a non-normalized Gaussian mixture intensity along each ray and converts it into interval-wise occupancy probabilities via a Poisson event formulation, allowing multiple occupied hypotheses to coexist without enforcing mutual competition across depth. The predicted mixture components are interpreted as occupancy hypotheses to initialize sparse 3D Gaussian primitives, which are refined and rasterized for semantic occupancy prediction. Experiments on the nuScenes benchmark show that RayOcc achieves state-of-the-art overall IoU and mIoU among the compared Gaussian-based occupancy methods.
[CV-53] OrientSAM: Mitigating Camera-Centric Shortcut in Multimodal Spatial Reasoning via Orientation-Aware Spatial Alignment
链接: https://arxiv.org/abs/2607.17657
作者: Wenxiao Fan,Hang Yin,Kan Li
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注:
Abstract:Multimodal large language models (MLLMs) still struggle with spatial reasoning that requires perspective transformation. In particular, they often rely on camera-centric cues rather than reasoning from the reference object’s viewpoint, leading to systematic errors in non-camera reference settings. In this paper, we first analyze this failure mode and show that object orientation is a key factor underlying such camera-centric shortcut behavior. To address this issue, we propose OrientSAM, an orientation-aware spatial alignment framework for multimodal models. OrientSAM injects explicit orientation information into multimodal representations through orientation-aware tokens and Fourier-based angle encoding, and further adopts a curriculum learning strategy to progressively improve perspective-aware reasoning. In addition, we build a spatial data construction pipeline to generate orientation-aware spatial supervision from large-scale images. Experiments on Spatial-MM, ViewSpatial, and 3DSRBench show that OrientSAM consistently outperforms strong baselines, especially on non-camera-view, person-centric, and orientation-sensitive tasks. The results further demonstrate that explicit orientation modeling is important for mitigating camera-centric shortcut behavior and enabling more robust allocentric spatial reasoning in multimodal models.
[CV-54] BMFA: Boundary-Minority Free-Energy Adaptive Screening
链接: https://arxiv.org/abs/2607.17656
作者: Wenyan Xu,Alizer Wong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages. Experiments include synthetic analysis, COCO, LVIS, closed-loop DeiT-Tiny evaluation, and ImageNet-1K classification
Abstract:Vision Transformers process spatially redundant tokens efficiently only when coarse token summaries preserve the evidence required by exponential attention aggregation. We identify a boundary-minority underestimation failure in which a spatially small, high-response region contributes dominant Gibbs mass while remaining nearly invisible to a block mean. We formalize the failure through the discrepancy between normalized log-mean-exp free energy and mean summarization, prove that minority Gibbs mass can remain non-vanishing as its spatial support and mean contribution vanish, and characterize the limitations of finite-order moment corrections. Building on the resulting analysis, we introduce Boundary-Minority Free-Energy Adaptive Screening (BMFA), which constructs a hierarchical piecewise-constant approximation and recursively refines blocks according to a computable lower-bound increment of local free energy. Controlled synthetic tests, COCO and LVIS diagnostic probes, closed-loop DeiT-Tiny evaluations, and ImageNet-1K experiments establish a consistent evidence chain. BMFA reduces the mean synthetic underestimate from 2.582 to 0.261 at a 5.794% leaf ratio, lowers the COCO image-edge mean gap from 2.254 to 0.526, and preserves 71.520% ImageNet Top-1 accuracy at a 55.861% leaf ratio. The current prototype evaluates selection quality after full QK computation; the reported leaf ratio therefore characterizes representation granularity rather than verified sparse-kernel speedup.
[CV-55] LFM: Leverag ing Foundation Models for Source-Free Universal Domain Adaptation
链接: https://arxiv.org/abs/2607.17653
作者: Jing Li,Pan Liu,Meng Zhao,Wanli Xue,Yanhong Yang,Xu Cheng,Fan Shi,Jianhua Zhang,Qinghua Hu,Shengyong Chen
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multimedia (cs.MM)
备注: Accepted by IEEE Transactions on Multimedia (2026)
Abstract:Source-free universal domain adaptation (SF-UniDA) adapts a pre-trained source model to an unlabeled target domain under both covariate and label shifts, without access to source data. However, existing SF-UniDA methods rely on inefficient techniques such as threshold tuning and clustering. Foundation models (FMs), known for their generalization and zero-shot capabilities, remain underexplored in SF-UniDA. In this paper, we propose a framework that leverages foundation models (LFM) for SF-UniDA. We use a vision-language model (VLM) to compute similarities between target samples and text labels, including those for unknown classes generated by prompting a large language model. The label shift type is determined by analyzing the coefficient of variation of a similarity-based sample-level score. Unknown samples are identified using a binary Gaussian mixture model fitted to another similarity-based metric. Under a consensus strategy, the pseudo-labels generated by the VLM are refined by the target model initialized with the pre-trained source model, integrating knowledge from both the source domain and foundation models. Finally, these refined pseudo-labels are used to train the target model. Extensive experiments across all possible label shifts and multiple benchmarks demonstrate the effectiveness and superiority of our proposed LFM framework. Our code is available at this https URL.
[CV-56] Direct Clinical Joint Angle Extraction from Parametric Body Model Rotation Matrices
链接: https://arxiv.org/abs/2607.17639
作者: Joey Páolo Kardolus,Daan Hendriks,Jaap Jansen
类目: Computer Vision and Pattern Recognition (cs.CV); Quantitative Methods (q-bio.QM)
备注: 31 pages, 9 figures. Code: this https URL
Abstract:Quantitative joint angles are rarely available in routine care because the tools are slow, costly, or confined to a laboratory. We show that clinical joint angles can be read directly from the per-segment rotation matrices a parametric body model already produces, with no inverse-kinematics or musculoskeletal-model fitting step. On the OpenCap LabValidation cohort, using the GEM-X body-model estimator on single-smartphone video, our pooled mean absolute error is 4.50 degrees over the fifteen joint angles that match the OpenCap Monocular reference set, the same accuracy range as OpenCap Monocular’s 4.8 degrees on the same cohort and reference standard, from a much simpler pipeline. The step that connects a body model to clinical angles is a small calibration table rather than an optimisation, so the same procedure transfers unchanged to other body models: repeating it on SAM 3D Body, changing only the table, gives 4.66 degrees, statistically indistinguishable from GEM-X, and runs in real time from a live single-camera stream. The method needs no per-recording inputs beyond the video itself: no participant height, no camera-intrinsics database, no per-subject model scaling. This broadens where movement analysis is practical, from in-clinic and at-home recording to telerehabilitation and large-scale decentralised studies.
[CV-57] Reviving Ancient Paintings via Poem: A Colorization Framework for Aligning Cultural Semantics
链接: https://arxiv.org/abs/2607.17638
作者: Junming Gao,Biao Zhu,Xiaosong Wang,Tan Tang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The irreversible fading of ancient paintings disrupts the “congruence between poems and paintings”, a core aesthetic principle where visual imagery harmonizes with literary inscriptions. Although diffusion models provide strong generative priors, restoring historically faithful colors remains difficult: visual restoration is inherently ambiguous, while direct text guidance often causes modern semantic bias, over-saturation, and cross-boundary color leakage. To address this, we propose PoemColor, a poem-guided ancient painting colorization framework. Our method aligns poetic cultural semantics with painting restoration through two key designs. First, the Poetic Painting Projector (P3) converts implicit poetic context into a classical color-aware condition via poem-to-palette pretraining, reducing the ambiguity of poem-to-color mapping. Second, Structure-Aware Semantic Attention (SASA) regulates how poetic color semantics are injected into the diffusion backbone by jointly controlling their propagation direction and regional injection strength. In addition, we construct a hybrid restoration dataset that integrates synthetic degradation with expert-restored artifacts, providing both scalable supervision and real classical color references. Extensive experiments demonstrate that our framework significantly outperforms state-of-the-art methods, delivering controllable colorization that revives both historical authenticity and poetic semantics.
[CV-58] MixDiffusion: Mixing Diffusion-based Uni-condition Text-to-Image Generation Models for Multi-condition Image Synthesis
链接: https://arxiv.org/abs/2607.17634
作者: Pengcheng Wan,Liang Han,Lin Xu,Bowen Xiao,Liqiang Nie
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 8 figures
Abstract:Recent advances in text-to-image (T2I) generation have enabled controllable image synthesis by incorporating conditions beyond text. However, most existing diffusion-based methods are limited to a single type of control condition (e.g., bounding boxes or keypoints), which restricts their flexibility. To address this limitation, we propose MixDiffusion, a training-free diffusion framework for multi-condition T2I generation. MixDiffusion theoretically supports an arbitrary number of control conditions, including bounding boxes, keypoints, sketches, depth maps, reference images, and text, by collaboratively integrating multiple pre-trained uni-condition diffusion models. The key insight of the proposed approach is to derive the predicted noise distribution in each denoising step of the diffusion-based multi-condition image generation model from the predicted noise distributions of multiple diffusion-based uni-condition models with a derived integration formula, which is supported by rigorous theory proof. Owing to its training-free nature, MixDiffusion is easy to deploy and readily extensible to new control modalities.
[CV-59] Brain-Aligned Multi-Stream Video Transformers with Sparse Self-Selection
链接: https://arxiv.org/abs/2607.17625
作者: Amir Hosein Fadaei,Mahyar Maleki,Mohammad-Reza A. Dehaqani
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 23 pages, 11 figures
Abstract:Modern video transformers typically ignore principles from primate vision and are rarely evaluated against neural data, limiting their biological interpretability. We introduce a sparse winner-takes-all token selection module that replaces dense self-attention to improve efficiency and approximate competitive routing observed in biological visual circuits. We further propose a neuro-inspired split-and-fuse video transformer which uses two complementary pathways: a high-resolution, low-frame-rate “what” stream and a low-resolution, high-frame-rate “where” stream, fused before classification. On Kinetics-400 and Something-Something V2, our best variant operates on the Pareto frontier of accuracy versus inference time among models of comparable scale and pretraining, and showing improved robustness to spatial perturbations. Using representational similarity analysis between model embeddings and time-resolved EEG recordings for the same video stimuli, our model attains a peak brain-model correlation of 0.18 (about 78% of the noise ceiling) and consistently outperforms strong video transformer baselines, suggesting that pathway specialization and sparse competition are useful inductive biases for efficient, brain-aligned video understanding.
[CV-60] Rarity-Aware Discrete Diffusion with Spatially Consistent Decoding for Photo-Realistic Image Super-Resolution
链接: https://arxiv.org/abs/2607.17612
作者: Ao Li,Yapeng Du,Yi Xin,Lei Zhu,Le Zhang,Guangtao Zhai,Ce Zhu,Xiaohong Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 5 tables, 6 figures
Abstract:Continuous diffusion models have become the dominant paradigm for photo-realistic image Super-Resolution (SR), but they typically formulate reconstruction as continuous signal-level denoising and incorporate semantic priors through external conditioning modules. This makes it less direct to exploit the unified token-based scaling paradigm of modern multimodal models. Autoregressive models provide a more native semantic representation by modeling images as discrete visual tokens, yet their causal decoding is inefficient for high-resolution reconstruction. Discrete diffusion offers a promising middle ground by enabling non-causal, parallel prediction over visual tokens. However, directly adapting discrete diffusion to SR remains non-trivial due to two task-specific challenges: (1) the long-tailed distribution of visual tokens, which under-represents rare but perceptually critical textures; and (2) spatially inconsistent parallel decoding, which may introduce isolated artifacts. To address these issues, we propose DiMOO-SR, a rarity-aware multimodal discrete diffusion framework for photo-realistic image SR. During training, Inverse Frequency Sampling (IFS) prioritizes under-represented but information-rich tokens. During inference, Spatial Consistency Ranking (SCR) refines token confidence using local neighborhood agreement to improve structural coherence. Extensive experiments on widely used real-world SR benchmarks demonstrate that DiMOO-SR achieves competitive perceptual quality with only a few parallel decoding steps, highlighting the potential of discrete diffusion for generative image super-resolution. The code will be released upon publication.
[CV-61] Coarse-to-fine Framework for Generative MEF via Implicit Neural Representation
链接: https://arxiv.org/abs/2607.17611
作者: Sangmin Han,Jinho Kim,Jinwoo Kim,Dongyoung Kim,Seon Joo Kim
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-exposure fusion (MEF) expands the luminance range beyond what a single exposure can capture. Combining images taken at different exposure levels requires handling geometric differences while naturally merging their complementary brightness information. It often demands generative completion where details are missing. Diffusion-based generative methods address these challenges, however, they are computationally expensive and struggle to preserve fine structures in saturated regions. We propose LIIFusion, a coarse-to-fine framework that balances fusion quality and efficiency in generative MEF. The coarse stage performs low resolution generative fusion, enhanced by an adaptive exposure correction that recovers structure lost in saturated over-exposed areas. The fine stage adapts a local implicit image function into a multi-exposure fusion function: conditioned on the HR OE/UE sources and the coarse output, it queries arbitrary target coordinates and fuses source evidence regardless of the HR input resolution. LIIFusion achieves up to 3.5 \times speed-up over existing generative methods while maintaining or improving structural fidelity and perceptual quality. We believe this framework provides an effective pathway toward making generative MEF more practical in real-world applications.
[CV-62] Semantic Color Naturalness Breaker: Preventing Illegitimate Colorization via Content-Aware Color Priors
链接: https://arxiv.org/abs/2607.17610
作者: Yuki Nii,Futa Waseda,Ching-Chun Chang,Isao Echizen
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: SMC 2026 Accepted
Abstract:Automatic image colorization enables large-scale and low-cost reuse of grayscale media (e.g., manga panels and archival photographs), facilitating unauthorized reuse and redistribution. Once released online, grayscale content can be readily turned into unauthorized colorized derivatives using off-the-shelf models, creating a practical need for proactive, content-side protection at publication time. Building on Uncolorable Examples (UE), which add imperceptible perturbations to released grayscale images to degrade unauthorized colorization, we propose Semantic Color Naturalness Breaker (SCNB) – a semantic-level UE framework that drives colorization outputs toward content-inconsistent colors while preserving the visual fidelity of the released grayscale media. We further introduce Content-aware Color Distributional Distance (CaCDD), a ground-truth-free, content-aware measure of color plausibility derived from semantic color priors, used both as the optimization objective of SCNB and as an evaluation metric. Experiments on ImageNet show that our method remains effective under small perturbation budgets and common post-processing, supporting practical deployment in real-world content-sharing pipelines.
[CV-63] ConsiSpace: Learning Geometric Consistency Matters for Video Spatial Reasoning ECCV2026
链接: https://arxiv.org/abs/2607.17599
作者: Ting Huang,Zhenyu Zhang,Wenyuan Huang,Jian Yang,Hao Tang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:Video spatial reasoning is essential for navigation-oriented perception and long-video question answering, where models must infer spatial relations across long horizons under changing viewpoints. However, existing multimodal large language models (MLLMs) remain largely semantic-centric, and often fail to reliably aggregate consistent spatial evidence from redundant video observations, leading to inefficient or unstable reasoning. To address these issues, we propose ConsiSpace, a geometry-consistency-aware framework for geometry-sensitive video spatial reasoning that turns spatial consistency into both an evidence organization principle and an explicit post-SFT learning signal. We build a geometry-consistent memory (GCM) including implicit evidence tokens and explicit geometric cues, and leverage efficient organization strategies to compactly preserve task-related spatial evidence. Furthermore, we utilize unified consistency self-supervised reinforcement learning (UC-SSRL) after supervised fine-tuning to improve cross-view stability, with answer-, metric-, and topology-consistency rewards. Extensive experiments on three spatial-reasoning benchmarks, VSI-Bench, OSI-Bench, and MMSI-Video-Bench, show consistent gains, improving the average score by 12.6 points over the strongest baselines.
[CV-64] Miles: Metric Learning with Expandable Subspace for Pre-Trained Model-Based Class-Incremental Learning
链接: https://arxiv.org/abs/2607.17593
作者: Kai Jiang,Zisong Lin,Hongyuan Zhang,Xueru Bai,Xuelong Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: This work has been accepted by IEEE Transactions on Image Processing
Abstract:Class Incremental Learning (CIL) aims to learn new concepts consistently from a data stream without forgetting. Unlike typical CIL methods which need to learn a model from scratch, pre-trained model (PTM) can easily adapt to a new task with fine-tuning. However, existing PTM-based CIL methods fail to achieve a trade-off between performance and computational expenditure, i.e., they either adopt the same parameter space so that leading catastrophic forgetting, or expand a new branch for each task but adding more computational cost. To this end, we propose MetrIc Learning with Expandable Subspace (Miles) to harness the prior information within pre-trained knowledge, thereby orchestrating an efficient expansion of the parameter space through guided optimization. Specifically, it decouples the learnable modules with the pre-trained model, exploiting prior information from intermediate features of the backbone network to enable more flexible parameter expansion. Then, a central loss is adopted to guide the new category to cluster towards the corresponding prototype in the new task subspace while incorporating an auxiliary distance regularization term to maintain metric equilibrium across tasks. Extensive experiments on six benchmark datasets demonstrate that Miles achieves state-of-the-art performance in various CIL settings.
[CV-65] Pixel-Space Diffusion Transformers
链接: https://arxiv.org/abs/2607.17585
作者: Renye Yan,Jikang Cheng,You Wu,Ling Liang,Wei Peng,Athanasios V. Vasilakos,Qingyu Zhao,Yu Zhang,Ehsan Adeli,Kilian M. Pohl,Guoying Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Latent diffusion models (LDMs) enable efficient high-resolution image synthesis by denoising in a VAE-compressed latent space. However, fixed visual tokenizers can discard fine textures and structural details, while separate representation and diffusion training creates a mismatch between reconstruction and generation objectives. These limitations have renewed interest in pixel-space diffusion, which models raw pixels directly, removes the VAE bottleneck, and supports end-to-end optimization. This formulation better matches the demands of high-fidelity generation but introduces challenges in high-dimensional modeling, including noise scheduling, loss weighting, token efficiency, and scalable architecture design. Pixel-space modeling also offers a promising basis for unified multimodal systems: raw pixels, text, and task conditions can be represented in a shared token space and jointly processed by a single Transformer, narrowing the gap between visual understanding and generation. This paper reviews Pixel-Space Diffusion Transformers (pDiTs) from the perspectives of model architecture, continuous generative mechanisms, and unified multimodal modeling. We summarize representative methods, identify key technical challenges, and discuss future directions toward high-fidelity, end-to-end vision foundation models that integrate generation and understanding.
[CV-66] Scalable Model-Assisted Multi-Target Estimation in Large Image Collections UAI2026
链接: https://arxiv.org/abs/2607.17581
作者: Max Hamilton,Jinlin Lai,Daniel Sheldon,Subhransu Maji
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 42nd Conference on Uncertainty in Artificial Intelligence (UAI 2026)
Abstract:Computer vision models are increasingly used as measurement tools to estimate population-level quantities from large image collections, but prediction errors introduce bias and the resulting estimates lack statistical guarantees required in scientific applications. Prior work uses a Monte Carlo framework to combine model predictions with ground-truth annotations by sampling some images for humans to label and is able to provide unbiased estimates with controllable accuracy, but primarily addresses single-scalar estimation. We study the more general problem of multi-target estimation, where many quantities (e.g., class counts or proportions) must be estimated simultaneously, and adapt sampling and estimation strategies from survey sampling to this setting. Evaluations on five detection and segmentation datasets with 7-80 classes show that importance sampling excels with moderate annotation budgets or fewer targets, whereas uniform sampling with control variates is superior when estimating many targets or operating with minimal labels. Additionally, a subset-based ratio estimator remains highly competitive across all regimes. Ultimately, our framework effectively combines biased model predictions and limited human labels into rigorous scientific measurements.
[CV-67] AGG: Jacobian-Aggregated Group Gradient for Efficient GRPO Training of Diffusion Models
链接: https://arxiv.org/abs/2607.17572
作者: Ruiyi Ding,Jie Li,He Kang,Ziyan Liu,Chengru Song,Yuan chen
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Systems and Control (eess.SY)
备注: 21 pages
Abstract:Group Relative Policy Optimization (GRPO) is a powerful reinforcement learning algorithm for aligning generative models with human preferences. While successful in large language models~\citeshao2024deepseekmathpushinglimitsmathematical, its extension to diffusion and flow matching models introduces a severe computational bottleneck: gradients must be back-propagated through the high-capacity DiT backbone at \emphevery timestep of the sampling trajectory, making high-resolution text-to-image (T2I) training prohibitively expensive. Training-free DiT inference acceleration methods (e.g., \Delta -DiT, ScalingCache) exploit the fact that DiT hidden states and velocity predictions vary \emphsmoothly and nearly linearly along the trajectory. We ask whether the same linearity can reduce the backward-pass cost of DiT RL training, and answer affirmatively with \textbfJAGG (\textbfJacobian-\textbfAggregated \textbfGroup \textbfGradient), which reduces full transformer backward passes from W to 2 per group of W consecutive steps. JAGG approximates intermediate-step Jacobians via t -weighted interpolation of the endpoint Jacobians, then aggregates per-step upstream signals into two composite gradients applied through a single joint backward pass. We prove this interpolation is \emphexact when the velocity is linear in (z,t) , and a cosine-similarity routing rule (\textttjagg_frac) deploys JAGG only where the assumption holds. Experiments on T2I benchmarks show JAGG delivers \sim 2 \times backward speedup with negligible quality degradation.
[CV-68] FlexiGrad: Adaptive Gradient Modulation for Hierarchical Fine-Grained Classification
链接: https://arxiv.org/abs/2607.17563
作者: Zilu Zhou,Dongliang Chang,Junhan Chen,Zhanyu Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Many fine-grained recognition tasks contain hierarchical labels such as order, family and species. Although this supervision should be beneficial, jointly optimising all levels often leads to unstable training because coarse and fine classifiers impose inconsistent gradients on the shared backbone. This hierarchical gradient conflict prevents the model from learning a coherent coarse-to-fine representation. In this paper, we propose FlexiGrad, a simple and parameter-free method that regulates gradient interactions during backpropagation. FlexiGrad removes only the harmful conflicting component when tasks disagree and reinforces the shared direction when they partially agree through a smooth hierarchy-aware weighting function. This produces stable optimisation and preserves both global structure and fine-grained discriminative cues. FlexiGrad integrates into existing architectures without modification while improves multi-granularity accuracy on CUB-200-2011, FGVC-Aircraft and Stanford Cars. The code will be available at PRIS-CV/FlexiGrad.
[CV-69] Hierarchy-Aware and Anatomy-Guided Learning for Lung Ultrasound Video Classification
链接: https://arxiv.org/abs/2607.17551
作者: Alya Almsouti,Lotfi Mecharbat,Noha Aboukhater,Yousef Alabrach,Siddiq Anwar,Andre Kumar,Ibrahim Almakky,Mohammad Yaqub
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Lung ultrasound (LUS) is a bedside tool for assessing pulmonary edema in patients at risk due to heart failure or impaired kidney function. However, automated LUS analysis remains challenging because of speckle noise, imaging artifacts, and operator-dependent acquisition variability. In this work, we present a deep learning framework for multi-class LUS video classification that explores two components: hierarchy-aware training, and anatomy-guided learning. Starting from a strong baseline, we introduce hierarchical training strategies and then introduce pleural line mask supervision to guide model attention toward anatomically relevant regions. We study four clinically relevant classes–healthy, B-lines, consolidations, and mixed B-lines with consolidations–using an open-access dataset of 1,886 videos from 219 patients, evaluated with patient-level five-fold cross-validation. Results show that hierarchy-aware training improves pathological separation relative to flat classification, while mask-guided attention supervision achieves the highest mean macro-F1 of 65.7% and produces more localized attention patterns. Transfer experiments on the external COVID-BLUeS dataset further show competitive and parameter-efficient adaptation while preserving pleural-focused attention behavior. These findings suggest that combining clinically structured objectives with anatomy-guided supervision is a practical approach to robust, interpretable LUS video analysis. Code and model implementations are available at this https URL.
[CV-70] DecoyFace: Beyond Obfuscation via Controllable and Imperceptible Identity Misdirection for Privacy-Preserving Face Recognition
链接: https://arxiv.org/abs/2607.17504
作者: Zhihan Ren,Lijun He,Xinyao Wang,Xinzhu Fu,Fan Li
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Split face recognition reduces client-side computation but exposes intermediate features to feature inversion attacks and unauthorized analysis by honest-but-curious (HBC) servers. Existing privacy-preserving face recognition methods mainly aim to resist unauthorized reconstruction, typically producing features whose inversion yields visibly degraded results, which may reveal the existence of protection and motivate adaptive attacks. To address this issue, we propose DecoyFace, an imperceptible decoy-oriented framework that steers unauthorized reconstruction toward a plausible but incorrect identity while preserving recognition utility. The key idea is to decompose the intermediate representation into a reconstruction-sensitive subspace and its complementary subspace. The client injects decoy identity cues into the reconstruction-sensitive subspace, while limited recognition-relevant evidence from the true sample is retained in the complementary subspace. On the server side, an authorized canonicalization module suppresses decoy-dominant components and recovers a recognition-friendly representation. This design addresses both attacker-side inversion from intercepted features and HBC server-side reconstruction from canonicalized representations. Experiments show that DecoyFace preserves competitive recognition accuracy while substantially reducing identity leakage to 2.93% under U-Net attacks and 0.74% under Flow-Matching attacks while yielding visually plausible and imperceptible reconstructions, with over 99.78% face validity on LFW dataset.
[CV-71] Generative Transmission: Rethinking Computation Bandwidth and Memory in Communication
链接: https://arxiv.org/abs/2607.17482
作者: Xiangyu Chen,Jixiang Luo,Yuankai Fan,Haibin Huang,Chi Zhang,Xuelong Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Submitted to WAICA on 30 April, accepted on 8 July, and awarded Best Paper on 18 July
Abstract:Under the AI Flow framework, communication is shifting from transmitting fidelity-oriented information flows toward delivering task-oriented and perception-oriented token flows across heterogeneous network resources. Video communication is a fundamental component of modern information networks. However, under ultra-low-bandwidth and weak-network conditions, conventional video coding and transmission methods, which are primarily optimized for pixel-level fidelity, often struggle to balance visual usability, transmission efficiency, and robustness to unstable links. With the rapid advancement of generativemodels, video communication is also moving from precise signal reconstruction toward receiver-side perceptual utility and system-level usability. In this paper, we propose Generative Transmission (GenTrans) for video communication under ultra-low-bandwidth and weak-network conditions. Built upon Generative Video Compression (GVC), GenTrans formulates video transmission as a joint optimization problem involving bandwidth, computation, and memory, rather than treating it merely as a signal coding task. By leveraging generative priors, cross-clip memory reuse, runtime state reuse, and weak-network-aware transport, GenTrans significantly reduces transmission overhead while enabling visually coherent and practically useful reconstruction. Experimental results show that GenTrans supports effective video transmission under ultra-low-bitrate and weak-network conditions, achieving improved transmission efficiency, decoding efficiency, and robustness while preserving perceptual quality.
[CV-72] raversRL: Traversable Pedestrian Pathway Generation With Reinforcement Learning ECCV2026
链接: https://arxiv.org/abs/2607.17479
作者: Bin Han,Robert Wolfe,Bill Howe
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ECCV 2026, main conference
Abstract:Automatically generating pedestrian pathways from aerial images requires producing a connected network suitable for routing, not just detecting where sidewalks appear. Sidewalks and crossings, in contrast to roads, may be partially occluded, implicitly defined, and exhibit complex connectivity patterns. Existing segmentation-based approaches focus on labeling pixels to infer segments, but often produce disconnected or fragmentary graphs that are unreliable for navigation. We introduce TraversRL, a vision-conditioned model that iteratively grows a pathway network from an aerial image, simulating a traveler navigating the built environment. TraversRL uses an action space of short and long direction-distance segments designed to adapt to complex patterns and span occlusions, and uses a combination of graph-level and step-wise rewards to balance overall connectivity with precise edge placement. Across three visual backbones and three intersection datasets, TraversRL substantially improves buffered IoU with the ground-truth graph relative to a state-of-the-art segmentation baseline, and more than doubles metrics of connectivity. Moreover, combining global and local rewards produces cleaner graphs with fewer spurious branches while further improving overall performance. These results demonstrate that modeling pathway extraction as a sequential decision process from the perspective of a traveler, while optimizing for final graph quality with reinforcement learning, produces significantly more reliable pedestrian networks.
[CV-73] DA-MergeLoRA: Hypernetwork-Based LoRA Merging for Few-Shot Test-Time Domain Adaptation ECCV2026
链接: https://arxiv.org/abs/2607.17467
作者: Siobhan Reid,Zhixiang Chi,Li Gu,Omid Reza Heidari,Ziqiang Wang,Yang Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: ECCV 2026, Code: this https URL
Abstract:Few-shot Test-Time Domain Adaptation (FSTT-DA) seeks to adapt models to novel domains using only a handful of unlabeled target samples. This setting is more realistic than typical domain adaptation setups, which assume access to target data during source training. However, prior FSTT-DA approaches fail to effectively leverage source domain-specific knowledge, relying on shallow batch normalization updates, prompt-based methods that treat the model as a black box, or ensembling strategies that do not capture cross-domain relationships. To address these limitations, we introduce a new FSTT-DA framework that integrates LoRA fine-tuning with model merging. In our approach, separate LoRA modules are fine-tuned on CLIP’s vision encoder for each source domain. Since LoRA modifies only a small fraction of the model’s parameters, it retains the base model’s generalized knowledge while internally learning domain-specific features. To adapt the learned knowledge to a specific target domain, we propose a hypernetwork trained via meta-learning that generates per-column merging factors to combine LoRA modules. Given a small batch of target images, the hypernetwork produces merging weights that fuse source LoRA modules into a single adapted representation. Our results demonstrate state-of-the-art performance across various domain adaptation datasets. Our code is publicly available at this https URL.
[CV-74] Bio-SFT: Asymmetric Cortical Guidance and Retinal Adaptation for Robust HDR Reconstruction
链接: https://arxiv.org/abs/2607.17456
作者: Tingyu Cheng,Ting Zhang,Chongyi Li,Zhaoqing Pan,Tiesong Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recovering high dynamic range (HDR) radiance from a single standard dynamic range (SDR) image is highly ill-posed. Extreme luminance variation and severe quantization in dark regions make accurate reconstruction challenging, often leading to visual artifacts and color distortions. To address this problem, we propose Bio-SFT, a bio-inspired spiking frequency transformer for single-image HDR reconstruction. Bio-SFT incorporates three biologically motivated components. First, a learnable Naka–Rushton retinal adaptation frontend stabilizes the input under complex lighting conditions. Second, an explicit Parvo–Magno split introduces asymmetric Parvo-to-Magno guidance, allowing high-frequency structural cues to modulate low-frequency reconstruction. Third, an event-driven SNN hard gating module applies all-or-none spiking to suppress dark-region noise while preserving structural details. The module is trained with a sparsity prior to encourage efficient feature utilization. Built for end-to-end training within a transformer backbone, these lightweight components provide strong parameter efficiency. Experiments on HDRTV1K show that Bio-SFT achieves competitive perceptual quality and consistently improves HDR-VDP-3 and \Delta E_ITP while reducing artifact propagation in symmetric guidance pipelines.
[CV-75] An Explainable FFT-Based Spatial-Frequency Fusion Framework for Deepfake Detection
链接: https://arxiv.org/abs/2607.17441
作者: Pamela Kirui,Cho Hyuk,Qingzhong Liu,Haodi Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Deepfake generation has raised growing concerns regarding digital media authenticity, misinformation, identity fraud, and public trust. Recent studies show that combining spatial and frequency features leads to stronger detection results than using independently. This paper presents MSCA-FFT, a Fast Fourier Transform (FFT)-based multi-scale cross-attention framework for image-level deepfake detection. The model combines a partially fine-tuned Xception spatial branch with an FFT-based frequency branch. The frequency branch processes the log-scaled FFT magnitude spectrum through shallow convolutional layers, avoiding inverse frequency-to-image reconstruction used in DCT-based pipelines. The spatial and frequency representations are refined by transformer encoders, fused through cross-attention, and passed to an MLP classifier for real/fake prediction. Experimental results show that MSCA-FFT achieves consistently higher performance than the DCT-based state-of-the-art spatial-frequency fusion method and the compared baseline models. The ablation study further indicates that the FFT-based frequency branch provides complementary spectral cues when fused with spatial features. In addition, FFT-based frequency analysis and Grad-CAM/LIME explanations show consistent evidence around manipulation-sensitive facial regions, including the eyes, mouth, nose, and facial boundaries.
[CV-76] meLens2: Generalist Video Temporal Grounding with Multimodal LLM s
链接: https://arxiv.org/abs/2607.17423
作者: Yuhan Zhu,Changlian Ma,Xiangyu Zeng,Xinhao Li,Zhiqiu Zhang,Songze Li,Jun Zhang,Tianxiang Jiang,Yuandong Yang,Ziang Yan,Zikang Wang,Xinyu Chen,Haoran Chen,Shaowei Zhang,Limin Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Technical Report
Abstract:Video multimodal large language models (MLLMs) can describe what happens in a video, but rarely identify when the supporting evidence occurs. We study generalist video temporal grounding, in which one model predicts a variable-cardinality set of evidence intervals across video lengths, domains, query forms, and viewpoints. Existing training strategies are misaligned with this set-valued task: long-video labels often rely on brittle one-pass annotation, while reinforcement-learning rewards either fail to distinguish non-overlapping predictions or require fragile segment matching. TimeLens2 treats temporal evidence as an interval set throughout supervision and optimization. TimeLens2-93K constructs reliable multi-span supervision through caption-derived proposals, independent localization, cross-agent consensus, semantic verification, and boundary refinement. Our temporal Wasserstein reward computes exact one-dimensional (W_1) between uniform distributions over merged interval supports, providing dense, matching-free feedback under unequal cardinalities and equivalent fragmentation; temporal IoU complements it with precise-overlap feedback. Across seven benchmarks, TimeLens2-2B outperforms all size-matched baselines on every benchmark, while the 4B and 8B variants achieve state-of-the-art performance, surpassing open-source models with up to 397B parameters. The 2B, 4B, and 8B variants improve over their Qwen3-VL backbones by 14.2, 13.0, and 18.1 mIoU points, respectively.
[CV-77] Does Super-Resolution Preserve Defect Evidence? A Low-False-Call Benchmark for Semiconductor Inspection
链接: https://arxiv.org/abs/2607.17401
作者: Shaoliang Yang,Jun Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 27 pages, 12 figures
Abstract:Super-resolution can make inspection images appear sharper without preserving the evidence needed to detect a defect. We study this failure mode with a benchmark that separates reconstruction from detection and evaluates both at a predeclared low false-positive rate. Ten end-to-end repetitions combine independently generated line/space and contact-hole images with model training, calibration, clean controls, weak defects, and a held-out defect morphology. Every reconstruction is scored by the same local residual detector, while direct and jointly trained detectors form a separate comparison track. Reconstruction fidelity and inspection utility diverge: the two learned reconstruction models attain the highest structural similarity yet detect fewer defect pixels than bicubic interpolation in every paired repetition. A direct DeepLabV3 detector reaches 0.1984\pm0.0385 recall at 0.000174\pm0.000084 false-positive rate and satisfies the held-out feasibility criterion in all ten repetitions. An illustrative joint model, DPU-WaferSR, passes independent clean calibration but exceeds the held-out limit in all ten repetitions, demonstrating that calibration success does not guarantee transfer. Weak-defect recall remains near zero for every feasible method. Applying the unchanged policies to 4,591 public Carinthia-S masks further reveals large method-dependent shifts on real SEM texture. These results support a simple conclusion: super-resolution for inspection should be judged by preserved task evidence and operating-point transfer, not reconstruction quality alone.
[CV-78] he PanAf-SBR Dataset: Social Behaviour Recognition for Wild Great Apes
链接: https://arxiv.org/abs/2607.17399
作者: Maciej Braszczok,Otto Brookes,Xiaoxuan Ma,Federico Rossano,Yixin Zhu,Mimi Arandjelovic,Hjalmar Kühl,Majid Mirmehdi,Tilo Burghardt
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Behavioural shifts in wild great ape populations, particularly the breakdown of social structures, can serve as an early indicator of population decline. Automating the detection of behaviours indicative of these shifts is therefore a critical task for conservation. Several valuable datasets have recently been introduced for the automated recognition of great ape behaviour, yet few include fine-grained social behaviour annotations, and those that do are captured either in captive settings or via aerial platforms such as UAVs. We address this gap by introducing PanAf-SBR, the first wild great ape camera trap dataset annotated with social behaviours. PanAf-SBR extends PanAf500 with 100 additional videos covering 36,063 frames. These come with 81,096 annotations including bounding boxes, segmentation masks, intra-video identities, and seven social behaviour classes defined under the action giver and receiver convention of ChimpACT. We use this data together with the AlphaChimp architecture to establish the first benchmarks for fine-grained social behaviour recognition in wild great apes from camera trap footage. We further conduct bidirectional transfer learning experiments between PanAf-SBR and the captive ChimpACT dataset, finding that cross-dataset pre-training is highly beneficial for specific classes rather than of uniform benefit. Finally, we examine the role of background context by inverting the segmentation masks to suppress non-ape pixels.
[CV-79] Automating Visual Recognition of Leprosy in Wild Chimpanzees
链接: https://arxiv.org/abs/2607.17395
作者: Katie I. Murray,Anna C. Bowland,Marina Ramon,Elena Bersacola,Aissa Regalla,Manmohan D. Sharma,Markus Mueller,Majid Mirmehdi,Dave Hodgson,Kimberley J. Hockings,Tilo Burghardt,Otto Brookes
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Leprosy (Mycobacterium leprae) has been confirmed in wild western chimpanzees (Pan troglodytes verus) in West Africa, presenting as clear and progressive visual symptoms. Manual review of camera-trap footage at landscape scale is infeasible, motivating the need for automated screening. We present the first deep learning pipeline for wildlife leprosy detection and contribute the PanLep300 dataset of 125,670 annotated bounding-box crops across 953 tracks from 303 camera-trap videos with ecologically-motivated splits that withhold whole individuals and camera installations. We benchmark spatial (2D), temporally aggregated (2.5D), and video-based (3D) classification approaches to investigate which approach is best suited to automated leprosy detection in wild apes. We find that simple aggregation of crop-level predictions consistently matches or outperforms both learned temporal models and end-to-end video architectures – consistent with leprosy’s static cutaneous presentation. We further find that performance is suppressed when tracklets contain frames of partially visible individuals – as commonly occurs at the start and end of a track – and demonstrate that this can be addressed through targeted construction and aggregation strategies.
[CV-80] SkyVLaM: Multimodal Large Language Model for UAV Video Understanding in Remote Sensing
链接: https://arxiv.org/abs/2607.17386
作者: Kaiwen Jing,Ruixu Jia,Bingyao Li,Ruizhe Ou,Ming Wu,Chuang Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by WAICA 2026
Abstract:Recent advances in Multimodal Large Language Models (MLLMs) have significantly improved remote sensing (RS) multimodal understanding. Language-conditioned segmentation is crucial for fine-grained target understanding in Unmanned Aerial Vehicle (UAV) videos. However, this task remains challenging due to the prevalence of small, visually ambiguous targets and dynamic aerial perspectives. In this paper, we propose SkyVLaM, a multimodal large language model for UAV video understanding. SkyVLaM constructs sparse tokens directly from patch-level video representations through a temporal basis perceiver, regularizes the sparse basis to encourage complementary temporal cues, and adaptively selects a temporally coherent dense segment for high-resolution inspection. The resulting sparse and dense tokens are jointly processed by a large language model for query-conditioned segmentation. We further build SkyVid, consisting of SkyVid-VGCG and SkyVid-RVOS for video grounded conversation generation and referring video object segmentation, respectively. SkyVid contains 101 videos, 33.6K frames, and 1.53M pixel-level object instances. Experiments show that SkyVLaM provides a more effective allocation of the visual token budget and improves language-conditioned video segmentation in UAV scenarios.
[CV-81] STAR: Skeletal Token Alignment and Rearrangement for Interaction Recognition
链接: https://arxiv.org/abs/2607.17342
作者: Yuhang Wen,Mengyuan Liu,Zixuan Tang,Junsong Yuan,Sirui Li,Beichen Ding
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted for publication in IEEE Transactions on Multimedia (IEEE TMM)
Abstract:Understanding physical human-robot and human-human interactions is a challenging yet emerging topic in 3D vision. While most existing methods rely on skeleton sequences–effective in low-light and privacy-sensitive environment–they face two major challenges: 1) learning and effectively exploiting interaction cues from skeletal data, and 2) compensating for the lack of visual information absent in skeletons alone. To address these challenges, we propose skeletal token alignment and rearrangement (STAR) for human-robot and human-human interaction recognition. It learns interaction-specific skeleton features and enriches them using visual cues by aligning skeleton and RGB video representations in a shared latent space. Specifically, STAR consists of three key components. First, we design a skeleton encoder that captures fine-grained interdependencies using Entity Rearrangement (ER) and Interactive Spatiotemporal Tokens (ISTs). Second, we present Visual Interaction Encoding that introduces a Focus on Interactions (FoI) strategy to attend to spatiotemporal regions relevant to interactions in RGB videos. Finally, these representations are aligned via a contrastive learning objective, with a refinement head further refines predictions. During training, STAR leverages both skeleton and RGB video data to learn robust, discriminative interaction representations. At inference time, it operates on skeletons alone, retaining visual-informed benefits while preserving skeleton-only efficiency. Extensive experiments on Chico, HARPER, NTU Mutual 11 and 26 datasets consistently validate our approach by demonstrating superior performance over state-of-the-art methods. Our code is publicly available at this https URL.
[CV-82] Understanding From Human Perspective: A Multi-agent System for Interactive Egocentric Medical Image Segmentation
链接: https://arxiv.org/abs/2607.17341
作者: Rongjun Ge,Dongyang Wang,Heng Zhu,Zhirui Li,Yang Chen,Yuting He
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Interactive egocentric medical image segmentation (IEMIS) plays an important role in smart-glasses-assisted medical image review, segmenting the medical targets a clinician refers to from their egocentric view. Once it succeeds, the object-level visual evidence it provides strengthens the review and underpins fine-grained analysis and clinical decision-making. However, the instruction and the video both come from the user’s egocentric perspective, which poses two challenges. (1) Semantic ambiguity leaves the model unable to confirm the user-intended target. (2) Visual variability makes the segmentation jump from frame to frame. In this paper, we propose EgoMed-Agent, a multi-agent system that understands the target from the human perspective through two workflows. (1) The \textitTarget Confirmation Workflow grounds the instruction against candidate targets with a reliability score, confirming the target when the grounding is reliable and asking the user to clarify when it is not, thereby confirming the segmentation target. (2) The \textitLocalization-Guided Propagation Workflow couples mask propagation with per-frame target localization, using the localized target to correct the propagated mask whenever the two diverge, so the segmentation stays on the target across the egocentric video. Extensive experiments show that EgoMed-Agent reaches 71.34% average Dice, far above the best text-prompted baseline (11.70%). Our code is available at \hrefthis https URLour project page.
[CV-83] Orthogonal Knowledge Refreshing for Domain-Incremental Object Detection ECCV2026
链接: https://arxiv.org/abs/2607.17340
作者: Aoting Zhang,Dongbao Yang,Chang Liu,Xiaopeng Hong,Can Ma,Yu Zhou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ECCV 2026
Abstract:Domain-incremental object detection (DIOD) requires models to continually adapt to new domains while preserving prior knowledge. Recently, parameter-efficient fine-tuning offers a promising avenue, wherein a pre-trained model is frozen and a small number of learnable parameters are injected for downstream tasks. However, these methods risk overwriting critical past knowledge, triggering inter-domain interference and performance degradation. To address this challenge, we propose Orthogonal Knowledge Refreshing (OKR), a simple yet effective framework for DIOD. OKR incrementally constructs independent domain-specific subspaces via dedicated low-rank branches for each domain, which are seamlessly fused for a holistic decision, enabling conflict-free capacity expansion without domain selection during inference. To minimize knowledge interference during fusion, we present a gradient-based orthogonal refreshing strategy that projects gradient updates of new domains onto the orthogonal complement of the fused historical subspace, supporting continual adaptation without forgetting. Moreover, to mitigate semantic fragmentation across domains, we enforce topology-aware consistency, aligning the semantic structures of old and new domains. Extensive experiments validate the superiority of OKR, outperforming the best exemplar-free method by significant margins of +5.6% and +6.5% mAP on the Pascal VOC and BDD100K series, respectively.
[CV-84] MIS-HCC: Hierarchical Channel Clustering for Efficient Medical Image Segmentation
链接: https://arxiv.org/abs/2607.17329
作者: Bo Zhao,Haoran Yu,Lifei Liu,Zongcheng Chu,Yining Liu,Chang Liu,Szu-Yu Chen,Zequn Xie
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Medical image segmentation models require both high accuracy and lightweight design to accommodate real-world medical applications. The deployment of these models on resource-limited medical platforms remains a significant challenge due to their high computational and parameter requirements. Existing pruning methods for model compression mostly overlook the intrinsic connections and similarity between the internal structures of complex deep neural networks. As a result, compressed models may not effectively retain the basic features of the pretrained network. To solve this problem, we propose a hierarchical clustering compression method for medical image segmentation models (MIS-HCC). This approach employs hierarchical clustering to partition channels and fuse their parameters efficiently. Specifically, it leverages the Wasserstein distance to represent similarity of channels within layers of pre-trained network, forming a similarity matrix that guides the clustering process. Channels within each cluster are then fused to produce a compressed network. Experimental results on three medical image datasets application demonstrate that MIS-HCC outperforms the state-of-the-art methods in both accuracy and compression efficiency, offering an effective solution for deploying medical image segmentation models on resource-limited medical platforms.
[CV-85] PACE: Polar Axis-Conditioned Estimation for PairUAV Relative Localization
链接: https://arxiv.org/abs/2607.17268
作者: Ze Rong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:PairUAV relative localization maps two UAV images to a polar navigation command. Although heading and range share the same pairwise pose context, treating them as homogeneous coordinates forces both outputs to use the same decoder evidence and optimization state. Controlled readout probes reveal a different structure: the two axes favor different decoder-depth combinations, their best checkpoints disagree on 80.8% of a validation trajectory, and range errors exhibit a distinct high-error tail. We introduce method, Polar Axis-Conditioned Estimation, which retains a shared Reloc3r-style pair representation while assigning axis-specific readout interfaces. Heading uses mid/late relational evidence, whereas range remains attached to a direct late metric path. On the official hidden test, the strongest released raw predictor scores 0.002460; the complementary PAAER predictor scores 0.002514 with a slightly lower angle error. Deterministic challenge packaging, reported separately from learned estimation, yields the final score of 0.001874. Code, checkpoints, predictions, and reconstruction tools are available at this https URL.
[CV-86] VecFontLLM : Anchor-Guided Direct Synthesis of Chinese Vector Fonts
链接: https://arxiv.org/abs/2607.17251
作者: Hao Yuan,Yuxuan Luo,Xing Chen,Zhouhui Lian
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages
Abstract:Direct generation of Chinese vector fonts is a challenging and ongoing problem. A Chinese vector glyph contains complex component structure, anchor layout, and Bézier curve details, which work at different scales, but a standard vector sequence writes them together in one long sequence, making the task of vector font synthesis challenging. Existing direct vector generators often fail on complex characters, while raster-domain methods must vectorize the synthesized glyph images afterward. To address the above-mentioned problem, this paper proposes VecFontLLM, an anchor-guided multimodal large language model for direct few-shot synthesis of Chinese vector fonts. Our key idea is to generate vector glyphs through anchors rather than a standard vector sequence. Specifically, the proposed VecFontLLM first predicts and refines an anchor scaffold that fixes the coarse layout of components and contours, and then completes Bézier control points to recover local curvature and style. At test time, a confidence-guided generation chain samples multiple component candidates and continues synthesis from the highest-confidence one, improving stability for complex glyphs. This work demonstrates, for the first time, high-quality few-shot synthesis of complex Chinese vector glyphs directly in the vector domain, without raster generation or vectorization. Experiments on several Chinese font datasets show substantial improvements over existing vector font synthesis methods, competitive glyph rendering quality against raster-domain baselines, and vector command distributions close to real fonts.
[CV-87] Induce to Empower: Improving Lightweight Baselines via Foundation Model Induction for Generalized Polyp Segmentation
链接: https://arxiv.org/abs/2607.17208
作者: Shivanshu Agnihotri,Snehashis Majhi,Deepak Ranjan Nayak,Dwarikanath Mahapatra,Debesh Jha
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Automated polyp segmentation in colonoscopy continues to pose challenges due to substantial appearance variations and indistinct polyp boundaries. Although emerging foundation models (FMs) such as DINOv2, SAM, and OneFormer, demonstrate remarkable generalization capabilities, their direct transfer to the polyp segmentation task and deployment in real-time clinical settings are difficult due to lack of large-scale labeled data and high computational demands. In addition, adopting multiple FMs together raises concerns, even though they encode complementary semantic and structural information. While lightweight models, including U-Net, PraNet and U-Net++, are computationally efficient, they often struggle to generalize across datasets due to limited representational capacity. To address this gap, we propose Lite-Polyp Inductor (Lite-Pi), a novel foundation model induction framework that significantly enhances lightweight polyp segmentation baselines. Our proposed framework generates FM-specific prototype representations and aligns them semantically with the corresponding foundation model priors through reconstruction-based supervision. Subsequently, transformer-based fusion is introduced to highlight the polyp relevant representations, including salient boundary information, while preserving complementary semantic cues. Extensive experiments across five polyp segmentation benchmark datasets demonstrate that Lite-\pi significantly improves lightweight baselines, achieving superior generalization performance with minimal computational overhead and thereby, offering a practical solution for generalized polyp segmentation. Our code is available at GitHub. this https URL
[CV-88] Cross-Coordinate Correspondence Pruning for Image-to-Point Cloud Registration
链接: https://arxiv.org/abs/2607.17200
作者: Xin Liu,Rong Qin,Huipeng Lin,Leizhi Shu,Jin Wu,Chi-Man Vong,Liang Lin,Jufeng Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Recent detection-free approaches have shown significant efficacy in image-to-point cloud (I2P) registration by employing a coarse-to-fine matching pipeline. In the coarse stage, down-sampled image features and voxelized point cloud features are typically fused to establish initial coarse correspondences for subsequent refinement. However, existing methods largely overlook the critical role of point cloud density, which fundamentally dictates the quality of coarse correspondences and the final registration results. Specifically, excessively sparse point clouds lead to an insufficient number of inliers, while overly dense ones often introduce a high outlier ratio. Consequently, this creates an inherent density trade-off, thereby significantly limiting the registration accuracy of current approaches. For mitigating this trade-off, we propose a novel Cross-Coordinate Correspondences Pruning (CCP) strategy to acquire sufficient inliers while ensuring a low outlier ratio. To minimize interference from inter-modal coordinate discrepancies, we first project cross-coordinate coarse correspondences to the 2D image coordinate system for spatial unification. Subsequently, a lightweight pruning network is responsible for predicting the inlier confidences, which are used to filter coarse outliers, from coordinate geometric and modal feature dimensions. To maximize inlier recall, we further design a Multi-Density Point Ensemble (MDPE) strategy that consolidates and deduplicates pruned coarse correspondences across varying point cloud densities. Our method achieves a significant performance improvement, surpassing existing state-of-the-art methods by at least 8.6% in Registration Recall across various benchmarks.
[CV-89] BanClickThumb: A Multimodal Dataset and Transformer Fusion Benchmarks for Clickbait Detection in Bengali YouTube Videos
链接: https://arxiv.org/abs/2607.17182
作者: Md. Ariful Islam,Md Tanvirul Islam,Md. Maruf Hossain Miru,Md Khalid Syfullah
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Clickbait, where video titles and thumbnails exaggerate or misrepresent content, reduces user trust, wastes attention, and promotes misinformation on video-sharing platforms. Detecting Bengali clickbait remains challenging because publicly available multimodal datasets are limited. To address this gap, we introduce BanClickThumb, a curated dataset of 7,147 Bengali YouTube thumbnail-title pairs from five content domains, annotated by ten annotators with high agreement (Cohen’s Kappa: 0.83-0.93). Using this dataset, we benchmark text-only, image-only, and multimodal approaches. Among unimodal models, BanClickTextFormer (XLM-RoBERTa) achieves 0.82 accuracy, while BanClickImageFormer (SwiftFormer) reaches 0.68. Our proposed multimodal model, BanClickFusionFormer, combines ViT and XLM-RoBERTa through intermediate fusion and achieves the best accuracy of 0.84. Error analysis shows that dense thumbnail text, figurative language, and culturally specific slang remain challenging. Our findings demonstrate the effectiveness of multimodal fusion for Bengali clickbait detection and provide a publicly available benchmark to support future research on low-resource multimodal content analysis.
[CV-90] VLA-ReID: Video-Level Association for Re-Identification in Multi-Object Tracking with Highly Similar Objects
链接: https://arxiv.org/abs/2607.17157
作者: Yanrong Qin,Xiaoyan Cao,Yao Yao
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 11 pages, 8 figures, 2 tables
Abstract:Multi-object tracking (MOT) aims to localize multiple objects in videos while preserving their identities over time. Long-term identity preservation remains difficult when objects are small, densely distributed, and highly similar in appearance, as in bee swarm scenes. Existing trackers rely on re-identification (re-ID) models trained through single-instance assignment (instance-level querying). At inference, however, MOT requires global assignment between multiple trajectories and detections, corresponding to video-level querying. This training-inference mismatch can cause identity switches among visually similar objects. Existing approaches also often require substantial additional annotations to enhance appearance discrimination. We propose Video-Level Association re-ID (VLA-ReID), which reformulates re-ID as video-level association modeling. It uses aggregated historical trajectory features as queries and all current-frame detections as candidates, enabling direct optimization of their global association at each frame. In addition, Frame-Common Appearance Estimation (FCAE) estimates a common appearance direction from current-frame detections, while Common-Appearance Suppression (CAS) removes the corresponding component along this direction from trajectory and detection features. This amplifies discriminative differences among highly similar objects without additional annotations. Experiments on BEE24 show that VLA-ReID improves HOTA by 1.1, MOTA by 0.3, AssR by 2.6, AssA by 0.7, and IDF1 by 0.8 over state-of-the-art trackers, while reducing identity switches by 28%. These results demonstrate the effectiveness of video-level re-ID modeling for appearance-based association in MOT.
[CV-91] Noise-Robust Box-Supervised Infrared Small Target Detection via Physics-Inspired Soft Label Optimization
链接: https://arxiv.org/abs/2607.17148
作者: Xizhe Zhang,Fan Shi,Mianzhao Wang,Jiangpeng Zheng,Xu Cheng,Shengyong Chen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 24 pages, 8 figures
Abstract:Infrared small target detection (IRSTD) commonly relies on pixel-level mask supervision. Such annotations, however, are costly and inherently uncertain because infrared targets have blurred boundaries and weak textures. We formulate box-supervised IRSTD as a problem distinct from generic box-to-mask segmentation and point-supervised IRSTD. Its central challenge is to construct stable pixel-level soft supervision from highly contaminated boxes. To this end, we propose Hotspot-Anchored Label Optimization (HALO). HALO localizes a radiometric anchor inside each box under local background-statistics constraints, then synthesizes a Physically Anchored Gaussian (PAG) soft label around the anchor. This turns noisy box supervision into continuous, pixel-level soft labels. The entire process is performed offline before training, remains decoupled from the detector backbone, and requires no online label updates. Experiments on public datasets show that HALO is competitive with representative box-supervised methods under standard tight boxes. Under looser or shifted box annotations that better approximate real scenarios, HALO is substantially more robust while remaining consistent across backbones. We further introduce a contamination-aware operating-regime analysis to characterize the effective boundary of this class of methods and reveal how intrinsic signal-to-clutter ratio relates to performance.
[CV-92] STBridge: Shared-Target Alignment for Bridging Understanding and Generation in UMMs
链接: https://arxiv.org/abs/2607.17140
作者: Ye Wang,Hongjun Wang,Hao Fang,Tongyuan Bai,Zuwei Long,Peixian Chen,Wei Liu,Weibo Gu,Xing Sun,Rui Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Unified multimodal models (UMMs) aim to integrate visual understanding and generation within a single architecture, but architectural unification alone does not ensure semantic consistency. A model may describe the intended target correctly while generating an inconsistent edit. This exposes an understanding-generation alignment gap: linguistic and visual outputs live in different spaces, yet should be governed by the same target semantics. We study this gap in image editing, where an instruction defines a target state that can be both described and visually realized. Given a source image and an edit instruction, we compare a UMM’s target caption with its edited image to test whether the two outputs converge on the same result. Our analysis shows that existing UMMs remain weakly aligned, especially for fine-grained entities, attributes, spatial relations, and local details, indicating that semantic unification is not achieved by architecture alone. To bridge this gap, we propose STBridge, a shared-target alignment framework that connects understanding and generation through a common target state. Here the target caption expresses the desired visual result, while the edited image realizes it visually, replacing separate task-specific paths with a shared information flow from target expression to target realization. STBridge follows an align-then-optimize strategy: supervised fine-tuning first establishes the shared-target channel, and sequential reinforcement learning further refines target-centered coordination. Across visual understanding, image generation, and image editing benchmarks, STBridge consistently improves over the initialization model. Alignment analysis confirms that STBridge narrows the gap between what the model describes and what it generates, demonstrating shared-target alignment as an effective post-training strategy for bridging understanding and generation in UMMs.
[CV-93] Denoising Models Develop Human-Like Perceptual Illusion Representations Across Architectures
链接: https://arxiv.org/abs/2607.17138
作者: Gautam Ranka,Paras Chopra
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Deep neural networks trained on natural images are shown to produce outputs consistent with human observers for brightness illusions. While this phenomenon has been documented across architectures, all evidence, to date, is measured at the output level: restored pixels, decoded trajectories, or classification decisions. Whether these models actually represent illusions internally, and if so where and how, remains unknown. We show that denoising models develop illusion-sensitive representations at specific internal layers, across varied architectures. Specifically, we identify the layers and channels that discriminate illusory from physically matched control regions. We show that the denoising objective is a more important driver of the effect than the architecture. On domain-appropriate stimuli, these activations track a validated psychophysical model of human brightness perception (FLODOG; Spearman \rho \geq 0.70 ) and scale monotonically with parametric illusion strength. Leveraging these findings, we provide causal evidence via channel ablation showing that illusion-sensitive channels specifically and substantially affect the internal signal. Yet injecting these representations into the generation pipeline produces no measurable pixel shift across all tested architectures; we term such representations perceptual phantoms: active in internal processing yet invisible to any output-based evaluation. While related internal-output dissociations have been characterized in language models, this is the first such characterization for perceptual representations in denoising vision models.
[CV-94] Learning Emotion from Motion: Kinetic Multi-Stream Skeleton Modeling with Metadata-Conditioned Weak Label Distributions
链接: https://arxiv.org/abs/2607.17121
作者: Sosuke Suzuki,Yijin Wei,Koichiro Kamide,Ran Dong,Haoran Xie,Chao Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ACII2026 workshop
Abstract:Skeleton-based emotion recognition from body motion remains challenging because emotional expressions are often characterized by subtle dynamic and relational motion cues, and hard labels may not fully capture ambiguity among related emotion categories. For the DIEM-A task in the MMAC ACII 2026 Challenge, we propose a multi-branch skeleton-based emotion recognition framework that combines a 6D rotation-based branch, a part-aware kinetic multi-stream branch, and a metadata-conditioned weak label distribution learning (LDL) branch. The branches are trained independently and fused by a probability-level ensemble at inference time. In 10-fold leave-performer-out cross-validation, the proposed framework improves Accuracy from 0.271 to 0.366 and Macro-F1 from 0.252 to 0.353 over the rotation-based baseline. Explainability ablations show that velocity and bone streams, as well as arm and leg regions, provide important cues for recognizing emotional body motion.
[CV-95] he generator is the tracker: Multi-object tracking by painting persistent identity colours
链接: https://arxiv.org/abs/2607.17120
作者: Haiyu Yang,Miel Hostens
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Multi-object tracking (MOT) is conventionally decomposed into detection followed by association, with object identity maintained as external state: track buffers, motion models, and appearance embeddings. We ask whether a video generator can maintain that state in pixels. We fine-tune a 22B text-to-video diffusion model (LTX-2.3) with a lightweight in-context LoRA to translate an RGB clip into an ID-map clip, a video in which every person is painted a flat, distinct color that persists over time: same color, same identity. Long videos are generated as chained windows, where each window is conditioned on the cleaned tail of the previous one. A brief continuation fine-tune teaches the model to extend a given coloring, after which identity flows through the chain with no tracker, no motion model, and no re-identification module. On the DanceTrack test server, our system, to our knowledge the first generative tracker evaluated there and the only entry with no detector and no tracking stack, reaches 40.3 HOTA. This is well below today’s specialist state of the art (=70 HOTA), but with a unique, inverted error profile: its association score (AssA 44.1) exceeds every tracker of the original benchmark suite while detection remains the sole deficit. Controlled comparisons show the mechanism matters: the same generated windows linked by classical post-hoc association score 2x worse (18.2 HOTA), and frame-to-frame IoU association fragments tracks that the generator’s colors keep whole. On 383 mined occlusion events, the generator re-acquires identities after gaps at a 42% conditional rate where appearance-embedding baselines score zero, including gaps longer than its temporal context, evidence that the generator’s color assignment functions as an emergent re-identification signal. We release code, checkpoints, and the full pre-registered experimental log.
[CV-96] DepthART: Scaling Foundation Monocular Depth to Tiny Models
链接: https://arxiv.org/abs/2607.17099
作者: Feng Xue,Wu Chen,Mingshuai Zhao,Guofeng Zhong,Anlong Ming,Haozhe Wang,Dianqiao Lei,Zhaowen Lin,Haiyang Zhang,Nicu Sebe
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted in ACM Multimedia 2026 ; Code: this https URL ; Project page: this https URL ;
Abstract:Recent geometric foundation models (e.g., Metric3D, Depth Anything and UniDepth) have substantially improved monocular depth estimation (MDE) in both cross-scene generalization and metric-scale prediction, yet these gains have not translated to tiny models. We bridge this gap with DepthART (Depth Anything Rethought for Tiny Models), which is a compact MDE model for on-device deployment across diverse scenes. We first identify two capacity-driven bottlenecks in tiny models: (i) overfitting to dataset-specific distribution bias and (ii) unstable metric adaptation under camera shift, where full fine-tuning easily damages transferable geometry. Accordingly, DepthART combines two simple but effective strategies: a bias-resistant data sampling scheme to reduce distribution bias under the same training budget, and a camera-conditioned fine-tuning protocol that freezes the distilled encoder and adjusts metric scale conditioned on intrinsics while better preserving cross-dataset generalization. Across datasets, DepthART consistently surpasses previous tiny baselines in both zero-shot generalization and metric accuracy (e.g., zero-shot \delta_1 =0.964 for DepthART-S on NYUD v2), and in some cases approaches heavy models. We further provide a scalable model family, with DepthART-S reaching 347/245 FPS (strict FP32) on an RTX A6000 at 224^2/448^2 , 102 FPS (TF32) on a Orin NX 8GB, and over 15 FPS (FP32) on a Jetson Nano 4GB.
[CV-97] HarmoHOI: Harmonizing Appearance and 3D Motion for Multi-view Hand-Object Interaction Synthesis
链接: https://arxiv.org/abs/2607.17097
作者: Lingwei Dang,Juntong Li,Zonghan Li,Hongwen Zhang,Liang An,Wei Min,Yebin Liu,Qingyao Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Hand-Object Interaction (HOI) synthesis is a cornerstone for animation production and embodied AI. Despite the strong priors of video foundation models, multi-view consistent HOI synthesis remains challenging due to complex hand motions and occlusions. We present HarmoHOI, a unified diffusion framework that jointly and harmoniously generates synchronized multi-view HOI videos and globally aligned 3D point tracks. Our core insight is that robust multi-view consistency fundamentally requires globally aligned 3D geometry and motion. To this end, we propose a Mixture of Multi-view Diffusion Transformer that co-models RGB videos and 3D point tracks. By representing point tracks as pseudo-videos, we align 3D geometric signals with the 2D latent space of foundation models, thereby minimizing the domain gap and easing adaptation of priors. To further ensure geometry consistency, we introduce Global Motion Aligning Diffusion, which refines coarse point tracks into metric-scale, globally aligned 3D trajectories. HarmoHOI enables on-the-fly co-evolution of 2D appearance and 3D motion during denoising. To overcome the scarcity of multi-view HOI data, we employ a hybrid data curriculum learning strategy that successfully transfers generic priors from single-view data to synchronized multi-view generation. Experimental results show that HarmoHOI achieves state-of-the-art performance in visual quality, motion plausibility, and multi-view geometric consistency. Project page available at this https URL.
[CV-98] Autoregressive B-Rep Shape Generation with Parametric Surfaces
链接: https://arxiv.org/abs/2607.17093
作者: Dafei Qin,Rui Xu,Zeyu Shen,Kaichun Qiao,Hongyang Lin,Qixuan Zhang,Huaijin Pi,Lan Xu,Jingyi Yu,Wenping Wang,Taku Komura
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Generative CAD modeling has broad design and application potential. Despite significant advances in Boundary Representation (B-Rep) generation, the dominant representation in CAD, existing methods largely depend on uniformly sampled point- or grid-based geometry representations, sacrificing native surface types and parameters and thereby limiting geometric fidelity and downstream usability. We present ParaCAD, an autoregressive framework for point-cloud-conditioned B-Rep generation that directly operates on native parametric surfaces. ParaCAD introduces a surface-centric tokenization that explicitly encodes each face by its exact surface type and continuous parameters, preserving the intrinsic semantics of CAD geometry. Our model first generates parametric surfaces with constrained UV domains, and then constructs a valid B-Rep by globally intersecting these surfaces to recover edges and vertices. ParaCAD places point-cloud-conditioned generation at the core of B-Rep synthesis, making it practical for user-guided reconstruction and seamless integration into existing 3D generation pipelines. Extensive experiments demonstrate that ParaCAD produces accurate B-Reps with faithful point-cloud alignment, outperforming point-based baselines in geometric precision, robustness, watertightness and downstream usability.
[CV-99] ALLUDE: A Unified Evaluation System for Configurable Attacks in Differentiable Environments
链接: https://arxiv.org/abs/2607.17077
作者: Mansi Phute,Alexander Greenhalgh,Matthew Hull,Haoran Wang,Alec Helbling,ShengYun Peng,Elliott Faa,Willian Lunardi,Martin Andreoni,Wenke Lee,Duen Horng Chau
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Adversarial attacks against vision models like object detectors are often evaluated under limited conditions, leaving their performance under-characterized. Bridging simulation and differentiable rendering enables more robust, end-to-end evaluation of these adversarial attacks, yet there is no easy-to-use, unified system that offers a rich set of customizable configurations for adversarial attacks across multiple scenes, objects, environmental and lighting conditions, and camera trajectories. We present ALLUDE, which addresses these gaps, offering first-of-its-kind evaluation capabilities across Linux and Windows. We comprehensively demonstrate ALLUDE’s evaluation breadth through a two-pronged strategy: (1) using Latin Hypercube Sampling, we draw a representative subset from 5,400 configurations spanning 10 scene-object pairs, 9 weather conditions, 4 optimizers, 5 camera trajectories, and 3 detection models; (2) we stress-test existing attacks (CAMOU, RAUCA, FCA) under diverse weather conditions and continuous camera trajectories, revealing degradation of attack success across every attack, exposing evaluation gaps in prior work. Through ALLUDE’s end-to-end differentiable rendering, adversarial attacks can be optimized against shifting real-world deployment conditions. Our cross-platform code is open source.
[CV-100] AdvSerial: Physical Adversarial Attacks on Infrastructure-mounted Pedestrian Detectors via Semantic Feature Suppression
链接: https://arxiv.org/abs/2607.17069
作者: Yuanhao Huang,Yilong Ren,Jinlei Wang,Xuesong Bai,Jinchuan Zhang,Haiyang Yu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:AI-based visual perception systems are increasingly deployed in infrastructure surveillance, including roadside monitoring units, highway cameras, and smart-city pedestrian management systems. The security vulnerability of these systems to physical adversarial attacks poses a direct threat to the reliable operation of transportation infrastructure. We propose AdvSerial, a dynamic 2D–3D joint optimization framework for generating continuous high-angle physical adversarial patches against pedestrian detectors in infrastructure-based scenarios. We UV-map a boundary-aware quilted texture onto 3D garments, combine 2D digital attacks with 3D sparse- and continuous-frame rendering, and explicitly suppress person-specific semantic features while enforcing temporal continuity. A Feature Smooth Quilting strategy reduces visible patch boundaries and bounds cross-seam feature discontinuities. A serial-frame loss encourages long uninterrupted sequences of detection failures. In physical world experiments, AdvSerial achieves a 74.8% attack success rate on YOLO-v5 and degrades mean detection confidence from 84.30% to 39.38%. Experiments spanning eight detectors with different architectures demonstrate strong transferability. Notably, it achieves an 89.71% attack success rate on YOLO-v2 and resists both patch-detection defenses (NapGuard) and 3D-temporal perception (Sparse4D-v3). The results reveal persistent, temporally consistent failure modes under high-angle surveillance, and motivate the design of motion-aware and 3D-aware defenses for security-critical infrastructure deployments.
[CV-101] DROID-ANCHOR: Odometry-Anchored Recurrent Metric Depth Estimation
链接: https://arxiv.org/abs/2607.17058
作者: Yuxuan Chen,Brook Du
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Precise metric depth estimation is fundamental for autonomous robot navigation, yet monocular systems inherently suffer from scale ambiguity and scale drift. While recent recurrent flow-based SLAM systems have demonstrated state-of-the-art robustness, they remain scale-ambiguous. In this paper, we propose Metric-DROID, an end-to-end recurrent architecture that anchors visual SLAM to physical reality by integrating proprioceptive odometry. Our framework introduces the following innovations: (1) A LSTM Update Operator that encodes high-frequency odometry sequences into spatial feature maps, providing a persistent metric bias for iterative refinement. (2) An Uncertainty-Aware Metric Backend ( BA_odom ) that treats odometry as a geometric anchor with learned heteroscedastic covariance. By regressing a time-varying metric uncertainty \Sigma_o , our system intelligently balances visual re-projection and metric translation residuals, effectively mitigating the impact of wheel-slip and sensor noise. (3) We further propose a selective residual fine-tuning strategy to preserve pre-trained geometric priors while enabling zero-shot metric alignment.
[CV-102] High-Capacity Robust Watermarking Technology for High-Resolution Images
链接: https://arxiv.org/abs/2607.17056
作者: Shaowu Wu,Wei Lu,Qing Qian,Mingsen Deng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Most existing watermarking techniques are primarily designed for low-resolution images, with few methods tailored for high-resolution images. Moreover, the embedding capacity is often limited to fixed lengths (e.g., 30, 100, 256 bits, etc.), which struggles to meet practical demands. To address these issues, this paper proposes a high-capacity robust watermarking method for high-resolution images, capable of embedding a watermark of 4 KB (32,768 bits) into images with a resolution of 1024*1024, achieving an embedding rate of 0.0313 bpp. Specifically, this paper adopts a block-wise strategy to effectively embed the watermark into local regions, enabling the network to train and learn normally even under low-resource conditions. The encoder and decoder structures respectively employ a reversible symmetric architecture with three convolutional and three deconvolutional layers, ensuring consistency in the coupling and decoupling of the watermark and image features. Additionally, the loss function combines global and local losses with weighted contributions. By incorporating constraints on the visual quality and robustness of local block regions, the overall imperceptibility and robustness of the image are further enhanced. Extensive experimental results verify that the proposed method is effective and feasible in high-resolution image scenarios with high-capacity watermarking, while demonstrating strong robustness against various noise attacks.
[CV-103] Searching for Task-Specific Vision Paths: Evolutionary Block Pruning Across Vision-Language Models
链接: https://arxiv.org/abs/2607.17052
作者: Tarun Tomar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 14 pages, 8 figures. Code and aggregate evidence: this https URL
Abstract:Vision-language models normally execute the same complete vision encoder for every question, even when OCR, counting, object, attribute, and spatial queries may not require identical computation. We study whether fixed-budget combinations of vision blocks can be skipped without fine-tuning. A shared K-block route skips one searched set of exactly K blocks for every question, while a capability-specific K-block policy selects one same-size route using a known capability label. We introduce a source-balanced evolutionary search and compare it with independent ranking, contiguous removal, and random routes at matched budgets. Experiments use Qwen2.5-VL-3B-Instruct, SmolVLM2-2.2B-Instruct, and an 876-example image-disjoint selection split. Search transfers across architectures: on SmolVLM2, the searched shared four-block route beats independent construction by 4.91 percentage points. Capability specialization is less stable. On Qwen, the six-block capability policy beats the shared route by 2.17 points, driven by a 7.10-point OCR gain. On sealed IIIT5K, however, the SmolVLM2 OCR-specific route trails its shared route by 13.6 points. Combinatorial search reliably improves route construction, but capability labels do not define universally transferable vision pathways.
[CV-104] EvoGUI: An Evolution-Aware Benchmark for GUI State-Transition Understanding
链接: https://arxiv.org/abs/2607.17050
作者: Yaohan Yang,Minglei Shi,Borui Zhang,Jie Zhou,Jiwen Lu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:GUI agents must reason about how actions transform interface states, but end-to-end success rates entangle this ability with perception, grounding, planning, and recovery. We introduce EvoGUI, a diagnostic framework that converts normalized GUI trajectories into three complementary visual question answering probes: temporal ordering, inverse action/value prediction, and contrastive one-step successor discrimination. Their labels are derived from trajectory order and logged actions, requiring no additional task-label annotation after trajectory normalization. We instantiate EvoGUI-Bench from Mind2Web and WebLINX, yielding 3,000 instances across 120 domains, and evaluate 28 vision-language model configurations zero-shot. The strongest model reaches only 60.4 EvoGain, while model scale and GUI specialization do not reliably predict performance. These results establish EvoGUI-Bench as a scalable diagnostic complement to end-to-end GUI-agent evaluation while exposing substantial headroom in state-transition understanding. The source code is publicly available at this https URL.
[CV-105] Articulated Humanoid Head for a Robot Receptionist Capable of Natural Human Interaction
链接: https://arxiv.org/abs/2607.17042
作者: Tharusha Fonseka,Charuka Bandara,Moshintha Hewavitharana,Melisa Arukgoda,Wageesha N. Manamperi,Udaya S. K. Perera Miriya Thanthrige,Peshala Jayasekara
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: This work is accepted at IEEE/ASME International Conference on Advanced Intelligent Mechatronics (AIM2026)
Abstract:Humanoid robots have become increasingly popular in applications such as social interaction, education, and service roles, which drives the need for more natural and efficient human-robot interactions. However, currently available humanoid heads often face limitations, including high costs, mechanical complexity, and limited adaptability across diverse environments. To address these challenges, we present an articulated humanoid robot head designed for a receptionist role, integrating a mechanical structure with 21 degrees of freedom (DoF), including mechanisms for the mouth, eyes, eyebrows, and neck, and covered with realistic silicone skin to achieve a human-like appearance and expression. The system integrates a model-based architecture that combines SCRFD, ArcFace, and ByTetrack for face recognition and Llama and Whisper for natural language processing, with hardware support enabling real-time operations and human re-identification. The conversational ability and re-identification capabilities of the humanoid robot head were quantitatively measured, while its emotional expressiveness and human likeness were evaluated through a user study, achieving an average human likeness score of 4.13 out of 5.
[CV-106] Automated Cardiac Adipose Tissue Segmentation in Computed Tomography: A Literature Review
链接: https://arxiv.org/abs/2607.16992
作者: Andreas W. Aspe,Jonas Jalili Pedersen,Andreas Ohrt Johansen,Klaus Fuglsang Kofoed,Kristine Aavild Sørensen,Rasmus Reinhold Paulsen,Josefine Vilsbøll Sundgaard
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: This preprint has not undergone peer review (when applicable) or any post-submission improvements or corrections. The Version of Record of this contribution is published in Image Analysis - 23rd Scandinavian Conference, SCIA 2025, Proceedings (Lecture Notes in Computer Science, vol. 15726), and is available online at this https URL
Abstract:This review provides an overview of recent advancements in automated segmentation methods on Computed Tomography (CT) for two types of cardiac fat: Epicardial adipose Tissue (EAT) and Pericardial Adipose Tissue (PAT). These fat deposits, separated by the pericardium, have been linked to various cardiovascular diseases, with EAT receiving the most research attention. Their complex anatomical context makes manual quantification highly time-consuming and prone to considerable inter-observer variability. Automated methods effectively address these complications, offering a more efficient and consistent solution. This study encompasses a broad range of methods, spanning AI as well as non-AI approaches. Additionally, it presents the remaining challenges, including the need for larger annotated public datasets and optimized attenuation thresholds for contrast-enhanced CT. It is demonstrated that automated methods are able to achieve segmentation results comparable to the quality of human annotation, proving their potential as a clinical tool for discovering new biomarkers and enhancing patient outcomes.
[CV-107] When Physical Preferences Meet Semantic Constraints: Physical and Semantic Direct Preference Optimization for Text-to-Video Generation ACM-MM2026
链接: https://arxiv.org/abs/2607.16947
作者: Siwei Meng,Yawei Luo,Shu Zhang,Ping Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: This work is accepted by ACM MM 2026
Abstract:Text-to-video (T2V) generation models have achieved strong visual realism, but improving physical plausibility can come at the cost of semantic consistency with the input text. This tension arises because physical preference is typically determined by comparing dynamics between two videos, without accounting for whether either video faithfully depicts the scene specified by the prompt, making physical-semantic conflict a systematic tendency under this supervision paradigm. We formulate this challenge as a constrained preference optimization problem and propose Physical and Semantic Direct Preference Optimization (PSDPO), which modulates each preference pair’s contribution based on the agreement between its physical and semantic signals. A gradient-level analysis shows that PSDPO bounds the semantic drift from conflicting pairs to a controllable residual, and further motivates a staged optimization protocol that provably reduces cumulative drift. The resulting method operates entirely within the standard DPO framework, requiring no auxiliary models or additional loss terms. Experiments show that PSDPO improves physical plausibility by up to 2\times over the baseline on VideoPhy-2, while maintaining strong semantic consistency on VBench, achieving a more reliable balance than existing preference-based methods.
[CV-108] Points as Tori: Fast Pointwise Signed Distance for Point Clouds
链接: https://arxiv.org/abs/2607.16946
作者: Nicole Feng,Ioannis Gkioulekas,Keenan Crane
类目: Graphics (cs.GR); Computer Vision and Pattern Recognition (cs.CV)
备注: Published in ACM Transactions on Graphics, volume 45 (2026). For associated presentations and code, see this https URL
Abstract:We describe a method for computing signed distance to point clouds that allows fast pointwise evaluation at arbitrary spatial resolution. As input, our method takes a point cloud with normals; as output, it provides an analytical parameterization that allows queries of signed distance to the approximate underlying surface at arbitrary points - simultaneously providing reconstruction and distance. Our key idea is to reconstruct shapes by locally fitting point clouds with tori, which have closed-form signed distance functions. Tori are fitted in a feed-forward manner, using a pre-trained network to output per-point curvature and shift parameters. Importantly, our method does not require costly global optimization or spatial discretization, and is easily parallelizable. Underlying our method is a new theory that unifies signed distance with the classic reconstruction methods of winding numbers and Poisson surface reconstruction. We use our method to compute signed distance to point clouds arising from photogrammetry, meshes, 3D Gaussians, and neural implicits. Our method allows point clouds to be used directly in applications, without explicit surface reconstruction: as examples, we take offsets of point clouds, apply morphological and Boolean operations, and directly visualize offset surfaces using sphere tracing.
[CV-109] Investigation of Polycystic Ovary Syndrome (PCOS) Diagnosis Using Machine Learning Approaches
链接: https://arxiv.org/abs/2607.16941
作者: Al Zadid Sultan Bin Habib,Md Asif Bin Syed,Md. Ekramul Islam,Tanpia Tasnim
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: Published in 2023 5th International Conference on Sustainable Technologies for Industry 5.0 (STI), December 2023, Dhaka, Bangladesh
Abstract:Polycystic Ovarian Syndrome (PCOS) is a widespread hormone problem for women of childbearing age. Women with PCOS may not ovulate; they might have high levels of androgens and have many small cysts on the ovaries. It can cause missed or irregular menstrual periods, excess hair growth, acne, infertility, and weight gain. Machine Learning (ML) can effectively diagnose this disease at an earlier stage as tons of medical data are available now. Traditional approaches to detect PCOS encompass a combination of clinical evaluation, medical history assessment, physical examination, and laboratory tests. These approaches aim to identify the characteristic symptoms and hormonal imbalances associated with PCOS. Physical examination requires good resources and costs time and money. In recent times, data-driven techniques have substantially advanced disease prediction within the medical field. We aim to utilize ML approaches, incorporating unique feature selection algorithms, to predict PCOS. This paper introduces a data-driven approach to PCOS diagnosis, combining Feature Engineering and ML. Several feature selection approaches have been considered to select sets of features for training the ML model, including CatBoost, Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LGBM), AdaBoost, Random Forest (RF). Results demonstrate that AdaBoost, with ten features selected by RF Feature Importance and Highest Correlation (HC), provides the highest test accuracy.
[CV-110] What Do They See? Interpreting Complex Road Scenarios Through the Eyes of Vision-Language-Action Models for Safe and Trustworthy Autonomous Vehicle Learning
链接: https://arxiv.org/abs/2607.16938
作者: Kalpana Panda,Wesley Maia,Vinti Agarwal,Ross Greer
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Robotics (cs.RO)
备注:
Abstract:End-to-end autonomous driving models are now able to navigate complex road scenarios, mapping raw sensor observations directly to observed paths for open-loop evaluation and often effective driving in closed-loop evaluation. Yet the internal logic of these safety-critical systems remains largely opaque, due to the complexity of traffic scenes. We propose a counterfactual ablation framework called Counterfactual Vision Action Analysis (CVAA) that systematically removes individual detected objects from front-camera images using photorealistic generative inpainting to prepare counterfactual sets to evaluate the difference in the model’s response. This isolates the causal effect of each object’s presence on the model’s planning behaviour. Applied to the Alpamayo 1 trajectory predictor across 210 nuScenes driving scenes, we create a dataset Counter -nuScenes, using which we see that vehicles and pedestrians within the model’s ‘path’ dominate causal influence as expected, while traffic lights, as expected, exert disproportionate effect relative to their image footprint. However, we also find cases where the model responds strongly to objects a human driver would consider irrelevant. This brings forth a deeper question: does the model itself view the scene as a sum of individual objects influencing the outcome, or does it encode an entirely different set of internal features that do not correspond to human-legible scene elements? To further understand this, we compare intermediate representations of original and inpainted image pairs using mechanistic interpretability techniques and examine the effect of the removal through the various model layers. Together, these two stages offer a path from behavioral auditing to representational understanding, creating explainable driving systems and solidifying human-AI trust.
[CV-111] Pediatric Bone Age Prediction Using Deep Learning
链接: https://arxiv.org/abs/2607.16936
作者: Al Zadid Sultan Bin Habib,Md. Ekramul Islam,Md Asif Bin Syed,Md Younus Ahamed,Tanpia Tasnim
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Published in 2023 26th International Conference on Computer and Information Technology (ICCIT), IEEE, December 2023
Abstract:Pediatric bone age prediction is a crucial task in clinical practice that can help diagnose endocrine disorders and provide insight into a child’s growth and development. However, conventional bone age prediction methods are often labor-intensive and require specialized radiological expertise. This paper presents a Deep Learning (DL)-based approach to pediatric bone age prediction using EfficientNet with Additive Attention, a state-of-the-art neural network architecture for image classification and regression tasks. The method utilizes over 12,000 X-ray images from the RSNA bone age dataset. It involves image preprocessing, transforming them into three-channel images, and training a Convolutional Neural Network (CNN) to automatically learn the features of hand bone images. This approach provides a more effective and accurate solution for predicting bone age, which is critical in diagnosing pediatric endocrine diseases. This work uses two variations of the EfficientNet model (B0 and B4), where EfficientNetB4 is also finetuned with the Additive Attention mechanism. These three models predict the age for the original age, and their comparison is shown in curves. The predicted ages depict that in most cases, EfficientNetB4 and EfficientNetB4 with Additive Attention (EN-AA) successfully predicted the bone ages more accurately regarding the original age, and their performance was better than the EfficientNetB0. Specific performance metrics are provided to underscore this improvement. Learning curves for training and validation loss confirm effective learning without overfitting or underfitting, further validating our approach’s efficacy in pediatric endocrine disease diagnosis.
[CV-112] Splat-based 3D Scene Reconstruction with Extreme Motion-blur
链接: https://arxiv.org/abs/2607.16926
作者: Hyeonjoong Jang,Dongyoung Choi,Donggun Kim,Woohyun Kang,Min H. Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We propose a splat-based 3D scene reconstruction method from RGB-D input that effectively handles extreme motion blur, a frequent challenge in low-light environments. Under dim illumination, RGB frames often suffer from severe motion blur due to extended exposure times, causing traditional camera pose estimation methods, such as COLMAP, to fail. This results in inaccurate camera pose and blurry color input, compromising the quality of 3D reconstructions. Although recent 3D reconstruction techniques like Neural Radiance Fields and Gaussian Splatting have demonstrated impressive results, they rely on accurate camera trajectory estimation, which becomes challenging under fast motion or poor lighting conditions. Furthermore, rapid camera movement and the limited field of view of depth sensors reduce point cloud overlap, limiting the effectiveness of pose estimation with the ICP algorithm. To address these issues, we introduce a method that combines camera pose estimation and image deblurring using a Gaussian Splatting framework, leveraging both 3D Gaussian splats and depth inputs for enhanced scene representation. Our method first aligns consecutive RGB-D frames through optical flow and ICP, then refines camera poses and 3D geometry by adjusting Gaussian positions for optimal depth alignment. To handle motion blur, we model camera movement during exposure and deblur images by comparing the input with a series of sharp, rendered frames. Experiments on a new RGB-D dataset with extreme motion blur show that our method outperforms existing approaches, enabling high-quality reconstructions even in challenging conditions. This approach has broad implications for 3D mapping applications in robotics, autonomous navigation, and augmented reality. Both code and dataset are publicly available on this https URL.
[CV-113] Pedestrian Archetypes Extension – More Pedestrian Models for Autonomous Vehicle Safety Testing
链接: https://arxiv.org/abs/2607.16922
作者: Taorui Huang,Namita Gaidhani,Ritvik Bansal,S M Jubaer,Regina Lim,Rhett Zhao,Gavin Rafael Selin,Sunnie Deng Gao,Hasnain N Syed
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Extended version of Pedestrian Archetypes paper (published in IEEE IV 2025)
Abstract:In our prior work, Pedestrian Archetypes, we defined pedestrian archetypes as collections of behaviors that uniquely identify a specific type of pedestrian. The first paper proposed 12 pedestrian archetypes, including the Wanderer, Drunk, Distracted, Flash, Indecisive, Blind, Flock, Jaywalker, Elderly, Kid, Eventful, and Parked Pedestrian. These archetypes were introduced to move beyond single behavior labels and provide a more natural way to describe how dangerous pedestrians actually behave progressively in real-world traffic scenarios. However, upon further annotation of YouTube dash-cam videos, we identified 7 additional pedestrian archetypes with observable and significant behavioral differences from the previously proposed ones. These new archetypes capture pedestrian behavior patterns that could not be fully explained by the original taxonomy. In this pre-print, we introduce each new archetype, define its essential and optional behaviors, explain how it differs from previously proposed archetypes, and provide video-frame evidence showing the archetype in action.
[CV-114] ransferable Low-Rank Convolutional Bases for Onboarding Unseen Medical Imaging Modalities
链接: https://arxiv.org/abs/2607.16888
作者: Ranat Das Prangon,Istiaque Ahmed,Shajid Hasan Naim,Waseem Mustak Zisan,Hossain Md Shakhawat
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Deploying a medical imaging model that must later accommodate a modality it has never seen is a recurring practical problem: retraining the shared representation is expensive and destroys performance on the modalities already in service. We study this \emphonboarding problem under a strict leave-one-domain-out protocol, in which a convolutional backbone is pre-trained on source modalities (Kidney CT and Brain MRI), frozen permanently, and then required to accommodate an unseen modality (Chest X-ray). Under this protocol we establish three findings. First, decision-layer parameter-efficient fine-tuning is insufficient when the backbone has never observed the target modality: a linear probe and fully-connected LoRA both fall well short, whereas convolutional LoRA recovers most of the achievable accuracy, showing that adaptation must reach the convolutional features. Second, and centrally, the low-rank convolutional \emphbasis learned on the source modalities \emphtransfers: freezing that basis and training only its up-projections onboards the unseen modality using just 0.78% of full fine-tuning’s parameters, at an accuracy 6.11 percentage points above a random basis of identical size, while an equivalent decision-layer basis exhibits no reliable transfer. Third, adapter-based onboarding leaves source-modality accuracy exactly unchanged ( \Delta = 0.00 pp), whereas full fine-tuning reaches the highest target accuracy only by catastrophically degrading the source modalities. A Mahalanobis score on frozen backbone features detects the unseen modality with high sensitivity at a strict source-retention threshold, providing a practical trigger for when onboarding is required. All results are reported over three seeds with paired bootstrap confidence intervals.
[CV-115] InfoDense: Density-Aware Regional Decisive Replay for Memory-Efficient Incremental Face Forgery Detection
链接: https://arxiv.org/abs/2607.16873
作者: Jikang Cheng,Hao Shen,Xueyi Zhang,Guangcheng Wang,Zhongyuan Wang,Renye Yan,Baojin Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The rapid evolution of face forgery techniques has introduced an increasing variety of manipulations. Incremental Face Forgery Detection (IFFD), which incrementally adds new forgery data to fine-tune previously trained models, has emerged as a promising approach to handle evolving forgery threats. However, conventional replay-based IFFD methods suffer from catastrophic forgetting. Storing full historical images under limited memory often either fails to preserve subtle forgery cues or introduces domain bias, reducing the model’s ability to learn intrinsic and transferable manipulation characteristics. In this paper, we propose a Density-Aware Regional Decisive replay strategy, termed InfoDense, to address these challenges. InfoDense prioritizes artifact-dense and forgery-critical regions, significantly reducing storage requirements while maintaining high-fidelity forgery evidence. We first introduce InfoDense Cut to localize decisive patches using CLIP-based embeddings. Then, InfoDense Select ranks candidate segments by combining latent-space representativeness and decisive patch counts, ensuring both diversity and information density in the replay buffer. Finally, InfoDense Fuse reconstructs unbiased training inputs by adaptively merging stored segments with current-task samples, enhancing knowledge retention and generalization. Extensive experiments on challenging incremental deepfake benchmarks demonstrate that InfoDense effectively mitigates catastrophic forgetting while improving cross-domain generalization.
[CV-116] InLiER: Learning-Free Heterogeneous LiDAR Place Recognition via Intermediate Mixed-Radix Structural Keypoint Tokenization
链接: https://arxiv.org/abs/2607.16862
作者: Nikolaos Stathoulopoulos,George Nikolakopoulos
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Accepted for publication in IEEE Robotics and Automation Letters (RA-L). 8 pages, 8 figures
Abstract:LiDAR place recognition supports loop closure, relocalization, and multi-agent map management. As robotic platforms increasingly combine LiDARs with different fields of view, resolutions, and scanning patterns, existing descriptors degrade because they are tightly coupled to sensor-specific characteristics. We present InLiER, a learning-free pipeline based on an intermediate tokenization step. Height-sliced keypoints from structural elements receive mixed-radix token IDs encoding height, radial distance, local shape, and azimuth from local 3D geometry, in a compact sub-2KB representation. The same vocabulary is reorganized across three retrieval stages: height-ceiling histogram intersection for fast rotation-invariant shortlisting, binary bitmask alignment for yaw estimation and reranking, and token-guided geometric verification for 6-DoF pose estimation. InLiER achieves state-of-the-art performance on the HeLiPR dataset and in real-world field experiments, among modern handcrafted methods and outperforms the learning-based baseline on most cross-sensor configurations.
[CV-117] Dataset Distillation by Influence Matching
链接: https://arxiv.org/abs/2607.16859
作者: Haoru Tan,Wang Wang,Sitong Wu,Xiuzhe Wu,Yangtian Sun,Chirui Chang,Shaofeng Zhang,Xiaojuan Qi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:We revisit dataset distillation from an outcome-centric perspective. Rather than aligning process surrogates (per-step gradients or training trajectories), Influence Matching (Inf-Match) aligns the final outcome of training: it learns a compact synthetic set whose effect on the converged parameters matches that of the full dataset. Concretely, we introduce a fully differentiable, sample-level influence estimator that quantifies parameter shifts from adding or removing data, without time-consuming inverse-Hessian products or convexity assumptions. The estimator runs in linear time by unrolling the optimization dynamics and applying a first-order Taylor approximation. We then learn the synthetic set by minimizing the mismatch between its influence and that of the real dataset, yielding outcome alignment rather than heuristic process imitation. Inf-Match delivers the best accuracy across standard classification benchmarks. For instance, on Tiny-ImageNet (IPC=10), Inf-Match attains 31.5%, a +4.7% improvement over NCFM. Beyond classification, Inf-Match scales to vision-language distillation on Flickr30K, outperforming strong process-matching baselines. For instance, with 200 to 1000 synthetic samples, our method achieved a leading impressive average on image/text retrieval tasks, higher than NCFM by 2.5%. The code will be released via this https URL.
[CV-118] Look Clearly Before Answering: Mitigating Hallucinations in LVLMs via Saliency-Driven Perceptual Realignment
链接: https://arxiv.org/abs/2607.16841
作者: Pengxu Chen,Yao Zhu,Guangming Zhu,Jun Sheng,Jincai Huang,Xiangyang Ji,Liang Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: Accepted by ACM Multimedia 2026
Abstract:Large vision-language models (LVLMs) have demonstrated remarkable capabilities in multimodal understanding. However, they remain prone to hallucinations, generating responses that are inconsistent with the visual evidence. Existing mitigation methods largely address language-prior bias or cross-modal imbalance, while progressive visual degradation across perception and memory remains underexplored. In this work, we propose Saliency-Driven Perceptual Realignment (SDPR), a training-free framework that mitigates the degradation of visual awareness throughout inference. Specifically, we first introduce saliency-driven attention redistribution to release attention hijacked by non-semantic sink tokens, thereby recovering critical visual evidence. Second, we identify spatial distortion in the KV cache and propose saliency-driven cache alignment to preserve query-relevant visual features during generation. Finally, we introduce prior-constrained contrastive decoding to penalize unfaithful predictions induced by dominant language priors. Our proposed SDPR is robust against hallucinations due to its holistic alignment of visual awareness across the entire generative trajectory. Extensive experiments across diverse LVLM architectures show that SDPR outperforms state-of-the-art methods on both hallucination and general-purpose benchmarks, requiring no additional training and incurring minimal runtime overhead. The code is available \hrefthis https URL\colorbluehere.
[CV-119] opoGS: Planar Reconstruction via Topology-aware 3D Gaussian Splatting ECCV2026
链接: https://arxiv.org/abs/2607.16838
作者: Shanshan Pan,Jiale Chen,Yilin Liu,Hui Huang
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: European Conference on Computer Vision (Proceedings of ECCV 2026); Project page: this https URL
Abstract:Extracting structured, parametric 3D representations from raw images remains a fundamental challenge in computer vision and graphics. While recent advancements in the 3D Gaussian Splatting (3DGS) pipeline integrate planar primitives to yield compact and editable geometry, these approaches typically treat planes as isolated, discrete sets. This lack of topological connectivity hinders robust geometric reasoning, leading to fragmented reconstructions and misaligned boundaries that fall short of the precision for rigorous spatial analysis and professional design workflows. To address this, we introduce TopoGS, the first 3DGS framework to explicitly integrate both planar and topological constraints for coherent 3D reconstruction. Specifically, we extract global 2D topological relationships from multi-view image segmentations and anchor Gaussian primitives to these structural elements. This formulation enables the joint optimization of plane parameters, rendering fidelity, and topological adjacency. By enforcing strict multi-view consistency alongside these topological constraints, our method significantly mitigates geometric misalignments and produces connected, structured 3D models. Extensive evaluations on the ScanNet++ dataset demonstrate that TopoGS achieves state-of-the-art performance, providing a highly robust solution for generating accurate, topologically sound, and visually faithful scene representations.
[CV-120] Robust PnP on a Neuromorphic Processor for Object Pose Estimation
链接: https://arxiv.org/abs/2607.16834
作者: Tam Ngoc-Bang Nguyen,Mohsi Jawaid,Tat-Jun Chin
类目: Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET)
备注: 12 pages
Abstract:Neuromorphic computing is gaining attention in robotic perception due to its higher energy efficiency. While neural network-based methods can more readily exploit the distributed and parallelized structure of neuromorphic computers, crafting neuromorphic solutions for non-learning tasks is less straightforward. This hampers the usage of neuromorphic computing for perception pipelines that depend on both learning and non-learning components, such as object pose estimation (OPE) where state-of-the-art methods use a deep network to predict 2D landmarks and nonlinear optimization to solve perspective-n-point (PnP). In this paper, we propose a novel neuromorphic-deployable formulation for robust PnP, where given outlier-prone 2D-3D correspondences, the object pose with the largest number of inliers is determined. Underpinning our method is a distributed algorithm for robust least squares estimation of rigid body pose that can be executed on a neuromorphic processor. We also design a spiking neural network (SNN) to predict 2D landmarks from event data, where the main layers of the SNN were designed according to the principles of spiking neurons. Overall, our work enables neuromorphic treatment of the major stages of an OPE pipeline, from event sensing and learned landmark prediction, to geometric optimization for robust PnP. Results on neuromophic hardware (Intel Loihi 2) indicate the higher energy efficiency our neuromorphic robust PnP, while achieving competitive accuracy.
[CV-121] UniNDM: A Unified Noise-driven Detection and Mitigation Framework Against Sexual Content in Text-to-Image Generation
链接: https://arxiv.org/abs/2607.16828
作者: Yao Huang,Yitong Sun,Huanran Chen,Ruochen Zhang,Shouwei Ruan,Ranjie Duan,Maoxun Yuan,Yinpeng Dong,Hui Xue,Xiaochun Cao,Xingxing Wei
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 10 figures, accepted by TPAMI
Abstract:Despite the impressive generative capabilities of text-to-image diffusion models, they remain vulnerable to implicit sexual prompts, where subtle cues disguised as benign terms or adversarial tokens unexpectedly generate the inappropriate content due to model biases or latent correlations in training data. Existing safety mechanisms face fundamental limitations: detection methods primarily identify explicit content and fail to capture implicit malicious intent, while mitigation approaches rely on static negative prompts inadequate for diverse implicit scenarios. To address these challenges, we propose UniNDM, a unified noise-driven framework that rethinks safety mechanisms through the lens of noise dynamics in diffusion processes. Our key insight is that early-stage predicted noise exhibits inherent separability between normal and sexually explicit content, which we theoretically demonstrates quadratically increasing semantic concentration with timestep. Leveraging this property, we develop a lightweight noise-based detector achieving superior accuracy with virtually no computational overhead. For mitigation, we introduce noise-enhanced adaptive negative guidance: dynamically generating context-specific negative prompts via large language models to handle diverse implicit content, while optimizing initial noise by suppressing attention concentration on explicit tokens to provide comprehensive protection. Besides the U-Net-based diffusion models, we further extend our framework to emerging Diffusion Transformer architectures through region-constrained semantic guidance tailored for their unified multimodal attention. Comprehensive experiments across U-Net models and DiT models on both natural and adversarial datasets demonstrate substantial improvements over state-of-the-art methods, including SLD, UCE, Safree, etc. Our code is publicly available at this https URL.
[CV-122] Spatially-Aware Class-Agnostic Object Counting
链接: https://arxiv.org/abs/2607.16826
作者: Robert Wijaya,Md. Tanvir Hossain,Amanda Kau,Ngai-Man Cheung
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Generalised object counting aims to estimate the number of instances of an arbitrary object category from a single image, but many recent methods can struggle on structurally complex objects due to limited spatial modelling. We present \textitUpCount, a class-agnostic counter designed to better preserve spatial structure. UpCount strengthens the visual representation by extracting multi-layer features from a ViT-B/16 encoder and reassembling them into a refined multi-scale pyramid that is spatially refined using Dense Prediction Transformers and FeatUp, yielding features with improved structural and spatial sensitivity; a proposal–verification counting head then identifies repeated patterns and produces a density map for the final count. On FSC-147, UpCount achieves 12.39 MAE and 100.89 RMSE on the test set, and it transfers effectively to vehicle counting on CARPK (6.27 MAE, 8.79 RMSE). Code: this https URL
[CV-123] st-Time Registers as Global Priors for Tokenized Image Generation
链接: https://arxiv.org/abs/2607.16824
作者: Cheng-Yao Hong,Yifan Wang,Yuewei Lin,Chenyu You
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 39 pages, 10 figures
Abstract:Attention-based models often develop attention sinks, where a small number of tokens repeatedly attract attention and accumulate unusually large activations. In vision transformers, these outliers are closely related to registers, which have been diagnostically linked to global, low-frequency image structure. Existing work has largely studied registers through interpretability analyses and linear probes, leaving open whether they can be operationalized as plug-and-play signals for generation without retraining. We revisit this question in tokenized image generation. Using OpenCLIP and DINOv2 on ImageNet, we find that test-time register features exhibit stronger low-frequency concentration than both [CLS] readouts and patch-mean features, and show a consistent (albeit moderate) correlation with pixel-space DCT low-frequency energy. Motivated by these diagnostics, we introduce RegToken, a training-free procedure that converts register structure into a small set of global prior tokens by (i) NFN-based layer localization, (ii) TokenRank-guided subspace extraction, and (iii) a projection-and-conservation update on the register subspace. Inserted into a frozen compact 1D token generation pipeline, RegToken improves ImageNet generation and alignment metrics (e.g., FID-5k 20.5 to 20.1, SigLIP 3.6 to 3.9) without modifying pretrained weights, and accelerates test-time optimization (Steps@ \tau 74 to 52). Overall, our results suggest that structures often viewed as attention artifacts can be repurposed as lightweight global priors for tokenized generation.
[CV-124] Scene-SAM3D: Multi-View Scene Asset Generation Without Fine-Tuning
链接: https://arxiv.org/abs/2607.16805
作者: Yuqi Zhang,Yadan Luo,Xiangyu Sun,Fengyi Zhang,Zi Huang,Xin Tan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 9 figures
Abstract:High-quality 3D scene assets are critical for embodied applications such as robotic manipulation, navigation, and simulation. Despite their strong object priors, recent single-image 3D generation models such as SAM3D remain insufficient for real-world scenes, where severe occlusions, redundant observations, and cross-view inconsistencies make reliable scene generation challenging. We introduce Scene-SAM3D, a training-free framework that extends SAM3D from single-view object generation to calibrated multi-view scene asset generation. Scene-SAM3D selects a compact set of complementary views, reducing observation redundancy while providing additional evidence for regions occluded in individual views. Based on the selected views, it performs step-efficient latent velocity fusion to integrate multi-view evidence and suppress cross-view conflicts in canonical space. Finally, a lightweight rigid-object Gaussian optimization refines the scene layout within 200 iterations while preserving the generated object geometry. Experiments on Replica and ScanNet++ demonstrate consistent improvements at both instance and scene levels, with our method reducing scene-level CD by 43.8% on Replica and 30.9% on ScanNet++, while cutting flow-model sampling FLOPs and wall-time latency by nearly 20% under the same multi-view setting. Code will be released at this https URL.
[CV-125] Value-Monotonicity Matters: A Concordance Loss for Deep Survival Prediction
链接: https://arxiv.org/abs/2607.16802
作者: Meixu Chen,Kai Wang,Jing Wang
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (stat.ML)
备注:
Abstract:Deep survival models are evaluated almost exclusively by the concordance index (C-index), yet they are commonly trained using likelihood objectives such as the Cox partial likelihood, discrete-time negative log-likelihood, and DeepHit likelihood. This mismatch is usually considered acceptable because the C-index can be recomputed on validation data during training. However, for end-to-end training of high-capacity encoders on small, heavily censored oncology cohorts, frequent C-index evaluation is computationally expensive, making the loss value itself an important signal for monitoring, early stopping, and model selection. We show that likelihood losses are unreliable for this purpose and propose a value-monotone concordance loss. We prove that every strictly proper survival likelihood admits directions where the loss decreases while the C-index remains unchanged, causing the loss value to decouple from ranking performance. We then study a sigmoid concordance loss (SCL), whose value approximates one minus the C-index up to a temperature term, ensuring that lower loss corresponds to higher C-index during optimization. The loss is architecture agnostic and reduces to a convex survival ranking support vector machine for linear models. Across eighteen datasets from four modalities using a unified five-fold cross-validation protocol, SCL achieves discrimination comparable to standard likelihood losses and is the best or within one standard deviation of the best C-index. Unlike likelihood losses, SCL maintains a strong correlation between loss value and C-index during training, with rank correlations of 0.96 to 0.99 compared with -0.03 to 0.53 for likelihood losses. Calibration measured by the integrated Brier score is comparable. SCL provides a value-monotone optimization objective whose value can serve as a reliable surrogate for the C-index during expensive end-to-end training.
[CV-126] HTT-Net: Hierarchical Text-guided Transition Modeling for Surgical Video Phase Recognition
链接: https://arxiv.org/abs/2607.16787
作者: Kunjie Deng,Jinghui Zhang,Weidong Chen,Ganbin Li,Xiangjun Lyu,Zhendong Mao,Yingchi Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Surgical video phase recognition is a fundamental task in computer-assisted intervention, supporting workflow understanding, intraoperative guidance, and surgical quality assessment. Although recent visual-temporal models have achieved promising progress, accurate and temporally coherent phase recognition remains challenging due to local visual ambiguity, transient prediction noise, and insufficient use of procedural semantics. To address these challenges, we propose HTT-Net, a Hierarchical Text-guided Transition modeling Network for surgical video phase recognition. The key idea is to introduce structured surgical semantic knowledge into phase-aware segment construction and semantic refinement. Specifically, we construct a hierarchical surgical semantic memory with intra-phase descriptions, inter-phase transition descriptions, and fine-grained semantic units. Based on this memory, the proposed Transition-Aware Segment Construction (TAS-Con) organizes frame-level evidence into coherent segment representations and handles boundary clips with inter-phase transition descriptions. Furthermore, we introduce Transition-Aware Segment Calibration (TAS-Calib), which calibrates phase-aware segment representations through hierarchical surgical semantics and improves discrimination under visual ambiguity without dense frame-level vision-language fusion. Experiments on Cholec80 and LCRS-100 demonstrate the effectiveness of HTT-Net for robust surgical video phase recognition.
[CV-127] Spatiotemporal Facial Action Unit Detection using Twin Cycle Autoencoders for Driver Monitoring
链接: https://arxiv.org/abs/2607.16760
作者: Sai Sidharth D
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Driver monitoring systems (DMS) increasingly rely on facial cues to infer drowsiness, distraction, and cognitive load in real time. Facial Action Units (AUs), grounded in the Facial Action Coding System (FACS), provide an objective and interpretable representation of such states, but their automatic detection in the driving context is complicated by low and variable illumination, partial occlusion, head-pose variation, and the subtlety and short duration of relevant AU activations. Existing AU detectors largely treat spatial appearance and temporal dynamics separately, limiting their ability to exploit self-supervisory signal from abundant unlabeled driving video. We propose the Twin Cycle Autoencoder (TCA), a spatiotemporal architecture composed of two coupled cycle-consistent autoencoder branches: a Spatial Cycle Autoencoder that disentangles AU-relevant appearance from identity through image-level cycle consistency, and a Temporal Cycle Autoencoder that enforces forward-backward consistency over latent AU trajectories to capture onset-apex-offset dynamics. The two branches are coupled through a cross-branch latent alignment loss and fused via an attention module before multi-label AU classification. We evaluate TCA on the DISFA and BP4D benchmarks and on an in-cabin naturalistic driving dataset, and observe consistent improvements over CNN-RNN, 3D-CNN, and graph-based AU baselines, particularly for low-intensity and rapidly transitioning AUs relevant to fatigue (AU45, AU43) and yawning (AU26). We further show the model sustains real-time throughput on an embedded Jetson Xavier NX platform, supporting its use in production-grade advanced driver assistance systems (ADAS).
[CV-128] Hybrid Machine Learning for Articulation Angle Estimation of Truck-Semitrailer Combinations ITSC
链接: https://arxiv.org/abs/2607.16758
作者: Qixuan Zhang,Jonas Boettcher,Simon F. G. Ehlers,Marvin Stuede
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: Accepted for publication at 2026 International Conference on Intelligent Transportation Systems (ITSC)
Abstract:Accurate articulation angle estimation of trucks with trailers is critical for autonomous driving and advanced driver assistance system (ADAS). Existing methods either require manual initialization, additional sensors, or prior knowledge and signals from trailers, or they lack real-world validation, limiting practical deployment. This paper presents multiple learning-based models to directly estimate articulation angles from visual and kinematic inputs, eliminating the need for dedicated driving maneuvers for initialization, bounding box annotations, trailer-mounted sensor signals, or prior knowledge of trailer parameters. Two learning-based models are integrated with a kinematic model within an extended Kalman filter (EKF) framework, and an adaptive weighting scheme based on uncertainty quantification is applied for measurements involving visual input. Extensive real-world experiments with different trailer types demonstrate the approaches’ robustness and generalization under out-of-domain conditions, including new trailers, varying colors, and lighting conditions. Results show that the hybrid method achieves accurate and reliable articulation angle estimation while maintaining reduced implementation requirements and practical deployment advantages.
[CV-129] VisionAssist: An Open-Source Smartphone Assistant for AI-Based Visual Accessibility
链接: https://arxiv.org/abs/2607.16750
作者: Ayşe Özlem Çalışkan,Jordi Sanchez-Riera
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: research work from Erasmus+ internship
Abstract:People with low vision often face challenges in performing everyday tasks that require interpreting visual information. We present \textbfVisionAssist, an open-source mobile application designed to improve independence by providing AI-powered visual assistance through a smartphone. The application integrates three complementary functionalities within a single interface. First, it enables users to locate specific objects by analyzing the live camera feed. Second, it generates spoken descriptions of captured images, allowing users to identify visual content such as food labels, documents, and everyday objects. Third, it integrates with the smartphone’s contacts and calendar to facilitate emergency calls and provide voice-based reminders. The application supports hands-free interaction through voice commands and delivers all feedback using text-to-speech synthesis, making it fully accessible to users with visual impairments. By combining multiple assistive services into a unified platform and releasing the project as open-source software, the proposed solution aims to encourage community contributions and accelerate the development of accessible technologies. The source code is publicly available at: this https URL
[CV-130] Multi-Dimensional Quality Assessment for AI-Generated Human-Centric Videos: Dataset and Model
链接: https://arxiv.org/abs/2607.16742
作者: Sijing Wu,Yunhao Li,Huiyu Duan,Yucheng Zhu,Xiongkuo Min,Patrick Le Callet,Guangtao Zhai
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted for publication in IEEE TCSVT, 2026
Abstract:AI-generated human-centric videos play a crucial role in a wide range of modern applications. However, they often suffer from quality issues and semantic mismatches, underscoring the importance of effective quality assessment for such videos. To this end, we extend our previous dataset HVEval with pairwise preference annotations, resulting in HVEval+, the largest holistic quality assessment dataset for AI-generated human-centric videos, which comprises 1k prompts based on a comprehensive taxonomy, 20k videos generated by 24 text-to-video (T2V) models, and extensive human annotations, including 60k mean opinion scores (MOSs) and 60k preference pairs across 3 dimensions (i.e., spatial quality, temporal quality, and text-video correspondence), as well as 20k category-specific question-answer (QA) pairs. Along with the HVEval+ dataset, we further propose MoE-Rater, a Mixture-of-Experts (MoE)-inspired and multimodal large language model (MLLM)-based all-in-one method that supports multi-dimensional quality rating, multi-dimensional pairwise comparison, and category-specific question answering within a single model. Specifically, we introduce Mixture of Projector Experts (MoPE) and Mixture of LoRA Experts (MoLE), together with a three-stage training strategy consisting of task-aware pre-training, task-specific adaptation, and adaptive routing optimization, to effectively unify multiple tasks, resulting in superior performance on both HVEval+ and Human-AGVQA datasets. Extensive experiments and comprehensive analysis demonstrate the significant potential of the HVEval+ dataset and the MoE-Rater method in advancing AI-generated video quality assessment and further facilitating the evaluation and optimization of T2V models.
[CV-131] Constraint-Anchored Reasoning Traces ACM-MM2026
链接: https://arxiv.org/abs/2607.16727
作者: Zehua Cheng,Wei Dai,Jiahao Sun
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, ACM MM 2026
Abstract:Autoregressive multimodal large language models (MLLMs) suffer from error snowballing: a single incorrect inference early in a chainof-thought (CoT) trace corrupts all downstream reasoning. We find that in state-of-the-art open-source MLLMs, once the first error occurs, the reasoning cascades into failure across all remaining steps in 65% of such cases (a metric we term the snowball rate). Existing mitigations-sampling multiple chains, post-hoc self-verification, or full program synthesis-either lack symbolic grounding, catch errors too late, or sacrifice the flexibility of natural language reasoning. We propose Constraint-Anchored Reasoning Traces (CART), a neuro-symbolic framework that trains MLLMs to interleave natural language reasoning steps with symbolic constraint assertions: lightweight, machine-checkable statements about visual content (e.g., count(red_objects) = 3). A dual-pronged Constraint Propagation Module-combining a learned neural grounding head with Boolean Constraint Propagation-continuously verifies these anchors against extracted visual features and checks their mutual logical consistency. When a contradiction is detected, a backtrack controller halts generation and reverts to the last consistent checkpoint, preventing error propagation. A variable-frequency emission mechanism allows the model to adaptively control anchor density, avoiding trace bloat. We construct 218K training instances by augmenting GQA, CLEVR-CoGenT, and VCR with ground-truth constraint annotations derived from scene graphs, and fine-tune open-source MLLMs (LLaVA-NeXT, Qwen2-VL) via LoRA. On five benchmarks, CART reduces the snowball rate from 0.65 to 0.14, improves GQA accuracy by +4.6 percentage points over trainingonly baselines, and achieves 89.1 F1 on POPE-all with at most 18% inference overhead.
[CV-132] Can Experts Adapt Without Training? On Test-Time Modality Generalization in MVLMs MICCAI2026
链接: https://arxiv.org/abs/2607.16726
作者: Raza Imam,Darakshan Rashid,Yutong Xie,Dwarikanath Mahapatra,Brejesh Lall,Mohammad Yaqub
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to MICCAI 2026
Abstract:Medical vision-language models (MVLMs) promise broad zero-shot generalization, yet their reliability collapses when confronted with unseen modalities and domains, precisely where clinical robustness matters most. To address this gap, we revisit test-time modality generalization from the perspective of Mixture-of-Experts (MoE) and ask: can experts route-and-adapt without any optimization during inference? We identify a fundamental specialization-generalization dilemma at test time, where blindly aggregating modality experts dilutes modality-specific knowledge, while selecting one highly confident expert risks mismatch under shift. To address this, we propose MoBE: a fully optimization-free framework that performs dynamic expert selection and adaptation at test time. MoBE combines entropy-guided dynamic routing in MoE settings with expert-wise Bayesian adaptation, enabling experts to update their confidence and adapt online without gradient updates. Without parametric updates, MoBE augments a static MVLM with test-time routing and online statistics, achieving average accuracy gains of +4.72, +7.17, and +4.3 over state-of-the-art TTA methods across seen, unseen, and heterogeneous medical benchmarks, highlighting the effectiveness of training-free expert adaptation for robust modality generalization.
[CV-133] OFD-Net: Teacher-Free Reliable Semi-supervised Medical Image Segmentation with Orthogonal Feature Disentanglement Net of Foreground-Background
链接: https://arxiv.org/abs/2607.16705
作者: Shao-feng Jiang,Zhe-yang Jing,Qin Lu,Huan-huan Shi,Zhen Chen,Cong-xuan zhang,Chen Yi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Semi-supervised learning (SSL) is an effective solution for medical image segmentation with limited annotations. Existing SSL methods mainly rely on pseudo-labels generated by teacher-student supervision or cross-network consistency. However, these methods lack an explicit structural reference for judging pseudo-label quality. Low-quality pseudo-labels may lead to unreliable training, error accumulation and confirmation bias when processing unlabeled data with substantial appearance variations. To address this issue, we proposed OFD-Net, a teacher-free single-network framework for reliable semi-supervised medical image segmentation. OFD-Net employs an Orthogonal Feature Disentanglement Module (OFDM) to capture OFD features for reliable SSL by disentangling unlabeled data into background and foreground representations with a reliable structural distribution, thereby effectively reducing error accumulation and alleviating confirmation bias among unlabeled data. Specifically, OFD-Net explicitly employs a Disentanglement Guidance Module (DGM) to inject the resulting structural priors of foreground-background into the decoder by deformable convolution processing, and outputs predictions with clearer foreground representations. Based on DGM and the OFDM, we further develop a reliability-aware pseudo-label learning mechanism that evaluates unlabeled supervision according to the structural consistency between the main prediction and the disentangled foreground-background responses, and then down-weights unreliable regions during training. Extensive experiments on four public medical image segmentation benchmarks, namely ISIC-2016, Kvasir-SEG, Synapse, and ACDC, validate the effectiveness of OFD-Net. These results confirm that orthogonal foreground-background disentanglement enables OFD-Net to establish an efficient and reliable training paradigm within a teacher-free single-network framework.
[CV-134] Foundation-Assisted Active Learning for Object Detection Annotation
链接: https://arxiv.org/abs/2607.16671
作者: Jinchang Zhang,Arnold Zumbrun,Jing Lin,Guoyu Lu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:The annotation cost for remote sensing object detection is high, while existing active learning methods still face several challenges in object detection scenarios, including the coupling of localization and classification uncertainty, severe localization noise in the cold-start stage, and pseudo-diversity caused by high-recall candidate proposals. To address these issues, we propose a foundation-model-collaborative active learning and semi-automatic annotation framework for efficient construction of remote sensing object detection datasets. We build a dual-source mechanism consisting of a reference localization source (SA-source) based on UPN+SAM2 and a detector prediction source (OD-source), and further propose a Foundation-model-enhanced Dual-Source Uncertainty estimation to improve sample selection quality in the cold-start stage by jointly modeling localization consistency and classification confidence. Furthermore, we propose Object-Centric Diversity Sampling, which constructs object-level representations using DINOv2 features and SAM2 masks to improve sample coverage while suppressing pseudo-diversity. To address geometric noise in the semi-automatic annotation stage, we design Dual-Source Box Switching, which replaces noisy detector boxes with matched refined boxes from the SA-source, thereby reducing the manual burden of box refinement. Experiments on DIOR, HRSC2016, DOTAv2, and FAIR1M show that our method achieves superior or comparable results under most annotation budgets, with notably stronger cold-start sample efficiency in the low-budget regime.
[CV-135] DORS: Dynamic Attention Routing for Diffusion-based Object Removal in Dense Scenes
链接: https://arxiv.org/abs/2607.16656
作者: Haitong Tang,Haipeng Liu,Yang Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 10 figures, to appear in ACM Multimedia 2026
Abstract:Object removal aims to eliminate target objects specified by a mask while preserving visual consistency with the surrounding regions. Existing methods typically rely on contextual information from surrounding regions. However, in dense scenes where the surrounding regions contain instances visually similar to the removal target, such reliance often leads to semantic interference, resulting in incomplete removal. This problem arises from erroneous information propagation in the attention space, where masked queries tend to align with such instances due to global similarity matching in self-attention. To address this challenge, we propose a Diffusion-based Object Removal framework for dense Scenes, dubbed DORS, built upon a Dynamic Attention Routing mechanism comprising two complementary components: Instance-Filtered Attention (IFA), which suppresses misleading semantic information from similar instances through dynamically constructed mask-guided attention constraints, and Context-Guided Routing (CGR), which dynamically routes complementary scene information to maintain visual consistency. We further introduce DOR-Bench, a benchmark tailored for object removal in dense scenes. Extensive experiments demonstrate that DORS outperforms state-of-the-art methods, particularly in reducing incomplete removal and duplicate artifacts. The code will be available at this https URL.
[CV-136] Position: Explanation Stability Is a Property of the Model Method Pair Not the Model ICML2026
链接: https://arxiv.org/abs/2607.16652
作者: Kabilan Elangovan,Daniel Ting
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted Position Paper at ICML 2026. this https URL
Abstract:This position paper argues that claims about explanation stability are scientifically invalid without cross method validation. Just as statistical significance requires the test statistic to be specified, stability should either be evaluated across multiple attribution paradigms or explicitly scoped to the computational objective of a single method. In controlled chest X ray experiments, DenseNet201, ResNet50V2, and InceptionV3 achieved AUC values above 99%, yet their stability rankings reversed across attribution methods. LayerCAM ranked InceptionV3 as the most stable model, with an IoU of 0.777, whereas GradCAM++ favored DenseNet201 and reduced InceptionV3 stability score by 17.3%. These findings demonstrate that explanation stability is an emergent property of the model method pair rather than an intrinsic characteristic of the model alone. We therefore argue that explanation based claims should be validated across multiple attribution methods and that regulatory submissions should explicitly specify the attribution operators used to avoid creating illusory safety assurances.
[CV-137] DRIFT: Difficulty-aware Rectified Flows for Through-plane MRI Super-Resolution
链接: https://arxiv.org/abs/2607.16649
作者: Yoonseok Choi,Eun-Gyu Ha,Daniel Kim,Mohammed A. Al-masni,Ming-Hsuan Yang,Dong-Hyun Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Magnetic Resonance Imaging (MRI) is often acquired with anisotropic resolution to reduce scan time, producing stair-step artifacts along the through-plane direction. In through-plane MRI super-resolution, an efficiency-fidelity trade-off arises: feed-forward regressors are fast but oversmooth at large slice-thicknesses, while sampling-based methods improve fidelity at high inference cost. We propose DRIFT, a two-stage thickness-conditioned rectified flow framework for through-plane MRI super-resolution with continuous input slice-thickness. Stage 1 employs an Anatomical Projection Network (APN) to map low-resolution patches to a coarse high-resolution manifold, providing a deterministic anatomical initialization that shortens the residual transport of Stage 2 and stabilizes slice-wise refinement. Stage 2 refines details via rectified flow and introduces a Physics-Aware Difficulty (PAD) metric derived from slice-thickness induced through-plane bandwidth deficit to guide an Adaptive Integration Scheduler (AIS), allocating ODE steps by thickness. A Consistent Endpoint Trajectory Alignment (CETA) loss enforces thickness-consistent reconstructions. Experiments show that DRIFT outperforms super-resolution baselines while reducing inference cost. Code, models, and interactive demos are available at this https URL.
[CV-138] DARA: Degradation-Aware Low-Rank Residual Adaptation with Original-to-Corrupted Distillation for Corruption-Robust Animal Re-Identification
链接: https://arxiv.org/abs/2607.16644
作者: Cynthia Xie,Talia Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Animal re-identification (Re-ID) relies on fine-grained identity cues that can be disrupted by blur, noise, compression, and other visual degradations. Existing robustness strategies based on degradation-augmented training or pixel-level restoration improve robustness indirectly, but do not explicitly repair shifts in the identity retrieval space. We study corruption-robust animal Re-ID as input-conditioned feature-space repair and introduce DARA, a lightweight retrofit for compact Re-ID models. DARA freezes the fine-tuned backbone and learns routed low-rank residual experts to adapt degraded-input embeddings without corruption-type annotations. To stabilize this adaptive repair, original-to-corrupted distillation uses an original-image teacher to preserve individual embeddings and retrieval relations. Experiments on ATRW, FriesianCattle2017, MPDD, and SeaStarReID2023 show that DARA improves corrupted-query retrieval over standard and augmentation-based fine-tuning, generalizes to unseen corruptions and cross-domain evaluation, and recovers 77.0% of the corrupted-query mAP gap to full corrupted fine-tuning while adding only 0.49% parameters and 0.05% FLOPs.
[CV-139] Diversity-Oriented Fine-Tuning for Uncertainty-Based Hallucination Detection
链接: https://arxiv.org/abs/2607.16643
作者: Qiuyuan Li,Hongliang Dai,Piji Li
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Existing hallucination detection methods are typically conducted at the inference stage, without making any modifications to the model itself. In this paper, we are interested in exploring fine-tuning strategies that enhance the detectability of hallucinations in the resulting model. Focusing on semantic-entropy-based detection, we observe that many erroneous outputs remain undetected because the model produces nearly identical incorrect answers across multiple runs. To address this, we propose diversity-oriented fine-tuning to encourage more varied generations. We introduce two specific strategies: one based on Supervised Fine-Tuning (SFT) and the other on Direct Preference Optimization (DPO). Extensive experiments are conducted to evaluate our approach and analyze the behavior of the models before and after fine-tuning. We find that after adopting our fine-tuning methods, the models become less likely to produce low semantic entropy responses for hallucinated answers, thereby improving the effectiveness of hallucination detection, eventually yielding results better than or comparable with state of the art methods. The code will be publicly released.
[CV-140] ale: Blending Multi-Instance LoRA Text Encoders and a Zero-Shot LLM Judge for Ambivalence/Hesitancy Recognition in Videos
链接: https://arxiv.org/abs/2607.16635
作者: Abdel-Karim Al-Tamimi,Ali Rodan
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:We present TellTale, a text-only approach to ambivalence/hesitancy (A/H) recognition in interview videos, evaluated on the BAH dataset as part of the 3rd A/H Video Recognition Challenge (11th ABAW Workshop, ECCV 2026). Although the dataset provides video, audio, facial crops, and transcripts, TellTale relies on the transcript alone and combines three probability streams. Two text encoders, multilingual-e5-large and mDeBERTa-v3-base, are fine-tuned with parameter-efficient LoRA adapters under a multiple-instance learning (MIL) objective, in which transcript chunks are scored individually and pooled with a smooth maximum so that only the video-level label is needed for supervision. The third stream requires no training: a quantized 14B instruction LLM is prompted, zero-shot, to rate each transcript for A/H. The three probabilities are combined by a weighted average and a single decision threshold, both selected on participant-grouped cross-validated predictions. On the organizer-scored private test set of 152 videos from unseen participants, TellTale achieves a Macro-F1 of 0.7364 and an average precision of 0.7940, compared with 0.2827 Macro-F1 for the official vision-based baseline.
[CV-141] SPARE-GS: Structural Parsimony and Resource Efficiency for 3D Gaussian Splatting
链接: https://arxiv.org/abs/2607.16624
作者: Zhang Chen,Shuai Wan,Fuzheng Yang,Jiazhi Xia,Weiyao Lin,Junhui Hou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:3D Gaussian Splatting (3DGS) achieves high-fidelity novel view synthesis in real-time; however its training efficiency and representation compactness are hindered by excessive primitive proliferation. To address this challenge, we formulate the structural evolution of 3DGS as a global budget-constrained optimization problem and derive an optimality condition, which requires the marginal utility of structural resources to be balanced across spatial regions under a finite primitive budget. Based on this formulation, we propose SPARE-GS, a general plug-and-play framework that dynamically aligns the distribution of 3D Gaussian primitives with regional representational demand. SPARE-GS estimates capacity-normalized regional demand, assigns adaptive target quotas, and uses regional budget deviations to coordinate densification, pruning and adaptive termination toward a more balanced structural allocation. Extensive experiments across standard, accelerated, and structure-enhanced 3DGS pipelines demonstrate that SPARE-GS reduces the Gaussian count and training time by an average of 30.38% and 23.81%, respectively, while improving the average PSNR. Moreover, the resulting compact representations reduce downstream processing time and improve the rate-distortion performance of diverse compression and pruning methods, demonstrating the broad applicability of global structural budget regulation.
[CV-142] PAVXploreRL: Physical-Action-Visual World Model Reinforcement Learning with Action Exploration
链接: https://arxiv.org/abs/2607.16602
作者: Han Wang,Zijun Wang,Shuoshuo Xue,Rui Cao,Fenjiao Cheng,Xiaodang Liang,Roy Ka-Wei Lee
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 9 pages, 5 figures
Abstract:Action-conditioned world models are a key component of embodied AI, serving as scalable policy evaluators that reduce reliance on expensive real-world rollouts. To accurately capture diverse action-induced dynamics, such models should satisfy three key objectives-Physical Plausibility §, Action Adherence (A), and Visual Fidelity (V), collectively referred to as PAV-while remaining robust to both in-distribution (ID) expert demonstrations and out-of-distribution (OOD) actions. However, existing methods primarily rely on ID action-video pairs and pixel-level reconstruction losses, which do not explicitly optimize PAV objectives and generalize poorly beyond expert data. To address this, we propose PAVXploreRL, a reinforcement learning framework built on a pretrained latent world model that explicitly optimizes PAV objectives through reward-driven training. To improve action generalization, our method jointly leverages ID trajectories and noise-driven OOD action exploration, without paired video supervision. Experiments show that PAVXploreRL consistently outperforms pretrained baselines, achieving a 5.6% average gain across benchmarks and producing higher-quality PAV properties. As a policy evaluator, it also yields more reliable performance estimates and reduces the overestimation bias of prior expert-only world models such as Ctrl-World. Code: this https URL
[CV-143] Digital measurement of droplet flame diameter in microgravity combustion images using Segment Anything Model 2 with automatic prompt selection
链接: https://arxiv.org/abs/2607.16587
作者: Minghui Xu,Chaoyi Zhou,Aaron P. Cecil,Xi Liu,Siyu Huang,Yuhao Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Flame diameter is a key measurable parameter in microgravity droplet combustion, but its extraction from self-illuminated frames remains difficult because soot tails, blurred luminous boundaries, chamber reflections, and droplet drift introduce substantial measurement bias and operator dependence. This work presents an AI-enabled digital measurement workflow for automated flame diameter from combustion images. The workflow integrates automatic prompt-point generation into Segment Anything Model 2, employing Random Sample Consensus (RANSAC)-based circle fitting. The automatic prompt strategy removes subjective manual point selection, while the video memory mechanism maintains temporal consistency for drifting droplets, and the RANSAC fitting rejects soot-tail pixels as geometric outliers. The method is validated by 19,537 flame images of n-heptane, n-decane, and n-octane droplets with varying initial diameters. Compared with manual-reference measurements, the proposed workflow achieves a mean relative agreement of 96.9%, a mean absolute percentage error of 3.1%, and substantially outperforms conventional Hough circle detection, which performed worse under the same evaluation conditions. The results also show that the measurement accuracy improves with increasing droplet size. The proposed workflow has a combined standard uncertainty of 8.54% and achieves approximately a 229-fold improvement in efficiency over manual measurement. These results demonstrate that the proposed SAM2-based workflow provides a reproducible, fully automated, and metrologically characterized digital measurement system for extracting flame diameter from challenging combustion images. The approach supports high-throughput combustion diagnostics and illustrates that AI-based segmentation can be integrated into quantitative measurement workflows for digitalized image-based metrology. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.16587 [cs.CV] (or arXiv:2607.16587v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.16587 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-144] Autonomous VR-Based Risk Detection for Situational Awareness in Dangerous Settings
链接: https://arxiv.org/abs/2607.16582
作者: Mohammad Eskandari,Murali Krishna Varma Indukuri,Stephanie M. Lukin,Cynthia Matuszek
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 7 Pages, Accepted to RO-MAN 2026
Abstract:In high-risk environments such as disaster response, situational awareness depends not only on detecting hazards but also on communicating them clearly to human operators. Vision Language Models (VLMs) have shown strong potential for scene understanding in safety-critical settings, yet their value as part of human-facing robotic systems remains underexplored. We present a VR-based Human Robot Interaction framework for studying how VLM-assisted robots can support situational awareness in simulated hazardous environments. In our system, a robot explores a virtual scene and queries a VLM to identify potential hazards and annotate user-facing points of interest. These annotations are presented to a human operator through an immersive VR interface. This framework enables controlled evaluation of both robotic hazard identification and the communication of safety-critical information to users. Results from our study indicate that the annotated VR interface was preferred over the unannotated baseline and that participants reported high clarity, usefulness, and comfort when interacting with the system. These findings suggest that combining VLM-based robotic perception with immersive visualization is a promising approach for supporting situational awareness in hazardous settings.
[CV-145] CNS-Edit: Category-Agnostic 3D Editing with Coupled Neural Shape Representation
链接: https://arxiv.org/abs/2607.16577
作者: Jingyu Hu,Weilong Yan,Zhengzhe Liu,Haipeng Li,Ka-Hei Hui, Hao (Richard)Zhang,Chi-Wing Fu
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注:
Abstract:This paper presents a latent-space 3D shape editing framework built upon a coupled neural shape (CNS) representation and a neural feature volume optimization. This work extends CNS-Edit, built on Coupled Neural Shape optimization, to CNS-Edit++, by generalizing the category-specific coupled representation to category-agnostic 3D shape editing with foundation models. The Coupled Neural Shape (CNS) representation couples a global latent code that captures high-level shape semantics with a 3D neural feature volume that provides spatial context for local shape manipulation. Then we formulate a coupled neural shape optimization procedure that co-optimizes these two components subject to a given editing operation. Our framework can be instantiated on both the category-specific 3D inversion model and category-agnostic 3D foundation models. We provide various shape editing operators, including copy, resize, delete, mix, point-wise drag, and region-wise drag, each of which is formulated as an objective to guide the CNS optimization. To preserve regions outside the editing area, we further introduce two complementary region-wise control mechanisms, i.e., KV-cache replacement and latent feature regularization. Extensive quantitative and qualitative evaluations across different 3D generative models demonstrate the strong capabilities of our approach over state-of-the-art solutions.
[CV-146] From Modalities to Propositions: A Language-Centric Framework for Multimodal Intelligence
链接: https://arxiv.org/abs/2607.16560
作者: Nadine Chang,Maying Shen,Shizhe Diao,Jialiang Wang,Jingde Chen,Thomas Breuel,Pavlo Molchanov,Rafid Mahmood,Jose M. Alvarez
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Multimedia (cs.MM)
备注:
Abstract:We propose a language representation for multimodal data in which any observation, whether image, video, or text, is expressed as a bag of atomic propositions, simple statements about the entities, actions, and relations in a scene. A global semantic codebook unifies these into a shared vocabulary of canonical atomic propositions, placing every modality and observation into one interpretable space that spans fine grained facts to high level concepts and composes into richer ones. This brings interpretability with reasoning, cross-modal understanding and retrieval, and compositionality that enables complex multimodal understanding, rich data curation and complex structured retrieval. We demonstrate the framework on autonomous driving and open-world data.
[CV-147] AffectFuse: Cross-Task Feature Fusion with Temporal Modeling for Multi-Task Affective Behavior Analysis
链接: https://arxiv.org/abs/2607.16546
作者: Dipit Saha,Mohammad Raihan Rashid,Shah Mohammad Abdul Mannan,Ahnaf Tahmid,Md. Mehedi Hasan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Affective behavior recognition in the wild requires joint prediction of continuous valence-arousal, categorical facial expression, and multi-label action units from unconstrained face images. We present our system for the Multi-Task Learning (MTL) track of the 11th Affective Behavior Analysis in-the-wild (ABAW) competition on s-Aff-Wild2, the static selected-frame version of Aff-Wild2. The method focuses on post-encoder adaptation: frozen AffectNet-supervised backbones provide multi-resolution features, while task-specific temporal heads and cross-task fusion modules select the useful signals for each target. For action-unit recognition, we adapt MAE-Face with Low-Rank Adaptation (LoRA) and use DISFA through per-unit expert routing rather than direct sequential transfer. Ablations over backbone, temporal, fusion, and AU-adaptation choices define the final configuration. The final system obtains P = 1.7302 on the official validation split, showing that post-encoder adaptation and task-wise modeling choices provide a strong MTL pipeline without training a new large-scale face foundation model.
[CV-148] Do Vision Encoders Exhibit Human-like Color Thresholds?
链接: https://arxiv.org/abs/2607.16540
作者: Engy Ehab,Pablo Hernández-Cámara,Nahla Belal,Jesús Malo,Javier Vazquez-Corral,Alexandra Gomez-Villa
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Understanding and characterizing human color perception is a longstanding research goal. One of the most traditional approaches is looking for the human color discrimination thresholds, the minimum chromatic differences perceptible to human observers. In recent years, deep neural networks have become the standard networks for computer vision tasks. In particular, deep vision encoders, foundation models trained on large-scale visual data, map images into latent feature representations. Despite the widespread use of deep vision encoders, few studies have investigated whether their internal representations exhibit human-like discrimination thresholds. In this work, we present a large-scale exploratory study probing the chromatic sensitivity of more than 50 pretrained vision encoders, including convolutional networks and vision transformers, against human discrimination thresholds. Using controlled chromatic stimuli at multiple chroma levels, we compare model-derived chromatic discrimination thresholds with human discrimination ellipses through a region-overlap metric (mIoU). Our analysis reveals generally weak alignment between model representations and human perceptual thresholds across all model families, with the best mIoU 0.25. Moreover, we find that self-supervised encoders consistently outperform supervised ones, while language-supervised models show the most polarized behavior, occupying both the top and bottom of the ranking. These findings suggest that human-like chromatic sensitivity does not emerge naturally from current large-scale visual training objectives for any of the analyzed architectures.
[CV-149] Geometry-Enhanced Portion Estimation for Multimodal LLM s
链接: https://arxiv.org/abs/2607.16514
作者: Lin Liao,Peng Li
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Image-based dietary assessment promises to replace costly, bias-prone manual recalls, but portion estimation remains a major blocker. Multimodal LLMs (MLLMs) recognize a wide range of foods zero-shot in uncontrolled photos, yet they are weak at portion estimation – a gap we measure across the current frontier (Gemini, GPT, and Claude flagships alike). We present a method that enhances a frozen, commercial MLLM with an accurate portion head: a small geometry-enhanced network on a frozen DINOv2 backbone with a structured softmax-ownership volume, consuming the MLLM’s per-food name, bounding box, and density range – no depth sensor, no MLLM fine-tuning. Evaluated fully open-vocabulary on three real-world benchmarks, the head cuts per-food portion error by 33-41% relative to the MLLM alone, outperforms every flagship MLLM’s direct estimates, and surpasses each benchmark’s originally published image-only model at its own reported metric.
[CV-150] WeedExpert-R1: Incentivizing Botanical Reasoning in MLLM s with Reinforcement Learning for Precision Weed Grounding
链接: https://arxiv.org/abs/2607.16492
作者: Zonglin Yang,Wei-Zhen Liang,Nevin Lawrence,Xin Qiao,Benjamin Riggan,Chi-En Chiang,Fuchen Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Precision weed control requires species-level identification and instance-level localization. However, conventional object detectors use a closed vocabulary, limiting their deployment across regions, and cannot explain their predictions in complex agricultural scenes. Multimodal large language models (MLLMs) offer visual grounding and reasoning capabilities, but insufficient botanical knowledge can cause hallucinations in fine-grained weed identification. This study introduces WeedExpert-R1, a multimodal model that learns visually grounded botanical reasoning through verifiable rewards. A domain-specific Chain-of-Thought synthesis pipeline combines a human-curated botanical trait dictionary with an Auditor-Synthesizer LLM workflow to generate reasoning data for supervised fine-tuning. Group Relative Policy Optimization is then applied with rewards for format, accuracy, instance count, and response length. Across 37 weed species from six datasets, WeedExpert-R1-4B achieved 75.82 percent exact-set precision at an IoU threshold of 0.5, 89.30 percent precision, and 87.81 percent recall. It outperformed proprietary models, including GPT-5.4 and Gemini-3.1-Pro, and larger open-source models, including Qwen3-VL-30B-Instruct and Gemma-4-31B-it. Results on unseen species further demonstrate its open-vocabulary capability and potential for deployment across diverse regions and crops without retraining.
[CV-151] Spectral-Morphological Attention U-Net: An Efficient Network for Active Wildfire Detection
链接: https://arxiv.org/abs/2607.16472
作者: Yugong Zeng,Jonathan Wu
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Over the past decades, the frequency of global wildfires has been increasing steadily. Therefore, if the fire can be detected and precisely located at an early stage, the potential hazards caused by it can be minimized to the greatest extent. The machine learning methods based on satellite images, due to their ability to automatically monitor extremely remote and vast areas, have shown great potential for application in the field of wildfire detection. To address this challenge, we proposed a new model named spectral-morphological attention U-Net(SMA-UNet), which includes a spectral attention module, a residual attention UNet backbone, a channel-spatial modulator, and a pair of differentiable morphological gates. We trained and evaluated this model with two datasets. These modules, excluding the backbone, are used to detect active fire events for the first time, especially the pair of differentiable morphological gates, which is innovatively developed. The proposed model achieved the highest scores in both datasets (e.g., intersection over union 75.16% in TS-SatFire, 22.50% in Sen2Fire). By conducting ablation studies of each module, we compared their independent contributions and tested their combinations. Ultimately, the integration of these modules yields a highly robust framework that significantly improves segmentation consistency across diverse and complex environmental conditions. Future work will focus on validating the proposed architecture across large-scale, multi-regional datasets from different satellite sensors to establish its broader generalizability for global wildfire detection.
[CV-152] DS@GT ARC at AnimalCLEF 2026: Species-Aware Graph Construction for Multi-Species Animal Re-Identification
链接: https://arxiv.org/abs/2607.16453
作者: Evan Sinclair Smith,Anthony Miyaguchi,Snigdha Palamari,Danté Evangelista
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 8 figures. Accepted to the CLEF 2026 Working Notes. Code: this https URL
Abstract:Automated individual animal re-identification is essential for large-scale biodiversity monitoring; however, field imagery complicates separating identity cues from nuisance variation in pose, illumination, background, resolution, and species-specific morphology. The DS@GT ARC submission to AnimalCLEF 2026 introduces a multi-species image-clustering system for re-identifying Eurasian lynx, fire salamanders, loggerhead sea turtles, and Texas horned lizards. Instead of relying on a single descriptor or nearest-neighbor retrieval, this approach formulates re-identification as species-aware graph construction over candidate image pairs. The pipeline integrates tailored preprocessing, global candidate retrieval, LightGlue-based local verification with multiple keypoint families, LightGBM pair scoring, conservative edge admission, and Leiden community detection. This design directly addresses a primary failure mode of clustering-based re-identification: high-scoring false pairs that act as bridge edges and merge distinct individuals through transitive closure. Across species, ablation studies demonstrate that local feature support, foreground-aware preprocessing, and species-specific backbone selection enhance pair evidence, while graph operating points determine the trade-off between fragmentation and over-merging. The selected submission achieved a public ARI of 0.733 and a private ARI of 0.674, ranking fifth among 230 teams. These results indicate that robust wildlife re-identification requires not only strong visual representations but also calibrated integration of global similarity, local identity markings, neighborhood context, and graph-level constraints. The code can be found at this https URL.
[CV-153] hink Plan Paint: Layout-Aware Reasoning for Controllable Image Generation in Unified Models
链接: https://arxiv.org/abs/2607.16409
作者: Junhao Liu,Jian-Wei Zhang,Tao Huang,Miles Yang,Zhao Zhong,Liefeng Bo
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 22 pages, 12 figures, 9 tables
Abstract:Unified Multimodal Large Language Models (MLLMs) offer a promising paradigm for unifying visual understanding and generation, yet they still struggle to follow complex spatial instructions and logical constraints in controllable image generation. To address this gap, we present ATLAS, a unified framework that equips MLLMs with a human-like “Think, Plan, and Paint” paradigm. We adopt layout as the shared representation that connects the three stages, enabling the model to reason about spatial requirements, plan explicit object arrangements, and render the final image. We further improve plan-to-image fidelity with reinforcement-learning-based layout alignment. We instantiate ATLAS at 7B and 80B scales, achieving state-of-the-art performance among MLLMs on image generation benchmarks and an average 65.31% improvement over existing layout-based unified MLLMs. On spatially related tasks, ATLAS obtains an average 23.06% gain over the base models. Through the same layout interface, ATLAS also supports instruction-guided editing and multimodal grounding. We further introduce ATLAS-Reasoning, a benchmark for evaluating generation under complex spatial instructions.
[CV-154] Apple-π: Benchmarking Thinking with Video Towards Law-Grounded Physical Intelligence
链接: https://arxiv.org/abs/2607.16401
作者: Runmao Yao,Kairui Hu,Yukang Cao,Ruisi Wang,Shulin Tian,Ziang Cao,Weichen Fan,Ziqi Huang,Yuhao Dong,Hao Li,Zhaoxi Chen,Zhongang Cai,Lei Yang,Ziwei Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Modern video generation models are increasingly hailed as emerging world models with an internalized grasp of physical law. Yet existing benchmarks largely evaluate physical plausibility only at the output level, without verifying whether the model arrives there through a faithful, law-grounded reasoning process. We introduce Apple-PI, the first benchmark that anchors video-model evaluation explicitly in physical laws. Apple-PI comprises three components. 1) Orchard: a dataset of 400 videos covering ten canonical tasks in classical mechanics. It separates single-law tasks for confounder-free diagnosis from multi-law tasks for probing generalization. 2) Benchmark Protocol: a three-stage protocol based on scientific reasoning, including Perception, Formulation, and Deduction. It uses chain-of-frames prompting on infographic-annotated first frames, treating the generated video as the model’s visible reasoning trace. 3) Evaluation Suite: a hybrid evaluation suite that combines MLLM-based subjective scoring with physics-law-grounded objective measures. This enables stage-resolved diagnosis of not only whether a model fails, but where it fails. Benchmarking 11 models shows that current video models remain far from reliable law-grounded world simulators, with the best video model scoring only 0.473. Our stage-, pillar-, and source-resolved analyses further expose a Perception-to-Formulation-to-Deduction bottleneck, weak multi-law state transfer, and a persistent Sim-to-Real gap. These findings position Apple-PI as a diagnostic foundation for guiding future video models toward world models with law-grounded physical intelligence.
[CV-155] Disentangling Model and Human Data Uncertainty in Apparent Facial Age Estimation
链接: https://arxiv.org/abs/2607.16378
作者: Andrei Foitos,Ivo Pascal de Jong,Matias Valdenegro-Toro
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 12 pages, 7 figures, ANNPR 2026 camera ready + appendix
Abstract:Estimating the apparent age of individuals from facial images is challenging due to the subjective nature of perception and the inherent variability of the data. We investigate the role of uncertainty estimation, attributing uncertainty jointly to a lack of knowledge (epistemic) or inherent noise/chance (aleatoric). Leveraging the APPA-REAL dataset, we train Bayesian Neural Networks on datasets of varying sizes using three BNN approximations: MC-DropConnect, Flipout, and Deep Ensembles using supervision on human aleatoric uncertainty available in the APPA-REAL dataset. Each model outputs both the predicted apparent age and the amount of aleatoric and epistemic uncertainty. Our results confirm the hypothesis that the inherent aleatoric uncertainty remains stable across dataset sizes, while epistemic uncertainty increases as training data decreases. These findings demonstrate that different sources of uncertainty can be quantified in face age estimation.
[CV-156] OmniStyle-INR: Universal and Multimodal Style Transfer for INRs
链接: https://arxiv.org/abs/2607.16362
作者: Rafał Kajca,Michał Miziołek,Kornel Howil,Rafał Tobiasz,Przemysław Spurek
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Style transfer remains a fundamental and highly important task across various data modalities, enabling creative manipulation conditioned by both reference images and textual descriptions. Recently, methods utilizing Gaussian Splatting have emerged as a unified representation for 2D images, video, 3D scenes, and 4D dynamics. However, representing videos and 2D images with Gaussian Splatting is structurally sub-optimal for dense continuous domains. The number of required Gaussians often approaches the total number of pixels, raising questions about the actual utility of such a representation for these specific modalities. In contrast, Implicit Neural Representations have established themselves as a much more popular and natural choice across all these data domains. Implicit Neural Representations naturally provide significant advantages, including data compression, inherent capabilities for super resolution, and seamless integration with deep generative models. To this end, we introduce OmniStyle-INR, a novel framework that leverages network-based continuous representations as a truly universal domain. Our approach successfully performs high-quality style transfer across all visual modalities, guided seamlessly by both text prompts and visual exemplars.
[CV-157] Spatial Transport of Integration Error in Generative ODEs
链接: https://arxiv.org/abs/2607.16361
作者: Songheng Yin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 21 pages
Abstract:A trained flow or diffusion model is usually run with only a handful of solver steps, and the integration error this leaves behind is unevenly distributed across the image. We ask where that error is injected and how it reaches the endpoint, and answer with a signed source-and-transport accounting of few-step integration error, tested to first order. A perturbation experiment on five models at 256^2 resolution shows the learned dynamics spread local disturbances widely: near the start of sampling, under 10% of the summed endpoint response remains at the source. Signed one-step truncation residuals, propagated through the model’s own linearized dynamics, reconstruct much of the endpoint error’s direction and regional structure (cosine 0.81-0.87), and a region’s error owes more to what arrives from elsewhere than to its own injection. Structure-destroying nulls, with protocols frozen before evaluation, locate what carries the account: randomizing contribution signs halves it, and reassigning which region receives each contribution, with content, norms, and signs intact, destroys it entirely. Where the injections land is readable from the model itself. The variation of its velocity or prediction field along the trajectory, a structure that emerges during training, predicts the final per-region gap (within-image rho of 0.57-0.70 on fine trajectories, weaker from the cheap solve alone). The prediction is partial because endpoint error depends not only on injected magnitude but on its sign, timing, and transport through the learned dynamics. A training penalty on the injected variation lowers few-step error, so the structure is one a model can be trained to change.
[CV-158] PhysAgent : Reflective Agent ic Physics Control for Physically Plausible Video Generation
链接: https://arxiv.org/abs/2607.16355
作者: Qirui Li,Jinkun Hao,Yibo Li,Ran Yi,Paul L. Rosin,Yu-Kun Lai
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: For project page, see this https URL
Abstract:Recent advances in physics-grounded video generation leverage physics simulation as a physical prior to guide video synthesis toward physically plausible outcomes. The simulation process is controlled by physical specifications, which are typically generated by a vision-language model in a single pass. Such one-shot prediction often fails to accurately translate user intent into executable simulations, particularly for fine-grained object dynamics, complex motion trajectories, and temporally structured interactions. In this paper, we propose PhysAgent, a reflective agentic framework that closes the loop among physical program generation, physics simulation, stage-specific verification, and targeted program repair. Beyond improving the control of coupled physical parameters, our framework enables the agent to progressively realize complex trajectories, multi-stage interactions, and precise event outcomes by treating each physical program as an executable hypothesis. In addition, we design a set of physics-control APIs to support more stable and complex motion behaviors. Extensive experiments demonstrate that PhysAgent produces more physically plausible videos, achieves better prompt alignment, and generalizes more effectively across diverse physical scenarios.
[CV-159] Privacy-Aware Synthetic Video Benchmarking and Relational Evaluation for Worker-Under-Suspended-Load Detection CVPR2026
链接: https://arxiv.org/abs/2607.16351
作者: Anshu Singh,Alejandro Seif
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted at the 3rd Workshop on Synthetic Data for Computer Vision (SynData4CV), CVPR 2026. OpenReview: this https URL . Dataset and code: this https URL
Abstract:Publicly shareable construction-video benchmarks remain scarce, especially for safety-critical hazards that are rare, dangerous to stage, and difficult to release. We study worker under suspended load, a relational hazard that depends on worker-load geometry and temporal persistence rather than object detection alone. We introduce SynthSite, a focused synthetic video benchmark of 55 clips spanning varied load configurations, viewpoints, clutter, occlusions, and surveillance conditions, together with a privacy-aware hybrid generation workflow that supports both publicly shareable benchmark creation and privacy-constrained synthetic video generation. We then ask whether worker appearance can be suppressed without undermining downstream hazard recognition. Under five whole-body privacy conditions, we evaluate worker and load retention, localization stability, and clip-level hazard recognition. We find that structure-preserving obfuscations retain substantially more downstream utility than appearance-smoothing baselines, and that preserving a raw visual reference alone does not guarantee the strongest agreement with human hazard labels. These findings suggest that privacy evaluation for construction safety analytics should assess not only appearance suppression, but also preservation of the geometric cues required for hazard reasoning. Our dataset and code are available at this https URL . Comments: Accepted at the 3rd Workshop on Synthetic Data for Computer Vision (SynData4CV), CVPR 2026. OpenReview: this https URL . Dataset and code: this https URL Subjects: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.16351 [cs.CV] (or arXiv:2607.16351v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.16351 Focus to learn more arXiv-issued DOI via DataCite
[CV-160] DMFNet: Dual-Backbone Multiscale Fusion Network for Urban Scene Classification
链接: https://arxiv.org/abs/2607.16338
作者: Anamitra Ghosh,Abhiroop Chatterjee,Susmita Ghosh
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Under review at InGARSS 2026. This work was conducted during the first author’s Master’s studies at Jadavpur University
Abstract:This article presents DMFNet, a dual-backbone multiscale feature fusion framework with residual feature propagation and spatial attention for remote sensing scene classification. Existing approaches often face challenges in effectively capturing multiscale feature interactions and learning robust feature representations from complex aerial scenes with high intra-class variability and inter-class similarity. To address these limitations, the proposed framework employs two pretrained backbone networks to extract diverse hierarchical feature representations. A multiscale feature fusion mechanism with residual feature propagation is introduced to enhance feature interaction across multiple resolution levels. In addition, a spatial attention module is introduced to emphasize informative spatial regions in multi-object scenes. Further, a two-stage training strategy consisting of backbone freezing followed by selective fine-tuning is adopted to ensure stable optimization and improved generalization. Experiments conducted on the benchmark AID dataset demonstrate that the DMFNet achieves an average accuracy of 97.46% \pm 0.14%. Ablative analysis further show the importance of various components in unison.
[CV-161] Local Brushstroke Quality Assessment via Vision-Language Feedback CVPR2026
链接: https://arxiv.org/abs/2607.16330
作者: Mio Mitamura,Hirokatsu Kataoka
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 6 pages, 8 figures. Accepted to the SAUAFG Workshop at CVPR 2026
Abstract:This paper investigates whether multimodal LLMs can evaluate local brushstroke quality in calligraphy and generate educationally useful natural language feedback. We construct an evaluation framework in which three multimodal LLMs (GPT-4o, Claude Sonnet 4, and Gemini 2.5 Flash) assess before-after image pairs of calligraphic works using a five-point ordinal scale, and compare their outputs against scores assigned by three expert calligraphers. We additionally examine a Retrieval-Augmented Generation (RAG) variant of Claude as a preliminary condition. Results show that all models achieve useful levels of absolute score accuracy (MAE), with GPT-4o performing best (MAE = 0.885). However, none of the models produce statistically significant overall rank correlations with human experts (Kendall’s tau). Vocabulary analysis of generated rationales reveals characteristic evaluative biases in each model, and RAG is shown to improve rank correlation while worsening absolute accuracy, constituting an important negative result for text-based rule injection.
[CV-162] Localization-Infused Vision-Language Semantic Fusion for Text-Guided Medical Image Segmentation
链接: https://arxiv.org/abs/2607.16327
作者: Songyue Han,Mingye Zou,Shuchang Ye,Lei Bi,Mingyuan Meng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Medical image segmentation is essential for modern computer-aided medicine. Recently, text-guided segmentation has shown promise by incorporating clinician-formulated textual reports as semantic guidance for image segmentation. These reports describe target appearance, location, and neighboring anatomy, providing explicit guidance for localization and delineation. Existing text-guided segmentation methods typically extract textual semantics implicitly through a pretrained text encoder and then integrate vision-language semantics via straightforward image-text feature fusion. However, these methods do not explicitly capture target-oriented information embedded in textual reports, particularly target location, and do not explore multi-level information fusion strategies beyond basic feature-level fusion, limiting the extraction and integration of critical textual semantics. In this study, we propose LoG, a localization-infused vision-language fusion framework for text-guided medical image segmentation. By jointly performing multi-scale target localization tasks, LoG explicitly captures target-oriented vision-language semantics and enables three-level localization-infused semantic fusion: (i) localization-guided feature fusion that directly infuses location-relevant semantics into visual features, (ii) localization-gated attention fusion that redirects multi-scale localization predictions to reinforce critical regions, and (iii) localization-constrained loss fusion that supervises segmentation based on spatial consistency with target localization. Extensive experiments on three benchmark datasets, involving three medical imaging modalities with paired textual reports, demonstrate that LoG achieves Dice scores of 91.59%, 80.71%, and 94.59% on QaTa-COV19, MosMedData+, and Kvasir-SEG, respectively, consistently outperforming state-of-the-art medical image segmentation methods.
[CV-163] CRISP: Pre-LLM Yet Text-Driven Visual Token Pruning for Efficient LVLM Inference ICME2026
链接: https://arxiv.org/abs/2607.16326
作者: Xu Li,Yi Zheng,Mengyang Zhao,Yuxuan Liang,Zhe Liu,Rui Zhu,Xiaolei Chen,Wei Zhou,Baoquan Zhao,Juncen Guo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by the 2026 IEEE International Conference on Multimedia and Expo (ICME 2026) as Oral
Abstract:Large Vision-Language Models (LVLMs) typically require processing hundreds to thousands of visual tokens, leading to substantial inference overhead. Existing visual token pruning methods either operate before the LLM using text-agnostic heuristics or prune inside the LLM at the cost of efficiency and noisy cross-modal attention. To address these limitations, we propose CRISP, a pre-LLM yet text-driven visual token pruning framework that preserves both instruction-relevant evidence and essential scene context. CRISP works in a two-stage pipeline: Stage 1 first identifies text-aligned visual tokens, and Stage 2 enhances contextual completeness through semantic diversity. Extensive experiments on LLaVA-1.5 and LLaVA-NeXT demonstrate that CRISP achieves superior performance retention under aggressive pruning ratios, maintaining up to 99.5% accuracy while reducing inference cost and latency by more than 2 times. CRISP serves as a practical solution for efficient LVLM inference, especially in resource-constrained scenarios.
[CV-164] RegionFM: Interpretable Region-Based Brain MRI Classification Using Foundation Model Embeddings
链接: https://arxiv.org/abs/2607.16325
作者: Wei Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:
Abstract:Foundation models provide powerful representations for brain MRI analysis, but their predictions remain difficult to interpret in anatomically meaningful terms. Clinical assessment of brain MRI is commonly organized around anatomically defined structures and regional abnormalities, whereas conventional explanation methods typically produce voxel- or patch-level importance maps that do not explicitly quantify the contributions of individual brain regions. To address this mismatch, we propose RegionFM, an interpretable framework that integrates anatomical segmentation with brain MRI foundation-model embeddings. RegionFM first divides each MRI scan into anatomical regions and constructs a separate MRI volume for each region. A frozen foundation model then encodes each region into an embedding, and a region-additive logistic model combines these embeddings such that every anatomical region contributes an explicit scalar term to the final prediction. This formulation supports both subject-level and cohort-level analyses of regional contributions. We evaluate RegionFM on cognitive-impairment classification using embeddings from multiple pretrained brain MRI foundation models. The results show that RegionFM maintains performance comparable to less interpretable fine-tuning approaches while providing anatomically grounded explanations. Randomized embedding ablations yield near-chance performance, indicating that the predictions rely on meaningful structure captured by the foundation-model embeddings rather than simple feature statistics. Overall, RegionFM better aligns model explanations with anatomy-based clinical reasoning while maintaining competitive predictive performance.
[CV-165] SGMCE: Segment-Grounded Morphological Concept Explanation for Malaria Parasite Species Identification in Thick Blood Smears
链接: https://arxiv.org/abs/2607.16324
作者: Ahmed Tahiru Issah,Charles B. Delahunt,Carine Mukamakuza
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Accepted for publication in the 30th Conference on Medical Image Understanding and Analysis (MIUA 2026), Dublin. To appear in Springer Lecture Notes in Computer Science (LNCS)
Abstract:Malaria diagnosis in endemic regions depends on species-level identification of Plasmodium parasites in thick blood smears, but deep learning detectors classify detections without providing morphological evidence for their predictions, limiting the ability of microscopists to audit those predictions at the case level. We present SGMCE (Segment-Grounded Morphological Concept Explanation), a post-hoc explanation framework that requires no additional training, no morphological annotations, and no labelled explanation data, yet produces per-detection natural-language explanations anchored in thick-smear morphology. For each detection, SGMCE extracts mask-guided crop thumbnails, computes fourteen handcrafted computer-vision morphological features (shape, colour, chromatin, haemozoin pigment) using adaptive within-mask thresholds, and queries GPT-4o with both visual evidence and computed measurements, conditioned on a thick-smear-specific knowledge base compiled from the World Health Organization bench aids. The primary output is a structured explanation identifying which morphological features support the detected species and why the competing species are excluded. Explanations are validated by four automatic metrics: Knowledge-Base Consistency (KBC), CV-Claim Faithfulness (CCF), Discriminativeness Score (DS), and LLM-as-Judge (LLMj). A sentence-level semantic scoring rule with species-aware negation filtering resolves the vocabulary mismatch between clinical prose and knowledge-base terms. Across 737 detections from 139 thick-smear images spanning four Plasmodium species and white blood cells, parasite-class mean KBC is 0.91, mean DS is 0.99, and mean CCF is 0.97, while a per-rule CCF breakdown confirms that the CV-grounded claims made by the vision-language model are consistent with the measurements they cite.
[CV-166] GMoT: Gated Motion-Aware Tokenization for Fine-Grained Micro-Gesture Video Reasoning with Multimodal LLM s ACM-MM2026
链接: https://arxiv.org/abs/2607.16322
作者: Taorui Wang,Wei Xia,Hui Ma,Zijia Song,Jiayu Zhang,Zeheng Wang,Yong Xu,Zitong Yu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to ACM MM 2026
Abstract:Micro-gesture recognition demands the detection of fleeting, spatially localized movements that are frequently overwhelmed by dominant static appearances and background noise. While Multimodal Large Language Models (MLLMs) excel at general video understanding, they inherently struggle with subtle kinematics and often rely on static posture priors. To this end, we propose GMoT, a Gated Motion-Aware Tokenization module that explicitly distills sparse kinematic evidence into a compact sequence prior to temporal modeling. GMoT dynamically spotlights action-relevant regions via spatially weighted pooling, extracts adjacent-frame temporal differencing to capture precise motion energy, and adaptively fuses these cues into the visual stream using a conservatively initialized semantic gate. To transition from simple classification to evidence-grounded reasoning, we further introduce a progressive reward-guided policy refinement paradigm, supported by a semi-supervised annotation pipeline that generates anatomically focused captions. Beyond achieving the best Top-1 accuracy among the compared methods on iMiGUE (67.32%) and SMG (73.11%), improving the Qwen3-VL-8B baseline by +6.80 and +3.11 points, our framework introduces Body-Region Grounding (BRG) Recall as an anatomical-grounding proxy conditioned on correct predictions, together with an overlapping-label cross-domain transfer protocol between iMiGUE and SMG. Extensive evaluations demonstrate that our GMoT-augmented model improves in-domain accuracy, retains clear gains under label-preserving corruptions, and improves accuracy-oriented cross-domain transfer under explicit small-split caveats while maintaining high anatomical grounding in its generated rationales.
[CV-167] he Devil is in the Dark Pixels: Toward Brightness Bias-Robust Denoising
链接: https://arxiv.org/abs/2607.16320
作者: Sungjun Cho,Zhuangzhuang Chen,Xiaomeng Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 6 figures, 4 tables. Code: this https URL
Abstract:In this paper, we reveal an important yet overlooked problem in image denoising: under signal-dependent camera noise models, dark regions suffer from inherently low Signal-to-Noise Ratio (SNR), as signal intensity decays far faster than noise variance diminishes, making detail recovery in dark areas fundamentally challenging. Yet rather than compensating for this difficulty, MSE-trained denoisers exacerbate it – reconstructing dark pixels up to 6x worse relative to their per-band noise floor. This bias stems from two compounding factors: signal-dependent noise inflates bright-pixel residuals, and the network’s Jacobian norm increases monotonically with brightness. Together, these cause bright regions to chronically dominate gradient updates at the expense of dark ones. To this end, we propose Brightness Bias-Robust Denoising (BBRD), a drop-in replacement for MSE loss that partitions pixels into brightness bands, normalizes per-band error by empirical noise variance, and applies Group Distributionally Robust Optimization (Group-DRO) to dynamically upweight whichever band is currently worst, with zero additional parameters or inference cost. Across 8 architectures and 2 datasets in our experiments, BBRD is the only method among 13 tested alternatives that improves each brightness band simultaneously, achieving up to +0.45 dB on dark bands, +0.32 dB on bright bands, and +0.65 dB aggregate Peak Signal-to-Noise Ratio (PSNR) on SIDD, with the largest per-band gains in the darkest regions where detail recovery matters most. Code is available at this https URL
[CV-168] Deep Learning-based Filtering for Video Coding: A Survey on Architectures Algorithms and Complexity Analysis
链接: https://arxiv.org/abs/2607.16319
作者: Young-Woon Lee,Byung-Gyu Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:As Ultra-High-Definition (UHD) displays and immersive media services become ubiquitous in the Internet of Things (IoT) and Consumer Electronics (CE) sectors, including 8K display and mobile devices, the demand for high-efficiency video coding is unprecedented. While Deep Learning-based Filtering (DLF) has emerged as a promising solution to mitigate compression artifacts inherent in standards like High Efficiency Video Coding (HEVC/H.265) and Versatile Video Coding (VVC/H.266), its deployment in CE devices is severely constrained by computational complexity, memory bandwidth, and power consumption. To bridge the gap between academic research and practical deployment, this paper presents a comprehensive, hardware-oriented survey of DLF techniques. We propose a systematic three-dimensional taxonomy classifying methods into (1) Integration Scheme within the Video Coding, (2) Coding Information Utilization, and (3) Network Design Strategy. Unlike prior reviews, this work critically analyzes the trade-offs between Rate-Distortion (RD) performance and hardware feasibility, highlighting the evolution from heavy, performance-oriented models to lightweight, hardware-friendly architectures targeting Neural Processing Units (NPUs). Furthermore, we incorporate the latest standardization activities from the Joint Video Experts Team (JVET) on Neural Network-based Video Coding (NNVC) to provide realistic guidelines. We also identify open challenges such as real-time inference latency and error propagation, providing a roadmap toward robust, low-power intelligent video coding in next-generation CE vision endpoints.
[CV-169] Monte Carlo Dropout Uncertainty and Entropy-Thresholded Selective Prediction for Architecture-Agnostic Brain Tumor MRI Triage
链接: https://arxiv.org/abs/2607.16317
作者: Medhansh Sharma
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 18 pages, 8 figures, 5 tables
Abstract:Deep networks now subtype brain tumors on MRI about as well as specialist readers, yet accuracy is not what keeps them out of the clinic. What matters at the point of care is whether a model’s confidence can be trusted to flag the cases it is likely to misclassify and defer them to a human. Deterministic estimates cannot: an auxiliary confidence head trained alongside the classifier collapses to a near-constant output that says nothing about correctness. This study proposes an uncertainty-first pipeline for four-class brain tumor MRI (glioma, meningioma, pituitary, no tumor) that reads predictive uncertainty from Monte Carlo (MC) Dropout over T = 20 passes and turns the resulting entropy into an explicit rule for deferring uncertain cases to a radiologist. We partitioned 7,200 images by perceptual-hash cluster, closing the near-duplicate leakage that inflates accuracy under naive splitting, and evaluated the pipeline on ViT-B/16 and ResNet-50 across five seeds along three axes: discrimination, calibration, and selective prediction. Both discriminate strongly (macro-AUC 0.994; accuracy 0.962 and 0.964), and no seed separates them (0 of 5 significant, p 0.05), so the result is driven by the uncertainty pipeline, not the network. A single temperature scalar pulls the deterministic softmax into tight calibration (expected calibration error 0.016-0.020), and deferring the most uncertain 5% of cases lifts accuracy on the rest to about 0.98 on both (area under the risk-coverage curve 0.010-0.011). MC-Dropout uncertainty here is thus calibrated, non-collapsing, and directly actionable through a concrete deferral rule, providing an architecture-agnostic basis for calibrated, defer-to-human brain tumor MRI triage under internal validation.
[CV-170] Eddy-VL 1.9B: Structural Pruning and Layered Distillation for Edge-Deployable Multimodal Embedding
链接: https://arxiv.org/abs/2607.16316
作者: HanYeong Cho,Changwoo Kim,Taeuk Chu,Jimin Park
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11pages,4figures,Model weights and inference code are available on Hugging Face
Abstract:In this report, we introduce Eddy-VL 1.9B, a compressed multimodal embedding model built on Qwen3-VL-Embedding-2B for offline, edge-deployable vision-language retrieval. Eddy-VL targets air-gapped forensic and investigative settings where cloud APIs are unavailable and low latency is essential. Compression combines (i) probe-driven structural pruning that removes four redundant text-decoder layers (28 to 24) ranked by adjacent-layer linear CKA, and (ii) layered knowledge distillation with hole-covering teacher-student mappings, mid-layer attention-map 1-CKA, and final-layer MSE and cosine losses with Matryoshka dimensions 128, 256, 512, 1024, 2048. The released model contains 1,926,188,032 parameters (3.85 GB bf16), representing approximately 9.5% fewer parameters than the 2.13B teacher model. Empirical evaluations on MMEB-V2 (78 tasks, VLM2Vec protocol) show that Eddy-VL achieves an overall score of 63.2 compared with 68.9 for the teacher, retaining 91.7% of the teacher’s performance while recovering 6.4 of the 12.1 points lost through pruning alone (56.8). Compositional reasoning performance remains close to the teacher on SugarCrepe (86.1 vs. 86.4), MR2-Bench (24.5 vs. 24.7), and ARO (59.5 vs. 60.4), while Winoground group performance (6.8 vs. 8.5) remains the primary limitation. Depth pruning also reduces forward latency by approximately 10% (150.0 to 136.4 ms per image on NVIDIA DGX Spark using FlashAttention-2). We present the architecture, compression methodology, training procedures, and evaluation results, demonstrating the effectiveness of Eddy-VL for multimodal retrieval under constrained edge deployment. Model weights and inference code are publicly available on Hugging Face.
[CV-171] Depth-Regularized JEPA World Models Learn More Transferable Representations from Real Outdoor Robot Data
链接: https://arxiv.org/abs/2607.16314
作者: Usman M. Khan
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO)
备注: 13 pages, 3 figures
Abstract:World models, especially based on JEPA architectures, have been shown to learn robust dynamics of various environments. However, learning from visually complex real-world data remains a challenge, especially in unpredictable outdoor environments. We introduce depth as a geometric prior during training in learning more robust latent dynamics directly from robot video data and handling visual complexity. This combines depth supervision with an isotropy-inducing latent regularizer (SIGReg), maximizing task-agnostic latent diversity while constraining how that diversity is organized, with the combined objective targeting the highest-entropy representation consistent with scene geometry. To satisfy this greater complexity without increasing inference time, we also add training-only overparameterization. Training an 18M-parameter model on video from a real agricultural robot, we evaluate with frozen-representation visual odometry probes, predictor-based surprise detection, and multi-step latent rollout fidelity. Compared to the baseline LeWM, our method lowers visual odometry probe error by 33%, substantially increases surprise-score separation both in-domain and on the out-of-domain TartanGround benchmark, and improves multi-step rollout fidelity under domain shift, with gains that grow with rollout horizon. Notably, we also see improvements in surprise-score separation on physics understanding that is not directly tied to 3D geometry, such as lighting and shadows. These results show that a lightweight training-time geometric prior makes a compact JEPA world model more useful and more transferable on real outdoor data with strong underlying representations, without adding inference overhead. Our work suggests that depth as a physically grounded prior can enhance world model generalization on a variety of tasks.
[CV-172] xperception – Making Robotic Grasping Easier
链接: https://arxiv.org/abs/2607.16312
作者: Matteo Bortolon,Andrea Caraffa,Alice Fasoli,Fabio Poiesi
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: White paper published for Ital-IA 2026
Abstract:The transition toward high-mix low-volume manufacturing demands flexibility in robotic manipulation. However, conventional vision systems remain a bottleneck, requiring extensive data collection and model retraining whenever a new object is introduced to the production line. To overcome this rigidity, we present xperception, a zero-shot 6D pose estimation technology that eliminates the need for object-specific fine-tuning and laborious data annotation. By directly utilizing typical CAD models and integrating the rich semantic features of foundation models (e.g. DINOv2, GeDi), xperception achieves millimeter-accurate 6D pose estimation. xperception showed robustness against severe occlusions in industrial tasks like bin picking and is engineered for deployment on industrial edge hardware, such as NVIDIA Jetson Thor. Validated at a TRL of 6, the core methodology behind xperception is based on the FreeZe algorithm, which won the international BOP Challenge 2024, paving the way for scalable, plug-and-play robotic automation in unstructured high-mix low-volume manufacturing industries.
[CV-173] Seeing What Is Actually There: PriVE-Bench and PriVE-Tools for Counterfactual Evaluation of Agent ic Visual Evidence in VLMs
链接: https://arxiv.org/abs/2607.16311
作者: Jingyu Sun,Jiachen Tu,Yuyang Xue,Yaoxin Jiang,Guoyi Xu,Zhengtao Yao,Rui Qian,Yizheng Sun,Hongpeng Zhou,Jingyuan Sun,Yan Lin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Vision-language models (VLMs) often answer visual questions using learned language and category priors rather than grounding their predictions in the image itself. Counterfactual images provide a natural diagnostic setting for this failure mode: when visible evidence contradicts what is usually true, a grounded model should answer from the pixels, while a prior-following model will produce a canonical but visually incorrect response. However, existing counterfactual benchmarks mainly ask whether such prior-following behavior exists. In this paper, we ask a further question motivated by the rise of tool-augmented and agentic vision systems: can additional visual evidence views help VLMs reason against their priors? We introduce PriVE-Bench, a Prior-vs-Visual Evidence Benchmark that uses paired original and counterfactual images to distinguish visually grounded answers from prior-consistent errors. We further introduce PriVE-Tools, a controlled agentic-vision-inspired extension that evaluates whether tool-derived visual evidence – including bounding boxes, crops, zoom panels, and contours – improves grounding under the same counterfactual conflicts. Across open- and closed-source VLMs, we compare raw, paired-image, and tool-conditioned inputs using accuracy, prior-following error rate, and other-response rate. Our results show that visual evidence tools can help in some settings, especially when models can use localized evidence effectively, but they are not a universal remedy: several models continue to follow language and category priors even when relevant visual evidence is explicitly provided.
[CV-174] SaaF: Scene-Specific Ambiguity-Aware 3D Language Fields towards Interactive Real-World Object Retrieval IROS2026
链接: https://arxiv.org/abs/2607.16309
作者: Yuga Yano,Daiju Kanaoka,Hakaru Tamukoh,Yasutomo Kawanishi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 8 pages, accepted as a conference paper for IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS2026)
Abstract:We propose Scene-specific Ambiguity-aware 3D Language Fields (SaaF), a novel Gaussian Splatting-based 3D language field designed for interactive object retrieval in a given real-world scene. Interactive object retrieval using natural language is a crucial capability for service robots operating in complex real-world environments. While recent 3D language field methods for object retrieval establish associations between rendered pixels and autoencoder-compressed CLIP features, they suffer from two limitations: (1) reduced discriminability among similar objects due to feature compression, and (2) poor handling of ambiguous queries, often resulting in unstable or incorrect retrieval. To address these limitations, SaaF introduces a metric learning strategy to construct a unified feature space that is both instance-discriminative and ambiguity-aware. (i) To enhance instance-level visual discrimination, SaaF employs metric learning that pulls image features from multiple viewpoints of the same object closer together in the feature space. (ii) To establish ambiguity awareness, the model jointly trains on multiple text labels generated by the proposed method from each tracked object image sequence, including ambiguous descriptions, to learn the semantic relationships between ambiguous and specific features in a target scene. This feature space enables fine-grained visual understanding while allowing the system to estimate query ambiguity and interactively request clarification when needed. Experimental results demonstrate that SaaF not only improves retrieval accuracy over previous methods but also robustly detects and handles ambiguity in the user text queries under open-vocabulary settings.
[CV-175] DAUPNet: Domain-Aware Uncertainty Modeling for Reliable Prototype Discrimination in Cross-Domain Few-Shot Semantic Segmentation
链接: https://arxiv.org/abs/2607.16308
作者: Lei Yuan,Zhongxu Hu,Jingyi Wen,Pengxing Yi
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Cross-domain few-shot semantic segmentation (CD-FSS) has predominantly been formulated as learning domain-invariant representations or improving support-query correspondence. Nevertheless, large domain shifts still make prototype matching unreliable: inconsistent hierarchical responses corrupt the support representation, deterministic prototypes cannot express boundary and appearance ambiguity, and treating prototypes with different reliability equally during optimization weakens foreground-background separation. We therefore propose DAUPNet, a unified framework that reformulates cross-domain prototype matching as uncertainty-aware prototype discrimination. DAUPNet first harmonizes hierarchical support-query features to provide stable evidence, then represents foreground and background prototypes probabilistically, and finally uses their estimated uncertainty to regulate contrastive optimization. On four standard target domains, DAUPNet achieves 72.6% and 76.7% average mIoU in the 1-shot and 5-shot settings, respectively, including substantial gains on the two medical domains. These results demonstrate that modeling prototype uncertainty and incorporating it into optimization provides a robust and interpretable approach to CD-FSS under severe domain shift. The code is available at this https URL
[CV-176] CoBind: Stage-Aware Compositional Binding for Training-Free Text-to-Image Generation
链接: https://arxiv.org/abs/2607.16307
作者: Kaijie Chen,Ethan Caldwell,Mira Vossen,Julian Hartwell,Serena Whitlock,Adrian Bellamy
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages
Abstract:Diffusion-based text-to-image models often fail on complex prompts involving multiple entities, attributes, and relations, producing object omissions, incorrect attribute assignments, or reversed spatial layouts. Existing training-free methods mainly strengthen token-level attention, but do not explicitly model which attributes belong to which entities or when different constraints should be enforced during denoising. We introduce \textbfCoBind, a training-free framework for stage-aware compositional binding. CoBind parses a prompt into a composition graph of entities, attributes, and relations. It first establishes the global layout using entity-completeness and relation constraints, then binds attributes to their target entities through contrastive cross-entity optimization. Structural guidance is gradually relaxed in later denoising steps to preserve textures and visual details. CoBind also adapts the guidance strength according to the current satisfaction of each constraint, reducing unnecessary latent updates. CoBind requires no retraining or additional annotations. Experiments on T2I-CompBench++, GenEval, and multiple diffusion backbones show consistent improvements in attribute binding, spatial relations, and complex compositional generation while maintaining competitive visual quality. Comments: 22 pages Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.16307 [cs.CV] (or arXiv:2607.16307v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2607.16307 Focus to learn more arXiv-issued DOI via DataCite
[CV-177] LookME: Lookup-Based Multimodal Embeddings for Layer Injection in Vision-Language Models
链接: https://arxiv.org/abs/2607.16305
作者: Zeyu Xu,Xingzhong Hou,Pengkai Guo,Siling Lin,Xiao Xu,Menghua Zhai,Haoyu Chen,Yunke Zhang,Fei Huang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Vision-Language Models (VLMs) have achieved strong progress in multimodal understanding. However, scaling dense or sparse Mixture-of-Experts (MoE) models to improve performance limits deployment in resource-constrained environments due to the trade-off between high memory usage from full loading and increased latency from on-demand loading. Recently, the Per-Layer Embedding (PLE) architecture addresses this by scaling models with large external embedding tables stored in ROM and performing lightweight lookup to retrieve relevant embeddings to enhance token representations. Nevertheless, existing PLE-style methods are primarily designed for text embeddings due to the convenience of ID-based retrieval, limiting their effectiveness in VLMs where multimodal embeddings contain richer information for visual tasks. In this paper, we propose LookME, the first framework that enables lookup-based enhancement for multimodal embeddings in VLMs while supporting partitioned storage and on-demand loading. To efficiently lookup arbitrary continuous multimodal embeddings from large-scale embedding tables, we propose a hierarchical two-level lookup method employing a coarse-to-fine strategy that performs lookups from the scene-level to the intra-scene primitive-level. Furthermore, we integrate the lookup method with a sparse injection strategy, which adaptively prioritizes critical embeddings over voluminous multimodal embeddings within layers, and facilitates embedding table reuse across neighboring layers, improving the trade-off among efficiency, model size, and performance. Experiments on multiple visual benchmarks show that LookME outperforms text-only PLE-style methods, validating the effectiveness of lookup-based multimodal embedding enhancement.
[CV-178] Dual-Domain Self-Supervised Artifact Removal Framework for Photoacoustic Computed Tomography
链接: https://arxiv.org/abs/2607.16304
作者: Yucheng Zhou,Shuang Li,Yu Zhang,Yibing Wang,Chulhong Kim,Seongwook Choi,Changhui Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 13 pages, 9 figures
Abstract:Photoacoustic Computed Tomography (PACT) often faces severe challenges from reconstruction artifacts due to sparse detection conditions. In this work, based on the distinct differences in artifact patterns between back-projection-based and Fourier-based reconstruction algorithms, we propose a self-supervised artifact removal framework that employs a lightweight Siamese Neural Network and a composite loss function integrating cross-domain fidelity and uncertainty-weighted consistency, effectively decoupling dual-domain features and filtering artifacts. Comprehensive validations using simulations, phantoms, in vivo rat and human experimental data demonstrate that the proposed method can significantly suppress image artifacts. Furthermore, enabled by the acceleration of the spatial-domain and frequency-domain inverse operator, this end-to-end approach also achieves exceptional computational efficiency.
[CV-179] Med-OPD: Improving Medical Vision-Language Models via Evidence-Aware On-Policy Distillation
链接: https://arxiv.org/abs/2607.16303
作者: Yunhang Qian,Jiaquan Yu,Jiawei Liu,Meng Wang,Hongwei Bran Li,Xiaobin Hu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Medical Vision-Language Models (Med-VLMs) require reliable reasoning from fine-grained visual evidence, yet existing models can produce plausible clinical answers by relying on language priors or medical templates rather than truly attending to diagnosis-critical regions. On-Policy Distillation (OPD) offers dense token-level supervision on student-generated trajectories and provides a privacy-compatible means of capability transfer without requiring the redistribution of raw patient data. However, standard OPD uniformly distills all tokens, causing sparse evidence-dependent tokens to be diluted by abundant clinical narrative tokens. Inspired by the success of OPD in the large language model community, we propose \textbfMed-OPD, to our knowledge the first unified post-training framework that integrates on-policy distillation with medical evidence-aware supervision for Med-VLMs. We introduce \textbfMedical Evidence Advantage (MEA), a teacher-grounded counterfactual signal that uses an answer-aware hint to focus teacher scoring on evidence supporting the target diagnosis, and measures each token’s dependence on medical visual evidence by comparing teacher likelihoods under the original and evidence-degraded imaging modalities. Based on MEA, Med-OPD redistributes the distillation signal at both the token and trajectory levels, emphasizing diagnosis-critical tokens and evidence-reliant rollouts. Experiments on OmniMedVQA subsets show that Med-OPD consistently outperforms SFT and standard OPD across CT, MRI, Disease Diagnosis, and Lesion Grading. These results demonstrate that evidence-aware distillation can better strengthen medical VLMs’ reliance on key visual evidence and improve reliable multimodal medical reasoning. The source code and data is publicly available at: this https URL
[CV-180] FedDP-PALD: A Privacy-Preserving Federated Latent Diffusion Framework with Prototype Aggregation for Medical Data Synthesis
链接: https://arxiv.org/abs/2607.16300
作者: Md. Sajeebul Islam Sk.,Khan Enaet Hossain,Md. Mehedi Hasan Shawon
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Medical images and physiological signals provide valuable information for accurate diagnosis. Developing diagnostic models often requires patient data from multiple institutions, although strict privacy regulations limit the sharing of sensitive clinical records. Federated learning enables multiple hospitals to train a shared model without exchanging raw data. However, existing methods face two problems: the information exchanged during training can reveal whether a patient’s data were used, and synthetic data meant to replace real records often fail to preserve their predictive structure, which limits clinical use. To address this issue, we propose FedDP-PALD, a privacy-preserving federated latent diffusion framework for multimodal medical data synthesis under formal privacy guarantees. It jointly processes chest X-ray images and electrocardiogram (ECG) signals through gated multi-head attention with modality-availability masks, remaining effective even when a modality is missing. We also introduce Differentially Private Prototype Mixture Aggregation (DP-PMA), which clips class-level latent prototypes and adds calibrated Gaussian noise before combining them on the server to maintain (\epsilon, \delta) differential privacy. We evaluate FedDP-PALD on PneumoniaMNIST, ChestMNIST, and MIT-BIH datasets, where differential privacy reduced summary-level attack AUROC from 0.6229 \pm 0.0026 to between 0.5016 and 0.5093 for privacy budgets from \epsilon = 1 to \epsilon = 8 . On the test data, synthetic-latent training achieved an F1 score of 0.8993 \pm 0.0006 and an AUROC of 0.9057 \pm 0.0503, close to the 0.9747 \pm 0.0132 real-latent training. These results show that FedDP-PALD generates private synthetic representations that preserve useful decision performance while strongly resisting membership inference.
[CV-181] Rethinking Feature Reliance Evaluation with Semantically Matched Suppression
链接: https://arxiv.org/abs/2607.16298
作者: Ning Jiang(1 and 2),Tianyi Luo(4),Zhengyong Huang(1 and 2),Yao Sui(1 and 2 and 3) ((1) Institute of Medical Technology, Peking University Health Science Center, Beijing, China, (2) National Institute of Health Data Science, Peking University, Beijing, China,(3) Institute for Artificial Intelligence, Peking University, Beijing, China,(4) School of Computer Science and Engineering, Sun Yat-sen University, Guangzhou, China)
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Understanding whether visual recognition models rely on shape, texture, or color is central to interpreting their behavior. Prior cue-conflict studies have strongly influenced the view that CNNs are texture-biased, yet such tests measure cue preference under artificial conflicts rather than feature reliance during natural recognition. We revisit this question through controlled feature suppression and show that performance drops are difficult to interpret unless different suppression operations impose comparable category-level damage. We introduce a semantically matched evaluation framework that compares shape and texture suppression at matched levels of category separability loss. Under this framework, ImageNet-trained CNNs show stronger degradation under texture suppression than under shape suppression, revealing greater texture reliance than suggested by unmatched suppression analyses. Extending the comparison across architectures, we find that Vision Transformers retain higher accuracy than CNNs under both shape and texture suppression. Brain encoding further shows that ViT representations exhibit smaller suppression-induced decreases in neural prediction performance under the tested suppression settings. These findings indicate that semantic comparability is essential for interpreting feature reliance from suppression experiments, and suggest that the robustness advantage of ViTs may be related to representations more compatible with human visual cortex.
[CV-182] Emergent Hierarchical Monosemantic Neurons from the Group-Contrastive Forward-Forward Algorithm
链接: https://arxiv.org/abs/2607.16295
作者: Yiming Tang,Qinglin Qi,Zhaoqian Yao,Harshvardhan Saini,Dianbo Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Mechanistic interpretability has made significant strides in understanding neural network representations, with sparse dictionary learning (SDL) methods, most prominently sparse autoencoders, as a central paradigm. However, recent work has reported several limitations of this paradigm: SDL objectives are non-identifiable; SDL methods rely heavily on the Linear Representation Hypothesis; and a growing body of evidence points to concepts that are encoded non-linearly and are therefore not expressible as any single direction. We hypothesise that a different route to monosemanticity is available. Biological visual systems exhibit highly selective neurons organised into hierarchies of increasing abstraction, and this organisation emerges from local, layer-wise learning rules rather than from a global error signal; we therefore ask whether a biologically plausible learning algorithm will likewise yield monosemantic neurons. To test this, we propose Group-Contrastive Forward-Forward (GCFF), a forward-forward training algorithm that combines class-specific routing with within-class contrastive objectives, reaching monosemanticity through architectural constraints rather than sparsity. Because GCFF attaches multiple non-linear layers to the representation under study, its neurons can therefore capture the non-linear concepts. On CLIP representations, a single trained GCFF module recovers monosemantic neurons whose abstraction increases progressively with depth, reaching environmental properties that hold independently of an image’s foreground, without any sparsity constraint or supervision of abstraction level. We further demonstrate that GCFF can train networks from scratch, achieving state-of-the-art performance among forward-forward algorithms on various image classification benchmarks.
[CV-183] A2BM: Alignment-Aware Bridge Matching for Image-to-Image Translation
链接: https://arxiv.org/abs/2607.16294
作者: Aimi Okabayashi(UBS Vannes),Georges Le Bellier(LIP, CEDRIC - VERTIGO),Nicolas Audebert(LaSTIG, IGN, CEDRIC - VERTIGO),Charlotte Pelletier(OBELIX),Thomas Corpetti(LETG - Rennes),Nicolas Courty(OBELIX)
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Paired image-to-image translation underpins a wide range of computer vision tasks, including image editing, sensor translation, and domain adaptation. Bridge matching and flow matching have recently emerged as powerful frameworks, extending diffusion models to arbitrary source and target distributions. However, their standard formulations assume perfectly aligned training pairs, treating all source-target correspondences as equally reliable. In practice, real-world applications often involve weakly aligned pairs due to changes of acquisition conditions, including e.g. asynchronous captures, different illuminations, or misregistration. In this work, we introduce Alignment-Aware Bridge Matching (A ^2 BM), a bridge matching method that leverages image pairs alignment during training. By incorporating alignment scores, the model learns to disentangle true semantic correspondences from misalignment artifacts. At inference time, we use the alignment score as a control variable over translation fidelity, with strongly aligned outputs obtained when prompting the model with the highest alignment score. We validate A ^2 BM on both controlled synthetic experiments and on challenging real-world tasks, including cross-sensor super-resolution and pixel-space unsupervised domain adaptation. In all settings, A ^2 BM consistently improves translation fidelity over strong GAN-, diffusion-, and Schrödinger bridge-based baselines, establishing alignment conditioning as a principled solution for image translation models with weakly aligned data.
[CV-184] SLT: Robust Quantum Neural Networks for Noisy-Label Medical Image Classification via Supermartingale-based Label Transition
链接: https://arxiv.org/abs/2607.16293
作者: Jun Zhuang,Mohammad Al Hasan,Yiyu Shi,Chaowen Guan
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Preprint, under review
Abstract:Noisy-label learning in small-scale medical image classification is challenging and hinders the superiority of deep neural networks. Recent studies suggest that quantum neural networks (QNNs) have shown potential in limited-data regimes, yet their use for noisy-label learning remains under-explored. A key obstacle is QNNs’ intrinsic “natural smoothness”, which may regularize training but also obscure high-confidence samples needed for noise-transition estimation. We propose Supermartingale-based Label Transition (SLT), an anchor-free loss correction framework for robust QNN-based medical image classification under noisy labels. SLT models entropy reduction in predictive distributions as a supermartingale and uses its monotonic behavior to identify stable transition-matrix refinement steps. This enables dynamic transition updates while reducing noise-driven oscillations during QNN training. We further provide a convergence analysis showing that the proposed transition-refinement process reaches a steady state. Experiments on multiple public small-scale medical image datasets demonstrate that SLT consistently improves QNN-based classification and stably outperforms classic noise-label learning baselines under synthetic and real-world label noise.
[CV-185] It Depends on the Dataset: When a Brain-Encoding Models Predicted Responses Beat Their Visual Backbone for Video Memorability
链接: https://arxiv.org/abs/2607.16292
作者: Carson Rodrigues
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 10 pages, 3 figures. Code and predicted-response arrays released for reproduction; source videos/scores not redistributed
Abstract:Brain-encoding foundation models predict fMRI responses to video, audio, and text well enough to win the Algonauts 2025 challenge. We ask whether their predicted responses, obtained with no scanner, are a useful feature lens for a downstream human-behavior task: forecasting the memorability of short videos. We project each clip into TRIBE v2’s predicted cortical space and forecast short-term memorability with ridge regression, against a matched control: the model’s own V-JEPA2 visual backbone taken before the brain projection. The answer is dataset-dependent, and cleanly so. Within Memento10k the backbone wins (Spearman 0.594 vs 0.544 for the brain projection); within VideoMem the brain projection wins (0.415 vs 0.368, delta +0.047, 95% CI [+0.009, +0.088]). Both within-dataset gaps have bootstrap intervals excluding zero, in opposite directions. Cross-dataset transfer inherits the split: trained on Memento10k and tested on VideoMem the brain projection beats the backbone (+0.076), while the reverse loses heavily (-0.311). Each representation transfers best onto the dataset it already fits better. The VideoMem advantage is not a sample-size artifact (it survives matched training size and a PCA-then-ridge pipeline) and not mere compression of the backbone (a compressed or heavily regularized backbone tops out below the brain projection, which also beats a transfer-tuned backbone, +0.053). So predicted-brain features carry a small but real memorability signal the backbone misses on one dataset and not the other: not a domain-general prior, but a dataset-specific representation. A vision-orthogonal component (partial Spearman 0.19, permutation p=2.5e-4) localizes to ventral occipito-temporal cortex. Code and predicted-response arrays are released; source videos and scores are not redistributed.
[CV-186] Beyond Target Scores: Measuring Off-Target Drift in Diffusion-Based Medical Image Editing
链接: https://arxiv.org/abs/2607.16291
作者: Todd Zhou
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Diffusion models can now edit medical images in visually plausible ways, but the standard evaluation question is too narrow: did the target score increase? In clinical imaging, target findings are entangled with co-morbidities, acquisition effects, and selection bias, so a model can appear successful by changing correlated non-target findings rather than isolating the intended pathology. We introduce CIB-Med-1, a trajectory-level benchmark for controlled biomarker editing in chest radiography. CIB-Med-1 evaluates directional pleural effusion editing through calibrated target progression, inversion rate, and off-target semantic drift over 14 clinically motivated nuisance axes. The benchmark exposes a reward-hacking failure mode in which diffusion editors increase effusion scores while simultaneously altering parenchymal, cardiomediastinal, pleural, chronic, or artifact-related findings. We further present a constrained diffusion guidance baseline that optimizes target progression subject to bounded off-target change. Across held-out radiographs, the constrained editor preserves target progression ( \rho_\mathrmtrend=0.88 vs. 0.90 for unconstrained guidance) while reducing median off-target drift from 0.46 to 0.20 and 90th-percentile drift from 0.98 to 0.33 . Drift magnitude tracks empirical target–off-target association, supporting the view that semantic instability is structured rather than incidental. A blinded human validation probe with radiology trainees further shows stronger agreement with intended progression orderings ( \tau=0.61 vs.\ 0.29 for Pix2Pix). These results argue that medical image editing should be evaluated as trajectory-level semantic control, not as endpoint score maximization.
[CV-187] Strength-Parity Ensembling with Parameter-Isolated Experts for Multi-Task Affect Recognition
链接: https://arxiv.org/abs/2607.16290
作者: Hong Hai Nguyen,Van Thong Huynh
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Leading entries on the multi-task track of the 11th ABAW challenge rely on heavy ensembling, yet which member is worth adding to an already strong ensemble is rarely made explicit. We study this question for joint valence-arousal estimation, 8-way expression recognition, and 12-way action-unit detection from a single unconstrained face, under partial, long-tailed labels and a rule that forbids pretraining on Aff-Wild2. Building on a shared affect-latent that marginalizes the missing labels across two affect-supervised backbones, we propose a strength-parity rule: an added member lowers the ensemble error only when it is both decorrelated from the current members and a near-peer of them in individual accuracy. The rule exposes a concrete obstacle, as on a single backbone re-seeding and even distinct fine-tuning curricula re-converge to a prediction correlation of 0.98 and add no diversity. Parameter-isolation removes it: confining each adaptation to a disjoint low-rank subspace of a shared backbone yields experts that stay decorrelated at 0.91 while remaining near-peers, the strongest of them an AffectNet-adapted expert. The resulting system raises the overall validation score to 1.6949, against the organizers ConvNeXt-with-MixAugment baseline of 0.45; with per-AU calibration and by pooling the shared-latent heads valence-arousal byproduct as a further near-peer, the strongest configuration reaches 1.7259.
[CV-188] A Synthetic 3D Gear Dataset for Manufacturing Quality Inspection (MFGNet-Gear)
链接: https://arxiv.org/abs/2607.16288
作者: Ruo-Syuan Mei,Chenhui Shao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Quality control in smart manufacturing increasingly relies on data-driven methods, particularly deep learning, to automate the inspection of manufactured parts. Recent advances in three-dimensional (3D) metrology have enabled fine-scale assessment of dimensional accuracy, surface quality, and shape conformity. However, deep learning methods for point-cloud-based inspection require large volumes of labeled data covering part designs and defect types, which are costly and time-consuming to obtain. Moreover, defective parts are intrinsically rare in mass production, and the resulting class imbalance can degrade model performance and make rare defect types difficult to detect. Synthetic data generation (SDG) offers a promising approach to address these challenges by producing large, balanced, and fully annotated datasets. Yet, applying SDG to precision components requires representing part geometry and defect morphology parametrically, so that design and quality can be co-varied. This article describes MFGNet-Gear, a publicly available synthetic 3D dataset comprising 24,000 paired polygon meshes and point clouds across 12 gear designs and 4 quality classes, with 500 instances per design-quality combination. Gear geometries are generated with parametric computer-aided design software, with dimensional parameters perturbed by \pm 0.0254 mm and defect parameters sampled from distributions representing defect morphologies. For each mesh, 100,000 points are uniformly sampled using Open3D and stored as N \times 3 coordinate text files. Metadata labels identify the gear design and quality class, supporting part design classification, geometric defect detection, representation learning, and dataset benchmarking. MFGNet-Gear provides an open-source dataset for deep learning-based 3D metrology, with a reproducible generation pipeline extensible to additional part designs.
[CV-189] Identity-Consistent Expression Fields: A Disentangled Neural Radiance Field Framework for Few-Shot Facial Expression Synthesis
链接: https://arxiv.org/abs/2607.16287
作者: Minh Tran
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:Neural Radiance Fields (NeRF) have enabled photorealistic novel-view synthesis of 3D scenes and, in the facial domain, have been extended to reconstruct and animate 3D face models from a small number of images. However, existing few-shot dynamic NeRF methods for facial expression editing typically warp a single learned feature volume conditioned on target expression parameters, which can cause identity-specific appearance details (skin texture, fine geometric structure) to drift when the model is driven toward expressions far from those seen in the few-shot input set. We propose Identity-Consistent Expression Fields (ICEF), a framework that explicitly disentangles a static, identity-specific radiance component from a dynamic, expression-conditioned deformation component, and introduces an identity preservation regularizer that constrains the deformation network to modify only expression-relevant regions while leaving identity-specific canonical appearance untouched. ICEF further incorporates a confidence-weighted conditional feature warping step that down-weights unreliable warps for target expressions that are far, in parameter space, from the observed few-shot inputs, mitigating artifacts observed in prior few-shot dynamic NeRF methods when extrapolating to novel expressions. We relate ICEF to prior few-shot dynamic NeRF, static 3D-aware face generation, and disentangled face-editing radiance field methods, and describe an evaluation protocol measuring both novel-expression rendering quality and, specifically, identity-consistency metrics across a range of expression-parameter extrapolation distances.
[CV-190] Depth Estimators Are Implicit Neural Fields for 3D Scene Geometry Inpainting and Reconstruction
链接: https://arxiv.org/abs/2607.16286
作者: Yingzhao Jian,Zihao Lin,Hehe Fan
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:The 3D geometry of real-world scene data is often incomplete. Mainstream methods use depth estimators to inpaint missing structure. However, their prediction results can be inconsistent with observed geometry, or unreliable on out-of-distribution data. To solve these problems, we propose Neural Depth Field (NDF). Our key insight is that a depth estimator can also be a scene-level implicit field. As an estimator, it adapts to the target domain by learning observed depth data. As an implicit field, it fits the existing geometry to maintain consistency. Under this view, NDF addresses both problems through a single test-time optimization. Experiments show that NDF produces high-fidelity and globally consistent geometry across diverse scene data, ranging from indoor scans to satellite imagery. It reduces cross-view inconsistency by 63.3% and improves inpainting accuracy by 23.1%, achieving state-of-the-art performance in 3D scene geometry inpainting. The code is available at: this https URL.
[CV-191] A Shared Latent for Partially-Labeled Multi-Task Facial Affect Recognition
链接: https://arxiv.org/abs/2607.16285
作者: Hong Hai Nguyen,Sy Phan Van,Soo-Hyung Kim,Van-Thong Huynh
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:Facial affect in the wild is naturally multi-task: valence-arousal, discrete expressions, and facial action units describe the same face. Yet real corpora annotate these tasks only partially and unevenly, so most systems mask the missing labels or impute pseudo-labels and forgo the cross-task signal. We instead cast partially-labeled multi-task learning as marginalization over a shared affect latent: one variational bottleneck mediates all three task decoders, so a frame annotated for one task shapes the representation the others use, and the masked objective reappears as the reconstruction term of an evidence lower bound. On s-Aff-Wild2, where only 37% of frames carry all three labels, the classes are severely imbalanced, and pretraining on the source data is disallowed, we isolate where this coupling acts. On a single backbone it lifts expression macro-F1 from 0.403 for a dedicated specialist to 0.446, which the masked-loss model does not reach; a second, near-peer backbone with decorrelated errors then breaks an action-unit ceiling that external action-unit data could not, while valence-arousal stays within noise. Every gain is disciplined by a matched-control negative; together these controls indicate that the rare-class failure is representational, not a matter of loss shaping. As each task’s source is chosen on the evaluation split, we report the assembled result, a combined multi-task score of 1.679 on validation, as an in-sample endpoint and rest our conclusions on the controlled comparisons; a small, regime-dependent transfer of the expression advantage to AffectNet and RAF-DB is presented as exploratory rather than conclusive.
[CV-192] MAC 2026: Advancing Micro-Action Analysis Towards Fine-Grained Understanding MICRO
链接: https://arxiv.org/abs/2607.16284
作者: Kun Li,Dan Guo,Jihao Gu,Pengyu Liu,Xiaobai Li,Haoyu Chen,Yanbin Hao,Guoying Zhao,Meng Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Multimedia (cs.MM)
备注: Challenge Summary Paper of the 3rd Micro-Action Analysis Grand Challenge (MAC 2026) at ACM Multimedia 2026
Abstract:Micro-Actions (MAs) are subtle and spontaneous human behaviors that provide important non-verbal cues in social interaction and affective communication. However, their short duration, weak motion patterns, and fine-grained semantic differences make them difficult to annotate, model, and evaluate in a standardized manner. To promote academic research on micro-action analysis, we proposed and have annually organized the Micro-Action Analysis Grand Challenge (MAC) as a public benchmark platform for this emerging field. The first two editions of MAC established standardized evaluation settings for micro-action recognition and detection, providing publicly accessible datasets and protocols. Building upon these editions, this paper presents the 3rd MAC, held in conjunction with ACM Multimedia 2026. Under the theme of moving from recognition to fine-grained micro-action understanding, this edition further expands the scope of the challenge beyond conventional recognition and detection. In particular, we introduce a new task named fine-grained micro-action understanding, evaluated with the assistance of multimodal large language models, aiming to assess models’ ability to capture fine-grained semantic cues and interpret subtle human micro-actions at a deeper level. We summarize the datasets, task settings, evaluation protocols, competition results, and representative solutions from top-performing teams. Finally, we discuss future directions for micro-action analysis and its broader role in human-centric video understanding.
[CV-193] GenSyn10: A Multi-Generative AI Dataset For Benchmarking Image Classification
链接: https://arxiv.org/abs/2607.16283
作者: Md Faraz Kabir Khan,Saeed Anwar,Ghulam Mubashar Hassan
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid advancement of generative AI has outpaced our ability to reliably detect its outputs, particularly when detectors encounter generators they have not seen before. We introduce GenSyn10, a CIFAR-10-aligned synthetic image dataset of 60,000 images (10 classes, 32 \times 32, 50k/10k split) generated using three architecturally diverse state-of-the-art models: FLUX.2-dev (Rectified Flow Transformer), HunyuanImage-3.0 (MoE Transformer), and Qwen-Image-2512 (Multimodal Diffusion Transformer), to advance research in AI-generated image detection. A central challenge in this domain is that detectors perform well on known generators but degrade on unseen ones. GenSyn10 addresses this limitation by curating data from multiple contemporary architectures under a standardized generation protocol, enabling controlled and systematic evaluation of out-of-distribution (OOD) generalization to novel generators. Images are generated using a template-based prompt engine and downsampled to ensure consistency. We evaluate 17 image classification models under a four-stage protocol: real-data baseline, zero-shot transfer, fine-tuning, and retention. Despite a measurable domain gap, CIFAR-10-trained models achieve up to 96.86% zero-shot accuracy on GenSyn10, increasing to 99.88% after fine-tuning. In binary real-vs-synthetic classification, fine-tuned models achieve 97-99.9% accuracy on seen generators but drop to 79-96% on images from an unseen generator, highlighting persistent limitations in OOD generalization. These results establish GenSyn10 as a controlled benchmark for studying synthetic image detection beyond single-generator settings, supporting research on robustness, domain adaptation, and cross-generator generalization.
[CV-194] Moving Like a Human: Ego-Motion-Normalized Temporal Signatures for Real-Time Aerial Person Tracking on Milliwatt-Class Hardware
链接: https://arxiv.org/abs/2607.16282
作者: Akbar Anbar Jafari,Cagri Ozcinar,Gholamreza Anbarjafari
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 20 pages, 10 figures
Abstract:Follow-me person tracking must run on the drone itself, where affordable companion computers offer only a few effective int8 GFLOP/s. At typical follow distances a person spans 10-60 pixels, indistinguishable from clutter and beyond the reach of single-frame appearance detectors. The missing evidence is temporal and belongs in the input representation, computed analytically, rather than in learned temporal machinery. EMTS-Det is a five-stage system that estimates ego-motion, converts each frame into ego-motion-normalized residual-motion channels, detects person centers with a 22k-parameter, 7.6-MFLOP network, tracks a locked target with a Kalman filter in stabilized coordinates, and verifies tracks with a 1-D convolutional classifier of human motion (ROC AUC 0.941). Training uses a synthetic-motion curriculum with motion channels generated by the deployed ego-motion code. Multi-seed ablations locate the value in generalization: on held-out VisDrone-DET a luminance-only variant collapses to 0.051 AP25 versus 0.415, as does YOLOv8n fine-tuned identically despite 1,100 times the compute, while the deployed int8 detector reaches 0.694 AP25 in-domain and 0.444 on this split. Temporal-shift modules lower accuracy, so the deployed detector is stateless. Silent int8 calibration failures are documented; min-max calibration with propagated caches matches float within 0.008 AP. On a Raspberry Pi Zero 2W the pipeline runs at 31.85 FPS with 0.462 AP25 and 0.714 recall over 1,000 real-world UAV videos, versus 1.95 FPS and 0.172 AP25 for YOLOv8n. A 57-second field sequence shows auto-lock at 1.3 s, 97.9% lock recall, and recovery from all nine occlusions with zero false re-locks.
[CV-195] 3D FaceShell: Attribute Transfer in 3D Face Avatars as a VLM Defense Mechanism ECCV2026
链接: https://arxiv.org/abs/2607.16280
作者: Weston Bondurant,Srijan Das,Hieu Le,Stephanie Schuckers
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ECCV 2026
Abstract:Photorealistic 3D face avatars are increasingly deployed as reusable digital assets across applications such as telepresence, animation, and personalized media. At the same time, vision-language models (VLMs) can infer sensitive attributes from rendered images with open-ended semantic reasoning without any fine-tuning. This creates a new privacy challenge: once a 3D face avatar is shared, any of its renderings can be analyzed to extract high-level facial attributes. Existing defenses largely operate in 2D image space and do not address identity-preserving semantic manipulation of 3D facial representations. We propose 3D FaceShell, a framework for steering VLM interpretations of faces rendered from 3D models while preserving geometric fidelity and facial identity. 3D FaceShell augments the original 3D representation with a learnable Gaussian shell that produces subtle, spatially distributed perturbations optimized through multi-view embedding alignment. The perturbations are designed to be visually inconspicuous yet sufficient to redirect VLM-based attribute inference in a view-consistent manner. Extensive experiments on reconstructed celebrity face avatars and multiple black-box VLMs demonstrate that 3D FaceShell significantly increases attribute injection and mismatch rates while maintaining high perceptual similarity and identity consistency. Our results show that it is possible to manipulate VLM-level semantic interpretation of 3D faces without compromising their human-recognizable appearance.
[CV-196] A Step Forward Towards Trustworthy Risk-Aware Facial Retrieval (RA-FR) ICPR2026
链接: https://arxiv.org/abs/2607.16279
作者: Muhammad Emmad Siddiqui,Muhammad Rafi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at the 28th International Conference on Pattern Recognition (ICPR 2026). Extended version with additional citations and expanded sections
Abstract:Facial image retrieval in unconstrained surveillance environments is a high-stakes challenge where missing a subject of interest – a single false negative – is simply not an option. Despite near-perfect performance on curated benchmarks, current recognition systems falter under real-world domain shifts such as low resolution, motion blur, and uncontrolled illumination (e.g., SCFace). Addressing this reliability gap, we propose Risk-Aware Facial Retrieval (RA-FR), a framework that moves beyond fixed Top- k retrieval to adaptive set generation, guaranteeing ground truth inclusion within a user-specified risk level ( \alpha ) and confidence level ( 1 - \delta ). Our approach integrates three core contributions: (1) reducing aleatoric uncertainty via a hybrid blind face restoration technique coupling Latent Consistency Models (InterLCM) and DiffBIR; (2) extracting discriminative, restoration-robust features via self-supervised DINOv1 ViT-B with GGeM pooling; and (3) employing conformal prediction with Hoeffding’s inequality to dynamically calibrate retrieval set sizes based on query uncertainty. On the IMFDB benchmark, it consistently satisfies a 5% risk target with an average retrieval set size of approximately 10 images. By unifying domain-specific restoration, robust representation learning, and provable decision rules, RA-FR offers a pipeline that makes facial retrieval in surveillance both reliable and auditable. The code is available at: this https URL.
[CV-197] he JEPA Predictor: A Transferable Operator for Occluded Feature Completion
链接: https://arxiv.org/abs/2607.16274
作者: William Nguyen,Christopher Nguyen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 3 figures, 5 tables
Abstract:Joint-Embedding Predictive Architectures (JEPAs) train a predictor jointly with their encoder, but downstream deployment discards the predictor and reads features from the encoder alone. The predictor is, by construction, a learned operator from visible-context features to features at masked positions, the structure a partial-view classifier needs. We show that this operator is portable across encoder families. We first establish that, at heavy mask, retaining the frozen predictor on a JEPA encoder substantially closes the accuracy gap against the strongest non-JEPA discriminative baselines. We then bolt the frozen predictors of I-JEPA and V-JEPA 2 onto four non-JEPA hosts (CLIP, DINOv3, DINOv2, MAE) through a single linear projection between feature spaces, fit in closed form on 500 ImageNet-1k images. Across both ImageNet-9 and Stanford Dogs and across three mask fractions, the lift over each host’s masked-encoder baseline grows monotonically with the mask fraction K in every host-donor pair. CLIP paired with the I-JEPA predictor recovers most of the accuracy that masking removed on ImageNet-9 at heavy occlusion, and lifts fine-grained Stanford Dogs from 15.9% to 52.1% (+36 pp). The mechanism is identifiable: the projection pays a fixed cost on visible patches and the predictor provides a growing benefit on masked patches; the benefit dominates the heavy-occlusion regime. At low K on fine-grained classification the projection cost exceeds the benefit, defining the boundary where the linear bridge breaks down. The frozen JEPA predictor functions as a portable operator for occluded feature completion across encoder families, requiring no retraining of either model while fitting matched linear probes per mask fraction.
[CV-198] ForensicNet: Lightweight Attention-Enhanced MobileNetV2 for Automated Face Identification
链接: https://arxiv.org/abs/2607.16273
作者: Savitha N J,Lata B T
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: DOI: this https URL Link: this https URL
Abstract:In forensic environments, automated identification of perpetrators is difficult due to pose changes, changes in light, occlusion, and lack of labeled data. This paper presents ForensicNet, a lightweight deep learning framework for forensic face recognition that enhances attention. The suggested model combines the MobileNetV2 backbone with Convolutional Block Attention Modules (CBAM) to improve the learning of discriminative features while maintaining computational speed. A two-phase transfer learning strategy with adaptive layer unfreezing is used to improve domain adaptation and reduce overfitting. This study used publicly available datasets such as LFW and SCFace, with 15,000 facial images spanning 68 identity classes. The proposed model outperforms baseline architectures such as AlexNet, ResNet-50, and MobileNetV2, with an accuracy of 92.4%, a precision of 90.8%, and a recall of 89.5%. Additionally, the framework requires only 2.1 GFLOPs per inference, and hence can be used in real-time forensic surveillance applications.
[CV-199] AdaSurvMamba: Dynamic Fusion and Semantic Scanning for Multimodal Survival Analysis MICCAI2026
链接: https://arxiv.org/abs/2607.16260
作者: Jialong Zhong,Tingwei Liu,Baokun Yue,Jingjing Li,Yongri Piao,Miao Zhang,Leiye Liu,Jiahong Jiang,Wei Ji,Huchuan Lu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: MICCAI 2026 Accept
Abstract:Multimodal survival analysis utilizing whole slide images (WSIs) and genomic profiles is fundamental for cancer prognosis. Recently, state-space models like Mamba have emerged as powerful tools for sequence modeling. However, translating this success to complex multimodal tasks is hindered by two critical limitations. First, conventional fusion strategies assume a static multimodal interaction strength, ignoring the fluctuating diagnostic importance of each modality across different patients and local regions. Second, the standard Mamba architecture processes tokens along predefined physical paths. This rigid scanning disrupts the semantic continuity of spatially scattered medical features and exacerbates long-range decay. To address these challenges, we introduce AdaSurvMamba as a novel adaptive framework for multimodal survival analysis. The framework features a Dual-Scale Importance-Aware Reconstruction (DSIR) module to dynamically modulate cross-modal interaction strength. It evaluates diagnostic importance at both the sequence and token levels to reconstruct the input representations. Furthermore, we propose a Semantic Aggregation Scanning (SAS) module to overcome contextual fragmentation. The SAS module dynamically reorganizes discrete tokens into semantically continuous sequences via a shared prototype pool. It explicitly modulates the state transition step size using global modality context and semantic priors to adaptively control the information absorption rate. Experiments across five TCGA cohorts demonstrate consistent gains over existing methods. Code is available at this https URL.
[CV-200] Self-Evolving Just-In-Time Memory for Proactive Embodied Safety
链接: https://arxiv.org/abs/2607.16247
作者: Bingrui Sima,Lizhong Wang,Xiaoya Lu,Kun He,Xiao Yang
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:While Vision-Language Models (VLMs) have empowered embodied agents to execute complex household tasks, they struggle to proactively handle dynamically emerging hazards during closed-loop interactions. Existing safety approaches often rely on runtime guardrails to block unsafe actions or induce excessive caution, which severely stalls task progress instead of actively resolving the underlying risks. To break this safety-progress trade-off, we introduce the Self-Evolving Just-In-Time Memory framework, which reframes embodied safety from progress-stalling guardrails to proactive hazard mitigation. The framework consists of a Risk-Sufficient Topological Belief Graph (RSG) for persistent safety-relevant state tracking under partial observability, an Agency-Grounded Factual Memory for precise hazard anticipation, and an Experience Memory that injects procedural Meta-Skills to guide executable, progress-preserving mitigation. Furthermore, we propose an automated Test-Verify-Write loop, allowing agents to continually refine their mitigation Meta-Skills from execution traces at test time. Experiments on IS-Bench demonstrate that our framework substantially boosts the Safe-Success rate across multiple VLM backbones (e.g., +30.3% on Qwen3-VL-8B), enabling agents to proactively mitigate hazards without stalling task progress. Code is available at this https URL.
[CV-201] What Makes Linguistic Representations Good Models of High-Level Visual Perception in the Human Brain?
链接: https://arxiv.org/abs/2607.16214
作者: Anna Bavaresco,Ina Klarić,Raquel Fernández,Marie-Francine Moens
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Supplementary Materials in the public GitHub repository referenced in the paper
Abstract:Image descriptions represented with language models (LMs) predict human brain responses to naturalistic images in high-level visual regions, but the factors driving this predictivity remain unclear. To investigate this, we systematically studied how images are described and which language models are used to embed those descriptions. For a common set of images, we considered six caption types – including human-annotated and multiple machine-generated captions – differing along several dimensions. Each caption was represented with five LMs, spanning autoregressive LMs trained to predict upcoming words and text embedders, i.e., LMs fine-tuned on semantic tasks requiring sentence/document-level representations. Machine-generated captions yielded significant brain predictivity and alignment, often surpassing human-annotated captions used in previous work. Across caption types, text embedders consistently outperformed autoregressive LMs, a pattern replicated when measuring behavioural alignment with image-similarity judgments. Analyses of caption representations from different model layers further revealed that both brain predictivity and behavioural alignment peak at intermediate network depth, shortly after a point thought to mark the emergence of syntactic and semantic structure. Altogether, our results demonstrate that both the content of image captions and the LM used to represent them influence brain- and behaviour-modelling performance, establishing caption embeddings as a useful tool for studying high-level visual perception.
[CV-202] Co-addition and Subtraction of Undersampled Images
链接: https://arxiv.org/abs/2607.18054
作者: Matan Schlanger,Barak Zackay
类目: Instrumentation and Methods for Astrophysics (astro-ph.IM); High Energy Astrophysical Phenomena (astro-ph.HE); Computer Vision and Pattern Recognition (cs.CV)
备注:
Abstract:In astronomical imaging surveys, repeated observations of the same sky patches are taken in order to obtain deeper images and detect new sources. This is the case in the search for many transient phenomena, such as supernovae, gravitational wave (GW) optical counterparts and other cataclysmic variables. In many such surveys some of the images are undersampled, meaning that the pixel size is too large, and the image suffers from aliasing. For undersampled images, both co-addition of the images and background subtraction are done in a non-optimal manner, which leads to reduced sensitivity and an increased rate of false alarms. We present a new method (named Linear Undersampled Transients \ Addition (LUTRA)) that performs both processes in a mathematically proven optimal way, which allows improved performance for many scientific applications. It also allows easy and direct performance of measurements such as photometry and astrometry in a simple manner, while providing results in super-resolution. We demonstrate the performance of the method on public ZTF data and show \times 1.25 higher SNR compared to current methods. We provide an open source Python implementation. Subjects: Instrumentation and Methods for Astrophysics (astro-ph.IM); High Energy Astrophysical Phenomena (astro-ph.HE); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2607.18054 [astro-ph.IM] (or arXiv:2607.18054v1 [astro-ph.IM] for this version) https://doi.org/10.48550/arXiv.2607.18054 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-203] Luminosity-Adaptive Contrast Enhancement Using CLAHE for Retinal Fundus Images with Quantitative Validation and Comparative Analysis
链接: https://arxiv.org/abs/2607.17691
作者: K. Mithra,Prem Kumar Santhanam
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV); Emerging Technologies (cs.ET)
备注: 11 pages
Abstract:Background: Retinal fundus imaging is central to the early diagnosis of sight-threatening conditions including diabetic retinopathy, glaucoma, and retinal vein occlusion. Clinical utility of fundus images is routinely compromised by non-uniform illumination, motion blur, and low contrast - artefacts that increase the risk of diagnostic error. Effective image enhancement is therefore a prerequisite for reliable computer-aided ophthalmic diagnosis. Methods: This study proposes a two-stage image enhancement pipeline combining luminosity correction via HSV colour space decomposition with Contrast Limited Adaptive Histogram Equalization (CLAHE) applied exclusively to the Value (V) channel. Experiments are conducted on the publicly available DRIVE dataset (40 retinal fundus images, 584 x 565 pixels, Canon CR5 camera, ophthalmologist-annotated ground truth). Quantitative evaluation employs Peak Signal-to-Noise Ratio (PSNR), Structural Similarity Index (SSIM), and Contrast-to-Noise Ratio (CNR). Baseline comparisons include standard Histogram Equalization (HE) and Adaptive Histogram Equalization (AHE). A binary masking step is subsequently applied to isolate hyper-reflective regions consistent with vascular pathology. Results: The proposed method achieves PSNR = 29.3 dB, SSIM = 0.91, and CNR = 3.12 - outperforming HE (PSNR = 21.4 dB, SSIM = 0.74) and AHE (PSNR = 23.1 dB, SSIM = 0.79) across all metrics, with an average processing time of 0.14 seconds per image. Conclusions: The combined luminosity-CLAHE pipeline yields measurably superior contrast and structural fidelity compared to established baseline methods, with processing speed compatible with clinical screening workflows. Limitations and directions for deep-learning-based comparison are discussed.
[CV-204] C3DIR: A Deep Learning 3-Dimensional Cloud Property Retrieval Scheme for Passive Satellite Imagers
链接: https://arxiv.org/abs/2607.16929
作者: Charles H. White,Yoo-Jeong Noh,John M. Haynes,Imme Ebert-Uphoff
类目: Atmospheric and Oceanic Physics (physics.ao-ph); Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注:
Abstract:We develop the Cloud 3-Dimensional Imager Retrieval (C3DIR), a deep learning model that estimates 3-D cloud properties for multiple passive satellite imagers trained to match retrievals from the Earth Cloud Aerosol and Radiation Explorer(EarthCARE) ACM-CAP product. This work is aimed towards moving AI/ML 3-D cloud algorithms closer towards operational use. C3DIR predicts the occurrence water content of ice, cloud liquid, and rain along the imager line-of-sight and uses a voxel-level collocation approach to account for the misaligned viewing geometries of passive imagers and active profiling instruments. This precise collocation methodology allows for constructing vertical profiles using voxels contained by multiple imager pixels to facilitate comparisons with active profiling instruments. Qualitative case studies show that C3DIR can accurately depict multiple distinct overlapping cloud layers, albeit with some smoothing. Quantitative evaluations illustrate that C3DIR overall excels at hydrometeor detection which intuitively tends to be a function of water content. However, detection of voxels classified as liquid cloud remains difficult due to the their small geometric thickness, finer horizontal scale, and frequent tendency to be obscured or embedded within ice clouds. In general, water content estimation is reasonably accurate, yielding the best results in ice clouds but uncertainties remain for liquid and rain water content. Column-integrated water paths are in tighter agreement with EarthCARE. Comparisons with the algorithms underpinning current NOAA operational products highlight several areas where C3DIR may offer improvement. Overall, these results demonstrate the potential for C3DIR to provide flexible 3-D output depicting vertically resolved cloud structure which can offer broader utility for aviation applications, numerical weather modeling, and climate research. Subjects: Atmospheric and Oceanic Physics (physics.ao-ph); Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV) Cite as: arXiv:2607.16929 [physics.ao-ph] (or arXiv:2607.16929v1 [physics.ao-ph] for this version) https://doi.org/10.48550/arXiv.2607.16929 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[CV-205] ECG-LLM : Foundation Model for ECG-Based Cardiac Reasoning
链接: https://arxiv.org/abs/2607.16323
作者: Alexander Selivanov,Friederike Jungmann,Jan Kehrer,Karl-Ludwig Laugwitz,Eimo Martens,Daniel Rueckert
类目: ignal Processing (eess.SP); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Image and Video Processing (eess.IV)
备注:
Abstract:Electrocardiography (ECG) is an inexpensive, standard-of-care test for cardiac symptoms, but front-line triage often lacks immediate access to definitive imaging such as echocardiography (ECHO) or cardiac magnetic resonance (CMR). Furthermore, most existing ECGAI systems are limited to fixed diagnostic labels or automated reports, constraining their use for patient-specific clinical reasoning. To address this gap, we introduce ECG-LLM, an ECG-conditioned large language model trained across four cohorts comprising 679,112 ECG studies from 186,409 patients. Using a novel multimodal-to-language supervision strategy, ECG-LLM is trained on clinically structured question-answer pairs derived from ECG signals, clinical context, CMR, and ECHO. This unified approach enables the model to answer diverse cardiovascular questions from a 12-lead ECG alone, spanning both conventional interpretation and phenotypes not directly visible on standard ECGs. ECG-LLM successfully recovers conventional ECG measurements, such as heart rate, and strongly predicts complex CMR-derived phenotypes, including ventricular and atrial volumes and ventricular function. Crucially, it detects vital echocardiographic phenotypes, including increased LV wall thickness, aortic stenosis, and right-ventricular systolic dysfunction. On standard ECG understanding tasks, ECG-LLM matches or exceeds existing baselines for diagnostic report generation and the ECG-QA benchmark. By moving beyond fixed-label prediction, this multimodal framework provides clinically valuable, question-driven cardiovascular reasoning to support general practitioner and front-line triage decisions when specialist review is delayed.
人工智能
[AI-0] Learning Adaptive Safety Margins for Visual Navigation
链接: https://arxiv.org/abs/2607.18200
作者: Junyi Hu,Shuaihang Yuan,Geeta Chandra Raju Bethala,Anthony Tzes,Yi Fang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Robots in cluttered indoor spaces often fail not because they cannot generate collision-free paths, but because a fixed safety margin is mis-calibrated: conservative margins cause detours and timeouts, while permissive margins lead to near-boundary shortcuts under perception bias. Diffusion-based planners propose diverse trajectory candidates from egocentric RGB-D, yet reliable selection remains the bottleneck. We propose a context-conditioned safety critic that learns an adaptive clearance preference for ranking diffusion proposals, decomposed into three complementary terms: (i) a safety term with a clearance-budget penalty and a control-barrier-function residual for waypoint- and transition-wise safety, (ii) an efficiency term combining a smoothness penalty with a safety-gated detour-ratio penalty that avoids detours without incentivizing risky shortcuts, and (iii) a distance-constraint matching term that anchors the learned budget to realized ESDF clearances to prevent margin collapse. We train the critic with privileged ESDF geometry in simulation and distill it into a perception-only selector via a two-stage teacher-student procedure. On PointGoal navigation in HM3D and MP3D, including cross-dataset transfer, our method achieves the highest success rate (SR) and success weighted by path length (SPL) among strong diffusion, optimization, and RL baselines. Trained purely in simulation, it transfers to a Unitree G1 humanoid and navigates cluttered indoor scenes without task-specific tuning.
[AI-1] A Continual Validation Updating and Decision-Making Framework for Self-Adaptive Digital Twins via Robust Model Predictive Control: A Case Study in Additive Manufacturing
链接: https://arxiv.org/abs/2607.18164
作者: Yi-Ping Chen,Ying-Kuan Tsai,Vispi Karkaria,Seul Lee,Daniel Apley,Wei Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Statistics Theory (math.ST)
备注:
Abstract:Digital Twins rely on surrogate models to mirror physical systems in real time, yet these models can degrade as operating conditions evolve, a phenomenon known as concept drift. Maintaining surrogate fidelity under drift, particularly when models must also capture aleatoric uncertainty, remains an open challenge. Existing adaptive frameworks lack principled mechanisms for detecting when updates are needed, for efficiently adapting models from limited streaming data, and for certifying that updates genuinely improve predictive performance. Here we present an adaptive Digital Twin framework that integrates a Fisher score–based multivariate drift detector, Low-Rank Adaptation (LoRA) for parameter-efficient continual learning, and a Mann–Whitney U test for online statistical validation. The framework monitors surrogate-model confidence via Fisher score vectors, triggers targeted fine-tuning of fewer than 1% of model parameters upon drift detection, and statistically certifies predictive improvement before deploying the updated surrogate. Applied to a stochastic linear system and a directed energy deposition additive manufacturing process as case studies, the framework successfully detects distributional shifts with short delays and restores both predictive accuracy and uncertainty quantification under abrupt and incremental drift. These results establish a statistically rigorous and computationally tractable pathway for sustaining the trustworthiness of neural-network–based Digital Twins throughout their operational life cycle.
[AI-2] OR Else: A Differentiable Trust Region for Policy Optimization
链接: https://arxiv.org/abs/2607.18163
作者: Chinmay Rane,Kanishka Tyagi,Michael Manry
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 22 pages, 5 figures
Abstract:PPO and the GRPO baseline studied here use clipped surrogate objectives whose favorable-direction saturation introduces an abrupt change in the scalar objective’s derivative. We ask whether Output Reset (OR), a smooth one-sided saturation rule, offers a useful alternative for large language model post-training. PPO-OR and GRPO-OR replace the clipped policy term with an OR squared-margin loss in rollout-relative token log-ratio space; the advantage sign determines the update direction, and a token contributes zero direct OR residual after crossing the favorable margin. We compare PPO-clip with PPO-OR under generalized advantage estimation (GAE), and GRPO with GRPO-OR under group-relative advantages, using \textttLlama-3.2-1B-Instruct on Anthropic \texttthh-rlhf with one shared reward model and three seeds per method. Under GAE, PPO-OR has a mean final training-time reward-model score 0.305 higher than PPO-clip, with a larger observed across-seed spread. Under group-relative advantages, GRPO-OR does not have a higher mean score, but shows a smaller observed spread, a near-zero terminal OR residual, and a declining overshoot fraction, while the matched GRPO clipped-objective trace remains variable. Both group-relative methods exhibit substantially larger rollout-to-current log-ratio displacement than the GAE methods, and OR does not consistently reduce it. Thus, OR changes optimization behavior in both matched comparisons, but the observed reward effect differs between them. At G=2 , the GRPO-OR diagnostics do not translate into a reward-score gain. Whether larger groups change this outcome remains open. The reported scores are training-time reward-model measurements, not held-out human-preference performance.
[AI-3] RIM: Reducing AI-Generated CodeSlop via Agent Trajectory Minimization
链接: https://arxiv.org/abs/2607.18161
作者: Alex Mathai,Shobini Iyer,Aleksandr Nogikh,Petros Maniatis,Franjo Ivancic,Junfeng Yang,Baishakhi Ray
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Operating Systems (cs.OS)
备注:
Abstract:Coding agents are increasingly used to accelerate code generation in many downstream tasks, such as fixing bugs, building applications, and prototyping. However, despite their value as coding assistants, agent-generated code tends to be larger and more verbose than the corresponding human-written implementation. In this work, we show that the cause lies in the agent’s own search process: while iterating toward a passing solution, an agent accumulates speculative edits, abandoned hypotheses, and temporary changes that persist into the final patch. This may seem harmless for a single patch, but the problem compounds as agents take responsibility for ever-larger portions of a codebase-a codebase that was once minimal and well-maintained slowly accumulates redundancy faster than it can be cleaned up, drifting to a state that is harder to maintain. Given the magnitude of this problem, we take a step towards alleviating this issue. First, we formally define this phenomenon as CodeSlop-the residual and functionally unnecessary edits commonly seen in AI-generated code. We then introduce our algorithm TRIM (Trajectory-guided Redundancy Identification and Minimization). Rather than minimizing CodeSlop directly, TRIM instead minimizes agent trajectories. As we show empirically, this indirect technique of minimizing CodeSlop is highly effective: TRIM cuts CodeSlop by 17.9%-32.9% across agentic scaffolds, with negligible performance regression. TRIM is also highly efficient, requiring roughly half the validation cost of algorithmic baselines such as Delta Debugging.
[AI-4] Differentiable Logic Gate Networks for Low-Latency EEG Classification on Edge Devices
链接: https://arxiv.org/abs/2607.18149
作者: Shyamal Y. Dharia,Stephen D. Smith,Camilo E. Valderrama
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Published in the Proceedings of the 39th Canadian Conference on Artificial Intelligence, PMLR 318, pages 377-391, 2026. Conference version: this https URL
Abstract:Real-time EEG classification on edge devices is bottlenecked by the floating-point arithmetic of conventional neural networks. We investigated Differentiable Logic Gate Networks (Diff-Logic) as a hardware-native alternative that compiles models into pure Boolean circuits executable via bitwise CPU operations. Through rigorous iso-parameter experiments across four EEG datasets spanning two classification tasks, binary dementia detection and 3-class emotion recognition, we compared Diff-Logic against matched-capacity Multi-Layer Perceptron (MLP) and Binarized Neural Network (BNN) baselines at four complexity tiers (50k-500k parameters). On dementia screening, Diff-Logic achieved 80.2% Macro F1, outperforming the MLP baseline by 6.8%. On emotion recognition, the MLP retained a moderate performance advantage but incurred a 2.3 \times higher latency and 14 \times larger model size when deployed on a power-constrained (7W) Nvidia Jetson Orin Nano CPU (Single-core). Critically, Diff-Logic inference time remained nearly constant across a 10 \times increase in model scale, achieving a peak speedup of 2.9 \times over MLPs at the largest complexity tier. Our results establish logic-based neural architectures as a practical paradigm for resource-constrained brain-computer interfaces, achieving competitive or superior performance while natively satisfying the latency and memory constraints of portable edge deployment. Code is available on GitHub: this https URL
[AI-5] LLM s and Agent ic AI Systems for Smart Grids: A Tutorial on Architectures and Applications
链接: https://arxiv.org/abs/2607.18147
作者: Daniela Rojas,Abdulwahab Albassam,Aidan G. Leung,Jett Ngo,Ryan Luo,Peter R. Quawas,Junpyung Kim,Kangkai Liang,Mansi Nanavati,Jonathan Mai,Meng-Chi Tsai,Yun-Tong Tsai,Yize Chen,Yuanyuan Shi
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI)
备注: 28 pages, 11 figures, 6 tables; plus supplementary material. Review/tutorial article
Abstract:Large language models (LLMs) and agentic AI systems have evolved from natural language tasks to using external tools to plan, retrieve, and act in technical domains. In smart grids, recent work applies agentic schemes to forecasting, optimization, and control, wrapping trusted solvers behind language interfaces and orchestrating multi-step workflows. The literature lacks a unified approach to designing and evaluating such systems. LLMs can produce numerically plausible yet physically infeasible outputs, evaluation protocols vary across tasks, and the boundary between what the model should and should not compute is implicit. This paper presents a solver-grounded design principle: a numerical result is reported only when it originates from a trusted tool and passes explicit verification. We review the building blocks of LLM and agentic AI systems for power systems: prompting strategies and agentic architectures. We instantiate the principle in four case studies: wind power forecasting, EV charging scheduling, power flow analysis, and contingency diagnosis, each comparing an LLM-only baseline against its solver-grounded counterpart on identical data and metrics. EVAgent reproduces the CVXPY optimum while reducing LLM-only unmet energy by 7.5-9.5x, and GridDebugAgent repairs 17/39 contingency cases while reducing total violations by 52.3%. We propose a four-group evaluation framework spanning task utility, solver-grounded correctness, faithfulness and safe failure, and cost and latency. A consistent division of labor emerges: the agentic system reliably orchestrates, retrieves, and explains, while trusted tools compute and a verification gate decides what is reported.
[AI-6] Can We Break LLM s Out of Self-Loops? Fine-Grained Reasoning Control with Activation Steering
链接: https://arxiv.org/abs/2607.18100
作者: Sheldon Yu,Tong Yu,Xunyi Jiang,Rohan Surana,Gagan Mundada,Sungchul Kim,Lina Yao,Julian McAuley,Junda Wu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Extended reasoning has become standard for frontier Large Language Models (LLMs), yet the trajectories these models produce remain largely uncontrollable. Existing methods for shaping how a model reasons are prompt based approaches and operate at the input level, offering no fine-grained control over the reasoning process itself. Related work analyzes and discovers latent transition dynamics in the reasoning traces from Large Language Models. Building on this, we statistically characterize these states, and show that failure trajectories get stuck in self-loops, exhausting the token budget without progress toward the final answer. To intervene on these failures, We propose SOPHIA: Steering Of reasoning Processes via Hidden-state Intervention and Activations. We treat each reasoning trace as a sequence of latent states rather than an unstructured texts, and investigate whether inference time interventions can provide fine-grained control over the self-looping reasoning process. We classify every prefix to a latent state, record step level transitions, and use them to construct a bank of steering vectors indexed by state pairs. At inference time, a controller infers the current state and, given a target state, retrieves the corresponding vector and can also detect self-loops online from the transition structure to prevent the model from sinking into a reasoning black hole. Through extensive experiments, our method reliably intervenes on self-loop failures, with steering vectors that generalize to different state pairs. End task accuracy and token efficiency indicate that fine-grained controllability results in better reasoning quality.
[AI-7] Judge-dependent safety gains and model-specific helpfulness costs of evidence-sufficiency prompting in clinical LLM s
链接: https://arxiv.org/abs/2607.18086
作者: Koyar Afrasyab
类目: Artificial Intelligence (cs.AI)
备注: this https URL
Abstract:Background: LLM judges increasingly score whether clinical language models give overconfident answers under incomplete evidence, yet whether a measured “safety gain” reflects real behavior change or the judge’s calibration is unresolved. Using a structured evidence-sufficiency prompt as a test case, we asked whether it reduces unsafe overconfident answers, how far that effect depends on the scoring judge, and what it costs in helpfulness. Methods: In a retrospective public-data benchmark (Real-POCQi, HealthBench, MedRBench), four models (GPT-5.5, Claude Opus 4.8, Gemini 3.5 Flash, Grok 4.3) answered a fully paired common panel (1,200 cells) with a standard prompt and the wrapper. The pre-specified endpoint was the paired reduction in unsafe overconfidence scored by the primary judge (GPT-5.4-nano); secondary analyses added a different-family judge (Claude Sonnet 5), a correctness judge, matched scaffold controls, and a blinded three-clinician review. Results: Unsafe overconfidence fell from 49.3% to 24.7%, a paired reduction of 24.7 points (95% CI 21.8-27.7; p0.001), robust in direction across models and paraphrases. Magnitude was judge-dependent: Sonnet agreed on direction but nearly halved the effect (+13.1 points), with one-directional disagreement. Blinded clinicians characterized the primary judge as a high-sensitivity (1.00), low-specificity (0.55) screen, not a calibrated rate. The gain carried a model-specific helpfulness cost (correct diagnosis 80.3% to 50.3%): near-free for GPT-5.5, near-total for Gemini (-58 points). Matched scaffold controls showed genuine behavior change, not judge circularity. Conclusions: LLM-judged clinical safety effects should be reported as directional and relative, anchored to human review and evaluated jointly with helpfulness, not as calibrated absolute rates. This does not establish clinical deployment readiness. Comments: this https URL Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.18086 [cs.AI] (or arXiv:2607.18086v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.18086 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Koyar Afrasyab [view email] [v1] Mon, 20 Jul 2026 15:53:02 UTC (1,542 KB)
[AI-8] Enhancing Rubric-based RL via Self-Distillation
链接: https://arxiv.org/abs/2607.18082
作者: Mingxuan Xia,Yuhang Yang,Chao Ye,Shuai Zhu,Shenzhi Yang,Guangcheng Zhu,Yuhang Zhang,Cheng Peng,Haobo Wang,Siqing Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Rubric-based RL has recently shown promise in improving LLMs on open-ended tasks. A widely recognized limitation of rubric-based RL is limited exploration: criteria that no rollout manages to satisfy (Unexplored Criteria, UC) receive no optimization signal. Recent methods address this by incorporating rubric information as external guidance during rollout, yet they introduce a train-inference mismatch: the policy is optimized on rollouts produced under external guidance while this guidance is absent at inference time, causing error accumulation through autoregressive decoding. Moreover, these exploration-focused approaches overlook a fundamentally different failure mode that we term Suppressed Criteria (SC) – criteria that are satisfied by some rollouts yet whose learning signals are lost during optimization because scalar reward aggregation assigns them non-positive aggregate advantages. Our analysis reveals that SC are remarkably prevalent: over 57% of samples exhibit this failure mode throughout training, with an average of 1.8 SC per sample. To simultaneously address both UC and SC without introducing training-inference mismatch, we propose Criterion-Distilled Policy Optimization (CriPO), which enhances rubric-based RL via on-policy self-distillation. For UC, CriPO constructs a criterion-injection self-teacher and computes a localized forward-KL loss to inject missing behaviors into the policy. For SC, CriPO employs a counterfactual self-teacher to locate criterion-relevant tokens in negative-advantage rollouts and flips their token-level advantages to positive values, preserving useful patterns that would otherwise be suppressed. Experiments on medicine and science benchmarks demonstrate that CriPO consistently outperforms rubric-based RL, achieving stronger final performance with approximately 2\times fewer optimization steps.
[AI-9] SelectInfer: Selective Neuron Loading and Computation for On-Device LLM s
链接: https://arxiv.org/abs/2607.18081
作者: Huzaifa Shaaban Kabakibo,Eric Schniedermeyer,Artem Burchanow,Lin Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Models (LLMs) have demonstrated remarkable capabilities across a range of Natural Language Processing (NLP) tasks, but their high computational and memory demands pose significant challenges for deployment on resource-constrained edge devices. Existing approaches to model compression and optimization often rely on coarse-grained pruning or quantization, which can compromise accuracy or require re-training and fine-tuning. In this work, we introduce SelectInfer, a neuron-level optimization framework that enables efficient LLM inference on edge devices through selective neuron loading and computation. By profiling and identifying both task-specific and general-purpose neurons using an offline LLM profiler, SelectInfer implements two key optimizations: selective loading, which reduces memory footprint by selectively loading a subset of neurons that were identified to be most important during the offline stage, and selective computation, which dynamically computes only the most relevant neurons at runtime. Evaluation across multiple datasets shows that SelectInfer achieves significant reductions in memory footprint and computation while preserving task performance, making it a practical step towards enabling LLM deployment on edge devices
[AI-10] Generalised Bellm an recurrence and three dualities in sequential decision-making
链接: https://arxiv.org/abs/2607.18077
作者: Fernando E. Rosas,David Hyland,Daniel Polani
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 27 pages, 2 figures
Abstract:What gives the Bellman equation its form? We show that the recursive properties of optimal value functions follow from three conditions: that the dynamics decomposes through sufficient statistics, that the return decomposes recursively, and that the aggregation of uncertainty is compatible with both. When all three conditions hold on a common state, the Bellman equation arises from their mutual consistency; when one fails, tractability can often be recovered by augmenting the state or by deforming return or dynamics. The same conditions are shown to give rise to three dualities: one between probability and return, one between return and aggregation, and one between aggregation and probability. Our framework reveals these dualities as arising from a single construction, unifying methods developed separately across reinforcement learning, control, and decision theory.
[AI-11] SGN: A Similarity-based Generative Network for Data Generation under Distribution Shift
链接: https://arxiv.org/abs/2607.18072
作者: Jiaqi Zhu,Xincheng Chen,Yuncheng Wu,Zhaojing Luo,Beng Chin Ooi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Generative models trained on a source domain often produce samples that are poorly aligned with shifted target domains, limiting their effectiveness for target-domain data augmentation. Although target-specific adaptation can reduce this mismatch, it typically requires additional optimization and domain-specific parameters. We propose a Similarity-based Generative Network (SGN), a reusable framework that is trained once on labeled source data and applied to new target domains without parameter updates. SGN learns a latent space structured by label-induced pairwise similarities while preserving reconstructive information through an encoder-decoder architecture. At generation time, a small labeled representative set from the target domain is encoded and combined in the learned latent space, allowing the generated samples to inherit target-specific characteristics while maintaining class consistency. We further analyze the realizability and dimensionality requirements of the proposed similarity structure. Experiments on image and tabular datasets demonstrate the effectiveness of SGN for target-guided data augmentation under source-to-target distribution shifts.
[AI-12] Human Grounded Evaluation of Large Language Models for Optical Network Automation
链接: https://arxiv.org/abs/2607.18068
作者: Kiarash Rezaei,Omran Ayoub,Paolo Monti,Carlos Natalino
类目: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models (LLMs) are increasingly adopted for network automation, yet their output quality and inference cost can vary substantially across LLM families. We present HuGLEN, a stepwise evaluation pipeline that uses an LLM-as-a-judge together with a small set of expert ratings to enable scalable and reproducible comparison of candidate LLMs, and to rank them using a quality efficiency score (QES). We demonstrate HuGLEN for translating outputs from an explainable artificial intelligence (XAI) model for the optical network quality of transmission (QoT) estimation task into operator-friendly explanations. Our results show that a medium-sized LLM (12B parameters) achieves the highest QES, indicating the best trade-off between explanation quality and efficiency. Overall, HuGLEN reduces the human-labeling burden while supporting consistent model selection for operator-facing automation tasks.
[AI-13] Autoresearch with Coding Agents : Generalizers and Metric-Maximizers on Quran Recitation Data
链接: https://arxiv.org/abs/2607.18064
作者: Nursultan Askarbekuly,Mohamad Al Mdfaa,Ahmed Helaly,Gonzalo Ferrer,Manuel Mazzara
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Coding agents can now be left alone to improve software against a score. In this pattern–recently popularized as “autoresearch”–the agent receives a dataset, an evaluation script, and one editable file, and iterates without supervision: modify the code, measure, keep the change if the score improves. But what does the agent actually optimize–the developer’s intent, or the literal number? We ran this loop on a real production task: deciding which Quranic verses appear in a noisy speech-recognition transcript and splitting the transcript by verse. Two frontier coding agents, Claude Code and OpenAI Codex, started from the same blank file with the same instructions, budget, and reasoning effort, three runs each. Both independently invented the same algorithm (canonicalization, n-gram anchoring, dynamic-programming alignment)–and then diverged. Claude stopped early with compact, general code. Codex drove the score ~10x lower, largely by memorizing answers to individual evaluation rows (19-41 hardcoded verse ids per run): a clean natural instance of specification gaming by a production agent. In a preregistered second study, we added a held-out test set and told both agents it existed. The memorization vanished, and the score gap vanished with it–yet Codex’s general core transferred better and more consistently (held-out detection+split 0.085+/-0.004 vs. 0.121+/-0.031), losing only on one missed rejection of non-recitation input. Two exploratory community arms (Cursor, Antigravity) are consistent with the pattern. Every agent’s held-out solution matched or beat the hand-engineered pipeline it was built to replace–the best by an order of magnitude–and now runs in production. From the ways agents exploited our harness–reading sibling runs through shared git state, leaving notes to “future runs” in persistent memory–we distill five design rules for evaluating autonomous agents.
[AI-14] Adaptive Adversaries: A Multi-Turn Multi-LLM Benchmark for LLM Agent Security
链接: https://arxiv.org/abs/2607.18063
作者: Devina Jain,David Hartmann,Chuan Li
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Second Workshop on Agents in the Wild: Safety, Security, and Beyond
Abstract:LLM-based agents process external content, exposing them to prompt injection and multi-turn manipulation. Most safety benchmarks evaluate defenders against fixed attack pools collected before evaluation, single-turn or multi-turn. We present a 21-scenario benchmark for \emphadaptive multi-round attacks against memoryless LLM defenders: an autonomous LLM attacker observes prior defender responses and pivots across rounds, while each defender response is evaluated as a fresh interaction. Holding the 21 scenarios, attackers, defenders, and structured-output scoring fixed, restricting scoring to the first attacker turn yields 0 - 1% attack success rate (ASR); allowing 15 rounds of adaptive attack yields 5.4 - 14.0% . Pooling three frontier attacker LLMs uncovers 1.4 - 2.2\times as many unique successful attacks as the best single attacker, and the generated attacks have low cosine similarity ( 0.02 - 0.14 ) to attacks in existing benchmarks. Claude Opus 4.6 and GPT-5.4 are tied in aggregate ( 5.4% each; overlapping 95% CIs), but their weaknesses differ sharply: on one scenario Opus reaches 60% ASR ( 95% CI 36 – 80% ) while GPT-5.4 and Gemini each stay at 7% (CI 1 - 30% ; the gap is preserved in a higher- N replication). 13 of 21 scenarios distinguish at least one defender pair, yet rankings disagree across scenarios (Kendall’s W = 0.19 ). We release the benchmark – 21 evaluation scenarios, 10 public development scenarios, the orchestrator, baseline harnesses, and a multi-attacker CLI – plus 945 transcripts from the 3 \times 3 frontier matrix, an attack-replay dataset, and 18,422 gpt-oss-20b battles from an open competition’s final scoring rounds.
[AI-15] AdaHome: An Adaptive Smart Home Assistant using Local Small Language Models
链接: https://arxiv.org/abs/2607.18034
作者: Eu Jin Lim,Zhaoxing Li,Sebastian Stein
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Smart home assistants interpret a wide range of user commands, from explicit device control to underspecified and preference dependent requests. While recent systems based on Large Language Models (LLMs) improve this capability, they often rely on heavyweight reasoning pipelines and cloud-based deployment, limiting their efficiency and suitability for resource-constrained environments, and raising privacy concerns. In addition, existing approaches provide limited support for stable long-term personalization. To address these issues, we present AdaHome, an adaptive smart home assistant designed for locally deployed small language models in smart home environments. Rather than applying complex reasoning uniformly, AdaHome introduces an intent-aware planning framework that dynamically routes commands either to straightforward prompt-based or lightweight reasoning-based components. For commands requiring interpretation, we adopt a Chain-of-Draft strategy to enable efficient and stable decision-making. To support personalization, we further propose a preference adaptation mechanism that learns from user feedback over time without requiring prompt augmentation or model retraining. We evaluate AdaHome against representative LLM-based baselines under a unified small model setting. AdaHome achieves substantially higher accuracy on direct commands (86.7%) while reducing latency by up to 3 \times . Furthermore, it maintains competitive performance on ambiguous inputs with lower computational cost. In multi-turn scenarios, AdaHome achieves 88% preference consistency, compared to 52.5% for a prompt augmentation baseline.
[AI-16] Natural Language Access to Domain-Specific Metadata: A Reusable Framework for LLM Query Generation
链接: https://arxiv.org/abs/2607.18029
作者: Blake G. Fitch,Cato Elia Kurtz
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:Researchers need to answer ad-hoc questions about the contents of domain-specific archives but often lack the expertise to write structured queries on the metadata. We show that when domain vocabulary and semantics are captured in a well-designed Web Ontology Language (OWL) ontology, Large Language Models (LLMs) can generate accurate structured queries zero-shot, without fine-tuning, retrieval augmentation, or multi-agent orchestration. We present the Natural Language Knowledge Graph Query (NLKGQ) system, a framework and development process that enables natural language access to metadata in such archives. The framework includes a web interface that helps researchers pose natural language questions, which a domain-agnostic harness translates to SPARQL via an LLM and executes against a knowledge graph. The development process begins with capturing domain vocabulary and semantics in a formal OWL ontology. Domain-specific code then extracts metadata from archive sources and imports it into a knowledge graph defined by the ontology. Both are designed for reuse across domains. We demonstrate the system on metadata derived from a large-scale neuroimaging research archive, evaluating multiple LLMs and ontology representations. The best configurations achieve 100% accuracy on a competence and regression question set developed with domain experts. An ablation study across eight ontology representations reveals that readable entity names and semantic annotations are the dominant factors in accuracy, more significant than model choice or prompt engineering. We also compare SPARQL to an auto-generated SQL database as query backends, showing that OWL’s structural features provide a substantial advantage over SQL DDL for LLM-driven query generation. Our demonstration domain also requires local LLMs on modest institutional hardware to address privacy concerns for human subject data.
[AI-17] Rethinking Heterogeneous LLM Merging: A Weighted Model Averag ing Perspective
链接: https://arxiv.org/abs/2607.18026
作者: Jiahe Fan,Yinghao Hou,Si Chen,Aiyuan Zhang,Hong Xie,Defu Lian
类目: Artificial Intelligence (cs.AI)
备注: 17 pages, 3 figures, 20 tables, preprint
Abstract:Can large language models with substantially different parameter spaces be merged by direct weighted averaging, without training or semantic alignment? Existing heterogeneous fusion methods typically introduce distillation, adapters, learned latent spaces, routing, or feature alignment, leaving open whether a simpler recipe can work for genuinely different billion-parameter checkpoints. We revisit this counterintuitive question through training-free dimensional adaptation followed by ratio-controlled interpolation. In union-style merging, we expand the smaller model into the larger parameter space; in intersection-style merging, we truncate the larger model into the smaller parameter space. Across Qwen-family model pairs and benchmarks covering mathematical reasoning, code generation, language understanding, commonsense reasoning, knowledge, and instruction following, deterministic expansion largely preserves the source model function, and small-ratio interpolation can improve over strong source checkpoints by transferring complementary capabilities. However, near-balanced interpolation often collapses, and task-level results reveal a seesaw effect in which gains on some capabilities coexist with regressions on others. These results show that simple parameter averaging, when paired with lightweight dimensional adaptation and carefully controlled ratios, is a surprisingly strong baseline for heterogeneous LLM merging, suggesting that the limits of direct weighted fusion may also bound what more complex heterogeneous merging methods can achieve at scale.
[AI-18] PAMD: Structured Adaptive Distances for Bisimulation Representations in Visual Reinforcement Learning ICML2026
链接: https://arxiv.org/abs/2607.18004
作者: Daegyeong Roh,Juho Bae,Han-Lim Choi
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 9 pages, 6 figures, plus appendix. Accepted to the 43rd International Conference on Machine Learning (ICML 2026)
Abstract:Many visual reinforcement learning (RL) algorithms learn representations by matching latent distances to a behavioral distance induced by reward and transition similarity. In practice, the choice of the latent distance can strongly affect performance: using a fixed, pre-specified global norms (e.g., \ell_p norms or other hand-designed metrics) may be overly restrictive to capture the behavioral distance. In contrast, unconstrained pairwise distances may admit degenerate solutions that drive the metric loss down without improving the representation. To address this gap, we introduce PAMD: Pairwise Adaptive Mahalanobis Distance, which parameterizes a positive-definite, pair-conditioned metric for measuring latent state similarity. PAMD is a simple plug-in for existing bisimulation-based methods, offering a more expressive yet structured alternative to fixed, pre-specified latent distances. We empirically validate our method on visual MuJoCo continuous-control tasks, where final performance of several recent bisimulation-based RL algorithms is substantially improved when equipped with the distance we propose. Comments: 9 pages, 6 figures, plus appendix. Accepted to the 43rd International Conference on Machine Learning (ICML 2026) Subjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG) Cite as: arXiv:2607.18004 [cs.AI] (or arXiv:2607.18004v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.18004 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-19] Harness Engineering for LLM -Driven GPU Kernel Generation
链接: https://arxiv.org/abs/2607.17979
作者: Yue Shui,Chenyu Ma,Hangfei Xu,Shengzhao Wen,Yanpeng Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 24 pages, 6 figures. Extended technical report on our submission to the MLSys 2026 FlashInfer AI Kernel Generation Contest. Code: this https URL
Abstract:Large language models (LLMs) can assist GPU kernel generation, but their practical effectiveness depends on whether generated code can be reliably constrained, validated, profiled, and selected. This paper presents a harness-centered system for LLM-driven GPU kernel optimization in the MLSys 2026 FlashInfer AI Kernel Generation Contest on NVIDIA Blackwell B200 GPUs. The system separates an evaluation harness from a profile-backed optimization controller: the harness enforces compilation, correctness, official-aligned timing, and artifact archival, while the controller turns profiler and workload evidence into bounded candidate-generation decisions. Human-authored skills capture operator constraints, references, profiling procedures, and promotion rules, while Codex and Claude Code agents generate candidate kernels inside those constraints. Across five operator definitions, the retained official-aligned artifacts achieved mean-latency speedups over supplied FlashInfer baselines of 1.62x, 18.05x, 29.68x, 1.12x, and 13.70x. The Agent-Assisted kernels outperform the Full-Agent artifacts across the evaluated definitions, indicating that expert-provided optimization directions, high-quality references, and workload context remain critical for reliable AI-driven kernel optimization.
[AI-20] SAGE: Subgoal-Conditioned Action Generation for Latent World Model Planning
链接: https://arxiv.org/abs/2607.17973
作者: Letian Cheng,Qi Zhang,Yisen Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Latent world models have emerged as a powerful planning paradigm by learning action-conditioned predictive dynamics and using them as internal simulators to imagine and evaluate candidate action sequences. However, as the planning horizon grows, performance becomes increasingly constrained by proposal quality: a fixed candidate budget must search an exponentially larger action space, making it difficult to expose the world model to high-quality candidate futures for evaluation. In this paper, we introduce a prior-conditioned planner that replaces random proposal initialization with structured guidance. At each planning stage, a goal-conditioned generator predicts the next reachable latent subgoal for a specified duration, which is then used to condition the generation of candidate action sequences. To capture semantic information across temporal scales, we use subgoals of varying durations as priors, balancing fine-grained local control with higher-level long-horizon progress. Then the frozen world model evaluates and refines these subgoal-conditioned proposals before execution. Experiments on PushT and OGBench Cube show that coupling latent subgoal decomposition with prior-conditioned action generation substantially improves long-horizon planning while preserving strong short-horizon performance. To be specific, when the target offset is 150 , it raises PushT success from 12.7% to 64.7% and OGBench Cube success from 26.7% to 67.3% .
[AI-21] OntoExtend: A Framework for Requirement-driven and Scalable Ontology Extension with LLM s
链接: https://arxiv.org/abs/2607.17963
作者: Anna Sofia Lippolis,Mohammad Javad Saeedizade,Stefan Schmid,Simon Blattner,Robin Keskisärkkä,Aldo Gangemi,Eva Blomqvist,Andrea Giovanni Nuzzolese
类目: Artificial Intelligence (cs.AI)
备注: Accepted in research track of Semantics 2026: this https URL
Abstract:Ontology extension refers to the process of enriching an existing ontology in response to emerging requirements, making it more complete. This task is a resource-intensive and error-prone process. Large Language Models (LLMs) have shown promising performance on generating ontologies from scratch, but current approaches rarely tie ontology extension explicitly to requirements or reusable core models, and offer limited, systematic evaluation of LLM outputs. This paper introduces OntoExtend, a requirements-driven framework for ontology extension with LLMs. It uses retrieval-augmented generation (RAG) over relevant input ontologies and requirements in the form of competency questions to propose grounded extensions. We evaluate OntoExtend on 39 CQs from two use cases: a public EU-project ontology, Onto-DESIDE, and an industrial ontology from Bosch. The generated fragments show few structural issues, satisfy all functional evaluation tests, and are rated by ontology engineers as requiring minor to moderate revision before integration. These results suggest that OntoExtend is useful as a drafting assistant for requirement-driven ontology extension in real world scenarios, while remaining sensitive to CQ specificity and modelling profile.
[AI-22] opological Signatures of Context-Level Reliability in TabPFN
链接: https://arxiv.org/abs/2607.17962
作者: James Hu,Mahdi Ghelichi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注:
Abstract:TabPFN is a transformer-based foundation model for tabular prediction that performs inference without task-specific training by conditioning on a support set and query inputs. Despite its strong empirical performance, its internal behavior on structurally difficult tabular geometries remains poorly understood. We study this behavior using zigzag persistent homology, treating TabPFN layer representations as evolving point clouds. We construct a controlled benchmark of synthetic tabular tasks with known true probabilities and varied intrinsic topology, including warped circles, tori, spheres, Hopf links, trefoil knots, and Swiss rolls. Across these tasks, we find that the topology of TabPFN’s internal representation geometry is strongly associated with dataset-level reliability; for example, the zeroth homology group H_0 fragmentation count correlates positively with mean absolute residual across controlled tasks, and this association strengthens in a high-resolution warped circle case study at large sample size. Harder geometries induce a dual topological signature: increased H_1 loop activity and increased H_0 fragmentation, while the H_1 persistence becomes shorter-lived. These descriptors correlate with Bayes error, mean absolute residuals, and overconfidence. Our results suggest that zigzag persistence diagnoses the reliability of the inferred in-context task geometry and provides a context-level view of when TabPFN operates in topologically stressed regimes.
[AI-23] RT-SHCUA: Real-Time Self-Hosted Computer-Use Agent for UAV Control
链接: https://arxiv.org/abs/2607.17951
作者: Di Lu,Bo Zhang,Xiyuan Li,Yongzhi Liao,Xuewen Dong,Yulong Shen,Zhiquan Liu,Jianfeng Ma
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Robotics (cs.RO); Systems and Control (eess.SY)
备注:
Abstract:Natural-language control offers a promising interface for unmanned aerial vehicles (UAVs), but directly applying self-hosted computer-use agents (SHCUAs) to UAV control introduces a structural mismatch. SHCUAs are designed for interactive host-side tool use, where delayed agent iterations are often acceptable. UAV control, however, is coupled with continuously changing physical states, strict timing constraints, safety risks, and security accountability. A stale, unauthorized, or tampered agent decision may therefore lead to unsafe or untraceable vehicle behavior. This paper proposes a real-time and security-oriented restructuring of SHCUA-based UAV control. Instead of allowing an SHCUA to directly issue flight commands, we transform its outputs into contract-bound UAV skill invocations with explicit timing, state, authority, fallback, and evidence semantics. Based on this abstraction, we design an architecture that separates semantic reasoning from onboard execution and security/safety enforcement. Slow cloud or edge reasoning is used for mission understanding, while onboard components validate and dispatch only timely, authorized, and state-consistent skills. Security-critical enforcement points can be protected by TEE-style or microcontroller isolation mechanisms without moving the full language agent or high-frequency flight-control loop into trusted components. Prototype evaluation shows that RT-SHCUA maintains bounded task-level responsiveness while supporting degraded handling, trusted admission, and auditable evidence preservation for SHCUA-mediated UAV actions. Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Robotics (cs.RO); Systems and Control (eess.SY) Cite as: arXiv:2607.17951 [cs.CR] (or arXiv:2607.17951v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2607.17951 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-24] A Geometric Perspective on Stabilizing Value Conflict Resolution ICML
链接: https://arxiv.org/abs/2607.17946
作者: Saket Reddy,Andy Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted to ICML Workshop on High-Dimensional Learning Dynamics
Abstract:Large Language Models (LLMs) often struggle to navigate value conflicts when trained with the compressed scalar rewards of Reinforcement Learning from Human Feedback (RLHF). To address this challenge, we investigate how chain-of-thought (CoT) reasoning can help improve performance in this domain. Geometrically, we show that CoT correlates with further smoothing the model’s loss landscape in its sharpest direction, helping resolve the optimization instability of traditional scalar rewards. We also demonstrate via relevant downstream benchmarks that value conflict-focused CoT may generalize to different kinds of moral reasoning, demonstrating that this CoT has the potential to be an effective mechanism for better moral reasoning. To capitalize on this potential, we create a new value conflict-focused CoT design that further smooths the sharpest direction of the loss landscape and increases moral reasoning performance. This finding shows that explicitly modifying and improving the design of reasoning dynamics offers a promising avenue for improving model performance on user requests with complex value conflicts, advancing pluralistic alignment in LLMs.
[AI-25] he Art of Not Forgetting
链接: https://arxiv.org/abs/2607.17944
作者: Ashmith Atmuri,Akshay Kumar,Yashaswini Rao Bhogarajula
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 23 pages, 12 figures
Abstract:We introduce CMP (Cognitive Memory Primitive), an architecture that represents inputs as sparse relational codes, stores them in a two-tier competitive memory, and learns entirely through local, gradient-free updates, with no backpropagation anywhere in the network. We use this architecture to test a specific hypothesis: that catastrophic forgetting, usually treated as a training-time defect to be patched with replay or regularization, is instead a structural consequence of how backpropagation assigns credit and that a learning rule that is local and sparse by construction should resist it without a patch. On a controlled domain-incremental protocol across 15 text domains, three-seed replicated, CMP’s backward transfer is 15-19x better than a matched-size Transformer trained with online EWC, and the result survives a domain-order control (reported as a range, +0.24 to +0.44, rather than a single figure). We report this alongside a real, substantial accuracy gap versus the Transformer baseline, a null result on a recognized vision benchmark, and a diagnosed, unresolved failure attempting to combine this architecture with a separate mechanism that improves raw accuracy, disclosed because an honest negative result is more useful than an omitted one. The central claim is narrow and falsifiable: local, sparse, non-backpropagation learning measurably resists catastrophic forgetting better than backpropagation with its standard fix, under conditions we state precisely.
[AI-26] he Aura in the Machine: Genealogy and the Status of the Work of Art in the Generative Era
链接: https://arxiv.org/abs/2607.17940
作者: Giorgio Presti
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注: Submitted to Minds and Machines on 08 Jul 2026
Abstract:This paper frames Generative Artificial Intelligence (AI) not as an unprecedented technological rupture, but as an industrial-scale manifestation of a deeply rooted historical process. Through a genealogy of generative arts, it shows how AI’s questions on authorship and creativity have precise historical precedents. A taxonomy of generative systems is proposed across three functional categories (medium, artwork, instrument), the attribution of which is editorial rather than ontological. From individual cognitive atrophy to Model Collapse, the systemic risks of creative automation are identified; environmental enrichment is proposed as an antidote. The role of the artist undergoes a radical metamorphosis: from craftsman of the object to entropic agent, systems designer, explorer, and negentropic curator. This pipeline-based taxonomy rests on a specific premise: the algorithmic system remains medium, instrument, or artwork, while creative agency resides in the humans distributed along it. Algorithmic Repetition is introduced as the aesthetic degeneration of aligned generative systems; the Benjaminian aura does not dissolve in the generative era but condenses upon the productive system. Manifestation is proposed as a third ontological status for generative works, transcending the dichotomy between original and copy. To support the proposed theses, two complementary aspects are examined: the radicalization of distributed authorship; and the reevaluation of older generative models, whose instability constitutes an aesthetic degree of freedom lost by recent ones. Comments: Submitted to Minds and Machines on 08 Jul 2026 Subjects: Computers and Society (cs.CY); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17940 [cs.CY] (or arXiv:2607.17940v1 [cs.CY] for this version) https://doi.org/10.48550/arXiv.2607.17940 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Giorgio Presti [view email] [v1] Mon, 20 Jul 2026 13:36:18 UTC (142 KB) Full-text links: Access Paper: View a PDF of the paper titled The Aura in the Machine: Genealogy and the Status of the Work of Art in the Generative Era, by Giorgio PrestiView PDFHTML (experimental)TeX Source view license Current browse context: cs.CY prev | next new | recent | 2026-07 Change to browse by: cs cs.AI References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[AI-27] PEARL: Auditable Repair for Scientific Reasoning Graph Extraction
链接: https://arxiv.org/abs/2607.17917
作者: Bohan Su,Pengze Li,Yuchen Lu,Xi Chen
类目: Artificial Intelligence (cs.AI)
备注: Accepted at WAICA 2026 Multi-Modal Agents for Science Workshop
Abstract:Scientific Reasoning Graph Extraction (SRGE) aims to recover explicit links among observations, evidence, intermediate claims, and paper-level conclusions. LLMs can produce graph-like scientific explanations, but their outputs often mix malformed syntax, drifting edge labels, incorrectly oriented roots, and weak source anchors. We propose PEARL (Peircean Extraction via Abstraction and Repair Layer), a training-free framework that turns noisy LLM graph responses into auditable reasoning graphs and repairs them toward strict semantic validity. PEARL first materializes explicit graph content under a closed Peircean schema, then uses matched evidence-grounded judge feedback to repair rejected edge types, local inference steps, and terminal roots while preserving an audit trail. On five 70-paper model archives from ARCHE, a benchmark for latent reasoning-chain extraction, PEARL raises strict gate passes from 0/350 for the LLM baseline to 300/350, with average REA improving from 0.339 to 0.906. The graphs provide a reliability layer for research-agent and AI scientist workflows that need inspectable reasoning traces rather than unconstrained graph regeneration. Code and audit artifacts are available at this https URL .
[AI-28] Stress Testing Concept Erasure with Large Language Model Agents
链接: https://arxiv.org/abs/2607.17890
作者: Yuyang Xue,Feng Chen,Zhihua Liu,Edward Moroshko,Jingyu Sun,Steven McDonagh,Sotirios A. Tsaftaris
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Concept erasure aims to remove semantic concepts from a trained generative model and is increasingly important for responsible AI deployment. However, verifying whether a model has robustly removed targeted concepts remains a critical challenge. Existing evaluation methods are typically pre-defined and static, failing to expose vulnerabilities under diverse natural-language probes and challenging conditions. Moreover, manually designed evaluation strategies can be biased and difficult to scale. We posit that concept erasure evaluation is best formulated as an adaptive hypothesis search, operationalised by agents that iteratively propose, critique, and verify tests to systematically expand coverage of failure modes. To this end, we propose Stress Testing Agents for Concept Erasure (STACE), a framework that autonomously stress-tests concept-erased models using multiple Large Language Model (LLM) agents, by iteratively generating and verifying stress-testing hypotheses grounded by external knowledge. We also introduce a suite of metrics for assessing the performance and efficiency of LLM-agent-powered stress-testing frameworks. Our extensive experiments show that STACE outperforms five LLM-based evaluation baselines on four concept categories. Further analysis across two T2I models, six concept erasure approaches, and various erasure strengths show that STACE is robust for different settings. We also show that STACE can be adapted beyond concept erasure evaluation to other problem domains, such as LLM jailbreaking. Our code is available anonymously.
[AI-29] ST-Veto: Spatio-Temporal Token Veto for Diffusion MLLM s via Taylor Prediction and Visual Grounding ICML2026
链接: https://arxiv.org/abs/2607.17884
作者: Keuntae Kim,Beomseok Lee,Hyunwoo Kim,Yong Suk Choi
类目: Artificial Intelligence (cs.AI)
备注: ICML 2026 - main
Abstract:Vision Language Models (VLMs) achieve strong reasoning with Chain-of-Thought (CoT) prompting but incur high sequential-generation cost, error accumulation, and limited self-correction. Diffusion Multimodal Large Language Models (dMLLMs) unmask tokens in an order-agnostic process, improving efficiency and enabling iterative refinement, yet their reasoning and how to enhance it remain underexplored. We propose a training-free method, Spatio-Temporal Token Veto (ST-Veto), which leverages the ability to observe all token positions at each diffusion step. Rather than relying only on current-step confidence, ST-Veto vetoes temporally unstable tokens via second-order Taylor prediction of confidence dynamics and filters weakly grounded tokens using image-attention mass, swapping them with safer candidates. Across multiple dMLLMs and multimodal reasoning benchmarks, ST-Veto consistently outperforms standard decoding policies and prior VLM reasoning methods, improving accuracy by up to 9% with no additional training or generation cost. Analyses show that ST-Veto steers generation toward higher-confidence, better-grounded paths.
[AI-30] Exploratory and Assimilating Reflection: Reflective Recall Cycle for Long-term Memory
链接: https://arxiv.org/abs/2607.17879
作者: Ganesh Senrayan,Moyuru Yamada,Ishan Jindal,Kiran Purohit
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:LLM-based autonomous agents require external memory to overcome their statelessness and limited context window for long-term interaction and dynamic knowledge reasoning. However, existing memory retrieval methods often lack adaptability and sample efficiency, and struggle to retrieve the right mixture of memories from heterogeneous stores. We propose Exploratory-Assimilating Reflection (EAR), a framework for high initial retrieval performance and sample-efficient adaptation. EAR combines two mechanisms: Exploratory Reflection, which performs iterative search to bootstrap retrieval and collect useful experiences for each query, and Assimilating Reflection, which replays these experiences from an Experience Buffer to refine a global reranker more efficiently than methods relying only on immediate rewards. Experiments show that EAR improves retrieval by up to 17.9% over the baseline retriever on two long-term dialogue benchmarks. We also show that EAR is highly sample-efficient and robust to noisy feedback.
[AI-31] ConceptTree: Bringing Semantic Transparency to Black-Box Decision Making for Robotic Manipulation
链接: https://arxiv.org/abs/2607.17861
作者: Yongyan Wen,Feifan Liu,Jinyi Chen,Bo An,Peng Liu,Siyuan Li
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Establishing interpretable decision-making processes in long-horizon robotic manipulation is critical for enabling reliable human oversight and intervention. However, existing approaches to robotic manipulation largely treat skill selection as opaque mappings from observations to actions, offering limited transparency into how decisions are formed. In this work, we propose ConceptTree, a framework that reframes high-level manipulation skill selection as reasoning over human-interpretable concepts, representing high-level policies as a sequence of concept-level predicates over visual observations. Rather than relying on implicit latent representations, our method learns a normalized concept space grounded in visual inputs, over which a decision tree is trained to predict high-level skills. This formulation yields a transparent decision process that is both traceable and intervenable, enabling direct inspection and modification of policy behavior. We evaluate our approach on a set of real-world robotic manipulation tasks with increasing complexity. Experimental results show that ConceptTree consistently outperforms existing concept-based baselines, particularly in complex, long-horizon scenarios. Furthermore, we provide qualitative case studies showing that our model supports fine-grained intervention by modifying individual concepts, enabling targeted correction of decision errors without retraining.
[AI-32] A Hardware-oriented Approach for Efficient Bayesian Inference Computation and Deployment
链接: https://arxiv.org/abs/2607.17855
作者: Nikola Pižurica,Matteo Risso,Nikola Milović,Alessio Burrello,Igor Jovančević,Conor Heins,Miguel de Prado
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Bayesian inference provides a principled foundation for reasoning under uncertainty, but its computational cost hinders deployment on resource-constrained edge devices. In this paper, we present a hardware-oriented methodology for accelerating discrete Bayesian inference on commercial off-the-shelf embedded GPUs. We identify that the latency of a broad class of variational message-passing algorithms is dominated by tensor contractions. Our approach restructures the memory layout of these operations using two complementary merging strategies that produce compact, regularly-shaped primitives better suited for efficient GPU execution. We then introduce optional sparse array representations and a tensor-clustering scheme to reduce the memory footprint. We instantiate the methodology and produce optimized variants of three message-passing algorithms for Hidden Markov Models (HMMs), namely variational filtering, variational message passing, and marginal message passing. Furthermore, we complement this with a machine-learning-based autotuner that automatically selects the best-performing algorithmic variant for a given generative model specification. Benchmarked on an NVIDIA Jetson Orin AGX across 770 randomly sampled realistic Partially Observable Markov Decision Process (POMDP) configurations, our implementations achieve speedups of up to 5x, with typical gains of 2-2.5x, while producing numerically identical outputs to the baseline implementations.
[AI-33] Phasor Attention: Mean Root Square Normalization for Phase Manifold Preservation
链接: https://arxiv.org/abs/2607.17822
作者: Sungwoo Goo,Hwi-yeol Yun,Sangkeun Jung
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:While Root Mean Square Normalization has become the de facto standard for accelerating modern sequence models, its reliance on the quadratic accumulation of independent scalars ( \sum x^2 ) inherently triggers outlier-induced numerical instability, gradient starvation, and anisotropic phase distortion. We introduce Mean Root Square Normalization (MRSNorm). By structurally pairing channels into 2D phasors, MRSNorm mathematically inverts the traditional scaling paradigm: it computes the localized L_2 magnitudes (Root Square) before aggregating them via a global L_1 average (Mean). This operational inversion strictly constrains activations to a phasor manifold, preserving conformal invariance. By sharing a single affine weight across phasor components, MRSNorm halves the total number of learnable parameters, proving that unconstrained spatial scaling in standard norms is a harmful redundancy. We analytically demonstrate that this geometric constraint yields a built-in, trigonometric gradient clipper governed by the Pythagorean identity, unconditionally equalizing the local gradient norm to ensure Gradient Homogeneity. Empirical evaluations on a ResNet with CIFAR-100 show that despite halved parameters, MRSNorm provides critical structural stability under rigorous stress tests. Under extreme hyperparameter settings where standard normalizations suffer from gradient divergence, MRSNorm successfully prevents numerical explosion and secures stable optimization trajectories. Our findings propose a fundamental paradigm shift toward phasor-based deep representation learning. The implementation of MRSNorm is available at Appendix C. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17822 [cs.LG] (or arXiv:2607.17822v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.17822 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-34] PGN: Design and Implementation of a Vision-Language Navigation System Based on Pangu Multimodal Foundation Model
链接: https://arxiv.org/abs/2607.17806
作者: Li Xian,Mingxi Li,Yizheng Wang,Yiming Shen,Qi Chen,Zhuoling Xiao
类目: Artificial Intelligence (cs.AI)
备注: 6 pages, 5 figures
Abstract:Vision-Language Navigation (VLN) requires an embodied agent to interpret a natural-language instruction and predict actions from temporally ordered visual observations. Adapting a multimodal large language model to VLN requires visual-language alignment, compact temporal inputs, action-space grounding, and stable training on the target hardware. This technical report presents PGN (Pangu Navigator), an offline VLN action-prediction system built on OpenPangu-7B. Training proceeds in two stages. First, PGMM aligns a frozen EVA-ViT-G/14 vision encoder with the frozen language backbone by training a Q-Former and a two-layer MLP projector. Second, PGN adapts the aligned model to expert navigation trajectories using five-observation windows, epoch-dependent temporal sampling, and a reasoning-then-action output format; this stage freezes the aligned visual pathway and updates three structural-token embeddings and LoRA adapters. The implementation combines mixed-precision computation, selective FP32 computation, and DeepSpeed ZeRO-2 on eight Ascend 910B NPUs. Under teacher-forced, open-loop evaluation on 500 held-out expert trajectories, V9 reports a 62.29% Normalized Action Match (NAM) and a 100.00% Non-empty Rate (NER). These metrics quantify offline expert-action alignment rather than closed-loop navigation success; evaluating error accumulation, path efficiency, and goal completion remains future work.
[AI-35] Financial Audit Assistance using Misinformation Detection and Explanation
链接: https://arxiv.org/abs/2607.17797
作者: Kshitij Madhav Jadhav,Sushodhan Vaishampayan,Manoj Apte,Sachin Pawar,Nitin Ramrakhiyani,Girish Keshav Palshikar
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Financial statements (FS) such as Balance Sheet (BS), Income Statement (IS) and Cash-flow Statement (CS) summarize the annual financial performance of a company. FS are widely used for evaluating corporate governance, credit appraisal, risk analysis, validate taxation, make investment decisions etc. Financial auditing is a complex and knowledge-intensive discipline whose one important aim is ensuring integrity, accuracy, fairness and absence of material misstatement in the published FS. Given the importance of FS, there are incentives to hide, omit or falsify information to misrepresent the true financial health of the company; e.g., reduce tax liabilities, or increase investor confidence. Given the complex, time-consuming and expertise-dependent nature of auditing, auditors would benefit from an AI-assisted system that automatically detects instances of misinformation in the given FS and identify likely sources of this misinformation in the financial data. In this paper, we present unsupervised techniques to identify misinformation in FS, and also generate explanations as to the financial variables that are likely sources of misinformation. The auditor can then explore in more detail the associated data sources and business processes to validate these suggestions. A crucial feature of our approach is the use of past corpus of FS and associated audit reports to generate insights, which help in providing assistance. We demonstrate the efficacy of these techniques on a large corpus of 11,460 FS over 5 years and associated audit reports. This paper integrates and adds more novel contributions over the previously reported research (Shinde et al., 2022)\citeSVAP22, (Vaishampayan et al., 2022)\citeVSPP22, (Pawar et al., 2023)\citePAPV23, which we have used as the foundation for our AI-assisted Auditor Assistance system.
[AI-36] Reasoning as a Double-Edged Sword: Architecture and Cross-Stage Robustness in Vision-Language-Action Models
链接: https://arxiv.org/abs/2607.17786
作者: Tuan Duong Trinh,Naveed Akhtar,Basim Azam
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Machine Learning (cs.LG)
备注:
Abstract:Does adding a reasoning step make a Vision-Language-Action (VLA) model more robust to perturbation? Intuitively, a policy that reasons before acting should absorb a perturbed input better than one that maps observations directly to actions. We test this premise head-on across three models that span the reasoning spectrum (no reasoning, a text chain-of-thought, and a latent iterative loop), perturbing each at the vision, reasoning, and action stages on LIBERO and SimplerEnv. Two questions organize the study: does the reasoning design shift robustness, and can the reasoning be read back at runtime as a safety signal? We find that the latent-iterative model is by far the least robust: under both stochastic noise and white-box perturbation its task success collapses, while the other two hold. This fragility is structural rather than cumulative: varying the reasoning depth at inference barely moves it. Reasoning outputs can in principle be monitored, but the monitors fail under fair tests. A plan–action consistency probe that looks near-perfect under naive evaluation falls to chance under adaptive attack. Under matched-FPR calibration, fusing it with an action-anomaly probe never lifts defended success above undefended. Scoped to these output-level behavioral probes under white-box vision-stage attack, this ceiling is a precondition that any viable defense must first satisfy.
[AI-37] Feature Attribution-Based Explainability Analysis of Deep Learning Models in Predictive Process Monitoring
链接: https://arxiv.org/abs/2607.17783
作者: Kseniya Sahatova,Rafael Seidi Oyamada,Xuefei Lu,Johannes De Smedt
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Predictive process monitoring supports the optimization and control of operational business processes by forecasting the future state or outcome of ongoing cases. While deep neural networks have achieved strong performance for these tasks by modeling sequential dependencies in event logs, their black-box nature limits trust and practical adoption. Feature attribution methods are often used to address this, but applying them directly poses a dilemma: event-level attributions impose high computational complexity for long traces, while explanations based on aggregated trace representations often fail to capture the underlying control-flow dynamics. To address this issue, we propose a local post-hoc explainability method for deep neural networks in outcome prediction. The method relies on a control-flow-aware segmentation algorithm that partitions a trace into meaningful segments and supports the computation of segment-level SHAP explanations. This makes it possible to identify which parts of a trace influence a prediction and which change points steer the case toward the predicted outcome. We assess the proposed segmentation method on a synthetic dataset with known process logic, where meaningful change points can be explicitly verified, and we demonstrate its usefulness on real-world event logs from a loan application process and an administrative process of a Dutch municipality.
[AI-38] Dynamic Defense Profiling Enables Cognitive Jailbreak of Text-to-Image Models
链接: https://arxiv.org/abs/2607.17779
作者: Dongdong Yang,Deyue Zhang,Zhao Liu,Zonghao Ying,Wenzhuo Xu,Jiankai Jin,Xiangzheng Zhang,Quanchen Zou
类目: Artificial Intelligence (cs.AI)
备注: 12 pages, 3 figures
Abstract:Text-to-Image (T2I) generative models have achieved remarkable progress in synthesizing high-quality visual content, yet they remain vulnerable to adversarial misuse, particularly in generating Not-Safe-For-Work (NSFW) images. Most existing jailbreak attacks primarily rely on heuristic prompt engineering or black-box optimization, treating model feedback as a binary signal (success or failure). This coarse-grained paradigm overlooks the rich information embedded in diverse failure modes, such as textual refusal, visual blocking, and semantic sanitization, resulting in inefficient exploration and severe semantic collapse. In this paper, we propose MIND, a cognitive jailbreak framework that reframes adversarial prompt generation as a belief-state inference problem over latent defense mechanisms. Instead of blindly searching for bypass prompts, MIND actively models the target system’s latent defense mechanisms by interpreting multi-modal feedback as high-density signals. Specifically, the framework integrates three core components: (1) a Multi-modal Judge for fine-grained feedback decomposition, (2) a Defense Profiler for iterative belief updating, and (3) a Meta-Memory module for retrieving historically effective attack strategies. These components are unified within a reasoning-driven evolutionary optimization process, enabling adaptive and semantically consistent jailbreak generation. Extensive experiments on the I2P benchmark demonstrate the effectiveness of MIND. Under six representative pre-processing and post-processing defense settings applied to the Stable Diffusion v1.5 model, MIND achieves an Attack Success Rate (ASR) of 95.62%, significantly outperforming existing methods. Additionally, the effectiveness of the proposed framework is validated across four widely used commercial T2I systems, achieving the highest ASR of 91.58% on Wan-2.5. Comments: 12 pages, 3 figures Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17779 [cs.AI] (or arXiv:2607.17779v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.17779 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-39] FIFA World Cup 2026 as a Contamination-Free Benchmark for LLM Forecasting Agents : Four Models a Bookmaker and 104 Matches
链接: https://arxiv.org/abs/2607.17765
作者: Jiacheng Ding,Cong Guo,Jason Xu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Databases (cs.DB)
备注:
Abstract:We introduce WC2026-Agents, a benchmark and dataset for evaluating large language models (LLMs) as autonomous forecasting agents on real, future events. For every one of the 104 matches of the 2026 FIFA World Cup, four frontier models – Claude Opus 4.8, ChatGPT (GPT-5.5, high reasoning), Gemini 3.1 Pro, and Grok (Expert Mode) – ran an identical search-act-reflect loop: gather evidence with a web tool, commit to a 1X2 (team-A win / draw / team-B win) distribution and a virtual 100-USD bet, and, after the match, reflect given only the final score. Because every match kicked off after the models’ training cutoffs, the benchmark is contamination-free by construction. Crucially, we pair the four agents with a fifth competitor drawn from the same information environment – the pre-match betting market – collected as per-match 1X2 odds, giving an economically grounded baseline and letting us score not just what an agent predicts but what it does with money. The release contains 416 forecasts and 414 reflections with verbatim reasoning, ground truth (including penalty shootouts), odds, and a reproducible evaluation suite. A reference evaluation surfaces findings that raw accuracy hides: the four agents issue an identical top pick in 92% of matches and none beats the market’s Brier score; indeed, a naive flat stake on the market favorite out-earns all four agents. Yet the agents diverge sharply as decision-makers: betting return-on-investment ranges from -18% to +10%, fading the market is unprofitable for all four, the share of forecasts that cite the market ranges from 12% to 100%, and self-reported error rates on wrong picks range from 36% to 86%. The benchmark thus measures calibration, decision quality, and self-knowledge – axes on which frontier models differ even when their predictions do not. Data and code: this https URL
[AI-40] me-Frequency Consistency Learning for Robust Speech Deepfake Detection ACM-MM2026
链接: https://arxiv.org/abs/2607.17761
作者: Jun Xue,Zhuolin Yi,Yanzhen Ren,Yihuan Huang,Jiayu Xiong,Yi Chai,Guanxiang Feng,Jiajun Liu,Tong Zhang
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: Accepted by ACM MM 2026
Abstract:Recently, speech deepfake detection (SDD) has achieved significant progress. However, its robustness evaluation remains largely confined to controlled additive noise scenarios, lacking systematic investigation of the complex distortions introduced by acoustic front-end (AFE) processing pipelines in real-world deployments. In this work, we simulate a unified AFE pipeline comprising acoustic echo cancellation, noise suppression, automatic gain control, and voice activity detection (VAD), and conduct a comprehensive evaluation of current state-of-the-art models. The results show that the nonlinear and time-frequency coupled distortions introduced by AFE significantly degrade detection performance. To address this issue, we propose a Time-Frequency Consistency Learning (TFCL) framework, which aims to learn invariant spoofing representations that remain stable before and after AFE processing. We observe that AFE not only introduces temporal misalignment (e.g., segment-level shifts caused by VAD), but also weakens or distorts critical frequency-domain cues. To this end, TFCL employs an attention-driven soft alignment mechanism to capture cross-temporal dependencies, along with frequency-domain structural consistency constraints to enforce feature invariance. As a result, the model is able to maintain stable representations under both temporal perturbations and spectral distortions. Extensive experimental results demonstrate that the proposed method effectively mitigates the performance degradation caused by AFE processing, significantly improving the robustness of SDD in real-world scenarios. The code is available at this https URL.
[AI-41] Generalize and Guide: Decomposing Rewards for Few-Shot Inverse Reinforcement Learning
链接: https://arxiv.org/abs/2607.17760
作者: Ziyi Liu,Grace Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注:
Abstract:Inverse reinforcement learning (IRL) provides a powerful framework for learning from demonstrations. However, real-world tasks often exhibit substantial natural variations (e.g., picking up mugs with varying shapes), making it impractical to collect demonstrations that fully specify a new task under every possible scenario. In practice, while demonstrations for the target task are limited, it is often easier to obtain datasets of heterogeneous but related behaviors. This motivates the problem of few-shot IRL with multi-task demonstrations (FM-IRL), where an agent must learn a new task with substantial variations from only a limited number of target-task demonstrations, together with sufficient demonstrations of related tasks and online agent experience. To do so, we must both recover the expert distribution of the new task and provide guidance when the agent deviates from it. We introduce Multitask discriminator Proximity-Guided IRL (MPG), which learns two complementary reward components: (1) a generalizable discriminator that transfers shared structure across related tasks to identify expert behavior in a new task, and (2) a proximity function that measures how far a state deviates from expert behavior and provides corrective guidance during exploration. We demonstrate the effectiveness of our method on multiple challenging navigation and manipulation tasks under significant variations (e.g., object configurations, table layouts, and initial robot poses), achieving an average success rate of 81.2%, outperforming the strongest per-task baseline by an average of 24.7 percentage points.
[AI-42] Seg2Grasp: A Robust Modular Suction Grasping in Bin Picking IROS2024
链接: https://arxiv.org/abs/2607.17757
作者: Hye-Jung Yoon,Juno Kim,Yesol Park,Jun-Ki Lee,Byoung-Tak Zhang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 7 pages, 6 figures, 2 tables. Published in the 2024 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2024)
Abstract:Current bin picking methods that rely heavily on end-to-end learning often falter when confronted with unfamiliar or complex objects in unstructured environments. To overcome these limitations, we introduce Seg2Grasp, a modular pipeline designed for robust suction grasping in dynamic and cluttered bin scenarios. Seg2Grasp is built on a three-step process: Segmentation, Grasping, and Classification. The Segmentation module employs a Transformer-based model to generate class-agnostic object masks from RGB-D images, ensuring accurate detection across various conditions. The Grasping module uses surface normals and mask proposals to determine the optimal suction points, enhancing grasp success. Finally, the Classification module leverages fine-tuned open-vocabulary Mask-CLIP for precise object identification, enabling versatile handling of diverse objects. Real-world robotic experiments demonstrate that Seg2Grasp outperforms existing methods in success rates and adaptability, establishing it as a powerful tool for automated bin picking in industrial settings.
[AI-43] Mobile Network Control with a World Model
链接: https://arxiv.org/abs/2607.17747
作者: Maxime Bouton,Ioanna Mitsioni,Simon Lindståhl,Jaeseong Jeong
类目: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI)
备注:
Abstract:The increasing complexity of mobile networks necessitates intelligent and dynamic control strategies for efficient, energy-conserving management. We propose a world model-based approach for network control that enables adaptive configuration of crucial parameters. The world model is trained from historical data and predicts the impact of its actions on future network states. Our controller leverages the model’s uncertainty estimate to robustly find optimal network configuration changes. Furthermore, the optimization objective can be changed dynamically without model retraining. We demonstrate the effectiveness of the approach in simulated closed-loop control of a mobile network energy-saving feature. Our results show improved performance in balancing energy savings with quality of service, compared to traditional methods and reinforcement learning approaches. Finally, we show the world model performance on real network data from, and evaluate counterfactual actions proposed by the controller under various throughput constraints.
[AI-44] WuYu-EnvLE-Bench: A Benchmark for Evaluating Large Language Models in Environmental Law Enforcement
链接: https://arxiv.org/abs/2607.17745
作者: Ziliang Yang,Yi Zhang,Kaijun Lin,Jiachao Ke,Haihong Xu,Zongguo Wen
类目: Artificial Intelligence (cs.AI)
备注: 98 pages, 45 figures,
Abstract:Large language models (LLMs) are increasingly considered for environmental enforcement, but their ability to produce traceable enforcement decisions remains unclear. We introduce WuYu-EnvLE-Bench, a benchmark built from real enforcement cases, regulatory standards, and expert review. It contains 2,521 benchmark instances, 14 tasks, and 12 pollution-medium subdomains across pre-enforcement, in-enforcement, and post-enforcement workflows. Using Absolute Environmental Enforcement Score (AES) and Intelligent Enforcement Index (IEI), we evaluate open-source and closed-source LLMs across capability, response quality, and resource efficiency. Results show that LLMs perform well on rule-bounded tasks but remain unreliable in evidence-chain construction, contradiction detection, multi-source integration, and procedural judgment. Model scaling also shows diminishing returns: medium-sized models approach leading models in structured tasks, while larger models do not reliably overcome evidence-reasoning bottlenecks. WuYu-EnvLE-Bench highlights the need for evidence-grounded, rule-aware, and task-adaptive enforcement reasoning.
[AI-45] Semantically Similar Logically Distinct: Diagnosing the Semantic-Answerability Gap in Table RAG
链接: https://arxiv.org/abs/2607.17742
作者: Jiaming Tian,Liyao Li,Wentao Ye,Haobo Wang,Lihua Yu,Zujie Ren,Gang Chen,Junbo Zhao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Tables are a critical knowledge source in retrieval-augmented generation (RAG), but a retrieved table may lack sufficient evidence to answer a query, a property we call answerability. While answerability broadly concerns whether a source or collection of sources contains sufficient evidence, retrieval models optimized for semantic relevance do not guarantee it even in the single-source case, creating a fundamental mismatch. To study this, we introduce TCR-Bench, a diagnostic benchmark for Table Content-level Answerability in RAG, built around sibling tables, i.e., tables with highly similar schemas but subtle content differences. On TCR-Bench, the dense retrievers we evaluate persistently exhibit a Semantic-Answerability Gap: they often retrieve the correct sibling group yet struggle to pinpoint the uniquely answerable table within it, dropping QA performance from 0.755 (oracle) to 0.330 (top-5 retrieved). Our analysis suggests this gap is associated with semantic accumulation, schema-level cue dependence, and weak row-column binding. As a diagnostic probe into the source of this gap, we test whether a lightweight two-stage pipeline, Answerability-Aware Reranking (AAR), applying direct query-table answerability judgment, can recover performance: it raises top-1 target retrieval from 18.2% to 57.4%, and this large gain is itself evidence that much of the observed failure reflects a missing answerability verification step, rather than an inherent limitation of model capacity alone.
[AI-46] MXSens: Sensitivity-Aware Mixed-Precision Quantization for Efficient LLM Inference
链接: https://arxiv.org/abs/2607.17733
作者: Simla Burcu Harma,Danila Mishin,Zhengyuan Su,Ayan Chakraborty,Elizaveta Kostenok,Dongho Ha,Babak Falsafi,Martin Jaggi,Yunho Oh,Amir Yazdanbakhsh
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:4-bit quantization enables efficient LLM inference, but suffers from significant accuracy degradation due to outliers. Prior work addresses this problem via data rotation or mixed-precision integer quantization, but often relies on software-managed scaling and frequent dequantization, incurring substantial overhead. Microscaling formats, such as MXINT, eliminate these inefficiencies by encoding scales in hardware, yet remain incompatible with rotation-based methods. Our analysis reveals that outliers vary in severity, from rare extremes to frequent mild deviations, and that quantization sensitivity is unevenly distributed across layers and columns. These insights motivate a fine-grained, sensitivity-guided approach. We introduce MXSens, a training-free method that assigns mixed mantissa bitwidths (4/6/8) based on column- and layer-wise sensitivity, naturally leveraging the block-wise structure of MXINT. MXSens outperforms state-of-the-art quantization methods across a range of models and tasks. Under the W4A4KV4 setting, MXSens achieves perplexities of 3.77 and 7.63 on LLaMA-2-70B and LLaMA-3-8B, respectively, substantially improving over existing baselines on WikiText-2. Our work establishes a new balance between accuracy and resource efficiency for LLM quantization.
[AI-47] Learning to Detect Cross-Modal Negation: An Analysis of Latent Representations and an Attention-Based Solution
链接: https://arxiv.org/abs/2607.17712
作者: Ali AbuSaleh,Leon Hammerla,Alexander Mehler
类目: Artificial Intelligence (cs.AI)
备注: This manuscript is an accepted version of the article (published at ICNLP2026). Published in IEEE Xplore, DOI: https://doi.org/10.1109/ICNLP69856.2026.11527861 document: this https URL
Abstract:Detecting high-level semantic concepts like negation across modalities remains a challenge for current multimodal systems. We analyze this as a fundamental representation learning problem, providing the first evidence that negation does not form a linearly or non-linearly separable class in the latent spaces of standard vision-language models (VLMs). We demonstrate that pretrained embeddings primarily encode modality-specific features, lacking a generalizable negation signal. To overcome this, we propose a novel cross-modal attention architecture that explicitly models inter-modal dependencies, achieving performance gains of up to +7.03% F1 over unimodal baselines. Our analysis reveals a key asymmetry: while textual negation often appears independently, visual negation is semantically dependent on linguistic context, a finding validated through our statistical analysis of 3,222 political video-text pairs automatically annotated via \textscQwen2.5-VL. By combining this analysis with self-supervised video representations (JEPA2), we advance the modeling of temporal negation. This work provides new methods and insights for learning robust, semantically-aligned representations in multimodal systems.
[AI-48] LaT: LLM -as-Trainer for Multi-Task Vehicle Routing Solvers
链接: https://arxiv.org/abs/2607.17708
作者: Yang Wang,Ya-Hui Jia,Wei-Neng Chen,Yi Mei,Wen Song,Zhiguang Cao
类目: Artificial Intelligence (cs.AI)
备注: 28 pages
Abstract:Multi-task neural solvers aim to handle multiple Vehicle Routing Problem (VRP) variants within a unified model, avoiding separate training for each constraint combination. However, VRP variants differ in optimization difficulty, while existing methods lack stage-wise feedback on their training status, making the model biased to some specific variants. Although meta-learning can support adaptive training, it typically requires bi-level optimization and additional gradient updates, increasing computational cost. To address this limitation, we propose LLM-as-Trainer (LaT), a plug-and-play training paradigm that uses a pretrained large language model as an external trainer. LaT periodically analyzes cross-task validation metrics to generate a stage-wise guidance vector. This vector is combined with the current task’s constraint vector and injected into each encoder layer, providing the neural solver with additional training information during subsequent policy optimization. Experiments on 16 VRP variants show that LaT improves the solution quality of several state-of-the-art multi-task neural solvers on both trained and unseen variants, supporting the effectiveness and generality of the proposed training paradigm.
[AI-49] ProEvent: An Event-centric Benchmark for Proactive Agents
链接: https://arxiv.org/abs/2607.17701
作者: Guanzhen Li,Liangming Pan,Leye Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Proactive agents are expected to anticipate user needs and provide autonomous assistance by perceiving environmental context without explicit instructions. A fundamental capability of such agents is to identify and track users’ upcoming events, enabling continuous and event-specific assistance. For example, by recording the time and location of a planned hike, an agent can deliver weather reminders in advance or provide navigation support before departure. However, existing works on proactive agents largely overlook event-centric assistance, and the open-ended nature of proactive assistance poses challenges for reliable evaluation. To bridge these gaps, we introduce ProEvent, the first event-centric benchmark designed to assess an agent’s ability to proactively maintain a user’s timetable based on ongoing instant messaging chats. ProEvent provides synthesized yet realistic chats that consider the dynamic interaction among users, concurrent chat threads, and noise in the real world, and evaluates proactive agents on response timing, single-step response correctness, and multi-step response correctness. Experiments on eight LLMs and pipelines reveal that current agents frequently overact and struggle with event cancellation. Notably, even GPT-5.1 only reacts correctly in 26.7% of scenarios. Further qualitative analysis reveals fundamental limitations of current LLMs as proactive agents, particularly in detecting implicit events and reasoning from the user’s first-person perspective. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17701 [cs.AI] (or arXiv:2607.17701v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.17701 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-50] Artificial Intelligence for Understanding and Managing Transportation Behavior in Sustainable Smart Cities
链接: https://arxiv.org/abs/2607.17694
作者: Junbiao Pang,Muhammad Ayub Sabir,Fatima Ashraf
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Urban transportation systems generate heterogeneous data, yet these data do not automatically become actionable management intelligence. This chapter adopts a behavior-centered perspective on artificial intelligence (AI), treating mobility records and passenger-generated text as behavioral evidence rather than behavioral truth. It examines four directions: bus arrival prediction for service reliability, taxi mobility pattern discovery for demand analysis and planning, abnormal behavior detection for accountable regulatory support, and passenger-perceived risk mining for service improvement. These directions are integrated through a closed-loop framework linking data input, behavior representation, AI inference, decision support, public value, and governance feedback. The chapter identifies data quality, privacy, fairness, interpretability, uncertainty, transferability, and human accountability as essential conditions for deployment. It thereby establishes a unified pathway from behavioral evidence to operational, planning, regulatory, and passenger-service decisions.
[AI-51] Integrating High-Level Requirements to Low-Level Tests with Machine-Readable VV Specifications
链接: https://arxiv.org/abs/2607.17686
作者: Mansur Arief,Nur Ahmad Khatim,Ali Akarma,Ahmad Alfan Alfian Irfan
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Modern software teams have mature tools for low-level testing, such as pytest, JUnit, and Jest, which make it inexpensive to write unit tests and run them on every commit. Systems engineering, in parallel, has developed rigorous principles for design verification and validation (VV), which has worked very well across engineering discipline to align user expecations and requirements with developers’ deliverables. In practice, however, the two rarely connect, and the link between users’ high-level requirements and the low-level tests that machines actually run is maintained by hand, if at all. This gap is increasingly costly for AI-enabled and cyber-physical systems, for which regulators now ask for traceable evidence that high-level requirements are met, while raw test results provide little of the structure such evidence requires. We introduce VNVSpec, an open-source framework that makes VV specifications machine-readable and executable. With this framework, users state high-level requirements directly or import them from catalogs derived from published standards. Then, the framework checks requirement quality, supports decomposition into module-level requirements with explicit metrics and acceptance criteria, links these requirements to test results through a traceability graph, and compiles the collected evidence into verdicts and audit-ready reports. We evaluate the framework by self-application, in which it is continuously assessed in CI against its own specification of 36 requirements verified by 449 tests, completed within limited time which scales linearly and thus can handle up to 10,000 requirements. We also discuss how the framework extends to testing black-box AI models and AI coding agents. The framework, its full test suite, the catalogs, and the benchmark scripts are available at this https URL.
[AI-52] Uncovering Latent Reasoning Strategies in Language Models
链接: https://arxiv.org/abs/2607.17674
作者: Awni Altabaa,John Lafferty
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:A language model p_\theta(y \mid x) trained on reasoning tasks learns to solve problems via multiple distinct strategies, yet these strategies are implicit and entangled within the model’s response distribution. We study the problem of decomposing the response distribution of a given pretrained language model into a structured, strategy-conditioned representation. Specifically, we learn a latent-variable factorization p_\theta(y \mid x) \leadsto (r_\phi(z \mid x), g_\phi(y \mid x,z)) , where a router r maps each input to a distribution over latent strategies z and a generator g produces the response conditioned on that strategy. A key challenge is that the generator, initialized from the base model, already represents p_\theta(y \mid x) without using z . Standard variational inference therefore gives the model no incentive to route information through z and can yield a severe form of posterior collapse. To address this, we propose a variational objective that measures fractional information gain relative to the base model’s response loss and concentrates reconstruction pressure on tokens with high base model surprisal, encouraging z to encode strategy-relevant response variation. We introduce a benchmark of multi-strategy algorithmic tasks and show that this objective recovers latent codes aligned with distinct reference strategies while preserving the base model’s response distribution.
[AI-53] Beyond Objective Expressivity: Geometry Preservation in Multimodal Contrastive Learning
链接: https://arxiv.org/abs/2607.17673
作者: Tillmann Rheude,Roland Eils,Benjamin Wild
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Contrastive learning is increasingly moving toward settings with three or more modalities instead of image-text pairs. Yet, extending models from pairwise to higher-order multimodal alignment can introduce optimization and representation challenges. We identify encoder Jacobian conditioning as a key factor in trimodal contrastive learning: poorly conditioned encoders exhibit collapsing or amplified singular-value spectra, leading to exploding Jacobian condition numbers and degraded multimodal alignment. We introduce geometry-preserving encoders (GPEs) by directly conditioning the Jacobian through regularization and demonstrating that simple modifications like LeakyReLU activations and residual paths recover these geometric benefits. Across a synthetic benchmark and four real-world datasets including missing modalities, improving Jacobian conditioning boosts retrieval and linear probe performance across multiple contrastive objectives, whereas expressive objectives yield little benefit in linear probes. More broadly, our results show that multimodal contrastive learning depends not only on objective expressivity, but also on the geometric and optimization properties of the underlying encoders.
[AI-54] Selectivity Matters: Source Node Influence Pruning for Unsupervised Graph Domain Adaptation
链接: https://arxiv.org/abs/2607.17668
作者: Ridong Han,Yawen Shen,Zhongnian Li,Tongfeng Sun,Xinzheng Xu,Abdulmotaleb El Saddik
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 12 pages,7 Figures
Abstract:Unsupervised Graph Domain Adaptation (UGDA) aims to facilitate knowledge transfer from a labeled source graph to an unlabeled target graph by mitigating cross-domain distribution shifts. Existing methods primarily focus on node-level feature alignment in latent spaces, relying on the implicit assumption that all source nodes contribute positively to the alignment. However, this assumption often fails because a node’s semantic information is intrinsically coupled with its topological graph structure. Due to structural shifts, source nodes with severe structural deviations (e.g., structural outliers) lack semantic counterparts in the target graph, and forcing alignment on them introduces severe noise and causes negative transfer. To bridge this gap, we argue that selective source node utilization is superior to full-graph training, thereby shifting the research paradigm from feature-level alignment to data-level refinement. To this end, we propose Source Node Influence Pruning (SNIP), a novel model-agnostic, data-centric refinement framework. Specifically, SNIP quantifies the structural discrepancy between individual source nodes and the target domain by integrating multiple centrality measures, assigning each source node an influence score. A rank-based normalization mechanism is further employed to eliminate scale variations across different measures, allowing SNIP to effectively identify and filter out structurally incompatible nodes with low influence scores. As a plug-and-play method, SNIP constructs a refined “sub-source” graph that is inherently more beneficial for subsequent alignment. Comprehensive experiments across eight transfer scenarios on five real-world datasets demonstrate that SNIP consistently outperforms competitive baselines and significantly enhances adaptation performance, validating the superiority of selective node utilization over full-graph training.
[AI-55] FlowBlock: Wavefront-Parallel Decoding for Self-Correcting Diffusion Language Models
链接: https://arxiv.org/abs/2607.17652
作者: Bing Tian,Haikun Liu,Xiaocheng Zhong,Zhuohui Duan,Zhaokai Luo,Huayi Jin,Zhiyong Wang,Xiaofei Liao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Block-wise diffusion large language models (dLLMs) decode sequentially at the block level, enabling effective KV-cache reuse across blocks but making inter-block decoding strictly serial. Prior work has attempted to unlock inter-block parallelism through post-training methods, but achieves only modest speedups and often degrades accuracy. We observe that self-correcting dLLMs offer a training-free alternative: token-to-token (T2T) editing can repair tokens drafted with a slightly stale upstream context, so a downstream block requires only an informative draft rather than a finalized predecessor. This turns block finality from a hard dependency into a scheduling resource. We propose \textbf\flowblock, a training-free parallel decoding framework built on two mechanisms. (i) \emphGated Wavefront Decoding admits blocks into a bounded wavefront only when a readiness gate is satisfied, jointly refines active blocks via T2T editing, and commits blocks in order under a windowed block-causal mask that preserves exact frozen-prefix KV caches reuse. (ii) \emphHeterogeneous Wavefront Packing assigns each request an independent wavefront while packing asynchronous windows into dense, shape-stable batched forwards. Across different benchmarks, \flowblock improves tokens per second (TPS) over LLaDA-2.1 and LLaDA-2.0, two serial block-wise dLLMs, by up to 2.95 \times and 4.01 \times , while reducing latency by up to 53.6% and 77.1%, respectively. It also improves average accuracy by 1.3 points. Compared with D2F, a training-based inter-block-parallel baseline, \flowblock achieves higher accuracy and up to 16 \times higher batched serving throughput.
[AI-56] Verify Repair Repeat or Stop? Robust Stopping for Noisy Verify-Repair Loops in LLM Agents
链接: https://arxiv.org/abs/2607.17641
作者: Yitao Wu,Si Shen,Rui Yang,Hong Peng,Bin Hu
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 18 pages, 10 figures, 10 tables. Under review
Abstract:Verify-repair loops are a standard means for large language model (LLM) agents to correct faulty plans in code generation, mathematical reasoning, and tool use. When both the verifier and the repairer are noisy, repair can damage already-correct plans, and reported acceptance keeps rising while true validity falls, so existing methods lack a principled basis for deciding when repair should stop. We propose VRR-Stop, a robust stopping framework for noisy verify-repair-repeat (VRR) loops. A four-parameter noise model separates verifier false acceptance and false rejection from the repair and damage behavior of the repairer. Belief filtering turns repeated verification votes into an estimate of committed validity, and the loop commits or repairs according to the sign of the true marginal gain, which requires only sign identifiability rather than accurate recovery of all parameters. When verifier discrimination approaches zero, calibration itself fails and estimation error can flip the stopping sign, so we pair VRR-Stop with VRR-Guard, an estimation-free fallback that replaces the incumbent candidate only under a sufficient verification margin. On a GSM8K stress setting, VRR-Stop improves final true validity by 60.6 percentage points over fixed five-round repair at an average cost of 0.72 repair rounds. Across settings, stopping reliability is governed jointly by verifier discrimination and the decision margin rather than by the absolute size of estimation error.
[AI-57] ypiCore: A Hybrid Active Query Strategy for Class-Incremental Learning on Time Series
链接: https://arxiv.org/abs/2607.17632
作者: Gabor Szucs,Samuel Jacsev,Marcell Nemeth,Davide Dalle Pezze,Gian Antonio Susto
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Time series data play a pivotal role across numerous domains, including healthcare and manufacturing. In real-world environments, models must cope with distribution shifts over time, a challenge commonly addressed through Continual Learning (CL) techniques. However, existing CL methods face a critical limitation: real-world data streams are rarely fully labeled, making annotation cost a major practical constraint. This paper investigates Active Class-Incremental Learning (ACIL) for multivariate time series, where a model must sequentially learn new classes while selectively querying labels under a fixed annotation budget. We present a systematic evaluation of a wide range of query strategies combined with multiple rehearsal-based approaches, assessing their impact on plasticity, stability, and label efficiency across four benchmark datasets. Our analysis reveals the limitations of uncertainty-based and distribution-aware methods in achieving strong performance under constrained labeling budgets. To address these shortcomings, we propose TypiCore, a novel hybrid query strategy that alternates between typicality-based and diversity-based sample selection across active learning cycles, enabling the construction of memory buffers that are both representative and diverse. Evaluated on the TSCIL benchmark, TypiCore delivers statistically significant improvements over all baselines and matches or surpasses fully supervised continual learning performance on multiple datasets while requiring a fraction of the available labels.
[AI-58] Mechanistic Attention Guidance for Agent Memory Refinement
链接: https://arxiv.org/abs/2607.17621
作者: Yechao Hong,Haiquan Qiu,Yaqing Wang,Quanming Yao
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Existing self-evolving memory systems mainly improve agent memory based on textual outputs, such as task trajectories and reflections. However, this text-based paradigm rarely incorporates internal mechanistic signals, leaving how retrieved memory is actually utilized during task execution underexplored. This limitation can lead to unreliable error attribution and hallucinated memory modifications. In this work, we show that retrieval-head attention provides a mechanistic signal for revealing segment-level memory utilization. By aggregating attention over memory segments and decision steps, we construct a context utilization matrix that exposes recurring memory-use patterns and indicates corresponding refinement strategies. Building on this observation, we propose Attention-Guided Memory Refinement (AGMR), a framework that uses utilization patterns revealed by attention to guide targeted segment-level memory updates. AGMR corrects or enhances memory for failed executions, simplifies memory for successful executions, and verifies each update through re-execution. Experiments on interactive decision-making benchmarks show that AGMR improves both task performance and memory efficiency over text-only memory refinement baselines. Code is available at this https URL
[AI-59] Re-Sonance: A Dysarthric Asynchronous Real-Time Speech Conversion System Based on a Three-Stage Cascaded ASR-LLM -TTS Architecture
链接: https://arxiv.org/abs/2607.17615
作者: Yuxuan Wu,Yifan Xu,Junkun Wang,Jiayong Jiang,Xin Zhao,Zhaojie Luo
类目: ound (cs.SD); Artificial Intelligence (cs.AI)
备注: Accepted by NCMMSC 2025
Abstract:Individuals with dysarthria face significant challenges in professional speaking scenarios such as conferences, presentations, and meetings, where real-time communication is crucial. While existing Augmentative and Alternative Communication (AAC) systems provide basic support, they often fail to meet the demands of professional speaking environments due to high latency and unnatural speech patterns. This paper presents Re-Sonance, a novel LLM-enhanced speech-driven AAC system designed for real-time professional speaking scenarios. By integrating Whisper ASR, Qwen LLM, and CosyVoice TTS, Re-Sonance achieves improved speech intelligibility and naturalness while maintaining real-time performance. Both subjective and objective evaluations using a Mandarin dysarthric speech dataset demonstrate that our speech reconstruction approach significantly improved intelligibility while preserving semantic coherence for speakers with mild to moderate dysarthria. Although performance remains limited for severe dysarthria cases, our findings validate the potential of LLM-based methods for enhancing speech-driven AAC systems, paving the way for more effective and accessible communication technologies.
[AI-60] rustworthy Protein-Ligand Binding Affinity Prediction via Reliability-Aware Multi-Engine Fusion
链接: https://arxiv.org/abs/2607.17601
作者: Yongchan Hong,Defu Cao,Wenjin Liu,Thomas Ku,Jordy Homing Lam,Emily Nguyen,Willie Neiswanger,Vsevolod Katritch,Yan Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate protein-ligand binding affinity prediction is central to computational drug discovery, yet modern docking engines frequently disagree without indicating which prediction to trust. Consensus scoring and ensemble methods improve mean accuracy but treat all predictions identically without interpretable confidence measures or uncertainty decomposition, ignoring the chemical context of each protein-ligand pair. To address this limitation, we introduce RELIABLE-BA (RELIABiLity-aware Evidential fusion for Binding Affinity), an evidential framework for multi-engine binding affinity prediction. Our model comprises three steps: (1) modeling each engine as an evidential expert via Normal-Inverse-Gamma distributions, (2) scaling epistemic uncertainty through learned reliability from molecular context while preserving each expert’s predictive mean, and (3) fusing experts through closed-form aggregation that captures both individual uncertainty and inter-engine disagreement. Experiments on the PDBBind and BDB2020+ benchmarks demonstrate competitive point prediction with substantially improved uncertainty calibration, and additional validation on the SARS-CoV-2 Mpro dataset and 5HT2A receptor demonstrates applicability to clinically relevant drug targets. Crucially, these uncertainty estimates enable reliable filtering of protein-ligand pairs, reducing prediction error by up to 25% when retaining only high-confidence pairs. To our knowledge, RELIABLE-BA is the first multi-engine binding affinity prediction framework to combine evidential fusion with context-dependent reliability, offering a principled path toward trustworthy AI-guided drug discovery. Our code is publicly available at this https URL.
[AI-61] Detection Attribution Narration: An End-to-End Pipeline for Explainable Money Mule Identification
链接: https://arxiv.org/abs/2607.17586
作者: Yuge Zhang,Yuanxing Zhang,Yichao Jin,Khairul Amsyar Mohd Razis,Nicholas Qi An Choo,Kai Yin Anders Wong,Xinyan Tang,Kenneth Zhu Ke,Wee Keong Dennis Lee,Jingyuan Zhao
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Money mule accounts are critical facilitators of financial fraud, yet detecting them at scale remains challenging due to the heterogeneous nature of transactional and behavioural data. We present an end-to-end pipeline for customer-level mule detection comprising three stages: (1) a LightGBM classifier trained on 280 engineered features spanning transaction patterns, account demographics, network topology, and temporal behaviour; (2) a TreeSHAP attribution layer that decomposes each prediction into feature contributions; and (3) a large language model (LLM) module that converts SHAP attributions into analyst-facing natural-language narratives. We evaluate across three open-weight LLM families and assess explanation quality through analyst feedback. In a live production deployment, the system achieves a yield rate of 89%, up from 61% under the incumbent rule-based system, with monthly alert volume expanding from 211 to 302, reflecting broader true-positive coverage rather than increased noise. This corresponds to a 60% incremental adverse detection beyond existing review workflows, substantially outperforming the rule-based approach. Qualitative feedback from analysts indicates that LLM-generated narratives reduce cognitive load during alert triage. We further discuss implications of deploying LLM-augmented explainability in regulated financial environments.
[AI-62] A Dual-Hypothesis Reasoning Framework for LLM Guardrails
链接: https://arxiv.org/abs/2607.17575
作者: Md Asiful Islam,Mihai Surdeanu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We propose ARBITER, a novel LLM guardrail framework that introduces two key ideas: (i) dual-hypothesis reasoning, a reasoning method for LLM guardrails that explicitly considers both safe and unsafe interpretations of a prompt before making a safety decision, and (ii) multi-component supervised fine-tuning (MC-SFT), a structured training loss for reasoning-based guardrails that decomposes LLM outputs into logical components and weights them according to their importance. Existing reasoning-based guardrails often rely on expensive procedures, such as generating reasoning traces using larger or closed-source teacher models and applying full-parameter fine-tuning. In contrast, ARBITER uses a cost-effective self-generation strategy for reasoning traces and LoRA-based parameter-efficient fine-tuning while still achieving better performance than these expensive approaches. Additionally, ARBITER provides faithful evidence-phrase explanations for unsafe decisions, enabling a more transparent and interpretable guardrail method. Experiments on three safety moderation benchmarks show that ARBITER outperforms existing reasoning-based and non-reasoning guardrail baselines, with clear gains in out-of-domain evaluations.
[AI-63] Predictive Training with Latent Imagination for Visual Quadruped Navigation
链接: https://arxiv.org/abs/2607.17574
作者: Yancheng Zhu,Wanli Ma,Chen Han,Irvin Haozhe Zhan,Bingfeng Qin,Yixin Xu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 10 pages, 9 figures
Abstract:Reinforcement-learning navigation policies for legged robots select actions reactively from current observations and short-term memory, with limited capacity to anticipate how moving obstacles will evolve in the near future. In dynamic environments, this reactivity causes the robot to respond too late because collision risk depends on short-horizon scene structure rather than on current obstacle positions alone. Lightweight predictive supervision applied to the policy’s recurrent state during training can encode anticipatory obstacle dynamics without modifying the inference-time controller. We augment a reactive LSTM-SRU navigation backbone with an auxiliary JEPA-style predictor and SIGReg regularization: during training, the predictor supervises the deterministic hidden state to anticipate its own next state; at inference, it is fully discarded, incurring zero additional computational cost. On simulated and real-world navigation benchmarks with dynamic obstacles, our method substantially improves navigation success while reducing collision rates through the predictive training signal alone, without additional inference-time parameters. Real-robot deployment on a Unitree Go2 demonstrates zero-shot sim-to-real transfer: the controller navigates cluttered indoor and dynamic outdoor environments without fine-tuning, with evasive behavior consistent with the collision reduction observed in simulation.
[AI-64] CoCurve: Cross-Module Co-Pruning Curvature for Training-Free Structured LLM Pruning
链接: https://arxiv.org/abs/2607.17568
作者: Zhiren Gong,Zihao Zeng,Zijie Wang,Tiantong Wang,Chau Yuen,Wei Yang Bryan Lim
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Structured pruning compresses large language models (LLMs) by removing whole computational units, such as attention heads and feed-forward (FFN) channel groups. Most training-free methods, however, rank these units independently, implicitly treating the loss from pruning a set as the sum of its individual losses. This view fails for Transformers, whose sublayers are coupled through a shared residual stream. Two individually weak units can thus be jointly indispensable, yet independent scoring is blind to such dependence and removes them together. We introduce CoCurve (Cross-Module Co-Pruning Curvature), a calibration-only, fine-tuning-free method that prunes attention and FFN units jointly. A second-order Taylor expansion of the token-level KL between the frozen model and its masked copy yields a single Fisher matrix whose diagonal is classical node saliency and whose off-diagonal entries are co-pruning curvature edges: the extra damage of removing two units together. Under a single-ablation additivity approximation this matrix reduces to a Gram product of single-unit ablation features, so the full M x M interaction is recovered from M forward passes, with no pairwise sweeps or gradients. Pruning then reduces to one budgeted quadratic program, solved in a single shot under a shared attention–FFN budget, with no labels, fine-tuning, or recovery.
[AI-65] ZifaMem: Structured Memory for Persona Preference and Emotional Continuity in AI Companions
链接: https://arxiv.org/abs/2607.17564
作者: Jingzhe Fang,Guozhi Xu,Yunfan Cui,Xiaochen Yang,Zhangyu Hua
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:AI companions are judged not only by single-turn fluency but by whether they sustain emotional continuity: remembering who the companion is, what the user prefers, and how the relationship has felt. We present ZifaMem, a structured memory system that organizes dialogue into session summaries, episodic memories, and a consolidated user model. Against a deployment-honest comparator that supplies the full raw dialogue history, and under a fixed LLM-as-a-judge protocol with route audits, structured memory raises pooled four-backbone emotional-intelligence scores by 11.4% (95% CI 6.3% to 17.1%), and persona grounding improves on all four backbones (Claude +42% relative). Multi-turn affect context wins a +39% net preference over a single-turn snapshot (exploratory), whereas an additional emotion state machine yields no measurable gain on any of five endpoints. Under an identical preregistered protocol, three memory systems (ZifaMem, Mem0, and filtered verbatim retrieval) each improve significantly over raw-history deployment, and ZifaMem and Mem0 are statistically equivalent within +/-5 points on the preregistered primary preference endpoint. The ZifaMem SDK, CLI, and portable Agent Skills are open-sourced at this https URL.
[AI-66] Reinforcement Learning: From Algorithms To Foundation Models
链接: https://arxiv.org/abs/2607.17560
作者: Zihan Ding
类目: Artificial Intelligence (cs.AI)
备注: Princeton University PhD Thesis 2026
Abstract:Reinforcement learning (RL) provides a framework for sequential decision making under explicit objectives. In its classical form, RL studies how an agent should act to maximise long-term reward in a dynamic environment. In richer settings, the problem extends beyond a single agent and fixed environment: intelligent behavior may require strategic interaction, adaptation to uncertainty, and reasoning over high-dimensional worlds. This thesis studies RL from two perspectives: algorithms in games and RL in the era of foundation models. The first part focuses on multi-agent RL in games. It examines how incentives, policies, and equilibrium concepts interact in competitive and general-sum environments, spanning two-player zero-sum games, large-scale video games, and multi-player settings with general structure. These works investigate learning in multi-agent systems and the behavior of RL methods in interactive environments. The second part studies RL with generative and foundation models, motivated by the idea that prior knowledge can enrich sequential decision making. Pretrained generative models and learned world models serve as representation tools and structured priors for planning, control, and policy optimization. The thesis develops diffusion-based world models, investigates RL for efficient video generation, explores generative models as policy classes, and studies interactive video world models in which actions shape future observations. It also addresses long-horizon modeling through architectures with memory. Together, these contributions present a unified view of RL as objective-driven adaptation in complex sequential domains. From strategic games to generative world models, the thesis highlights how RL connects decision making, environment modeling, and emerging foundation-model capabilities, offering a broader perspective on the principles underlying intelligent behavior. Comments: Princeton University PhD Thesis 2026 Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17560 [cs.AI] (or arXiv:2607.17560v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.17560 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-67] COLIP-2: Olfaction-Vision-Language Embeddings
链接: https://arxiv.org/abs/2607.17559
作者: Kordel Kade France
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
备注: Model Card
Abstract:The Contrastive Olfaction-Language-Image Pre-training 2 (COLIP-2) model is a multimodal embeddings space that places olfaction as a first-class citizen among vision and language. Molecular structure, gas-sensor readings, odor-descriptor language, and images are all trained into a single shared representation space, so that a robot can localize a detected aroma to objects in a scene probabilistically. No ImageNet-scale datasets of paired image-scent examples exists which warrants the need for their collection. Our intent with the release of COLIP-2 is to demonstrate the limit of what can be built for robotics with open-sourced olfactory data in order to ground the argument for why new methodologies and datasets are necessary in order to enable advanced olfactory-oriented perception capabilities. We enumerate results from internal testing of the COLIP-2 architecture and make necessary optimizations to run the model at the edge for real-time robotics applications. While developed with robotics in mind, the design of COLIP-2 has been influenced by experts across many disciplines of science in academia and industry, and we hope that the model can be useful in any multimodal domain requiring olfactory intelligence.
[AI-68] Why Does Feedback-Augmented Self-Distillation Fail to Improve Retrieval-Interleaved Search Agents ?
链接: https://arxiv.org/abs/2607.17558
作者: Fan Yang,Rui Meng,Yuxin Wen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:On-policy self-distillation (OPSD) offers a promising approach for training large language models without relying on a separate teacher model. However, its effectiveness on complex agentic tasks remains largely unexplored. In this work, we instantiate Feedback-Augmented Self-Distillation (FA-SD), a self-distillation algorithm for agentic search that leverages successful demonstrations as privileged information. We identify that models can rely on recurring reasoning-and-search output templates, producing trajectories that appear diverse but are largely agnostic to the input question, making the KL-based self-distillation signal uninformative. We term this phenomenon decoding collapse, a failure mode that can be missed by existing evaluation metrics. To understand its underlying cause, we show that although the self-teacher achieves stronger performance, learning remains inherently unstable due to inconsistent supervision signals. We further decompose this inconsistency into model inconsistency and prompt inconsistency, and show that the latter can significantly degrade the quality of the supervision signal, limiting the effectiveness of self-teacher learning. To mitigate this inconsistency, we introduce an exponential moving average (EMA) teacher to stabilize the self-teacher and provide more consistent supervision signals. Although the EMA teacher requires a warm-up phase during which performance may temporarily regress, it ultimately improves model performance by providing more stable supervision.
[AI-69] Retain or Consolidate? Budget-Dependent Operator Selection for Language Agent Memory
链接: https://arxiv.org/abs/2607.17545
作者: Qingcan Kang,Mingyang Liu,Shixiong Kai,Kaichao Liang,Zhentao Tang,Yuqi Cui,Tao Zhong,Mingxuan Yuan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Language agents depend on memory across interactions. However, the limited context windows of large language models (LLMs) and their inference costs constrain how much memory can be used at once. Existing systems mainly follow two strategies: memory retention and memory consolidation. Retention keeps raw records and preserves exact details, but relevant evidence may not fit under a tight budget; consolidation compresses and combines records, improving coverage per token but risking the loss of query-critical details. Neither strategy is universally preferable. This raises two central questions: when should consolidation replace retention, and which operator – Merge, Abstract, or Rewrite – should be selected? We formalize this decision by decomposing each operator’s utility into a coverage effect on evidence omitted by retention and a signed replacement effect on raw evidence that already fits. Their balance explains why the preferred action changes with relative budget pressure. We implement this mechanism with Offline Abstraction-Safety (OAS), a lightweight learner that estimates action utilities from pre-generation features with held-out harm calibration. The public LongMemEval and LoCoMo benchmarks show the same budget-dependent pattern. On LongMemEval, consolidation improves absolute accuracy by up to 48% under tight budgets, whereas retention is preferable under loose budgets; LoCoMo replicates this crossover at a smaller budget, consistent with its shorter evidence. On both datasets, cross-note abstraction and merging generally outperform local rewriting when compression is necessary.
[AI-70] CommitLLM : A Fine-Tuned Pipeline for Git Commit Message Generation
链接: https://arxiv.org/abs/2607.17532
作者: Md Rafid Haque,Poojan Narendrabhai Patel,Meetkumar Vijaybhai Raychura
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 7 pages, 4 figures
Abstract:Developers frequently write uninformative git commit messages such as “fix” or “update stuff”, degrading the value of version-control history for code review, debugging, and onboarding. We present CommitLLM, a three-stage pipeline that generates concise, Conventional Commits-compliant messages from code diffs using a fine-tuned small language model. The system combines (1) QLoRA fine-tuning of Mistral-7B-Instruct-v0.2 on the CommitPackFT dataset, (2) constrained decoding to enforce brevity, and (3) deterministic post-processing to strip conversational artifacts and enforce format. On a 50-sample evaluation, CommitLLM achieves 98% format compliance (vs. 22% for vanilla Mistral), reduces average output length from 154.8 to 37.9 characters, and improves LLM-as-a-Judge scores from 1.97 to 3.68 out of 5. Notably, the post-processing layers contribute more to quality improvement than the fine-tuning itself, suggesting that for structured-output tasks, treating the LLM as a component in a deterministic pipeline is more effective than optimizing the model alone. The entire system runs on a single consumer GPU (NVIDIA T4, 16 GB VRAM).
[AI-71] Can AI Agents Really Complete RTL-to-GDS? Lessons from Benchmarking Tool-Interactive EDA Workflows
链接: https://arxiv.org/abs/2607.17528
作者: Jinyuan Deng,Zhengrui Chen,Xufeng Wei,Tianyu Xing,Chenyi Wen,Cheng Zhuo
类目: Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR); Machine Learning (cs.LG)
备注:
Abstract:LLM-driven agent systems have emerged as a promising paradigm for electronic design automation (EDA), demonstrating strong potential for automating complex design workflows. However, existing evaluations primarily examine individual language models on isolated EDA tasks, providing limited insight into how different agent systems perform across complete EDA flows. In this work, we present FluxBench, a systematic evaluation of AI agents on end-to-end EDA workflows under unified prompts, tool environments, and technology library settings. Our evaluation covers representative scenarios, including RTL generation with open-source toolchains and an RTL-to-GDS flow using closed-source commercial EDA tools for industrial applications. Through these workflows, we assess agents’ capabilities in RTL code generation, iterative repair, tool-feedback utilization, logic synthesis, placement and routing (PR), and Engineering Change Order (ECO) automation. To further characterize the efficiency of agent systems, we introduce Token ROI, a cost-efficiency metric that measures effective improvements in EDA artifacts relative to token usage and runtime cost. Experimental results show that, even when built on the same foundation model, different agent system architectures can exhibit performance gaps of up to 86.27%. Moreover, among systems with comparable task performance, Token ROI can differ by as much as 105.92\times . In the RTL-to-GDS flow using PicoRV32 as a case study, FluxEDA achieves an end-to-end score of up to 97.94, outperforming Claude Code equipped with domain-specific EDA skills by up to 8.39\times . These results indicate that domain-specific skills alone are insufficient to improve agent performance in large-scale EDA scenarios. Instead, both agent system design and foundation model capability play critical roles in enabling effective automated EDA workflows.
[AI-72] One-step lowest-variance selection in a Gaussian random-field model motivated by masked diffusion: Total correlation and a square root collision threshold
链接: https://arxiv.org/abs/2607.17522
作者: Linjun Li
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Theory (cs.IT); Probability (math.PR)
备注: 27 pages; welcome comments
Abstract:Motivated by confidence-guided parallel unmasking in masked discrete diffusion, we study a single selection step in a stylized Gaussian random-field model. A locally dependent nonnegative score field represents position wise uncertainty, and the scheduler selects the K positions with the smallest scores. Dependence among the selected positions is measured through a distance-dependent Gaussian correlation model. This separation provides a tractable framework for quantifying how the geometry of low-score locations affects the dependence cost of factorized parallel decoding. We establish two complementary results. In a conservative sub-square-root regime, the conditional Gaussian total correlation of the selected block vanishes in probability. At the square-root scale, it remains non-negligible with positive asymptotic probability and admits a strictly positive expectation lower bound. Synthetic experiments support the predicted finite-size behavior. These results provide a rigorous stochastic-geometry baseline for understanding how budget size, score dependence, and spatial correlation jointly shape one-step confidence-based selection in masked discrete diffusion.
[AI-73] Retrieval-Augmented Interpretable Learning: Towards Task-Specific Zero-Shot Models in Healthcare
链接: https://arxiv.org/abs/2607.17508
作者: Sazan Mahbub,Caleb Ellington,Zhiyuan Li,Yixin Yang,Souvik Kundu,Ben Lengerich,Eric P. Xing
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:We introduce Retrieval-Augmented Interpretable Learning (RAIL), a probabilistic meta-learning framework for zero-shot generation of task-specific interpretable models that synthesizes coefficient-space structure from natural-language task descriptions and a memory of previously learned task-specific predictors. RAIL retrieves related source tasks, transfers structure through coefficient space, and generates a new predictor in the original diagnostic-feature space, enabling zero-shot and few-shot clinical procedure prediction with feature-level explanations. Its probabilistic formulation provides uncertainty over retrieval, model coefficients, and predictions, supporting reliability-aware deployment: uncertain predictions or unstable explanations can be flagged for additional clinical review rather than treated as automatic decisions. This makes RAIL particularly suited for healthcare settings, where prediction tasks are highly long-tailed, new clinical targets arise frequently, and models must remain inspectable, uncertainty-aware, and compatible with human oversight. Across long-tailed clinical procedure prediction tasks, RAIL maintains reliable performance across data-availability regimes: it achieves 73.4% accuracy in the held-out zero-shot settings, where no supervised task-specific model can be trained, and remains near 73.2% accuracy in the extreme few-shot regime with only 2-4 examples, where supervised task-specific models perform close to chance. RAIL further benefits from clinically informed task representations and yields retrieval, uncertainty, and coefficient-level diagnostics that make model behavior more transparent. These results suggest a path toward scalable clinical prediction systems that can adapt to new tasks while preserving interpretability and reliability.
[AI-74] Pailitao-MMSearch: Building Native E-Commerce Multimodal Search Foundation
链接: https://arxiv.org/abs/2607.17499
作者: Xiaohan Ye,Xu Chen,Zihan Gong,Jian Ding,Lianyu Du,Baicheng Chen,Yunmeng Shu,Jingqian Zhao,Zhixiang Zhao,Shuaiqi Jia,Chong Ma,Shuwen Xiao,Xiangheng Kong,Yuan Gao,Jun Song,Jinsong Lan,Xiaoyong Zhu,Bo Zheng
类目: Artificial Intelligence (cs.AI)
备注: Technical Report: Pailitao-MMSearch
Abstract:The evolution of e-commerce has fundamentally transformed how users search for products, shifting from simple text-based keyword queries to complex multimodal interactions that seamlessly combine product images, natural language descriptions, and mixed-intent instructions. However, existing approaches face a critical dilemma: single-modal specialist models, deployed independently for text retrieval, visual search, and voice recognition, operate in isolation and cannot handle cross-modal queries, while general-purpose vision-language models lack the domain-specific knowledge necessary for fine-grained product understanding, user behavior modeling, and commercial intent reasoning. In this work, we present Pailitao-MMSearch, one native e-commerce multimodal search foundation model designed to bridge this gap. Our approach introduces three key innovations: (1)HybSID (Hybrid Semantic ID);(2)a two-stage continual pre-training strategy; and (3)a hybrid reasoning post-training pipeline. Built upon Qwen and deployed on Taobao’s Pailitao multimodal search platform, Pailitao-MMSearch achieves substantial improvements in online A/B testing, including up to +13.61% in Gross Merchandise Volume (GMV) and +8.21% in transaction volume compared to traditional multi-modal search pipeline, demonstrating the effectiveness of our native e-commerce multimodal search large language models.
[AI-75] SALT: Salience-Aware Lexical Trie for Long-Context Compression
链接: https://arxiv.org/abs/2607.17486
作者: Oteo Mamo,Hyunjin Yi,Joydhriti Choudhury,Shangqian Gao,Weikuan Yu
类目: Performance (cs.PF); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:As large language models (LLMs) process increasingly longer prompts, computation and KV-cache memory costs have emerged as major bottlenecks in inference systems. Existing input-level prompt compression methods address this, but rank each sentence by a scalar relevance score, treating the document as an unstructured pool of words and sentences. Under tight budgets, this causes theme collapse, where the dominant theme(s) of a document consumes the budget, discarding less-frequent yet task-relevant themes. Preserving thematic coverage instead requires allocating the budget across recurring themes rather than scoring sentences in isolation. To this end, we propose SALT, a model-agnostic extractive framework that organizes per-sentence keywords into a trie ordered by sentence frequency (SF), a lightweight, reusable proxy for document thematic structure. This trie-based organization smooths memory allocation and prevents dominant themes from monopolizing the budget. Multi-anchor retrieval activates trie nodes labeled by query keywords at any depth, and the trie persists across dialogue turns, supporting multi-turn use without re-encoding the document. By preserving document themes, SALT reduces the prefill computation and memory cost of long-context prompts while remaining composable with KV-cache methods that target decoding-time latency and memory.
[AI-76] Panache: One-Pass Motif Discovery at Every Window Length
链接: https://arxiv.org/abs/2607.17481
作者: Tej Sanibh Ranade
类目: Artificial Intelligence (cs.AI); Databases (cs.DB); Machine Learning (cs.LG)
备注: 23 pages, 4 figures, 19 tables
Abstract:Motif discovery, the search for recurring patterns within a time series, is a core primitive of exploratory data analysis. A pattern, however, is defined by its duration, which analysts rarely know in advance. To resolve this unknown duration, an interval of window lengths is defined, and the accepted method is to try every length in that interval. Existing pan matrix profile (PMP) methods compute one z-normalized matrix profile per length, so L lengths cost L quadratic self-joins over the same series. We introduce Panache, to our knowledge the first one-pass streaming algorithm for z-normalized PMP motif discovery. It replaces the repeated self-joins with a single scan whose runtime is near-linear in the series length. The key observation is that mean-centering a subsequence changes only its DC Fourier coefficient, so the non-DC spectrum of every z-normalized subsequence can be maintained online by sliding-DFT recurrences and running statistics. This spectral state is the key under which similar subsequences collide in an occupancy-controlled hash directory and, through Parseval’s theorem, yields a lower bound that rejects most colliding pairs before any exact computation. Panache computes every data-dependent parameter itself, leaving only a resource budget to tune. At the default budget, it recovers all top-20 pan-motifs against exact fixed-exclusion ground truth on 17 UCR configurations, and is faster than every CPU and GPU baseline benchmarked in this paper. On Wafer at five million samples over 51 lengths, Panache completes one pass in 2.9 minutes and emits the exact motifs in 6.0 minutes, against 7.95 hours for the fastest exact CPU baseline and 38.3 minutes for SCAMP on an H100 GPU.
[AI-77] AEC-DS: Adaptive Erasure Coding with PDP-Triggered Reputation and QoS-Aware Migration for Decentralized Storag e
链接: https://arxiv.org/abs/2607.17460
作者: Shuaiwen Li,Weihang Yu,Ke Wang,Meng Han
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:In decentralized storage systems, audit results are often not used directly to guide later redundancy and shard-placement decisions, which can lead to inefficient resource allocation and delayed recovery. We propose AEC-DS, a closed-loop adaptive erasure coding mechanism driven by Provable Data Possession (PDP) feedback. PDP audits continuously update node reputation, while a QoS-aware migration policy adjusts shard placement according to node reliability and data priority. The policy moves high-priority shards from unstable nodes to more reliable nodes in the cold tier and penalizes unstable nodes in subsequent placement decisions. Simulations with 800 nodes and 500 files show that AEC-DS maintains 100% data durability under the evaluated fault model with a redundancy factor of 1.25x. Compared with Static-EC, Dynamic-EC, and DRD-EC, AEC-DS reduces cumulative recovery operations by 66.8%-75.2%. Ablation results further show that class migration plays a major role in preventing data loss, improving the measured loss-prevention capability by 176.8%. These results indicate that PDP feedback can connect integrity auditing with redundancy and placement adaptation, providing a practical path toward self-healing decentralized storage while accounting for the additional cost of migration.
[AI-78] Empirical Grounding Improves the Realism of LLM Agents Simulating Human Behavior During Disruptions
链接: https://arxiv.org/abs/2607.17437
作者: Chen Xia,Zexi Kuang,Yuqing Hu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model (LLM) agents offer a generative approach to simulating human behavior under conditions that may have few or no direct historical analogues, a common challenge in disaster and infrastructure-disruption planning. However, this generative capacity creates a validity problem: individually plausible agent reasoning may fail to reproduce empirical population behavior. We evaluate whether empirical grounding improves the statistical realism of LLM-agent simulations during disruptions. Specifically, we develop an empirically grounded LLM-agent framework that embeds demographic profiles from the American Community Survey, baseline routines from the American Time Use Survey, and urban spatial context into agent initialization, memory, decision prompts, and activity execution. An independent household survey conducted during the July 2024 Philadelphia heatwave is reserved as an external validation benchmark. Compared with an ungrounded LLM-agent baseline, the grounded model improved reconstruction of normal daily routines, increasing mean correlation with empirical activity profiles from 0.528 to 0.912 and reducing mean squared error from 0.066 to 0.008. Under heatwave conditions, the grounded model better reproduced survey-derived activity profiles, increasing mean correlation from 0.349 to 0.836 and reducing mean squared error from 0.098 to 0.012. The grounded model captured 46.4% of observed heatwave response amplitude, compared with 20.6% for the ungrounded baseline. These findings show that empirical grounding can make LLM agents more statistically credible simulators of population behavior while revealing remaining gaps in modeling human adaptation during disruptions.
[AI-79] Intermittent Control Is Not Diluted Control: A Switching Effect in Artificial Agency
链接: https://arxiv.org/abs/2607.17432
作者: Veronique Ziegler
类目: Artificial Intelligence (cs.AI)
备注: 20 pages, 10 figures, 4 tables
Abstract:Adaptive agents do not always regulate under the same timing conditions. Sometimes stabilization can begin before a disturbance has fully entered the internal state; at other times the agent can only recover after disruption has taken hold. A simple expectation is that an agent moving between these conditions should behave like a weighted average of the two fixed cases: the more time spent in reactive recovery, the greater the regulatory burden. This paper shows that expectation can fail. In a simulated adaptive agent with retained state history, at an operating point where sustained reactive control is more costly than sustained anticipatory control, intermittent access to anticipatory control reduces the mean regulatory burden below the value predicted by a fixed-mode mixture. The effect appears under both periodic and stochastic switching schedules: losing anticipatory access does not simply dilute its benefit, and restoring it intermittently can reorganize the later regulatory burden. High-statistics runs (N = 1000 matched replicates per schedule) resolve a negative nonlinear switching penalty across every tested schedule. The effect is small but consistent: about half a percent of the mean gain, with 63-68% of replicates falling below zero. Late-window diagnostics reveal no unresolved upward accumulation of regulatory burden. The result identifies a design-relevant timing principle. In history-dependent adaptive systems, the burden of remaining organized is not set only by how much time an agent spends in each mode; the order in which disturbance and recovery enter the state can change the subsequent burden. Intermittent anticipatory control may therefore act less like a partial failure of regulation than like a mechanism for reducing the long-term burden of recovery. Comments: 20 pages, 10 figures, 4 tables Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17432 [cs.AI] (or arXiv:2607.17432v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.17432 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-80] Kernelized Linear Attention: Breaking the Capacity Wall with Symmetric Cones
链接: https://arxiv.org/abs/2607.17419
作者: Ayoub Ghriss,Sourav Chakraborty
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Theory (cs.IT); Machine Learning (stat.ML)
备注: 39 pages, 4 figures, 10 tables. Code: this https URL
Abstract:Linear attention promises constant-time recurrent inference but degrades sharply on associative recall. We formulate attention recall as a spherical-packing problem and introduce Kernelized Linear Attention Activations (KATA), a framework whose feature maps are derived from first principles by certifying nonnegative attention weights through a self-dual homogeneous cone. Building on this observation, we show that rank-one positive semi-definite (PSD) features offer a favorable capacity–interference tradeoff. KATA recovers a parameter-free convex output gate and characterizes associative capacity through the Welch interference floor. For tolerances above this floor, KATA enlarges the state without adding parameters and admits spherical codes with exponentially many keys in the projection dimension. We implement KATA as fused Triton kernels at two operating points: a flash-attention-style forward up to \sim1.6\times FlashAttention-2 throughput, and an exact O(T) chunked-state form that reaches \sim11\times FlashAttention-2 forward throughput at 131 k tokens. An associative scan of the first-order feature lowers the inter-chunk recurrence depth to O(\log(T/C)) for chunk size C and averages \sim2.4\times the throughput of a matched sequential linear-attention baseline. On long-range MQAR and repeated-key overwrite, several KATA variants outperform Gated DeltaNet, with parameter counts and state sizes reported alongside accuracy. Induction preserves near-perfect recall, while kernel benchmarks show that the maps can be implemented efficiently. KATA retains 0.985 MQAR at a 16\times out-of-distribution length, approaching the softmax with roughly one quarter of the KV-cache entries. Experiments on 340M-parameter LLMs reveal a feature-dependent fluency trade-off and clarify how positional embeddings, delta rules, and decay gates interact with feature geometry.
[AI-81] CoEvoPR: Co-Evolving Placement Objectives with Routing Feedback via Large Language Models
链接: https://arxiv.org/abs/2607.17398
作者: Ruogu Chen,Weihua Xiao,Ramesh Karri,Jie Han
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
备注: 7 pages, 4 figures, 3 tables. Corresponding author: Weihua Xiao
Abstract:Analytical placers rely on differentiable objective functions to guide placement, typically combining intermediate surrogate metrics such as half-perimeter wirelength (HPWL) and cell-density penalties. However, these placement-stage surrogates remain misaligned with downstream routed and timing quality. Prior work reduces this gap with human-designed terms or learned black-box surrogates, but the former requires expert retuning and the latter is difficult to explain, debug, or deploy in analytical placement flows. CoEvoPR addresses these limitations with a large language model (LLM)-based framework that automatically evolves analytical placement objectives. At each generation, the prompt combines the restricted objective interface, baseline context, and archived prior candidates with routing-related feedback from placement, timing proxy, and routing tools. The LLM proposes readable differentiable objectives, which are embedded and validated in DREAMPlace, evaluated through a timing proxy and an actual router, and stored with their feedback to guide later generations. Across eight ChiP-Bench Nangate45 designs and three seeds, CoEvoPR reduces post-route routed wirelength and congestion by 16.9% and 36.7%, with gains of 0.70 ns in worst negative slack and a 912 ns reduction in total negative slack magnitude over native DREAMPlace. Across eight ICCAD 2015 Superblue designs, it reduces post-route routed wirelength and congestion by 5.4% and 23.2%. Code is available at this https URL.
[AI-82] A Phased Development Framework Enabling Islanded Operation of Sustainable AI Data Centers With Onsite Grid-Following and Grid-Forming Energy Architectures
链接: https://arxiv.org/abs/2607.17391
作者: Soham Ghosh,Nabil Mohammed,Mohammad Ashraf Hossain Sadi
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI)
备注: 26 pages, 13 figures, 9 tables
Abstract:As hyperscale and colocation AI data centers continue to expand, the electric grid is increasingly required to support large, concentrated loads, with individual facilities ranging from 500 MW to 2 GW. Current projections estimate that approximately 50 GW of AI data center capacity will require grid connectivity in the United States by 2030. While prior research has extensively examined the environmental and operational impacts of AI data centers, as well as their potential role as grid-interactive assets, limited attention has been given to the challenges associated with their scalable deployment through engineering, procurement, and construction (EPC) processes. This manuscript addresses this gap by proposing a phased development framework for AI data center expansion. The approach is designed to enable developers to meet aggressive time-to-market objectives while navigating multi-year constraints associated with interconnection approvals and lead times associated with the procurement of component equipment. A modular construction architecture is presented, along with a detailed analysis of integrated energy systems and the role of hybrid on-site generation in supporting incremental capacity growth. Electromagnetic transient simulations (EMT) are used to evaluate system performance, demonstrating that a combination of on-site natural gas generation and grid-forming energy storage can reliably support data center operations during early and intermediate deployment phases. The study further examines the transition to full grid interconnection, including the capability of the data center to operate in islanded mode during grid disturbances. Finally, the manuscript compares grid-forming control strategies for system reconnection and restoration under varying conditions.
[AI-83] Self-Modifying Lean Proof Agents with Verifier-Grounded Benchmark Coevolution
链接: https://arxiv.org/abs/2607.17352
作者: Yuqing Li,Zeguan Wu,Yu Gan,Junyu Liu
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Quantum Physics (quant-ph)
备注: 22 pages, 2 figures, 3 tables
Abstract:Designing effective Lean proof agents is a central challenge in formal mathematical reasoning. Beyond building stronger provers, recent work emphasizes the workflow around Lean: how an agent decomposes proof obligations, uses tools and compiler feedback, diagnoses failures, repairs proofs, and maintains structured proof context. Motivated by code-level self-evolving agents, we study whether such workflows can be evolved rather than hand-designed. We present a self-evolving Lean proof agent in which a small fixed, trusted runtime wraps a fully mutable workspace: the proof workflow, prompts, and tools. Unlike most self-evolving systems, which optimize against a fixed external benchmark, our system coevolves the agent and its benchmark. Between generations, the highest-scoring agent (the champion) revises the active task distribution through a mastery-throttled curriculum update that introduces harder proof obligations only after the current level is mastered, and a single-anchor recalibration re-runs the champion on the updated benchmark to keep scores comparable as difficulty rises. All evolution stays inside a Lean-grounded verification loop: however the agent rewrites itself, a success counts only when its behavior yields Lean-verified proofs under a trusted snapshot, and each attempt must emit a machine-readable, Lean-grounded proof context whose representation may evolve but whose groundedness is enforced. We run the coevolving trajectory and a fixed-benchmark baseline for 15 active generations and compare them on a held-out miniF2F test split. The best coevolving agent reaches a 45.1% held-out solve rate, versus 12.7% for the seed and 32.0% for the best fixed-benchmark agent, showing that verifier-grounded self-evolution can improve Lean proof workflows under a coevolving benchmark.
[AI-84] DeeperRadar: End-to-End MIMO Radar Design and Multi-Modal Fusion for Autonomous Vehicle Perception IROS2026
链接: https://arxiv.org/abs/2607.17351
作者: Eli Goldenshluger,Barak Pinkovich,Chaim Baskin
类目: Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注: Accepted for publication at the IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS 2026)
Abstract:DeeperRadar is a radar-centric, sensor-stack-conditioned framework that co-designs radar sensing and multi-modal 3D detection for autonomous mobility by learning a sparse acquisition pattern end-to-end with the fusion model. A learnable MIMO design module is trained end-to-end within a fusion network that operates directly on raw radar ADC data together with camera images and LiDAR point clouds. During training, the design module is supervised by the other sensors, enabling the system to learn both which receiver antennas to activate and the effective number of them. At deployment, the design module is removed and replaced by the learned sparse subsampling mask, leaving the downstream model architecture unchanged. Evaluated on the RADIal dataset, DeeperRadar discovers sparse, task-aware radar configurations that match or exceed full-array baselines while using fewer receivers, potentially reducing radar cost and integration complexity. These results show that learned optimal MIMO radar design depends on the fusion stack and the downstream perception task.
[AI-85] APAS: Throughput-adaptive Perception for Autonomous Systems
链接: https://arxiv.org/abs/2607.17317
作者: Aman Vyas,Vasista Kodumagulla,Zain Taufique,Pasi Liljeberg,Anil Kanduri
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 15 Pages, 20 Figures, Accepted at ACM/IEEE International Conference on Codesign of Embedded Systems at Embedded Systems Week (ESWEEK-CODES), 2026
Abstract:Autonomous systems rely on a perception module to navigate through dynamic environments. In real-world scenarios, the perception module’s throughput requirements vary at runtime due to changes in scene complexity. However, existing perception strategies assume a fixed FPS and static model-to-cluster mapping, resulting in either over/under provision of throughput requirements or unnecessary energy consumption across diverse scenes. Addressing this challenge requires tightly coupled \textitscene complexity awareness to estimate an appropriate FPS target and \textitdynamic model-to-cluster mapping to deliver the required throughput at minimum energy. We propose a throughput-adaptive perception strategy for mobile/edge platforms, enabling intelligent runtime resource allocation based on varying FPS targets. We use Reinforcement Learning (RL) with RRM (Reward Reasoning Model) and a GRU (Gated Recurrent Unit) agent to orchestrate perception tasks across heterogeneous mobile/edge platforms. We evaluate TAPAS on Jetson Orin NX across KITTI and unseen nuScenes. On the \textitKITTI dataset’s test sequences, TAPAS achieves 93-100% throughput met rate while saving energy by 76%. On the unseen \textitnuScenes dataset, TAPAS maintains 97% throughput met rate with 64% lower energy compared to \textitSOTA approaches, proving its robustness.
[AI-86] Learning-Driven Adaptive Audit Scheduling: A Sequential Decision Approach to Off-Chain Data Integrity
链接: https://arxiv.org/abs/2607.17305
作者: Changting Lin,Fan Li,Weihang Yu,Keyang He,Mingyuan Yan,Yourong Chen,Meng Han
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注:
Abstract:We model cryptographic auditing of off-chain data as a Constrained MDP (CMDP) under partial observability: the storage node’s hidden type and corruption state make the problem a POMDP, while a miss-rate ceiling rho imposes an explicit security constraint. We propose DRQN-CMDP, a Deep Recurrent Q-Network whose GRU layer maintains a belief over the latent node type, paired with Lagrangian dual ascent that adapts the miss-rate penalty lambda automatically. A pairing-free homomorphic-MAC primitive supplies O(1) on-chain verification cost. Across 13 methods–four DQN variants, PPO, A2C, PPO-Lagrangian, a stateful Bayesian heuristic, three fixed-rule baselines, and an oracle-informed heuristic–DRQN-CMDP achieves a favourable balance: 83% lower gas than fixed high-frequency auditing, single-digit miss rate (7.5%), and moderate detection latency–a combination no other method matches across all three objectives simultaneously.
[AI-87] Lookahead Branching for Neural Network Verification IJCAI2026
链接: https://arxiv.org/abs/2607.17290
作者: Liam Davis,Duo Zhou,Huan Zhang,Guy Katz,Clark Barrett,Haoze Wu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: Accepted to IJCAI 2026. Lookahead branching is part of the Marabou and α-β-CROWN verifiers
Abstract:In this work, we investigate the effect of lookahead branching strategies in neural network verification. We present a general recipe to integrate lookahead into any branch-and-bound verifier and demonstrate how one of the current state-of-the-art branching heuristics, FSB, can be viewed as a special instantiation of the lookahead branching strategy. We also describe how, in addition to improving the quality of branching decisions, lookahead can generate additional lemmas that accelerate verification. We instantiate the method in two representative branch-and-bound-based verifiers (Marabou and \alpha - \beta -CROWN), and demonstrate that lookahead leads to consistent speedups in verification time and up to 57% more solved instances. Code is available at this https URL.
[AI-88] SAGA: Synthetic Agent ic Graph Architecture for Temporal Benchmark Generation
链接: https://arxiv.org/abs/2607.17288
作者: Jiacheng Ding,Xiaofei Zhang
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:
Abstract:High quality temporal graph benchmarks with rich semantics and ground-truth anomaly labels are essential for training graph neural networks, yet remain scarce due to privacy constraints and annotation costs. We present SAGA (Synthetic Agentic Graph Architecture), a system for generating large-scale, semantically rich temporal graphs via a four-phase pipeline. Our Skeleton-First, Semantics-Second architecture decouples structure from semantics: (S) an O(1)-per-edge skeleton generator produces power-law graphs; (A) a dispatcher partitions causally ordered time blocks for parallel execution; (G) LLM agents inject domain semantics using RAG-based rule bases across four domains; and (A) a state alignment engine resolves conflicts via temporal replay, yielding anomaly labels as natural byproducts. Unlike structural generators (e.g., LDBC SNB, Kronecker/R-MAT) or purely LLM-based approaches, SAGA achieves structural realism, semantic richness, and automatic anomaly labeling in a unified framework. On a single H100 GPU with vLLM batching, SAGA generates 500,000 temporal edges with controlled anomalies in under 90 minutes, scaling to 100,000 nodes while maintaining clustering coefficients above 0.99. The system supports real-time pipeline visualization, interactive multi-domain tuning (Finance/AML, Network/IDS, Cyber/APT, Transportation), and a CLI for large-scale GPU-based experiments.
[AI-89] Lossless but Not Free: An Empirical Anatomy of Speculative Decoding on Consumer Hardware
链接: https://arxiv.org/abs/2607.17283
作者: Param Chordiya
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 15 pages, 5 figures, 4 tables. Code, benchmark harness, and all result artifacts: this https URL
Abstract:Single-stream autoregressive decoding of large language models is bound by memory bandwidth: each generated token requires one full forward pass through the target model, and successive passes cannot be parallelized. Speculative decoding restructures this computation: a small draft model proposes K tokens autoregressively, the target model scores all of them in one batched pass, and a rejection-sampling rule provably preserves the target model’s output distribution. We present a from-scratch, device-agnostic (CUDA/MPS/CPU) implementation and an empirical study across five draft/target backend configurations on a consumer Apple-silicon laptop. Distribution equivalence is verified at three levels, culminating in a two-sample test over roughly 9,200 real-model tokens per method ( \chi^2 = 162.5 , dof = 200 , p = 0.976 ) and exact greedy-sequence agreement. The best configuration reaches a measured 1.61\times wall-clock speedup at K=6 , on an acceptance profile declining from 69.7% at K=1 to 37.8% at the optimum, while three of five configurations decelerate, either because the draft fails to out-speed a small target or because the quantized Metal backend executes “parallel” verification serially, an effect we isolate and quantify. The failures are as instructive as the successes: speculative decoding pays off only when verification is genuinely batch-parallel and the draft/target latency gap is real.
[AI-90] AIGB-R1: Self-Evolving Generative Auto-Bidding via Hierarchical Planner-Executor Optimization
链接: https://arxiv.org/abs/2607.17281
作者: Yuejia Dou,Hesong Wang,Xinyu Zhang,Tianyu Wang,Zhilin Zhang,Chuan Yu,Jian Xu,Bo Zheng,Qi Qi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Auto-bidding plays an essential role in online advertising, automatically adjusting bids for advertisers to optimize their commercial goals. The emerging AI-Generated Bidding (AIGB) paradigm widely adopts generative modeling to optimize bidding strategies, yet suffers from the limited mode coverage of offline datasets and inadequate task-state understanding, hindering effective exploration of optimal strategies. Large Language Models (LLMs), with prior world knowledge and reasoning capabilities, offer a promising approach to overcome these limitations. However, directly applying LLMs to auto-bidding tasks faces inherent challenges in limited numerical precision, hallucinations, and inference latency. To address these limitations, we propose AIGB-R1, a hierarchical self-evolving auto-bidding framework aiming to enhance AI-Generated Bidding via LLMs’ Reasoning capabilities, comprising a high-level Planner module for macro-level strategy planning and a low-level Executor module for fine-grained decision-making. Building upon this, we design an experience-driven self-evolving loop, enabling autonomous strategy exploration and optimization from accumulated experience. We adopt a two-stage pipeline of offline pre-training and post-training alignment, and build an interactive bidding simulation environment for strategy rollout. Furthermore, we propose Decoupled Group Relative Policy Optimization (D-GRPO) to achieve end-to-end optimization via advantage decoupling. Experimental results on a large-scale public dataset demonstrate the effectiveness of AIGB-R1.
[AI-91] Between Safe Boundaries: Exploiting Temporal Consistency for Jailbreaking Text-To-Video Generation Models
链接: https://arxiv.org/abs/2607.17279
作者: Xingkai Peng,Jun Jiang,Jiayang Liu,Kejiang Chen,Weiming Zhang
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Multimedia (cs.MM)
备注:
Abstract:Recently, text-to-video (T2V) models have been widely deployed, sparking growing concerns over their robustness against jailbreak attacks. Existing jailbreak methods, mostly adapted from text-to-image attacks, suffer notable drawbacks when applied to T2V systems. They fail to fully leverage temporal consistency, an inherent characteristic of video generation. Besides, these methods demand heavy video query optimization, which is infeasible in practical black-box scenarios. Their adversarial prompt search is also driven by heuristic local signals, lacking principled structured exploration strategies. To tackle these limitations, we propose BSB, a structured, query-efficient jailbreak framework for T2V models. BSB harnesses temporal consistency by encoding harmful intent as the transition between two individually harmless boundary states. Under this paradigm, the attack targets boundary-state pairs whose interpolation tends to produce unsafe intermediate frames during video generation. Directly evaluating all candidate pairs within the video space incurs prohibitive computation cost. Instead, BSB conducts Monte Carlo Tree Search (MCTS) in a cheaper textual proxy space and regularly calibrates search outcomes with sparse video-level evaluations. We conduct comprehensive experiments on mainstream commercial T2V models including Veo 3.1, Sora 2, Seedance and Kling v1. Results show BSB surpasses all existing jailbreak baselines, delivering an average 18.6% relative gain in attack success rate over the strongest competitor across evaluated models. Our findings identify temporal consistency as an understudied yet vital attack surface for T2V models and verify that structured search facilitates effective vulnerability discovery under constrained query budgets.
[AI-92] An Explicit World Model Based on Data-First Ontology: DaoQL Multimodal Storag e Validation and Counterfactual Reasoning Evaluation
链接: https://arxiv.org/abs/2607.17269
作者: Zhanbo Li,Shifeng Wu,Xiangjin Meng,Wenjie Cai
类目: Artificial Intelligence (cs.AI); Databases (cs.DB)
备注: 20 pages, 2 figures. Code: this https URL
Abstract:Large language models encode world models implicitly in neural weights, which exposes four structural risks in high-precision domains such as medicine and finance: hallucination, frozen knowledge, poor explainability, and poor modifiability. This paper proposes data-first ontology: LLMs are treated as reasoning and language engines, while deterministic knowledge is moved into an explicit multimodal database, DaoQL. We formalize an explicit world model and show that, under rule independence, deterministic evaluation, and fixed conflict resolution, explicit models provide a sufficient condition for composable counterfactual decomposability; implicit models lack atomic read/delta semantics and therefore provide no comparable architectural guarantee. The implemented system focuses on DaoQL’s verified storage layer and explicit Eval path, integrating graph, column, vector, and full-text engines within one process. KVCache graph nodes, expert hot updates, and the DaoQL-Agent runtime remain future work. On an embedded same-machine setup, DaoQL reports graph BFS at 1.20 ms, HNSW at 83.1 us, and a Fluent hybrid query at 105.8 us; these results indicate engineering potential but must be interpreted with deployment-shape differences from client-server systems. Exploratory measurements on LDBC SNB SF1 and ANN-Benchmarks further show 34/34 query coverage with interactive-class queries mostly in the sub-millisecond to millisecond range, but only 1.8 QPS overall due to long-tail BI/IC queries; ANN-Benchmarks reaches Recall@10 = 99% at thousand-level QPS after a bridge-edge protection fix. In a five-domain counterfactual experiment (n = 1250), DaoQL+GPT-4o achieves 94% composable counterfactual decomposability, 49 percentage points above GPT-4o alone. The paper explicitly separates provable structure, preliminary empirical evidence, and architectural roadmap claims.
[AI-93] Coordinated Disentanglement with Iterative Mode Discovery Under Hidden Correlations ICML2026
链接: https://arxiv.org/abs/2607.17264
作者: Rong Hu,Ling Chen
类目: Artificial Intelligence (cs.AI)
备注: Published at ICML 2026
Abstract:Disentangled representation learning is a powerful paradigm for robust attribute prediction. While recent methods address attribute correlations, hidden correlations remain underexplored, where data under the value of a certain attribute exhibit underlying modes correlated with other attributes. To preserve mode information and achieve disentanglement, we jointly discover modes and enforce mode-based conditional independence. Yet, the interdependency between these two modules may lead to error amplification under naive iterations. We propose Coordinated Disentanglement with Iterative mode Discovery (CoDID), an end-to-end framework featuring a dynamic architecture that adapts to evolving number of modes, and a coordination mechanism that mitigates error amplification via meta-optimization. Empirical results demonstrate the state-of-the-art performance on diverse tasks.
[AI-94] Asynchronous Multimodal Diffusion Policy Composition via Latency-Aware Guidance Fusion
链接: https://arxiv.org/abs/2607.17257
作者: Zihao He,Hongjie Fang,Shirun Tang,Cewu Lu,Haoshu Fang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Diffusion policies have shown strong potential for robotic imitation learning, and recent extensions incorporate additional modalities to improve manipulation performance. However, these modalities often differ not only in information content but also in sensing rates and inference latencies. Existing multimodal diffusion policies typically rely on synchronous fusion or manually designed multi-frequency architectures, which either slow down high-frequency feedback or limit extensibility to new modality combinations. We propose LAG-Fusion, a latency-aware guidance fusion framework for asynchronous multimodal diffusion policy composition. LAG-Fusion allows modality-specific policies to operate at their native inference rates and contribute denoising guidance whenever available. To make asynchronous composition consistent, we derive a reference-frame rebasing rule for diffusion variables under relative action representations, enabling delayed guidance to be aligned before fusion. We instantiate LAG-Fusion in contact-rich manipulation by composing a low-frequency vision policy with a high-frequency force policy. Experiments under heterogeneous modality latencies show that LAG-Fusion improves policy responsiveness and task performance over synchronous fusion and specially designed force-aware baselines.
[AI-95] Distilled Reinforcement Learning for LLM Post-training
链接: https://arxiv.org/abs/2607.17247
作者: Chen Wang,Zhaochun Li,Jionghao Bai,Yining Zhang,Hexuan Deng,Ge Lan,Yue Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language model (LLM) post-training is essential for improving reasoning, adaptation, and alignment. Existing methods mainly follow two paradigms: reinforcement learning (RL) and on-policy distillation (OPD). However, RL relies on coarse-grained outcome supervision, resulting in difficult credit assignment and limited capability to acquire new knowledge. OPD, meanwhile, unconditionally matches teacher logits through KL divergence, which creates a dilemma: similar teachers provide little new knowledge, while substantially different teachers often yield ineffective guidance, largely restricting OPD to within-family distillation. We propose Distilled Reinforcement Learning (Distilled RL), which integrates teacher supervision into the RL objective to provide fine-grained guidance, selectively transfer new knowledge and avoid unconditional imitation. Distilled RL contains three components: reverse importance sampling with clipping, negative sample reset, and sequence-level geometric normalization. Through a concise and interpretable case study, we demonstrate that Distilled RL can effectively transfer previously unavailable knowledge from a teacher model to a student model. Extensive experiments across both within-family and cross-family distillation settings show that Distilled RL substantially outperforms standard RL and OPD in terms of both pass@1 and pass@k. Our code is available at this https URL.
[AI-96] LenGuard-GPC: Length Guarding with Guided-Prompt Consistency for Spatial Reasoning Reinforce Learning
链接: https://arxiv.org/abs/2607.17243
作者: Xingjian Tao,Yiwei Wang,Yujun Cai,Jing Tang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-view spatial reasoning requires vision-language models to compare visual evidence across images, align object correspondences, and infer spatial relations over long visual contexts, a setting where chain-of-thought reasoning tends to grow verbose without becoming more accurate. Reinforcement learning with verifiable rewards is a natural fit for this task, but standard GRPO reward relies on sparse outcome-level feedback and gives no signal about where a reasoning trajectory goes wrong, nor any control over its length. We propose LenGuard-GPC, a dense reward framework that addresses both problems together. For each sampled trajectory, it compares the token-wise predictive distributions under a standard prompt and a guided prompt, and uses the resulting token-sum KL divergence as a dense reward signal. Since this KL penalty accumulates over tokens and would otherwise reward shorter responses regardless of their quality, we introduce a staged length bonus that keeps reasoning length within a controlled range without simply encouraging brevity. On six multi-view spatial reasoning benchmarks, LenGuard-GPC improves accuracy over vanilla GRPO while reducing average response length.
[AI-97] A Large-Scale Measurement of AI Bill of Materials Completeness in Hugging Face Models
链接: https://arxiv.org/abs/2607.17242
作者: Md Erfan,Ahmed Ryan,Md Rayhanur Rahman
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 12 pages
Abstract:Pretrained machine learning (ML) models help developers build ML-intensive software systems without training models from scratch. However, model repositories often provide incomplete machine-readable documentation about model provenance, licenses, datasets, limitations, and external references, creating transparency and governance gaps across the AI supply chain. Artificial Intelligence Bills of Materials (AIBOMs) address these gaps by documenting AI artifacts, including models, metadata, licenses, datasets, model-card information, and external references. Taking public Hugging Face (HF) model repositories as a case study, this paper empirically investigates AIBOM completeness, defined as the extent to which repositories provide AIBOM-relevant information for machine-readable AI supply-chain documentation. We examine approximately 97.5K AIBOM artifacts to assess the extent to which generated AIBOMs: (i) contain required structural and metadata fields, (ii) represent model identity, license, and external-reference information, (iii) capture model-card documentation such as datasets, limitations, safety-risk assessment, and environmental information, and (iv) vary in documentation coverage across repository and artifact characteristics such as task, license availability, dataset declaration, model family, and paper reference. Results indicate that generated AIBOMs provide complete coverage of required AIBOM structure but limited AI-specific documentation completeness. Required fields are fully represented, but model-card, metadata, responsible-use, environmental, limitation, and meaningful-description fields remain weakly represented or missing across generated artifacts. Our findings motivate improved model-card practices, repository-level traceability, and automated AIBOM validation to advance the generation and adoption of more complete AIBOMs.
[AI-98] Constrained Path Reasoning : Measuring When Committed Stages Earn Their Cost
链接: https://arxiv.org/abs/2607.17240
作者: Honglin Li(ShanghaiTech University)
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:When does a committed intermediate stage in an LLM reasoning pipeline earn its cost? Constrained Path Reasoning (CPR) pairs a source-aware path hypothesis with stage-level accounting. Search generates provisional states; trusted or validated invariants can constrain hard, while other proposals remain soft and revisable. CPR predicts that task-compatible commitments can factor transitions, concentrate candidate mass, induce regularity, and expose feedback when their gains exceed propagated error and execution cost. The formalism covers discrete commitments and continuous flows and measures effective branching, endpoint concentration, and cost per usable output. Across 1,180 generated QCQPs and 40 engineered degenerate polynomial instances (2,140 endpoints), residual triage recovers 63.0% of repair-all’s additional feasible yield with 17.7% of its attempts. Fixed-LLM accounting (270 unique calls shared across nested arms) finds usable yield of 41.1% direct, 90.0% after formalization and deterministic execution, 20.0% after one-shot convexification, and 21.1% for the full path. In 120 paired-condition calls, a two-action rollback rule reaches 90% usable yield versus 36.7% for the feedback-conditioned selector. Two endpoint probes separate source from validation: a 72-output cross-trajectory transplant reduces entropy and acceptable mass; a 24-output same-call self-proposal pilot gives unchanged two-repeat collision entropy, 25.0% versus 8.3% usable yield, and 1/8 deterministically confirmed endpoint checks. Model-generated states supply hypotheses; trusted execution earns constraint strength.
[AI-99] A RFID Based Campus Wide Payment System
链接: https://arxiv.org/abs/2607.17233
作者: Miraj Uddin Chowdhury,MD Khairul Islam Prime
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 5 Pages, 5 figures
Abstract:This work titled “RFID Based Campuswide Payment System” introduces an innovative cashless payment solution for educational institutions. It uses RFID cards and a Raspberry Pi to enable hassle free payments for various campus services, such as cafeteria purchases, tuition fees, and library fines. A centralized database ensures real-time updates on transactions and account balances, accessible through a simple and user-friendly web interface. This work is involved in designing a secure system with object-oriented principles, setting up databases, and integrating hardware like RFID readers with a Raspberry Pi. The systems are proved to be a cost-effective and efficient alternative to traditional payment methods, enhancing convenience and security for students and administrators. The study also explored similar RFID applications, like smart parking and attendance systems, to identify challenges and improvements. Looking ahead, it envisions features like wearable RFID devices, voice-activated payments, and blockchain integration to boost security and usability. Results show that this system simplifies campus payments and has the potential for broader adoption in similar environments.
[AI-100] Specifying the Delegated-Autonomy Boundary: Requirements Engineering for Agent ic AI
链接: https://arxiv.org/abs/2607.17225
作者: Chetan Arora,Andreas Vogelsang,Abbi Sharma
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Agentic AI systems do not just predict or recommend; they plan, maintain state, and act in external environments with varying degrees of autonomy. This changes the requirements engineering problem in a specific and under-addressed way: it introduces what we call the delegated-autonomy boundary – the set of decisions about what may be delegated to the system, under what graduated authority, with what oversight, and how control is returned. Current practices bury these decisions inside prompts, tool schemas, and runtime policies, even though they are requirements-level commitments. This paper proposes two complementary artifacts. First, an Agency Justification Record (AJR) helps teams decide when an agent is warranted over simpler alternatives. Second, an Agentic Delegation Policy (ADP) captures what must be specified for safe and effective development: purpose, authority, information, coordination, assurance, and evolution. Crucially, authority in the ADP is modelled as graduated, i.e., a tiered structure. We illustrate the framework with two contrasting examples: a safety-critical hospital discharge coordination agent and an automated code review agent.
[AI-101] A Systematic Evaluation of Trajectory Data Curation for LoRA Fine-Tuning of Code Agents
链接: https://arxiv.org/abs/2607.17205
作者: Yunze Han
类目: Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 12 pages, 2 figures. Author’s accepted manuscript of a paper published in the ICIC 2026 proceedings (oral presentation)
Abstract:Supervised fine-tuning (SFT) of open-weight LLMs on expert agent trajectories has emerged as a prominent approach to building capable code agents without reliance on proprietary models. A central yet underexplored question is how trajectory quality and quantity jointly shape model performance. We present a systematic empirical study of trajectory data filtering for LoRA fine-tuning of Qwen2.5-Coder-7B-Instruct on the SWE-trajectory dataset (67,074 trajectories, of which 32,161 are resolved). We propose a two-axis quality scoring framework – Efficiency and Style – and evaluate it through 16 controlled experiments spanning strategy, scale, and ablation analyses. Since 7B-scale models attain near-zero SWE-bench resolve rates, we adopt cross-entropy (CE) loss on held-out trajectories as the primary metric, validated via first-action generation: CE loss and ROUGE-L are perfectly rank-correlated (Spearman \rho = -1.00), with limited-sample evidence supporting but not conclusively establishing this proxy. Our results reveal a scale-dependent quality-quantity trade-off: at small scales, doubling the dataset (500 to 1,000) yields ~12.7% CE-loss reduction whereas the TopQ-Random gap stays 1% (Mann-Whitney p 0.10); at 2,000 trajectories this same gap widens to 3.6% (p = 0.016). Ablation further identifies error-retry rate as the dominant sub-dimension, performing comparably to the full composite ( \Delta 0.2%). Together, these findings establish trajectory-level quality scoring as a viable but scale-sensitive lever for code-agent SFT and offer a proxy-validated evaluation protocol for the regime where end-to-end resolve rate is statistically infeasible.
[AI-102] oward Anthropomorphic Dialogue: A Closed-Loop Framework for Human-Like Chat Generation Evaluation and Preference Alignment
链接: https://arxiv.org/abs/2607.17191
作者: Wentao Liu,Siyu Song,Xi Chen,Youjia Li,Xiaokun Wang,Min Ji,Ji Wang
类目: Artificial Intelligence (cs.AI)
备注: 17 pages, 2 figures, 11 tables
Abstract:Human-like private chat requires more than fluent response generation: a system must preserve persona, relationship, memory, bounded knowledge, medium-specific timing, and a coherent multi-turn arc. We present AnthroDial, a closed-loop framework that formulates anthropomorphic dialogue as a joint problem of system architecture, executable evaluation, and diagnostic alignment. It combines (1) a role-conditioned scheduled dialogue runtime with persona and scenario cards, long-term memory, virtual time, and single-draft message decisions; (2) an executable benchmark with an L0 validity gate, five per-turn dimensions, and five dialogue-level dimensions; and (3) a post-training pipeline that filters 16,436 scheduled-decision examples for SFT and applies GRPO with a cognitive-diagnostic, ZPD-aware reward. The reward maintains Kalman-filtered capability estimates for each behavioral dimension, upweights dimensions with larger capability deficits, and uses rollout scores as task-level ZPD matches to focus optimization on learnable weak skills. On a benchmark with 55 personas, 50 scenarios, 50 persona-scenario bindings, and 100 role-conditioned cases per model, we evaluate 16 systems spanning frontier baselines, open models, thinking/no-think variants, and SFT/RL ablations. The strongest non-trained baseline reaches 32.00% strict ACC, while Qwen3.6-27B-SFT+RL reaches 39.00% strict ACC and a 98.5 overall score. In the 9B no-think setting, SFT and RL improve strict ACC from 0.00% to 13.00% and 18.37%. These results show that anthropomorphic dialogue benefits when generation, evaluation, and reward shaping share the same behavioral dimensions.
[AI-103] Is Your Model Thinking or Just Stagnating? PUMA: Diagnosing Reasoning Pathology via Phase-Momentum Alignment
链接: https://arxiv.org/abs/2607.17188
作者: Cheng Yan,Guangyang Ye,Wuyang Zhang,Fan Xu,Zhijun Fan,Xiang Xia,Yanyong Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Test-time scaling empowers Large Reasoning Models (LRMs) to tackle complex tasks via extensive Chain-of-Thought (CoT). However, this often induces the “overthinking” paradox, where redundant reasoning increases computational overhead without guaranteeing accuracy. Existing test-time efficiency optimization methods primarily fall into two categories: information-theoretic approaches, which are prone to “deceptive convergence” where low uncertainty masks hallucinations, and latent representation analyses, which are often post-hoc, lacking the real-time sensitivity for dynamic reasoning. To bridge this gap, we first posit the Phase-Momentum Alignment Hypothesis, asserting that reasoning correctness hinges on the temporal synchronization between geometric momentum and uncertainty resolution. We then theoretically formulate the Cognitive-Energy Model to characterize these dynamics through two orthogonal dimensions: Geometric Cognitive Effort, quantified by latent velocity and tortuosity, and Entropic Cognitive Uncertainty. To operationalize this, we introduce PUMA (Phase-Uncertainty Momentum Alignment), a training-free framework employing a tiered diagnostic architecture. By coupling lightweight phase monitoring with event-triggered geometric analysis, PUMA effectively distinguishes active exploration from passive stagnation, enabling precise interventions through adaptive truncation or corrective measures. Extensive experiments on LRMs spanning 1.5B to 32B demonstrate that PUMA consistently outperforms state-of-the-art baselines across diverse benchmarks, achieving a superior accuracy-efficiency trade-off and robust cross-domain generalization.
[AI-104] alaria: Session-Aware Serverless Serving of Hundred-Billion-Parameter LLM s
链接: https://arxiv.org/abs/2607.17181
作者: Utopia Meng,Unicornt Zhao,Derek Li,Goalen Gao,Frank Du
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI)
备注: 19 pages, 15 figures
Abstract:Serverless multi-model LLM systems multiplex popularity-skewed model catalogs over shared GPU pools, yet typically schedule each request independently. Tool-using agents break this abstraction: a session repeatedly calls an LLM across short tool gaps, carries a long reusable KV prefix, and is judged by session completion time (SCT). Load-only routing can separate a continuation from both its model and KV state, while round-based model multiplexing can delay even a correctly placed continuation until the target model’s next slot. Both failures are especially costly for hundred-billion-parameter models: their weights constrain residency, while long-context KV is expensive to reconstruct or move. We present Talaria, a session-aware serverless multi-model serving system that makes session continuity a joint placement-and-admission decision. Its router ranks placements by model residency, KV locality, and instance pressure, while soft reservations account for likely returns in the last serving instance’s admission budget. Session-prefill (SP) admits budget-eligible continuations before the active model slot closes. An instance-local substrate keeps HBM addresses stable, preserves host-restorable KV, and stages weights across model switches. On a single TP=8 server, we replay 30 SWE-Bench model-sessions (960 calls) over three models, each with more than 100B total parameters. Against an otherwise identical round scheduler with SP, host-KV restoration, and D2D staging disabled, Talaria cuts p50 SCT from 1000 s to 189 s and p95 from 2296 s to 867 s, speedups of 5.3x and 2.6x. Comments: 19 pages, 15 figures Subjects: Distributed, Parallel, and Cluster Computing (cs.DC); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.17181 [cs.DC] (or arXiv:2607.17181v1 [cs.DC] for this version) https://doi.org/10.48550/arXiv.2607.17181 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-105] DADIR: Density-Aware Data-level Imbalanced Regression Framework
链接: https://arxiv.org/abs/2607.17178
作者: Shermin Shahbazi,Hossein Mohammadi,Mohsen Afsharchi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Imbalanced learning addresses predictive modeling problems with underrepresented regions of the data distribution. Although widely studied in classification, imbalanced regression remains challenging because of continuous target variables and heterogeneous density distributions. Existing data-level methods often rely on fixed target partitioning or synthetic sample generation without jointly considering density variations and local feature-space structure. We propose DADIR, a Density-Aware Data-level Imbalanced Regression framework that exploits density information throughout the balancing process. DADIR comprises three components: (1) Density-Aware Adaptive Partitioning (DAAP), which recursively partitions the target space according to density variations; (2) a Density-Regularized Conditional Variational Autoencoder (DR-CVAE), which preserves sparse-region representations while learning latent features; and (3) latent-space data balancing, which combines feature-level clustering with oversampling to generate structurally consistent synthetic samples. Together, these components identify minority regions more effectively, preserve sparse-region information, and generate realistic synthetic data. The resulting balanced dataset can be used directly with existing regression models without modifying their architecture or learning objective. Experiments on diverse imbalanced regression datasets demonstrate consistent improvements in predictive performance, particularly in underrepresented regions, while also improving overall accuracy.
[AI-106] A Diagnostic Framework for AI Agent Behavior
链接: https://arxiv.org/abs/2607.17149
作者: Xichen Zhang,Yingjie Zhang,Tianshu Sun
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注:
Abstract:AI agents increasingly act within the same clinical, political, scientific, and social systems that behavioral scientists study. Evaluating these systems requires source-level diagnosis: the same behavioral pattern may arise from an agent representational substrate or from the roles, objectives, interaction structures, and governance rules that shape its expression. This Perspective proposes a diagnostic framework for AI agent behavior: layer attribution. The foundational computational layer defines what behaviors are possible through architecture, memory, perception, attention, and representation. The behavioral modulation layer shapes how those capacities are expressed through identity, resources, objectives, social interaction, institutional constraints, and governance. The framework clarifies three consequences: surrogate validity is a model-task-layer relation, human-AI divergence provides diagnostic evidence, and governance requires source attribution before intervention. Treating AI agents as behavioral actors therefore requires evaluation methods that determine where behavior originates before deciding how to explain, validate, or govern it.
[AI-107] Evidence Interfaces Shape How Retrieval-Augmented Readers Use Support
链接: https://arxiv.org/abs/2607.17108
作者: Junchi Liao,Jiawen Deng,Fuji Ren
类目: Artificial Intelligence (cs.AI)
备注: 18 pages; includes references and appendices
Abstract:In multi-hop RAG evaluation, a top-k answer score can hide two different failures: the retrieval window may drop part of the support chain, or it may contain support in a form the adapted reader does not use well. We call this reader-facing form of retrieved evidence an evidence interface. Using three support-annotated multi-hop QA benchmarks, we compare matched adapted readers trained with raw context, retrieval windows, and gold-support diagnostic renderings. These comparisons distinguish support-availability failures from remaining reader-interface effects. Top-k windows become interpretable only after checking whether the complete annotated support chain survives: when it does, short ranked windows can match or improve over raw context; when it does not, missing support explains much of the loss. Gold support-first improves matched readers; on 2Wiki and MuSiQue, a support-supervised ranker raises coverage and recovers raw-context quality at lower prompt cost, while retaining gold headroom. Support-removal checks further show that the gains rely on exposed evidence, not only answer priors. On support-annotated evaluations, top-k answer scores should therefore be reported together with complete-support coverage.
[AI-108] Fourier Geometric Wind Power Forecasting with Numerical Weather Prediction
链接: https://arxiv.org/abs/2607.17095
作者: Shiyuan Piao,Fan Zehui,Yang Liu,Hong Cheng,Juepeng Zheng,Jie Zhou,Fugee Tsung
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate short-term wind power forecasting is essential for grid stability and operational planning, yet remains challenging due to the complex interactions between atmospheric conditions and turbine dynamics. However, existing methods fail to effectively incorporate weather forecasting with wind turbine data (i.e., SCADA), leading to suboptimal solutions. To address this, we introduce a multimodal framework that integrates historical point-based SCADA data with grid-based Numerical Weather Prediction (NWP) forecasts, which is challenging due to heterogeneous input and the complex physical wind-turbine interactions. Our approach first explicitly decomposes inputs into scalar and vector features to better capture both site-specific and geometric dependencies and then incorporates a geometric encoder to extract rotation-invariant features from wind vectors. We further leverages a Fourier Neural Operator (FNO) architecture, which performs global convolutions in the frequency domain to efficiently model long-range spatiotemporal relationships. Extensive experiments on three real-world wind farms, with weather forecasting data, demonstrate that our model consistently outperforms state-of-the-art baselines, highlighting the effectiveness of its physically-informed design. The core implementation of our method is publicly available at: this https URL.
[AI-109] hAME: 3D Memory-Enabled Heterogeneous Accelerator for LLM Mixture of Experts
链接: https://arxiv.org/abs/2607.17074
作者: Pratyush Dhingra,Pramit Kumar Pal,Janardhan Rao Doppa,Partha Pratim Pande
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Distributed, Parallel, and Cluster Computing (cs.DC)
备注: Accepted for Publication in IEEE/ACM Embedded Systems Week (ESWEEK-26)
Abstract:Mixture of Experts (MoE) architectures have emerged as a dominant paradigm for scaling Large Language Models (LLMs). However, MoE inference on conventional hardware is constrained by three fundamental bottlenecks. These encompass the massive memory bandwidth required to fetch non-contiguous expert weights, the non-deterministic scatter-gather traffic generated by input-dependent token routing, and the tail-latency dependency imposed by synchronous expert output aggregation. To address these challenges, we propose ThAME, a three-dimensional (3D) heterogeneous multi-chiplet architecture for MoE inference. ThAME employs Ferroelectric Field-Effect Transistor (FeFET)-based non-volatile and DRAM-based volatile memory chiplets with a co-designed compute mapping strategy that aligns the distinct computational profiles of attention mechanisms and expert routing. Furthermore, we design a specialized Network-on-Chip communication backbone optimized to mitigate the bottlenecks associated with non-deterministic token routing traffic across the combinatorial space of input-dependent MoE traffic patterns. Experimental results demonstrate that ThAME outperforms state-of-the-art counterparts by up to 15.7x in terms of speedup and improves energy efficiency by up to 9.8x.
[AI-110] Bridging the Information Gap: Semantic Densification and Hindsight Distillation for Cold-Start Prediction
链接: https://arxiv.org/abs/2607.17070
作者: Hao Duong Le,Yifei Gao,Huan Li,Lun Jiang,Chen Bai,Ke Xing,Chen Zhang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:New-user cold-start is a critical bottleneck for e-commerce platforms: predicting user lifetime value (LTV) and conversion rate (CVR) for users with sparse interaction history. Two prior directions – LLM-based semantic augmentation and learning using privileged information (LUPI) – each face a key limitation. First, LLM augmentation produces unstructured rationales that are noisy and hard to operationalize in production. Second, naive student-teacher distillation can be brittle due to an information gap between the privileged teacher and the sparse student; moreover, this gap is heterogeneous across users. We propose SemRaD, a Semantic Reasoning-aware Distillation framework addressing both limitations. First, a Structured Semantic Reasoning Pipeline replaces free-form rationales with a structured schema built via a discover-curate-audit workflow, producing per user a Densified Semantic Profile (consumed by the deployed student via a Semantic-Gated Encoder that focuses on the most informative dimensions) and a Hindsight Distillation Target reconciled from pre- and post-conversion reasoning (used only at training). Second, to bridge this gap and handle its heterogeneity, a Hindsight-Aware Distillation Network transfers privileged knowledge via the hindsight target, with Distillation Experts improving transfer under per-user variability. On a large-scale industrial dataset, SemRaD lifts +1.9% LTV (Gini) and +1.0% CVR (AUROC) over a production-grade base; a four-week online A/B at Keeta confirms +1.0% LTV / +0.43% CVR. SemRaD also matches the production system’s LTV using only 9% of the training data while improving CVR by 0.8%.
[AI-111] When LLM s Over-Answer: Measuring and Mitigating Quality Issues in LLM -Based Hardware Description Language Question Answering
链接: https://arxiv.org/abs/2607.17063
作者: Ziteng Hu,Jiachi Chen,Wenhao Lv,Huan Zhang,Yingjie Xia
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:The rapid advancement of large language models (LLMs) has led practitioners to increasingly rely on them for answering questions about hardware description languages (HDLs). Because HDL is ultimately synthesized into physical hardware, an imprecise or redundant answer can propagate into timing violations or non-synthesizable logic that surface only late in the design flow, making the quality of HDL answers especially consequential. However, the quality of LLM-generated responses, particularly in comparison with answers provided by human experts, remains unclear. To investigate this question, we collect 6,246 HDL QA posts with accepted answers from Stack Overflow and curate them into a dataset, organized into a taxonomy of four main categories (Conceptual, Debugging, Generation, and Optimization) and ten subcategories. Using this dataset, we design a user study conducted with 19 HDL engineers with one to three years of experience. Our findings reveal a pervasive over answering tendency: LLMs supply correct content but bury it under redundant alternatives (65.7%) and verbose padding (69.1%), while nearly half of answers (49.0%) fail to fully align with expert answers yet participants still preferred LLM responses for readability (58.3%). Motivated by these findings, we propose a multi-agent framework for improving LLM-based HDL question answering. We evaluate answer quality using an LLM-as-Judge and two structural metrics: the number of core answers, which reflects redundancy since LLMs often provide multiple alternative solutions, and the length of non-core content, which reflects verbosity. Evaluated on the four mainstream LLMs, our framework increases the average core-answer quality score from 3.71 to 4.67 (+0.96) and the non-core content quality from 3.72 to 4.23 (+0.51), on a five-point scale.
[AI-112] Solver-Hard Is Not Model-Hard: A Hardness-Controlled Diagnostic for LLM Constraint Reasoning
链接: https://arxiv.org/abs/2607.17047
作者: Lucky Verma
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: 13 pages, 2 figures, 5 tables. Code and aggregate reproduction data: this https URL
Abstract:LLM constraint reasoners are often evaluated near the random-SAT phase transition, confounding density and solver hardness. We test instance-level transfer while near-matching clause density. At aligned size bins, with near-matched density and matched maximum clause width, we compare proof-hard expander-Tseitin and proof-easy ladder-Tseitin formulas, pigeonhole anchors, and density-mismatched controls. Theory separates their resolution hardness; a solver-specific Glucose mean-conflict proxy differs by up to 51\times , and five other solvers preserve the direction. Across three included models (243 instances each; a fourth is excluded for abstention), the near-matched-density accuracy gaps range from -32 to +20 points, with a pooled gap of +1.7 points ( p=0.74 ) and a wrong-signed correctness-versus-conflict association ( r=+0.15 ). A proof-preserving relabeling lowers accuracy in all five clusters for one model (mean -93 points) but not another, exposing model-surface sensitivity. In a preregistered extension, provider-reported completion-token spend does not consistently increase with the proxy after accounting for formula length and censoring. At 16k, the reasoning model spends more on proof-easy matched formulas and exhausts its budget on the solver-easiest UNSAT family; the 32k C1 gap is absent. These scoped dissociations concern verdict accuracy and observed token spend, not certificate solving, exact proof length, or allocation efficiency.
[AI-113] Where Does Agent Reliability Come From? A Cross-Benchmark Decomposition of Verification Loops Specialist Models and Scaffolding in a Production Enterprise Agent
链接: https://arxiv.org/abs/2607.17044
作者: Arunabh Dastidar(for the Leni Team)
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 19 pages, 5 figures, 5 tables. Evaluations conducted March-April 2026. Run-level evaluation record and audit scripts: this https URL
Abstract:Multi-step enterprise agent tasks fail in a characteristic way: single-pass inference has no checkpoint between deciding an answer and committing to it. We study one production system (Leni) whose architecture installs such checkpoints: verification loops (execute, observe, compare, correct) staffed by lightweight task-specialized post-trained models. We evaluate the unmodified production configuration on three public benchmarks stressing distinct failure modes: SpreadsheetBench Verified (silent computation error), BullshitBench v2 (premise confabulation), and the GAIA validation split (cascade error over long tool chains). The full system improves over its frontier base model by +11.0 percentage points on SpreadsheetBench (91.25% vs 80.25%, n=400, p0.001), +7 to +10 percentage points on BullshitBench (98% vs 91%, n=100), and roughly +15 points on GAIA validation (75.2% pass@1, n=165; 83.0% best-of-k). Our central contribution is a decomposition of that uplift: most of it comes from scaffolding, routing, and specialist models rather than from the verification step itself, whose isolated contribution is small (+1.5 points) but concentrated at the top of the score distribution, where it converts otherwise-failing tasks. We instrument the loop end-to-end, yielding an empirical verifier confusion matrix (catch rate about 0.20, fix rate 0.75, no false-alarm regressions) that grounds a compounding-reliability model. Specialist-swap ablations suggest that the loop’s value depends on who observes it: replacing the small trained verifier with the generating frontier model eliminates most rescues. A valid-premise control shows zero over-rejections in 100 expert-level questions.
[AI-114] Reward-Driven LLM Agent Workflows: Synthesizing POMDP Routing and Self-Correction for Autonomous Decision-Making
链接: https://arxiv.org/abs/2607.17038
作者: Amez Amanj Ali,Kuo-Kun Tseng
类目: Artificial Intelligence (cs.AI)
备注: 29 pages, 1 table, native TikZ/pgfplots diagrams. Code available at this https URL
Abstract:This paper addresses key technical challenges in current large language model (LLM) agent applications, including long-horizon planning, sparse reward attribution, and dynamic environmental interaction, by designing and optimizing an intelligent agent workflow. The proposed architecture is based on the synthesis of core AI paradigms: Visual, Language, Generative, Graph, Multimodal, Reinforcement, and Agent Intelligence. Unlike conventional baseline models that rely on static prompting and lack robust perception-action loops, our approach introduces a Partially Observable Markov Decision Process (POMDP) routing mechanism. This mechanism is augmented with an internal, self-correcting reward model that evaluates decision trajectories before execution. By integrating multimodal inputs and advanced reinforcement learning principles (such as proximal policy optimization and value function approximation), the agent maintains long-term structural memory and dynamically adapts its reasoning pathways to mitigate error accumulation. Empirical experiments on the ALFWorld embodied simulation environment and the WebShop online navigation benchmark demonstrate a 24.5% absolute improvement in task success rate and trajectory efficiency over mainstream baselines like the standard ReAct framework. Comprehensive ablation studies confirm the significant contribution of the reward-driven critique module in suppressing hallucination rates. This research bridges theoretical foundations of reinforcement learning and graph-based memory with autonomous agent workflows. Ultimately, the resulting architecture offers a practical, scalable reference framework for developing artificial intelligence technologies in complex, multi-step autonomous systems. Code is available at this https URL.
[AI-115] Scalable Causal Imitation Learning
链接: https://arxiv.org/abs/2607.17003
作者: Eylam Tagor,Mingxuan Li,Elias Bareinboim
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Reinforcement Learning Journal 2026 (Also RLC 2026)
Abstract:Imitation learning enables learning a policy in an unknown environment with a latent reward signal using expert demonstrations, but it struggles when the imitator’s and expert’s observations are mismatched and unobserved confounders are present in expert demonstrations. By identifying appropriate adjustment sets via the sequential \pi -backdoor criterion, causal imitation learning (CIL) provides a framework for approximating the expert’s policy from confounded data. However, existing CIL methods, Causal Behavioral Cloning (Causal BC) and Causal Generative Adversarial Imitation Learning (Causal GAIL), are designed for short-horizon, low-dimensional settings. When applied to continuous control tasks with long horizons and high-dimensional state-action spaces, these methods exhibit poor performance: Causal BC suffers from compounding errors, Causal GAIL is unstable and sample-inefficient, and sequential \pi -backdoor adjustment becomes impractical. We introduce Causal Soft Q Imitation Learning (SQIL) and Causal Inverse soft-Q Learning (IQ-Learn), two off-policy causal imitation learning algorithms that combine the causal adjustment framework with state-of-the-art inverse reinforcement learning objectives. Both algorithms operate on causally-adjusted state representations produced by an efficient approximation of the sequential \pi -backdoor criterion, exploiting the causal structure of continuous control environments to reduce the full-horizon adjustment to a fixed-size sliding window. We evaluate all methods in a suite of confounded environments and find that Causal SQIL and Causal IQ-Learn substantially outperform prior CIL algorithms on long-horizon tasks, sometimes surpassing the expert, whereas all causally unaware imitation methods fail to learn meaningful behavior.
[AI-116] Counterfactual Shapley Credit Assignment
链接: https://arxiv.org/abs/2607.16999
作者: Mingxuan Li,Kaizhan-Lee,Elias Bareinboim
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Reinforcement Learning Journal 2026 (Also RLC 2026)
Abstract:The Credit Assignment Problem (CAP) is fundamental to developing efficient and explainable Reinforcement Learning (RL) agents. Existing frameworks, whether relying on temporal contiguity or hindsight-conditioned reward reweighting, frequently fail to attribute properly between an agent’s policy (skill) and environmental stochasticity (luck). A principled approach to CAP must isolate the true causal drivers of observed outcomes from spurious correlations and environmental randomness. We introduce Counterfactual Shapley Credit Assignment, a novel framework grounded in causal theory that attributes credit and blame via the Counterfactual Shapley Value ( \phi -value). By redistributing environmental rewards, \phi -values enhance temporal credit assignment across three critical dimensions: sparse causality, high stochasticity, and delayed rewards, all while preserving the optimal policy. We derive a consistent estimator that computes \phi -values efficiently, enabling a new class of policy gradient methods, \phi -PPO, combined with Prioritized Trajectory Replay (PTR). Empirical results demonstrate that \phi -values align precisely to the ground truth causes of task rewards with superior sample efficiency in challenging environments where prior state-of-the-art methods fail to converge.
[AI-117] PriorProof: A Point-in-Time Measure of Technique Novelty for Formal Proofs
链接: https://arxiv.org/abs/2607.16997
作者: Neel Somani
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Logic in Computer Science (cs.LO)
备注: 17 pages, 2 figures. Code: this https URL
Abstract:Mathematicians distinguish proofs that explain, simplify, or introduce a nonstandard route, but these judgments are difficult to operationalize. We study a deliberately narrower construct: time-relative proof-route nonstandardness in formal mathematics. For a Lean theorem, PriorProof extracts the dependency footprint of its elaborated proof term and scores the weighted surprisal of that footprint under a retrieval-conditioned, hierarchically smoothed prior built only from an earlier quarterly snapshot of Mathlib. The method requires no hand-built technique ontology and no human labels: statement retrieval is learned from proof-derived contrastive pairs, while the scored object is read mechanically from proof terms. In a blinded topology study, 100 presentations collapse to 76 distinct underlying pairs: 12 canonical contrasts shown three times for consistency screening and 64 distinct stratified pairs. Against the majority of three retained domain raters, PriorProof agrees on 53/76 pairs (69.7%, Wilson 95% CI 58.7-78.9%), including 11/12 canonical pairs (91.7%, 64.6-98.5%) and 42/64 stratified pairs (65.6%, 53.4-76.1%). Score-gap quartiles are nonmonotone after repeat collapse; the endpoints are 12/19 (63.2%, 41.0-80.9%) in the smallest-gap bin and 16/19 (84.2%, 62.4-94.5%) in the largest, supporting an endpoint-calibration tendency rather than a resolved staircase. The best language-model condition agrees on 60/76 pairs (78.9%, 68.5-86.6%); on paired outcomes, PriorProof alone is correct on 8 pairs and the model alone on 15 (exact two-sided McNemar p = 0.210), so the difference is not established at this sample size. We therefore present PriorProof not as a replacement for expert or model judgment, but as a decomposable, time-anchored signal whose score gap provides an interpretable reliability indicator.
[AI-118] Expected Free Energy as Belief-Dependent Utility for rho-POMDPs
链接: https://arxiv.org/abs/2607.16981
作者: Patrick Cooper,Alvaro Velasquez
类目: Artificial Intelligence (cs.AI)
备注: 41 pages, 12 figures. Paper accepted as a spotlight at the 2026 International Workshop on Active Inference (IWAI); the workshop version will appear in the Springer Communications in Computer and Information Science (CCIS) series. Code and experiment data at this https URL
Abstract:An agent acting under partial observability must decide when to gather information and which observations are worth their cost. Standard POMDPs value information only through its eventual effect on reward. The \rho -POMDP framework instead rewards uncertainty reduction directly, through a belief-dependent utility \rho , but in practice both the choice of \rho and the weight placed on it are tuned by hand for every task. We show that active inference removes this tuning entirely. Minimizing Expected Free Energy (EFE) is exactly equivalent to solving a \rho -POMDP whose utility is expected information gain, and the exploration weight is fixed at w=1 because the variational bound expresses pragmatic and epistemic value in the same units (nats). We prove this equivalence for observe-then-commit POMDPs and extend it to factored observation POMDPs, a broader class that covers interleaved observe-act problems such as non-destructive testing and mobile sensing, where gathering information leaves the hidden state unchanged. Experiments support the theory. Across environments ranging from the classic Tiger problem to RockSample and a new Structural Inspection benchmark with over 65,000 states, the untuned weight matches or outperforms reward-only planning at the same horizon, avoids the over-exploration of bonuses tuned per task, and sits near the reward-maximizing knee of the success-reward Pareto frontier. The practical payoff is an exploration objective that works out of the box. In applications such as fault detection and medical screening, where every test has a price and every missed fault has a cost, EFE supplies a belief-dependent utility that is derived rather than tuned.
[AI-119] Lomekwi: Resource-Bounded Tool Discovery in LLM Agents
链接: https://arxiv.org/abs/2607.16961
作者: Roshan Klein-Seetharaman,Daniel Wang,Andrew Xu
类目: Artificial Intelligence (cs.AI)
备注: All authors contributed equally. 17 pages, 6 figures. Presented at the 2026 Conference on Learning Theory Workshop on Learning in an Agentic World
Abstract:Existing tool-use benchmarks report a single success rate for complex, multistep tasks. Inspired by ideas from cognitive science, we distinguish tool use from tool discovery and decompose the latter into curiosity (the model’s ability to discover the parts needed to build the tool), recognition (the model’s ability to discover the process of creating the tool), and efficiency (the model’s use of the tool after creation). We show that this framework can be applied to existing discovery tasks, such as Voyager. In addition, we provide evidence that recognition inversely scales with model size, and we introduce and analyze a class of combinatorial games that demonstrates this. We further observe inverse scaling in a separate environment designed to emulate real-world tasks.
[AI-120] CADENCE: Closing the Reasoning Gap via Coverag e-Adaptive On-Policy Distillation
链接: https://arxiv.org/abs/2607.16955
作者: Satyam Kumar,Saurabh Jha
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:On-policy knowledge distillation transfers reasoning from large teachers to compact students, but existing approaches suffer three compounding failure modes: (i) cold-start collapse, where a fresh student assigns near-zero mass to teacher-preferred tokens; (ii) state-agnostic divergence scheduling, where time-only forward/reverse-KL interpolation ignores the student’s coverage state; and (iii) binary reward sparsity, where pass/fail signals discard information from partially correct traces. We present CADENCE, a unified framework with a targeted fix for each. Its DRIFT mechanism schedules a per-token convex mixture of forward-KL and reverse-KL surrogate objectives on student-sampled trajectories (per-token surrogates, not sequence-level KL gradient estimators). Six components extend it: (A) COVA, a coverage-adaptive \beta schedule accelerating the forward-to-reverse transition; (B) FTB, a forking-token boost concentrating gradient at high-entropy positions via a globally-normalized entropy reference; © CCD, a dense reward adding numerical-proximity partial credit for incorrect-but-close traces; (D) LAP, brevity-preferential correct-rollout reinforcement; (E) EMR, an entropy-matching calibration regularizer; (F) BSD, a bootstrapped self-distillation phase. On GSM8K and MATH-500 (corrected 512-token protocol, 5 seeds, reported std), CADENCE distills a 0.5B student from a 1.5B teacher to 69.8 \pm 0.5% GSM8K pass@1 (from 48.7% pretrained; 63.2% of the teacher gap closed) and to 72.1 \pm 0.4% with a 3B teacher (76.2% closed), beating the strongest matched-compute label-using baseline (DRIFT+binary reward) by +4.4 \pm 0.7 points. All experiments run on a single Apple Mac Studio (M-series, 64GB unified memory), showing principled distillation reaches strong reasoning quality without datacenter-scale hardware. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.16955 [cs.LG] (or arXiv:2607.16955v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16955 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-121] A Multi-Agent System for 5G Throughput Prediction in Multi-Operator Urban Environments
链接: https://arxiv.org/abs/2607.16930
作者: Muhammad Kabeer,Rosdiadee Nordin,Nadiva Nuriftitah,Sian Lun Lau
类目: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI); Signal Processing (eess.SP)
备注:
Abstract:Throughput prediction is foundational for artificial intelligence-driven 6G resource orchestration. Conventional monolithic machine learning models struggle to generalize across diverse operators, mobility modes, and traffic types, leaving a critical stochasticity gap between signal conditions and achievable throughput. To overcome these constraints in heterogeneous urban environments, we propose a Tiered Multi-Agent System (TMAS) that dynamically routes edge telemetry to context-aware Domain Micro-Agents, validated on a dataset of 48,618 samples collected in Sunway City, Malaysia, with Nemo Handy drive test software, spanning three Tier-1 mobile network operators, three mobility modes, namely (i) elevated pedestrian walkway, (ii) ground-level shuttle bus, and (iii) elevated bus rapid transit; and three traffic profiles, namely (i) persistent download, (ii) persistent upload, and (iii) adaptive video streaming. Our evaluations reveal that TMAS overcomes predictability bottlenecks, achieving a coefficient of determination (R2) of up to 0.931 and a Mean Absolute Error (MAE) as low as 0.53 Mbps. The system demonstrates high operational efficiency, with rapid micro-agent training times, low inference latencies, and agentic routing overhead of 0.004 to 0.126 ms. These latency characteristics indicate the architecture is a promising candidate for the response times required by next-generation wireless networks.
[AI-122] PREFAIL: Identifying Precursors to Failures in Robotic Lift-and-Place Tasks to Improve Task Execution Performance
链接: https://arxiv.org/abs/2607.16921
作者: Zeyu Shangguan,Rajas Chitale,Rutvik Patel,Satyandra K. Gupta,Daniel Seita
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Non-prehensile manipulation enables flexible material handling with part carriers, but friction-based support makes high-speed motions failure-prone, while slower operation increases cycle time. Proactive failure prediction is therefore essential for efficient and reliable performance, yet existing approaches remain limited by key constraints, including sensitivity to dynamic actions and high dependence on known policy structures. Furthermore, existing methods and datasets lack a precise characterization of the latest intervention time, leaving it unclear whether a detected failure can still be prevented through timely intervention. In this paper, we investigate lift-and-place tasks for non-prehensile material handling manipulation and propose a more effective approach to predicting precursors to failures (PREFAIL) by analyzing the relative motion of target objects with respect to the carrier. We further introduce a dataset that precisely identifies the latest intervention time for risky manipulations, enabling rigorous evaluation of whether a failure prediction is actionable. We validate our approach on both simulation and real-world datasets. Our experimental results demonstrate that PREFAIL substantially improves both the accuracy and timeliness of responses to failure precursors.
[AI-123] Environment-free Synthetic Data Generation for API-Calling Agents
链接: https://arxiv.org/abs/2607.16900
作者: Seanie Lee,Sanjoy Chowdhury,Chao Jiang,Cheng-Yu Hsieh,Ting-Yao Hu,Alexander T Toshev,Oncel Tuzel,Raviteja Vemulapalli
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Training API-calling large language model (LLM) agents demands massive amounts of high-quality trajectories. However, collecting such data at scale typically requires fully implemented environments with executable APIs and realistic, pre-populated backend databases, creating a major bottleneck for scalability. To overcome this, we propose an environment-free synthetic data generation approach that leverages LLMs as on-the-fly digital world models. Given only API specifications, our method generates trajectories mimicking interactions between an agent and a stateful environment. Specifically, an LLM first generates diverse tasks solvable with the provided APIs. A teacher agent then iteratively solves each task while an LLM simulator generates coherent synthetic API responses conditioned on the task context and simulation history. Finally, an LLM judge filters the trajectories to ensure the quality of the resulting dataset. We evaluate our approach on the challenging AppWorld and OfficeBench benchmarks, which include both information-retrieval and state-changing tasks. Fine-tuning models on our synthetic data yields significant performance gains, demonstrating that effective supervision for API-calling agents can be generated without any executable environment. Our results establish LLM-based API simulation as a practical, scalable solution for training agents across diverse API ecosystems.
[AI-124] Certified-Gap Dual-Price Policies for Real-Time Truckload Bid Acceptance with Relocating Clock-Constrained Resources
链接: https://arxiv.org/abs/2607.16891
作者: Aswin Chandrasekaran
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注: 14 pages, 1 figure. Companion benchmark paper: arXiv:2607.07343
Abstract:A truckload carrier must accept or reject each load tender within seconds. The decision depends on fleet state, hours-of-service (HOS) clocks, and appointment windows. We model this as a weakly coupled dynamic program in which the resources relocate and carry clocks: serving a request moves the truck to a new market and depletes its clocks, and whether a truck can serve a request depends on its state. Occupancy-based reusable-resource models do not cover this setting. We build a real-time dual-price policy from the same Lagrangian relaxation that gives the problem’s upper bound. Policy and bound come from one object, so every run reports a certified optimality gap. We prove three things. First, the certificate is valid for any duals, any discretization, and any surrogate quality. Second, the policy’s same-time spatial-gradient rule is exactly fluid complementary slackness, and the policy is asymptotically optimal in the subcritical fluid regime; the fitted prices are also portable across sample paths, by linear-programming basis stability. Third, certificates have limits: per-resource Lagrangian slack can stay bounded away from zero at every fleet size. We exhibit a three-truck kernel with an exact rational certificate and a replication lemma. On a public closed-loop benchmark with thirty paired seeds, the policy – which needs no rollout labels, only one offline dual solve – beats a rollout-trained surrogate on two of three scenarios (tight: +2.0 pp, 95% CI [+0.5, +3.6], Wilcoxon p = 0.023; mild: +3.5 pp, CI [+2.4, +4.5]) and ties the third. It decides in 0.04-0.09 ms, three orders of magnitude faster than the Monte Carlo rollout teacher. Its certificates are stable across ten bounded instances per scenario, at 57-64% of optimal, within 3-6 points of what the 1000x-slower teacher certifies.
[AI-125] Do Speech Tokens Leak Voiceprints? Speaker Inversion Attacks Against End-to-End Speech Language Models
链接: https://arxiv.org/abs/2607.16870
作者: Ye Lu,Yihan Yan,Zhaoyang Zhang,Zhitao Ou,Runze Liu,Li Liu,Shen Wang
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注:
Abstract:End-to-end speech language models increasingly represent user speech with speech tokens rather than relying exclusively on cascaded ASR–LLM–TTS pipelines. Although these tokens support expressive and low-latency spoken interaction, they may also preserve sensitive speaker characteristics. We investigate whether exposed speech tokens leak voiceprints and formulate this risk as a speaker inversion attack. We introduce Audio BERT (AuB), a trainable model that constructs token embeddings from discrete codebooks and aggregates them into speaker-sensitive representations, and propose SpInv, a two-stage inversion method built on AuB to recover embeddings in the space of an attacker-specified speaker encoder. We evaluate Moshi, Higgs3, Kimi-Audio, and Qwen3-Omni using speaker-disjoint protocols on the VoxCeleb dataset. Extensive experiments show that, with only three seconds of frontend output, SpInv achieves cosine similarities above 0.70 in the attacker-specified speaker-encoder space.
[AI-126] Beyond Semantic Equivalence: Logical Graphs for LLM Uncertainty Quantification
链接: https://arxiv.org/abs/2607.16868
作者: Yanni Dong,Minghua Liu,Meiling Zhu,Xiaowei Huang,Lijun Zhang
类目: Artificial Intelligence (cs.AI)
备注: 14 pages, 5 figures, 5 tables. Under review
Abstract:Large Language Models (LLMs) often produce confidently stated yet unreliable outputs, posing critical challenges for deployment in safety-sensitive applications. Existing uncertainty metrics such as semantic entropy capture agreement at the level of semantic equivalence, but largely ignore the logical relationships between distinct answers. As a result, they tend to overestimate uncertainty and falsely flag hallucinations in settings where generated responses are diverse in form yet logically compatible (e.g., differing only in granularity or specificity). We propose Logical Graph Uncertainty (LGU), a framework that explicitly models implication and incompatibility among answers. LGU aggregates probability mass along entailment chains, computes entropy over logically maximal hypotheses, and penalizes mutual incompatibility among them. Across multiple question-answering benchmarks, LGU consistently improves uncertainty estimation over existing methods, and outperforms the semantic entropy baseline by up to +7.1% AUROC and +3.5% AUARC across datasets.
[AI-127] Principled Direction-Free Intrinsic Motivation through Model-Free Epistemic Free-Energy Estimators
链接: https://arxiv.org/abs/2607.16858
作者: Alireza Furutanpey,Schahram Dustdar
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted for a Spotlight Presentation at IWAI 2026
Abstract:Across environments with mixed sources of uncertainty, unsupervised reinforcement learning requires intrinsic motivation that does not precommit to a particular direction of surprise. Surprise minimization is scoped by design to ``unstable’’ environments. Prediction-error curiosity rewards total expected surprise, including irreducible noise. Bandit or mixture switching between surprise-minimizing and surprise-maximizing rewards reintroduces non-stationarity by construction. We propose a single intrinsic reward, stationary within each window, derived from the novelty contribution of a preference-free Expected Free Energy objective, expressed in reward-maximization form. Our claim is that parameter information gain, the expected surprise of the next state minus its irreducible part, is the appropriate intrinsic signal in both high-entropy and low-entropy components of the state space. Maximizing it seeks exactly the surprise the model can explain away. In regions of unresolved dynamics, this epistemic term drives exploration. As dynamics become resolved, the epistemic term vanishes, while an aleatoric penalty favors lower-variance transitions, all without fitting an explicit next-state predictor. A pseudocount supplies epistemic value, a probe-based penalty captures aleatoric variance, and a short-horizon gate protects informative successors. A window-based freeze of all reward-defining objects yields a stationary Bellman operator, explicit bounds on learning targets, and a conditional uniform-concentration result for the nonparametric estimators under mixing, smoothness, bandwidth, and capacity assumptions. In active-inference terms, the agent is preference-free where novelty is retained, standard likelihood ambiguity vanishes under full observability, a nonstandard transition-entropy penalty is added, and surprise minimization emerges in resolved regions of the state space.
[AI-128] Agent Brew: Lifelong Knowledge Brewing from Strong Teachers to Weak LLM Agents
链接: https://arxiv.org/abs/2607.16851
作者: Yangqin Jiang,Chao Huang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Deploying LLM agents typically requires a compact test-time student, even if a stronger teacher is available during training. We study knowledge brewing: distilling a teacher’s interactive experience into a persistent external memory for the student. Crucially, this requires no weight updates, expert demonstrations, ground-truth labels, or test-time teacher access. This setting poses two challenges: environments provide only sparse, binary feedback, and teacher-authored notes must be inherently tailored to be concretely executable by a substantially weaker student. To address these hurdles, we propose AgentBrew, comprising two coupled components. First, a failure-triggered teacher–Ralph Loop mitigates sparse feedback by transforming student failures into environment-validated notes. Second, student-aware synthesis calibrates teacher knowledge to the weak executor’s operational granularity, yielding model-specific, actionable guidance. Extensive evaluations and comprehensive ablations across coding, math, and tool-use tasks demonstrate that this asymmetric, training-free brewing paradigm produces highly capable yet deployable LLM agents.
[AI-129] Beyond Memory Leaderboards: Evaluating Scientific Memory as Budgeted Context Restoration
链接: https://arxiv.org/abs/2607.16848
作者: Maksim Sheverev,David Finkelstein,Sergey Nikolenko
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 40 pages, 6 figures, 8 tables
Abstract:Long-term memory is becoming a core component of LLM agents, but most memory benchmarks evaluate conversations or compact summaries, while research agents need to restore evidence from full scientific papers. We introduce two full-text scientific-memory benchmarks, Public AI Memory (PAIM; 81 papers, 66 questions) and Public Transformers (PTr; 252 papers, 98 questions). We evaluate eight memory/retrieval systems, including our own proposed Theoria, plus a no-retrieval baseline. Our results show that memory leaderboards are not interpretable without the full protocol: ingestion granularity, raw-text preservation, retrieval budget, retrieval modality, rubric audit, and judge choice all affect the outcome. For example, on PAIM Graphiti wins convincingly but uses 2.6M characters of retrieved context per query, and after controlling for retrieval budget the lead disappears. On PTr, for the systems where BM25 retrieval can be added cleanly, the sparse-dense hybrid is the single most significant intervention: hybrid variants of Simple RAG, Mem0, and Theoria tie for the lead within 0.03 points. Multi-judge and human side-by-side calibration show that LLM-as-a-judge rankings are consistent across frontier judges and agree with human evaluation, with an effective resolution of roughly one point on a ten-point scale. We argue that scientific memory should be evaluated as budgeted, modality-aware context restoration rather than as an unconstrained architecture leaderboard, and we release the datasets, harness, raw outputs, judgments, and scripts to reproduce our results and serve as tools for such evaluation. Our code is available at this http URL , and the datasets are available at this http URL .
[AI-130] From Overload to Insights: How AI Agents Can Support Scientists in Analyzing Complex Data
链接: https://arxiv.org/abs/2607.16845
作者: Tim Fuchs,Luca Gelisio,Steffen Hauf,Walid Maalej
类目: Artificial Intelligence (cs.AI)
备注: Accepted for publication at the 42nd IEEE International Conference on Software Maintenance and Evolution
Abstract:Scientists at European XFEL conduct experiments that generate very large and complex datasets. The subsequent data analysis is challenging as scientists must combine their domain expertise with facility- and software-specific knowledge scattered across documentation, tools, and support channels. To address this problem, we designed and evaluated an agentic AI system tailored to the scientists’ needs and integrated with the high-performance computing environment of European XFEL. Using a design science research approach, we conducted a rapid literature review, a systematic evaluation of 16 AI tools, multiple interviews, a focus group, and a user study with experts at European XFEL to develop and evaluate two prototypes. Our study identifies key knowledge challenges in scientific data analysis, derives requirements for an AI agent that supports knowledge retrieval and source code generation, and proposes design recommendations for a specialized system adaptable to the evolving AI tool landscape. These findings provide guidance for developing maintainable AI support in highly specialized scientific environments.
[AI-131] First-Order Predictable but Pairwise Frag ile: Local Task Adaptation in Trained Transformers
链接: https://arxiv.org/abs/2607.16821
作者: Irina Piontkovskaia,Sergey Nikolenko
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 36 pages, 11 figures, 10 tables
Abstract:Task arithmetic, sequential fine-tuning, activation steering, and first-order random search all operate through relatively small perturbations around an already trained checkpoint, and they rely on different local approximations: individual perturbations should be first-order predictable, task updates should compose with controlled interference, useful tangent structure should be stable and possible to estimate, and weight edits should have counterparts in representation space. We measure 8 such properties with the same harness around a multitask LoRA operating point, on 9 transformers (82M-7B), with a prospectively registered property list, thresholds, and test split. We find a shared one-direction validity window up to the tested scale 10^-2 , but no universal radius for pairwise composition or update ordering. Along individual directions, changes of the probe loss remain first-order predictable throughout the grid: a perturbation’s effect on the loss is essentially its projection onto the gradient, which is also what makes local random search work. Pairwise structure, however, proves to be far more fragile: on over a third of the measured (model, task pair) combinations, two-update order sensitivity sets in strictly inside that window; task-gradient subspaces rotate within tens of steps; additivity under our fixed activation probe fails at full task-vector scale on several models, including both held-out 7B models; and no model median passes the registered global mean-vector weight-to-steering correspondence bar. For two sequential task-gradient steps, the leading order-dependent term is the Lie bracket H_B\textbfg_A-H_A\textbfg_B ; its normalized prediction c(\eta)=\eta\kappa+O(\eta^2) tracks the measured defect at median ratio 1.002, while the onset scale \eta^\dagger\approx0.10/\kappa spans three orders of magnitude across models and task pairs.
[AI-132] FUSAR-R1: A Large-Scale Reasoning Model for Intelligent Interpretation of SAR Images
链接: https://arxiv.org/abs/2607.16819
作者: Yi Yang,Xiaokun Zhang,Yuxuan Li,Ruyi Zhang,Xinpeng Zhou,Haipeng Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:In recent years, large-scale vision-language models have been driving a paradigm shift in intelligent remote sensing image interpretation. By incorporating textual semantic information, the cognitive expression, semantic understanding, and human-computer interaction capabilities of interpretation models have been significantly improved, achieving initial progress in the field of Synthetic Aperture Radar (SAR) image interpretation. However, SAR images are affected by factors such as coherent imaging mechanisms, complex scattering characteristics, speckle noise interference, and target-background coupling, resulting in complex and variable image features with significant uncertainties and specializations. Existing SAR vision-language models do not yet possess the step-by-step analysis, logical judgment, and self-correction capabilities of human experts, making it difficult to support reliable intelligent interpretation in complex scenarios. To address this issue, this paper proposes a large-scale reasoning model, FUSAR-R1, for intelligent interpretation of SAR images. The model first constructs explicit chain-of-thought reasoning data by simulating the interpretation process of human experts and uses this data to guide instruction learning, thereby endowing the model with basic reasoning capabilities. Subsequently, a reinforcement learning strategy is introduced to optimize the model’s outputs based on inference results, enabling self-correction and more reliable reasoning. Experimental results demonstrate that FUSAR-R1 consistently outperforms existing multimodal large-scale models across various SAR interpretation tasks, including target detection, target counting and classification, and land-cover category recognition.
[AI-133] Explainable Lightweight Compact Deep Models for Speech Emotion Recognition ICML
链接: https://arxiv.org/abs/2607.16803
作者: Nelly Elsayed
类目: ound (cs.SD); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
备注: Accepted in the IEEE ICMLA 2026 Conference
Abstract:Speech Emotion Recognition (SER) is an important component in a wide range of human-centered applications, including healthcare, customer service, and human-omputer interaction. In medical and decision-support settings, there is increasing interest in models that not only achieve accurate emotion recognition but also support transparent predictions and efficient deployment. However, many existing SER approaches rely on complex deep learning architectures that limit interpretability and increase computational cost. This paper presents an explainable and lightweight speech emotion recognition framework based on a compact convolutional neural network architecture. The proposed approach utilizes log-Mel spectrogram representations to capture spectro-temporal speech characteristics and employs attentive statistics pooling to emphasize emotionally salient temporal segments. To improve model transparency, gradient-based class activation mapping (Grad-CAM) is incorporated to visualize the time-frequency regions that influence the model’s predictions. Experimental evaluation on the SAVEE emotional speech dataset demonstrates that the proposed framework achieves competitive recognition performance while maintaining a compact architecture with significantly fewer parameters than many existing SER models. The results indicate that efficient convolutional architectures combined with interpretable analysis can provide a practical balance between recognition accuracy, computational efficiency, and model transparency.
[AI-134] Supporting Autonomous Process Execution within a Multi-Perspective Constraint Frame via Numeric Planning
链接: https://arxiv.org/abs/2607.16738
作者: Paul Wittlinger,Giacomo Acitelli,Anti Alman,Fabrizio Maria Maggi,Andrea Marrella
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AI-Augmented Business Process Management Systems (ABPMS) enhance traditional BPMS by leveraging advanced AI techniques to define, execute, and monitor complex process structures. Within this landscape, Framed Autonomy denotes the capability of a system to autonomously advance the execution of a Business Process (BP) instance while strictly adhering to a predefined frame, i.e., a set of constraints that may span multiple perspectives. Existing research on framed autonomy has predominantly focused on control-flow constraints, either declarative or procedural, and typically relies on their transformation into automata-based representations. In this study, we extend this line of work by introducing a novel tool for what-if analysis that augments the process frame with multi-perspective constraints, including data-aware and temporal conditions. Given a partial process execution, the proposed approach exploits this enriched frame to recommend optimal continuations in compliance with the underlying process specifications. We additionally report an empirical evaluation demonstrating the scalability and effectiveness of the technique, thereby highlighting its potential for supporting autonomous and constraint-aware decision making in ABPMS.
[AI-135] RECON: Benchmarking Agent Memory for Compositional Reasoning over Long Contexts
链接: https://arxiv.org/abs/2607.16716
作者: Mihir Shriniwas Arya
类目: Artificial Intelligence (cs.AI)
备注: 25 pages
Abstract:Large language models and LLM-based agents are widely used as personal chat assistants, enterprise copilots, and autonomous workflow agents. In all these applications, memory (the ability to retain, access, and reason over information accumulated over long contexts and multiple interactions) plays a crucial role in determining the reliability of any agent. We introduce RECON (Reasoning over Extended Contexts with Obfuscated Narratives), a benchmark for evaluating compositional reasoning over long contexts. RECON spans 24 case files across three domains (criminal, medical, and financial), each ranging from 50k to 100k tokens, and tests agents on six memory intensive tasks: reconstructing multi-hop evidence chains, propagating cascading invalidations, resolving source conflicts, counterfactual reasoning, satisfying temporal constraints, and temporal fact retrieval. Recent memory benchmarks evaluate whether agents can retrieve scattered facts or detect if a fact has changed whereas RECON evaluates what happens after the change, whether agents can trace which downstream conclusions are affected, which survive through independent support, and how alternative timelines would have unfolded. Our evaluation reveals substantial limitations across current architectures: even the strongest non-Oracle system reaches only 22.4% Accuracy, with retrieval and reasoning each surfacing as challenges.
[AI-136] ractable Query Answering under Epistemic Confidentiality Policies in DL Ontologies (extended version) ISWC2026
链接: https://arxiv.org/abs/2607.16715
作者: Lorenzo Marconi,Daniela Rieti,Riccardo ROsati
类目: Artificial Intelligence (cs.AI)
备注: Accepted at ISWC 2026
Abstract:We study Controlled Query Evaluation (CQE), a declarative approach to confidentiality-preserving data access, in the context of Description Logic (DL) ontologies, and for confidentiality policies expressed through Epistemic Dependencies (EDs). We first address the problem of answering queries (specifically, Boolean unions of conjunctive queries) under known semantics for CQE (GA- and IGA-entailment). Our results show that if the TBox is expressed in \textDL-Lite_\mathcalR , CQE is computationally intractable in general. Moreover, in the presence of EDs, the IGA semantics has recently been proven not to satisfy an important confidentiality preservation property known as indistinguishability. With the goal of defining computationally easier and confidentiality-preserving forms of CQE, we introduce a new semantics for CQE, based on the notion of minimal policy violation (MPV). We show that the new semantics provides a sound approximation of the previous ones, while satisfying the indistinguishability property. We also prove that, in the case of \textDL-Lite_\mathcalR ontologies, query entailment under the MPV semantics can be decided in polynomial time in data complexity. Finally, we present a software implementation of our framework that we used to evaluate the feasibility of this new approach using an existing benchmark for OWL 2 QL.
[AI-137] Building a Neural Network from Scratch: Implementation Evaluation and Optimization
链接: https://arxiv.org/abs/2607.16682
作者: Yuanzhe Jia
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The widespread adoption of high-level deep learning libraries, while accelerating model development, has increasingly abstracted away the internal mechanics of neural networks, creating a gap between practical usage and fundamental understanding. To address this, the paper presents a self-contained neural network framework implemented entirely from scratch – without relying on automatic differentiation or pre-built deep learning modules. The implementation encompasses all essential components, including multi-layer architectures, diverse activation functions, regularization techniques, and state-of-the-art optimizers. Beyond serving as a pedagogical instrument that demystifies forward/backward propagation, gradient dynamics, and optimization landscapes, the framework demonstrates robust performance when applied to a multi-class classification task, successfully validating its correctness, numerical stability, and generalization across varied configurations. The extensible design and clean modularity further position it as a reliable baseline for educational purposes and future research exploration.
[AI-138] Falsification-Based Verification of LLM -Generated Optimization Models: Sound Test Batteries and Their Detection Limits
链接: https://arxiv.org/abs/2607.16646
作者: Haifeng Li,Mo Hai
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Large language models now translate natural-language descriptions of decision problems into solver-ready optimization models, but they fail silently. A generated model often runs and still formulates the wrong problem. This paper develops a theory of falsification-based verification for this setting. Every numeric quantity in the description is a typed slot, and a candidate model is tested only through solver calls on slot-transformed instances; no reference model or label is consulted. From duality, comparative statics, and polyhedral limit arguments we derive a battery of test classes covering directions, curvature, crush probes, prohibitive limits, annihilation, and exchange. Every test is sound, so a violation certifies unfaithfulness and the false-positive rate is zero by design. We characterize what such verification can never see, give conditions under which the canonical error classes are detected with certainty, and prove that no fixed-threshold perturbation tester is simultaneously sound and nontrivial. Experiments on 326 ground-truth models from NL4OPT and four benchmark families confirm the theory. The battery attains a 0.0% false-positive rate against 54.9% for a threshold tester, detects 70.0% of certified conditional-class mutants, convicts 40.4% of the mutants invisible to execution-accuracy scoring, and reproduces the predicted detectability pattern including its zeros.
[AI-139] opoTuner: Topological Finetuning of Large Language Models
链接: https://arxiv.org/abs/2607.16637
作者: Abdulkadir Erol,Yash Mahajan,Vepaul Hariprashad,Baha Rababah,Santu Karmaker,Cuneyt G. Akcora,Mubarak Shah
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Full fine-tuning remains a strong way to adapt pretrained LLMs, but it updates all weights and can be expensive. LoRA reduces the number of trainable parameters, but it does not directly answer which pretrained components should be trained and which can be frozen during adaptation. We introduce TopoTuner, a topology-guided fine-tuning framework for selective freezing of attention projection matrices. \method treats each projection matrix as a row cloud and uses Wasserstein distances between persistence diagrams to measure how its topology changes during fine-tuning. TopoTuner learns a reusable freezing profile from a source dataset and transfers it to efficiently fine-tune models on out-of-domain datasets, evaluating whether task-specific topological drift generalizes across question answering and sentiment analysis tasks. Across LLaMA-3.1-8B, Mistral-7B-v0.3, and Qwen3-8B-Base, TopoTuner is competitive with full fine-tuning while training only 1-2% of the model parameters, and outperforms LoRA in 7 out of 9 model-dataset settings, which can change up to 39.57% of the projection parameters. Along with minimized updates, TopoTuner reduces training time by 20.4% relative to full fine-tuning and 5.5% relative to LoRA on average. TopoTuner opens a new direction for reusable freezing profiles, where fine-tuning behavior learned on one dataset can be shared across multiple tasks. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2607.16637 [cs.AI] (or arXiv:2607.16637v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.16637 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-140] CLOSER-Bench: Evaluating Budgeted Cross-Stage Design Closure for Hardware Agents
链接: https://arxiv.org/abs/2607.16632
作者: Peilong Zhou,Zhirong Chen,Cangyuan Li,Haoyu Gao,Kaiyan Chang,Ziming Qu,Ying Wang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 6 pages, 2 figures, 4 tables
Abstract:Hardware engineering exposes coding agents to a form of long-horizon work that is difficult to capture with pass-at-k: progress is continuous, tool feedback is delayed and heterogeneous, and a backend failure may require revising RTL rather than tuning another physical-design parameter. Existing benchmarks measure RTL generation, repository repair, verification, PPA evolution, or physical implementation, but their different designs and oracles make it hard to determine where an agent succeeds or fails across abstraction boundaries. We introduce CLOSER-Bench, a controlled evaluation protocol for budgeted cross-stage design closure. For one design and one hidden objective, it pairs spec-to-RTL, RTL-to-GDS, and spec-to-GDS tasks, records every simulator, synthesis, STA, and place-and-route invocation, and measures final quality, anytime progress, tool cost, and cross-stage recovery. The benchmark is built on open-source Verilator, Yosys, OpenROAD, KLayout, Sky130, and the Harbor agent harness. A ten-task pilot spanning RTL repair, mutation-based verification, coverage, PPA optimization, design-space exploration, cross-model debugging, and security establishes the executable harness and exposes a sharp completion–closure gap: three agents solve a localized AXI repair task, while the matched verification-closure task separates a frontier agent from two otherwise successful baselines. We further validate a full RTL-to-GDS flow and construct a macro-based AXI/DMA streaming accelerator for the stage-paired evaluation. These results motivate treating hardware closure as a budgeted sequential decision problem rather than a collection of independent code generation tasks.
[AI-141] A Research Prototype for Closed-Loop Generative Design of Customized Foot Orthoses via Semantic-Physics Alignment
链接: https://arxiv.org/abs/2607.16631
作者: Rui Wang,Byungwon Min,Suxing Liu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Translating unstructured clinical prescriptions into patient-specific foot orthoses (FOs) is hindered by a semantic-physical misalignment: high-level clinical intent is not mapped deterministically onto the 3D geometric parameters of the orthosis, and existing design workflows remain dependent on manual expertise with no instantaneous biomechanical validation. We present TANS-FO, a research prototype-a modular pipeline with closed-loop feedback for computational design automation of customized FOs, not a clinically validated therapeutic device. A Text-Aligned Neural Surrogate (TANS) uses cross-attention to project clinical-text embeddings onto a continuous lattice-density field, while a Graph Neural Network (GNN) surrogate predicts plantar stress in real time as a substitute for Finite Element Analysis (FEA). The framework is anchored on the open-access PicoFoot-5K anthropometric database (5,230 subjects; 30+ anatomical parameters). Under standardized quasi-static loading, the GNN surrogate agrees with an Abaqus reference solver (R^2 = 0.94), and the full pipeline synthesizes manufacturing-ready lattice insoles within minutes. On the Male 18-40 cohort, the proposed system attains a surrogate-predicted peak-pressure reduction of 34.7% over parametric CAD, with a fit error of 0.42 mm. Separately, an exploratory feasibility observation (n = 12; 2-week follow-up; no control group) using VAS pain reporting indicates short-term comfort improvement (VAS 6.4 - 2.1), but this data is explicitly classified as preliminary observational evidence only-not evidence of clinical efficacy.
[AI-142] Privacy Cost as Equity Input: A Group Fairness Criterion for Differentially Private Machine Learning
链接: https://arxiv.org/abs/2607.16620
作者: Rakshit Naidu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR); Computers and Society (cs.CY)
备注: AIES 2026 (accepted)
Abstract:Differential privacy (DP) is increasingly deployed to limit membership inference risk in machine-learning systems. Prior work has shown that DP-SGD can widen accuracy disparities across demographic groups, but this framing treats fairness as a purely outcome-side concern. We argue that privacy cost, the information leakage borne by each group, is itself a form of harm, and adopt a compensatory-fairness framework in which a group that involuntarily bears greater privacy exposure is owed proportionally greater benefit from the system. From this principle we derive the \emphPrivacy-Cost Equity Ratio (PCER), a group fairness metric defined as a group’s positive prediction rate normalized by its per-group overfitting gap. By a standard membership inference bound, this overfitting gap upper-bounds each group’s vulnerability to inference attacks, making PCER a conservative measure of benefit relative to exposure. PCER needs only per-group train and test accuracy (no shadow models), making it a practical post-hoc audit tool. We evaluate PCER alongside standard fairness metrics across six benchmark–attribute combinations spanning tabular and NLP domains, under DP-SGD at a range of privacy budgets, and validate the overfitting-gap proxy against a direct threshold membership-inference attack. The results reveal patterns that outcome-based metrics miss. On COMPAS, PCER uncovers a persistent double disadvantage: the protected group bears both greater privacy exposure and worse predictive outcomes, something demographic parity gap masks entirely. Sensitivity analysis shows very strong privacy guarantees collapse both groups’ overfitting to a numerical floor, rendering exposure-based audits uninformative in that regime. Together, these findings show that fairness audits of privacy-preserving systems must account for who bears the cost of protection, not only who benefits from its outcomes.
[AI-143] DataFlow-Harness: A Grounded Code-Agent Platform for Constructing Editable LLM Data Pipelines
链接: https://arxiv.org/abs/2607.16617
作者: Runming He,Zhen Hao Wong,Hao Liang,Zimo Meng,Chengyu Shen,Xiaochen Ma,Wentao Zhang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 13 pages, 2 figures, and 5 tables. Technical report
Abstract:Large language models (LLMs) are increasingly used to automate data-processing workflows, yet coding agents typically produce scripts that are not automatically materialized as persistent, editable platform artifacts. We call this disconnect the \textitNL2Pipeline gap. To bridge it, we introduce \textscDataFlow-Harness, a platform that guides an LLM agent to construct platform-native directed acyclic graphs (DAGs) through typed, incremental mutations rather than free-form scripts. The platform combines \textscDataFlow-Skills for procedural guidance, a Model Context Protocol (MCP) layer that exposes the live operator registry and current pipeline state, and \textscDataFlow-WebUI, which synchronizes conversational authoring with a visual DAG editor. On a 12-task data-engineering benchmark, \textscDataFlow-Harness achieves a 93.3% observed end-to-end pass rate. Relative to Vanilla Claude Code, it reduces measured monetary cost by 72.5% and generation latency by 49.9%; its observed pass rate is within 0.9 percentage points of the Context-Aware Claude Code baseline while its cost is 42.8% lower. Per-task analysis indicates that Skills are most useful when construction depends on implicit procedural knowledge. These results show that live platform grounding can produce persistent, editable workflow artifacts with an observed reliability close to script-generation baselines and with lower measured construction cost and latency.
[AI-144] Just A Rather Very Intelligent Spoken Agent
链接: https://arxiv.org/abs/2607.16610
作者: Chen Chen,Zhehuai Chen
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Long-horizon AI agents are becoming increasingly capable, yet their interaction with users remains surprisingly thin. In most workflows, users give an initial instruction, receive only selective textual updates, and lose a clear sense of what the agent is doing or when to step in. This leaves a missing part in the current agent ecosystem: an always-on Jarvis-style mediator that keeps the agent continuously reachable to the user. Such a mediator should support real-time spoken interaction with the user, answer questions without interrupting the worker, proactively report progress or confusion, and inject user guidance back into the agent’s execution when useful. In this work, we introduce JarvisBench, a benchmark for measuring the dual value of mediation in long-horizon agent workflows. JarvisBench contains two complementary tracks: an agent-collaboration track that measures whether mediation improves downstream task completion, and a user-interaction track that measures whether mediation makes ongoing execution more understandable, responsive, and accessible to users. We instantiate the benchmark with a modular reference Jarvis prototype and evaluate it on 34 text-only WildClaw tasks executed in OpenClaw. Preliminary results with GPT-5.5, Claude Opus 4.7, Gemini-based, and GPT-based worker agents suggest that Jarvis-style mediation can provide trace-grounded responses to user questions and improve task performance when sparse user guidance is injected at appropriate moments. The results also show that effectiveness depends strongly on the mediator’s LLM brain, highlighting both the promise of this missing middle layer and the need for broader community effort. Demo page this https URL
[AI-145] FST.ai 2.5: Explainable and Uncertainty-Aware AI for Olympic and Para-Taekwondo Decision Support Athlete Digital Twins and Federation-Scale Analytics
链接: https://arxiv.org/abs/2607.16597
作者: Keivan Shariatmadar,Ahmad Osman,Ramin Rey
类目: Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Probability (math.PR)
备注: 14 pages, 4 figures
Abstract:The rapid digitalisation of elite sport has created new opportunities for integrating artificial intelligence (AI), performance analytics, and decision-support systems into athlete development and competition management. However, existing solutions remain fragmented, typically addressing isolated tasks such as performance analysis, athlete monitoring, or referee support. This paper presents \textbfFST \cdot ai~2.5, an explainable, uncertainty-aware, and secure AI framework for Olympic and Para-Taekwondo. \textbfFST \cdot ai~2.5 introduces a unified digital ecosystem integrating athlete intelligence, competition analytics, federation-scale data management, AI-assisted decision support, athlete and event digital twins, explainable performance indicators, and adaptive training recommendations. The framework supports World Taekwondo (WT), Member National Associations (MNAs), coaches, referees, analysts, and athletes through transparent, secure, and federation-aware governance. By combining multi-source competition data, athlete-performance information, and contextual evidence, \textbfFST \cdot ai~2.5 provides tactical diagnostics, longitudinal athlete monitoring, performance forecasting, personalised development planning, and federation-wide benchmarking using explainable and uncertainty-aware AI. Prototype deployments demonstrate the feasibility of the proposed framework. Although developed for Olympic and Para-Taekwondo, the methodology is broadly applicable to explainable AI, digital twins, and trustworthy decision support in combat sports and other high-performance sporting environments. Comments: 14 pages, 4 figures Subjects: Artificial Intelligence (cs.AI); Optimization and Control (math.OC); Probability (math.PR) MSC classes: 68T45, 68T37, 68T10, 68T07, 68T08, ACMclasses: I.2; I.4; I.5; I.6 Cite as: arXiv:2607.16597 [cs.AI] (or arXiv:2607.16597v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.16597 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-146] Learning from World Feedback: Why Model Uncertainty Fails as a Risk Signal in Model-Based RL ICML2026
链接: https://arxiv.org/abs/2607.16591
作者: Zhaohui Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted at the ICML 2026 Workshop on Reinforcement Learning from X Feedback (RLxF). 14 pages
Abstract:The RLxF programme argues that learning signals should come from world feedback rather than from internal model proxies. We instantiate this position in safe model-based control and distil it into three concrete design principles. Empirically, across four world-model architectures spanning a 2x MSE range, MPC planning is statistically equivalent (TOST, n=200), and dynamics-based uncertainty penalties increase collision rates from 26% to 34%: the standard MBRL safety proxy is anti-correlated with safety in this regime. Replacing the model-internal proxy with three world-feedback signals (a sensor-derived margin via minimum lidar, a temporal signal via time-to-collision, and an outcome-supervised feedback model g_psi trained on prior collision labels, structurally analogous to outcome-trained reward models in RLHF) reduces collisions to 1-14% without retraining the world model or the planner. The mechanism is structural: model uncertainty has support over state-prediction space, whereas task risk has support over constraint boundaries, with empirical correlation r 0.15. From this we extract three RLxF principles (ground risk in world outcomes, validate proxies before deployment, and substitute outcome-trained feedback models when direct world signals are unavailable) and argue they apply equally to model-based control and to verifier-based or RLHF approaches in LLM alignment.
[AI-147] Exact Network Surgery: Functional Invariance and Gradient Plasticity in Reactive Computational Graphs
链接: https://arxiv.org/abs/2607.16568
作者: Abdallah Khemais(ISITCOM, University of Sousse)
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Programming Languages (cs.PL)
备注: Companion paper: “Cost Accounting for Reactive Computational Graphs” (submitted concurrently)
Abstract:Function-preserving network growth techniques such as Net2Net and progressive stacking expand a model’s capacity without destroying its learned function, but existing formulations either tolerate numerical perturbations or require a full rebuild of the training program. We formalize Exact Network Surgery: the in-place insertion of a residual block into a live computational graph such that (i) the network function is preserved – bit-exactly under explicit floating-point hypotheses – and (ii) inserted parameters remain trainable immediately after insertion. We prove an identity-morphism theorem for gated residual blocks, a structural-locality theorem showing that a reactive invalidation engine recomputes exactly the downstream cone of the insertion point, leaving every other node’s value and optimizer state untouched, and an escape-from-initialization proposition showing that the Gradient Shadowing gate alpha, initialized at zero over a randomly initialized branch, receives a generically non-zero gradient at insertion time. We identify a degenerate configuration – zero-initialized output projections combined with a zero gate – that is an exact saddle point gradient descent cannot escape. Every claim is validated on the reference implementation in NeuroDSL, a reactive graph engine in Julia: grafting is bit-exact on every logit tested (0 mismatches out of 1600); the gate escapes zero at the first optimizer step and unlocks branch gradients at the second, exactly as predicted; the degenerate configuration exhibits gradients identically zero for the entire 600-step run; surgery cost tracks downstream cone size with r = 0.9992 while graft-plus-invalidation bookkeeping is constant (about 0.75 ms) across insertion depths; and training resumes bit-identically across a real process restart. A flagged preliminary appendix reports first single-seed observations on post-insertion gate dynamics.
[AI-148] ReqGenX: An Empirical Study of Atomic Decomposition Artifact Regeneration and Reconstruction for Legacy SRS Documents
链接: https://arxiv.org/abs/2607.16564
作者: Ragib Shahariar Ayon,Rayed Fahmi,Sumon Biswas,Shibbir Ahmed
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: Accepted at the International Symposium on Empirical Software Engineering and Measurement (ESEM) technical papers (main) track 2026
Abstract:Background: Evaluating automated Software Requirements Specification (SRS) generation is challenging because few datasets provide fine-grained traceability between source requirements, intermediate elicitation artifacts, and generated specifications. Aims: We aim to study whether legacy SRS documents can be transformed into traceable synthetic pre-SRS artifacts that support fine-grained evaluation of LLM-based SRS generation. Method: We conduct an empirical study using ReqGenX, a controlled pipeline that decomposes SRS sections into source-grounded atomic statements, routes atoms to standards-inspired artifact types through multi-LLM plurality voting, and generates artifacts using constrained prompts with iterative judge-guided refinement. We evaluate ReqGenX on seven PURE SRS documents using grounding, quality, information retention, and downstream reconstruction analyses. Results: ReqGenX produces faithful and usable atoms, with median AlignScore values typically between 0.96 and 0.99 and Prometheus scores ranging from 4.34 to 4.85. Generated artifacts remain strongly grounded in their source atoms, with AlignScore values typically between 0.80–0.94 and judge pass rates near 100%; stricter Prometheus evaluation yields pass rates from 54.8% to 97.1%. In a downstream SRS reconstruction case study, artifact-backed atoms remain recoverable from generated SRSs, with SBERT means between 0.69 and 0.75 and AlignScore medians between 0.76 and 0.84. Conclusions: Traceable synthetic pre-SRS artifacts can support more fine-grained evaluation of LLM-based SRS generation, while exposing tradeoffs among faithfulness, information retention, and artifact completeness.
[AI-149] Mitigating Compiler Fusion-Induced Power Bursts in Mobile NPU Inference as the Battery Depletes
链接: https://arxiv.org/abs/2607.16555
作者: Ryoga Yuzawa,Masayoshi Tomizuka
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: 16 pages, 6 figures
Abstract:Mobile devices increasingly rely on real-time NPU inference for camera and perception workloads. Under low-voltage conditions, however, a single inference can induce an instantaneous voltage droop in the power-delivery network, causing the power management integrated circuit to invoke dynamic voltage and frequency scaling (DVFS) and increase latency. We present a measurement study of this effect on a commercial smartphone. We show that aggressive operator fusion in a mobile NPU compiler can create monolithic superlayers whose concentrated execution produces large peak-current bursts. These bursts shift the DVFS-onset voltage upward and reduce the low-voltage operating margin. We further evaluate a practical black-box mitigation: a measurement-guided, pre-compilation graph rewrite that inserts barriers at selected peak-to-average power ratio hot spots to prevent harmful superlayer merging in the vendor NPU compiler. On Snapdragon 8 Gen 3 with MobileNetV4 at 768 x 768 resolution on ImageNet-1k, this method reduces peak current from 3.12 A to 1.94 A with 3.76% latency overhead, preserves stable latency deeper into the low-voltage regime, and shifts the inferred DVFS margin by approximately 173 mV.
[AI-150] Capacity and Redundancy Trade-offs in Multi-Task Learning UAI
链接: https://arxiv.org/abs/2607.16554
作者: Asif Khan
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Information Theory (cs.IT)
备注: Accepted in 42nd Conference on Uncertainty in Artificial Intelligence (UAI) 2026
Abstract:In multi-task learning (MTL) negative transfer is often considered as an optimization artifact, but it can also be viewed as a consequence of limited shared capacity and weak task redundancy. We investigate this effect through a Capacity–Redundancy (CR) identity that decomposes the sum of per-task predictive informations into joint predictive information that includes label redundancy defined via total correlation (TC), and a residual coupling term that quantifies interference left unresolved by the shared representation. Additionally, we show two key results: (i) a clustering-gap decomposition that gives a necessary and sufficient condition for clustered sharing to outperform global sharing, and (ii) a gradient–TC bridge in a Gaussian multi-task model that formally justifies gradient cosine similarity as a proxy for redundancy ordering. Empirically, we estimate the residual coupling \Delta from validation residual correlations, showing that clustered LoRA substantially reduces \widehat\Delta , outperforms size-matched random partitions, and results in statistically significant gains with multi-seed confidence intervals.
[AI-151] Building2Building: A Large Scale Benchmark for Generalizable Real-World Reinforcement Learning
链接: https://arxiv.org/abs/2607.16534
作者: Vincent Taboga,Justin Veilleux,Doseok Jang,Anushree Rankawat,Pierre-Luc Bacon
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 22 pages, 7 figures, to be published at RLC 2026
Abstract:Reinforcement learning (RL) has achieved strong results in control, yet learned policies remain brittle to changes in dynamics, action spaces, observation spaces, or goals, a critical limitation for real-world deployment. Existing benchmarks offer limited diversity and complexity, making it difficult to rigorously study transfer, multi-task learning, and meta-learning in RL. We introduce Building2Building (B2B), a large-scale suite of realistic Heating, Ventilation, and Air Conditioning (HVAC) control environments built on EnergyPlus, a state-of-the-art building simulator. B2B is fully compatible with the Gymnasium interface and features a parametric building generator, enabling the systematic generation of diverse building configurations with heterogeneous observation and action spaces. Based on this suite, we define benchmark tasks targeting key open challenges in RL, including goal adaptation, dynamics adaptation, action-space shifts, and cross-domain transfer. By providing a large-scale, diverse, and physically grounded testbed with standardized evaluation protocols, B2B enables systematic investigation of generalization and transfer in continuous control. Beyond advancing research on generalization in RL, this new benchmark also carries significant societal implications by enabling improved HVAC control at scale, one of the most energy-intensive systems in buildings.
[AI-152] Nonuniformity Principle in Human-AI Coworking
链接: https://arxiv.org/abs/2607.16530
作者: An Luo,Jie Ding
类目: Artificial Intelligence (cs.AI); Methodology (stat.ME)
备注:
Abstract:As generative AI is increasingly applied to automate multi-step and high-stake workflows, human judgment and involvement remain essential for ensuring the quality of AI-generated outputs. In practice, while it is desirable for human experts to provide oversight on AI regularly, often by reviewing intermediate outputs, giving feedback, making corrections, and steering subsequent steps, such oversight is constrained by the time and resources that humans can afford. This creates a tension between the need for human oversight and AI’s efficiency in delivering more output with less intervention. An important but underexplored question, then, is how to optimally engage humans in human-AI coworking. This work was originally motivated by our empirical observation that in long AI workflows, human oversight often improves user satisfaction while reducing unnecessary rework and token consumption. From there, we formulate the problem of where to place oversight stages in human-AI coworking. Under reasonable assumptions, we then develop the nonuniformity principle, which states that the optimal schedule places oversight stages with non-decreasing gaps along the workflow. We empirically validate this principle in two common AI agent workflows: writing literature reviews and constructing websites.
[AI-153] SEER: Supervised Learning to Control Energetic Reasoning
链接: https://arxiv.org/abs/2607.16523
作者: Sascha Van Cauwelaert,Michele Lombardi,Pierre Schaus
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:One of the main strengths of Constraint Programming is the ability to reduce the search space via propagation. However, propagation is a double-edged sword, with more pruning power coming at the price of larger computation time. For each problem constraint, the best propagator depends on the specific instance and may change at search time. In the literature, Machine Learning (ML) techniques and activity-based heuristics have been applied respectively for choosing (statically) the propagators for a batch of problems and to adapt (dynamically) the propagation strength. We propose to merge those efforts by using an oracle function, obtained via ML, to decide whether to run complex propagators for a target constraint. A combination of design choices makes the approach flexible and easy to embed in state-of-the-art solvers. In this paper, we focus on investigating the feasibility of building an oracle for the Energetic Reasoning propagator. Our experiments show that high prediction accuracy can be obtained, provide suggestions for classification features, and highlight important issues to address when building such an oracle.
[AI-154] K-IPO: Kendall-constrained Importance Preserving Oversampling for Imbalanced Tabular Data
链接: https://arxiv.org/abs/2607.16478
作者: Marios Tyrovolas,Argiris Sofotasios,Dimitris Metaxakis,Georgios Mermigkis,George Georgoulas,Panagiotis Hadjidoukas,Chrysostomos Stylios
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 24 pages, 5 figures, 9 tables
Abstract:Oversampling is widely used to address class imbalance in tabular classification, but existing methods can distort the feature importance ranking underlying model explanations. Although recent studies have quantified this distortion by comparing real and synthetic data, none have actively sought to prevent it. In this paper, we introduce Kendall-constrained Importance-Preserving Oversampling (K-IPO), a generator-agnostic, “generate-then-select” framework that preserves the original data’s feature importance ranking during augmentation. K-IPO iteratively generates minority-class candidates and accepts them only if their inclusion maintains a user-defined minimum Kendall’s tau (\tau) correlation with the reference ranking. Optionally, stricter constraints can be applied to the highest-ranked features. We evaluated K-IPO on 20 imbalanced binary classification datasets using three classifiers and multiple explanation methods. In most cases, K-IPO achieved the best or tied-best results in feature importance preservation, explanation consistency, and class separability. It also generally improved predictive performance while maintaining competitive computational overhead.
[AI-155] When to Use Which? Benchmarking Optimisers for Configurable Systems under Varying Budgets
链接: https://arxiv.org/abs/2607.16476
作者: Chao Jiang,Yulong Ye,Tao Chen,Miqing Li
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:Software configuration tuning is crucial for optimising system performance, and various optimisers have emerged over the last decade. Yet, the time required during the tuning process may vary across systems. In some systems (e.g., PostgreSQL), it may take a few minutes to measure a configuration, whereas in some others (e.g., MariaDB), it can take several hours. Moreover, even within the same system, users may have varying budgets and preferred settings. This naturally raises a question – Given a budget level, which optimiser is the best choice for SE practitioners? This matters because optimisers usually have their own ``comfort zone’’ and may perform very differently under distinct budgets. In this paper, we aim to answer this question. We systematically evaluate eight well-established optimisers across 22 configurable systems under varying budget levels. We find that, unsurprisingly, model-based optimisers (e.g., SMAC) are well-suited under tight budgets, and model-free optimisers (e.g., GAs) become superior with more generous budgets. However, interestingly, there is one optimiser, FLASH, that performs consistently well on most systems regardless of budgets. We lastly investigate the reasons behind this phenomenon and find that many systems possess good local optima (with large basins of attraction), allowing greedy optimisers (e.g., FLASH) to achieve strong performance. Source code, data, and supplementary materials of this work are available at this https URL. Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.16476 [cs.SE] (or arXiv:2607.16476v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2607.16476 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[AI-156] Berkeley and Heiserman as an Unexhausted Architecture for Embodied Machine Intelligence
链接: https://arxiv.org/abs/2607.16465
作者: Christopher A. Tucker
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Edmund C. Berkeley is usually remembered as a writer who helped connect symbolic logic to computing machinery. That description is correct, but incomplete. Read across Berkeley’s machine-oriented writings and projects, the central concern is broader: to show that symbolic logic can serve as a practical design language for machines that acquire information, retain it, respond appropriately to changing conditions, and organize their behavior over time. This paper takes Symbolic Logic and Intelligent Machines as the principal text in that effort, while treating it as representative of a larger Berkeley program that links logical form, circuitry, control, and intelligent behavior. On this reading, Berkeley does not merely treat intelligence as abstract symbol manipulation. He repeatedly defines intelligent machines in operational terms, ties logic to hardware realization, and describes machine behavior through the coordinated interaction of inputs, outputs, memory, calculation, control, states, and events. His treatment of robots and machine activities accordingly moves beyond static logical form toward temporally extended, environment-coupled behavior. David L. Heiserman’s machine-intelligence work extends this program toward adaptive creature architectures built around memory, confidence, and generalization. Taken together, Berkeley and Heiserman can be read not as exhausted historical episodes but as contributors to a still under-tested architectural approach to embodied robotic cognition.
[AI-157] Retrieval is Enough: Training-Free Interpretability with a Tool-Using Agent
链接: https://arxiv.org/abs/2607.16448
作者: Sriram Balasubramanian,Soheil Feizi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 25 pages, 7 figures
Abstract:Interpretability methods for neural network activations span a wide cost spectrum, from cheap, training-free techniques (such as linear probes, PCA, SVD) to more expensive training-based ones (such as SAEs and activation oracles). Training-based methods are typically more powerful, in part because they leverage large activation datasets during training. This raises a natural question - do they actually surface insights that go beyond what is recoverable from the training dataset itself? To address this, we equip an LLM agent with a vector database of activations paired with their textual contexts, along with tools for manipulating activations - projecting out directions in latent space, computing activation differences and averages. The agent iteratively queries the database, forms hypotheses from the retrieved samples, and validates them by constructing linear probes. We call this method HARP, for Hypothesis-driven Agentic Retrieval and Probing. Despite not involving any training, HARP outperforms both activation oracles and SAE-based agents on concept discovery, concept detection, model steering, and secret elicitation. The training-free design also makes HARP substantially cheaper and more flexible: new datasets can be indexed on demand whenever existing ones prove insufficient. More broadly, our results suggest that current training-based methods do not yet extract insights beyond their training data, and motivate benchmarks that explicitly require interpretability methods to demonstrate such insights. We release our code at this https URL
[AI-158] When to Plan: Learning to Select Between Reactive Control and Deliberative Planning
链接: https://arxiv.org/abs/2607.16421
作者: Adam Labiosa,Josiah P. Hanna
类目: Artificial Intelligence (cs.AI)
备注: RLC 2026
Abstract:It has long been recognized that humans have the ability to switch between fast, reactive decision-making and slower, deliberative planning. In this paper, we study the question of how to learn this ability, known as meta-reasoning, in artificial agents. We model reactive decision-making as a policy that directly maps state observations to actions. Such policies can be trained with reinforcement learning (RL) or imitation learning, but may generalize poorly outside of their training distribution. Alternatively, model-based decision-time planning is more likely to produce good actions across a broader set of states but requires additional computation time, which delays acting. In this work, we introduce an RL method for training a meta-reasoning policy that allocates computation by conditioning on a reactive-policy uncertainty score. This score enables it to predict when the reactive policy is likely to perform poorly and when planning is needed. We conduct an empirical study on motion planning and navigation environments, showing that this design enables the meta-reasoning policy to learn when the reactive policy provides a good-enough action versus when decision-time planning is needed. Additionally, we show that our design enables the meta-agent to shift toward fully reactive control as the reactive policy improves.
[AI-159] Signal-based Model Access Risk Analysis for AI System Operations Security
链接: https://arxiv.org/abs/2607.16414
作者: Maria Mahbub,Steven Young,Amir Sadovnik,Edmon Begoli,Chris Rugenstein,Donald Coulter,Anthony Ayodele
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Artificial intelligence (AI) systems are now ubiquitous across domains such as security, finance, healthcare, consumer technology, and large-scale cloud services, where they process massive volumes of data and make consequential decisions daily. This widespread adoption has created a broad attack surface through which adversaries can manipulate, evade, extract information from, or otherwise subvert deployed models. Depending on system design and exposure, attackers may have very different forms of access: some observe only final decisions, while others receive confidence scores, intermediate representations, or even full model parameters. While previous surveys typically organize evasion attacks into white-box, gray-box, and black-box categories based on the attacker’s knowledge of model internals (architecture, parameters, gradients), this taxonomy often conflates different deployment scenarios that provide vastly different output signals, all labeled as ``black-box’’ despite enabling fundamentally different attack strategies. Understanding how evasion attack strategies adapt to the specific information signals returned by deployed systems is critical for organizations making procurement and deployment decisions. To address this gap, we introduce the Signal-based Model Access Risk Taxonomy (SMART), a deployment-oriented framework that classifies attacker access according to the nature and richness of the information signals available from deployed AI systems. Using this taxonomy, we provide a structured overview of evasion attacks across progressively richer levels of information exposure, highlighting how deployment interfaces influence attack capabilities and informing more secure AI deployment and procurement decisions.
[AI-160] Interactive Task Alignment as a POMDP
链接: https://arxiv.org/abs/2607.16412
作者: Andy Dai,Zexue He,Zhenyu Zhang,Alex Pentland,Jiaxin Pei
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Current benchmarks for language models primarily evaluate execution on fully specified tasks. However, real user tasks are often ambiguous. Users arrive with incomplete, exploratory, or even inconsistent goals, requiring the assistant to first determine the intended task before carrying it out. We study this problem as task alignment: the ability to align with a user on their intended task. We introduce a general framework for converting specified tasks into underspecified interactions, formalized as a POMDP in which the model must infer a latent task from partial and evolving user intent. We validate our user simulator post hoc with a human user study. Across shopping, coding, and professional work settings, we find that while models often perform well once the task is specified, models still struggle with task alignment: current models act prematurely, interact ineffectively, and fail to resolve ambiguous requests. Models on average recover the user’s intended task only 22-32% of the time under ambiguity. In a human study in the same setting, humans reach 48%, outperforming all evaluated models. We show that post-training with supervised fine-tuning and reinforcement learning improves task alignment, but models still lag behind humans in resolving uncertainty through interaction. Together, our results suggest that current models still lack key interaction abilities required for reliable agency.
[AI-161] Fantastic Adaptive Taxonomies and How to Use Them
链接: https://arxiv.org/abs/2607.16387
作者: Mert Cemri,Andrei Cojocaru,Melissa Pan,Shu Liu,Shubham Agarwal,Alexander Krentsel,Jay Tang,Kannan Ramchandran,Joseph E. Gonzalez,Matei Zaharia,Alex Dimakis,Ion Stoica
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:
Abstract:An agent system’s execution traces record how it fails, and procedures that improve such a system without changing model weights (trajectory selection, prompt and workflow optimization, runtime monitoring) read these traces for feedback. Yet raw traces are a poor medium for accumulating that feedback: long, instance-specific, and lacking a stable vocabulary for recurring failures. We argue that an agent system should instead maintain an explicit representation of how it fails, induced from its own behavior and reusable wherever failure feedback is needed. AdaMAST builds this representation by converting a target system’s traces into a compact, evidence-grounded failure taxonomy: named failure codes organized along three fixed axes (system-level, role-specific, and domain-specific), with every name, definition, and evidence pattern induced from the traces; no code is hand-authored, no trace human-annotated. The taxonomy is not merely a post-hoc diagnostic but a shared feedback interface, improving agents in three ways. In agent-system search, taxonomy-coded diagnoses of failed candidates outperform free-form reflection on all five benchmarks we test. At runtime, taxonomy feedback raises SWE-agent’s resolution on SWE-bench Verified Mini from 60% with free-text reflection to 70%, and improves Claude Code from 64.0% to 70.7% as a runtime skill. In trajectory selection, AdaMAST-Judge, a verifier built on the induced codes, improves best-of-5 accuracy on Terminal-Bench 2.0 by 8-15 points over Pass@1. The vocabulary itself is compact (an order-of-magnitude compression that preserves trace distinctions), human-faithful (matching expert failure annotations more closely than a hand-crafted reference vocabulary), and adaptive (taxonomies induced for different domains share few codes). Adaptive failure taxonomies close the loop between the traces agents produce and the procedures that improve them.
[AI-162] AoA: Theorem Proving Agent over Abstract Syntax Tree of Redesigned Language
链接: https://arxiv.org/abs/2607.16372
作者: Qiyuan Xu,Joshua Ong Jun Leang,Renxi Wang,Wenda Li,Haonan Li,Luke Ong,Conrad Watt
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Programming Languages (cs.PL)
备注: 13 pages
Abstract:Interactive theorem proving (ITP) underpins program verification and formalized mathematics, but its manual effort limits scalability. LLM-based proof agents promise to ease this effort, but their heavy token consumption and API cost remain a major obstacle. We trace this cost to a shared root: current agents operate on serialized concrete syntax, emitting proofs as source text and recovering proof states through separate, line-number-based queries, so every edit shifts later lines and forces repeated relocation of errors and states. This same dependence on concrete syntax also blocks adoption of Minilang, a recent proof language that reaches SOTA on LLM-based proving but is too new for LLMs’ training corpora. We address both problems by lifting the agent off source text and onto the abstract syntax tree (AST): the model supplies proofs as JSON representations of Minilang’s AST – native to tool-calling LLMs – and drives the prover through a tree-edit model that fuses proof operations and states into one proof tree, so each operation carries its own subgoal’s state, readable directly off the tree. We realize this design in \emphAgent over AST (AoA). Against Amazon’s Isabelle Agent on miniF2F and NTP4VC-Pearl common success sets, AoA cuts API cost by 2.3–4.7x (normalized input-cache accounting), uses 2.9–6.9x fewer tokens and 3.9–8.9x fewer tool calls, and finishes 1.4–2.0x faster – while also solving far more problems on the harder verification benchmark.
[AI-163] PRISM: Multimodal Terrain Mapping for Rover Navigation in Unstructured Environments
链接: https://arxiv.org/abs/2607.16366
作者: Raul Castilla-Arquillo,Carlos Perez-del-Pulgar,Levin Gerdes,Alfonso Garcia-Cerezo,Miguel A. Olivares-Mendez
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:
Abstract:Robotic navigation in unstructured environments requires robust situational awareness to safely traverse hazards such as steep slopes and rocky terrain. To address this challenge, perception systems increasingly rely on multimodal sensor fusion. Specifically, integrating thermal imagery with standard optical and depth sensors enhances terrain differentiation, directly improving the reliability of mapping algorithms. This paper presents PRISM, a multimodal perception system for terrain mapping in unstructured settings. PRISM leverages a custom sensor suite to capture aligned RGB, depth, and thermal (RGB-D-T) imagery. At its core is OmniUnet, a novel vision transformer-based network specifically designed for multimodal semantic terrain segmentation. We validated the proposed system using two newly annotated datasets (BASEPROD and LAENTIEC) and demonstrate its real-world applicability through physical field experiments. Deployed on a resource-constrained embedded computer, PRISM efficiently generates traversability maps that directly enable autonomous navigation via a rover’s Guidance, Navigation, and Control (GNC) subsystem.
[AI-164] Reliable Remediation Impact Prediction for Black-Box Security Ratings
链接: https://arxiv.org/abs/2607.16357
作者: Nada Hanad,Mehdi Acheli,Ali NourEldin,Mohamed Sellami,Walid Gaaloul
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:
Abstract:Security rating platforms summarize externally observable cyber exposure and are expected to help organizations prioritize remediation. A platform may want to tell an organization how a candidate remediation action would affect its score, but repeatedly exposing exact score responses can reveal information about the hidden scoring engine. We propose a surrogate based approach for remediation score impact prediction that is designed to respect this opacity constraint. The surrogate predicts scores from organization configurations while explicitly representing checkpoint (i.e., a security check) applicability and the observed checkpoint set. A main challenge is that such predictions are not uniformly reliable: they depend on the amount and structure of the observable checkpoint evidence available for a given configuration. To address this, the approach combines applicability-aware surrogate construction, sensitivity analysis under controlled checkpoint restriction, a reliability layer for identifying unstable predictions, and score-impact prediction for supported remediation actions. Explicit modeling of checkpoint applicability is central throughout: it improves score prediction and provides the feature basis used by the reliability layer to identify unstable cases. We evaluate the approach on a real-world dataset of 5,188 organization configurations from a commercial security rating platform. The results show that the applicability-aware surrogate improves score prediction over simpler feature representations. For remediation, the surrogate predicts the score impact of supported actions, while the reliability layer helps identify cases in which these predicted impacts should be interpreted cautiously.
[AI-165] A Predict-then-Correct Loop Based on Few-Shot Continuous Contextual Bandit for Demand Forecasting
链接: https://arxiv.org/abs/2607.16354
作者: Zhiwei Lei,Benedict Jun Ma,Ilya Jackson
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Retail demand forecasting remains difficult when demand shifts faster than static forecasting models can be retrained, especially in early demand cycles where newly observed labels are sparse. To address this, this study aims to improve adaptive retail forecasting by proposing a predict-then-correct (PtC) framework that retains a first-stage machine learning (ML) forecast and applies a few-shot continuous contextual bandit correction policy with similar-SKUs augmentation and top-p masked updating. Across Walmart retail data and an exclusive beverage dataset, PtC delivers statistically significant reductions in MAPE, MAE, and RMSE across stable high volume, stable low volume, and erratic intermittent demand patterns, improves average RMSE by 9.52% over the ML-only baseline in the ablation study, and yields lower inventory costs than base-stock, proximal policy optimization, and soft actor-critic policies under the tested lead-time settings. These findings show that online forecast correction can bridge offline demand learning and real-time retail decision-making by adapting to sparse feedback without fully retraining the base forecasting model.
[AI-166] Boundary-Seeking GAN-Augmented TabTransformer for Adversarially Robust Intrusion Detection
链接: https://arxiv.org/abs/2607.16348
作者: Raihan Sultan Pasha Basuki,Aliyah Kurniasih
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Preprint. Submitted to Kuwait Journal of Science
Abstract:Machine learning-based intrusion detection systems (IDSs) often suffer from class imbalance and vulnerability to adversarial attacks, leading to degraded detection performance and reduced robustness. This study proposes a TabTransformer framework augmented by the Boundary-Seeking Generative Adversarial Network (BGAN) for flow-based intrusion detection using the CICIDS2017 dataset. BGAN serves a dual purpose by generating synthetic minority-class samples to mitigate data imbalance and producing adversarial samples to evaluate model robustness. Experimental results demonstrate that BGAN augmentation improves TabTransformer’s Macro-F1 score from 82.96% to 86.50%, with the largest class-wise improvement observed for Web_Attack (F1 score: 0.29 to 0.61). Robustness evaluation shows that all non-augmented models experienced a 100% Performance Drop Rate (PDR) under adversarial testing, whereas all BGAN-augmented models achieved negative PDR values, indicating improved resilience. Furthermore, the augmented TabTransformer maintained stable and low False Triggered Rate (FTR) values (1.51%-2.92%) across all noise levels, compared with the BGAN-augmented Decision Tree, which reached 49.09% under benign perturbations. These findings demonstrate that BGAN consistently enhances both class balance and adversarial robustness, while the proposed BGAN-TabTransformer framework provides an effective and adaptive intrusion detection solution for adversarial network environments.
[AI-167] AEVAL: From Anecdotal to Deterministic Testing for Agent ic Skill Workflows
链接: https://arxiv.org/abs/2607.16345
作者: Tejas Singh Anand,Yuet Ying Christina Wang,Wanting Jiang,Steve Masson,Tian Zheng,Bingjie Zhou
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Performance (cs.PF)
备注: 8 pages, 1 figure, 1 table
Abstract:Modern agentic systems increasingly rely on skills: installable packages of natural language and code that teach an LLM agent to perform a domain task. As skill repositories grow, developers need automated quality signals on every change, yet evaluation today is largely anecdotal: a developer asks an agent to “try the skill,” watches a demo, and forms a subjective impression. This yields neither reproducibility across runs nor comparability across versions, and scales poorly to marketplaces where one regression can silently break dozens of downstream workflows. We present AEVAL (Agentic Evaluation), a CI-integrated framework that replaces this practice with a deterministic, reproducible test pipeline for agentic skills. Every skill change triggers a test event: the skill runs against a developer-declared evaluation contract (this http URL) inside an automated executor, emitting a structured, evidence-grounded quality signal that downstream CI can route on. A key ingredient is a structural separation between executor and grader, preventing a subtle but pervasive failure mode: an agent that silently self-corrects during execution and then grades its own patched outputs as passing. Our contributions are: (i) a deterministic, change-triggered evaluation protocol with per-skill contracts and per-run artifact schemas; (ii) a formalization of self-correction bias as a distinct failure mode of naive agentic evaluators; (iii) an executor/grader separation with a first-attempt grading rule and explicit self-correction tracking; and (iv) a tiered, grounded-evidence fix-suggestion scheme (LV1 causal, LV2 quality) posted as inline merge-request comments. Validated on real skills in a production agentic stack across multiple agent SDKs, AEVAL converts spurious 100% pass rates into reproducible first-attempt fail signals with an auditable record of every executor fix. Comments: 8 pages, 1 figure, 1 table Subjects: Software Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Performance (cs.PF) Cite as: arXiv:2607.16345 [cs.SE] (or arXiv:2607.16345v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2607.16345 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Tejas Singh Anand [view email] [v1] Thu, 16 Jul 2026 21:33:05 UTC (109 KB)
[AI-168] LaCache: Exact Caching and Precision-Adaptive Inference for Diffusion Large Language Models
链接: https://arxiv.org/abs/2607.16339
作者: Xingru Chen,Zelang Liang,Yongjia Ma,Jiqing Zhan,Shuling Yang,Lian Wen,Kun Zhan
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Diffusion-based Large Language Models(DLLMs) enable parallel generation via Semi-Autoregressive (SAR) decoding in text generation. However, current methods suffer from severe operator-level redundancy: they recompute the entire sequence during denoising steps, ignoring that the prefix and masked suffix remain invariant within a block. We propose LaCache, a training-free acceleration framework that alleviates this redundancy through lossless caching and mixed precision. Specifically, LaCache employs Lossless State Memoization (LSM) by caching three types of intermediate results: (i) EmbedCache for embedding outputs, (ii) RoPECache for token-wise pre-attention states, and (iii) FACache for the online softmax statistics within FlashAttention. These caches allow the model to skip redundant computation on unchanged tokens without altering the output. To further alleviate memory-bandwidth bottlenecks, LaCache inegrates a per-group FP8 quantization strategy for FFN layers, tailored to step-dependent activation distributions across the diffusion process. Experiments demonstrate that LaCache alone achieves approximately 1.3X end-to-end speedup over vanilla DLLM. When combined with existing acceleration methods, LaCache reaches up to 40.2X end-to-end speedup while maintaining comparable task accuracy.
[AI-169] Autonomous mechanistic discovery of colorectal cancer vulnerabilities via multi-scale AI swarms
链接: https://arxiv.org/abs/2607.16262
作者: Christopher Baker,Tianyu Ren,Karen Rafferty,Hui Wang,Simon McDade
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The acceleration of automated scientific discovery has been fundamentally bottlenecked by the epistemic gap between the semantic reasoning of large language models (LLMs) and the deterministic physics of mammalian biology. While recent multi-agent frameworks have achieved autonomous hypothesis generation and in vitro experimental analysis, they lack the mathematically grounded, causal constraints required for multi-scale clinical translation. Furthermore, while algorithmic clinical digital twins successfully forecast biological states, they rely on black-box latent spaces, sacrificing mechanistic interpretability for predictive accuracy. Here, we introduce the Multi-Scale Autonomous Discovery Engine (Octopus), a neuro-symbolic architecture that unites zero-leakage, local LLM swarms with strict algorithmic physics engines. Rather than stopping at isolated cellular assays, the system autonomously generated therapeutic hypotheses against in vitro CRISPR dependency data (CCLE), traced dynamic causal cascades using mechanistic interpretability (XGBoost SHAP vectors), and orthogonally translated the emergent vulnerabilities in silico to predict in vivo mammalian tumor trajectory (PDX) and human overall survival (Marisa). In a fully unsupervised sweep of colorectal cancer transcriptomes, the pipeline autonomously identified Insulin-like Growth Factor 2 (IGF2) as a strictly bounded vulnerability to 5-Fluorouracil resistance. The discovery maintained significance after rigorous Benjamini-Hochberg false discovery rate correction (q=0.0292, Log-Rank p=0.0007 ) and successfully predicted significant in vivo tumor volume shrinkage in an independent mouse cohort (Mann-Whitney p=0.0373). By bridging the chasm between multi-agent reasoning and mathematically bounded clinical survival, this framework establishes a verifiable, zero-leakage paradigm for automated, end-to-end biomedical discovery.
[AI-170] Reducing Per-Sample Harm in Stochastic Optimization
链接: https://arxiv.org/abs/2607.16261
作者: Apostolos Avranas
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML)
备注:
Abstract:Modern optimizers combine gradients from the current mini-batch with historical optimization state, such as momentum or adaptive moments. While highly effective, aggregating across the batch and incorporating this history can produce parameter updates that increase the loss of individual samples. We term this effect harm and formalize the parameter update as an optimization problem that explicitly minimizes the conflicting impact of both batch averaging and past optimization state on current data. Because the exact formulation is intractable, we introduce a highly efficient proxy. We first reduce the problem’s dimensionality to the batch size, and then drastically cut memory and speed bottlenecks by successfully restricting the optimization to the last linear layer. This hinges on the unexpected finding that this layer alone reliably captures the second-order statistics of the per-sample gradients. The resulting surrogate problem integrates readily into standard optimizers like SGD and AdamW, and can be solved using a small number of GPU-friendly iterations. Crucially, the method exhibits favorable scaling properties, as the relative computational overhead shrinks as the model size or input grows. Experiments on image classification benchmarks confirm reduced per-sample interference and improved generalization. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Machine Learning (stat.ML) Cite as: arXiv:2607.16261 [cs.LG] (or arXiv:2607.16261v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16261 Focus to learn more arXiv-issued DOI via DataCite
[AI-171] Neural Controlled Differential Equations for EMT-Level Surrogate Modeling of Grid-Forming Inverters
链接: https://arxiv.org/abs/2607.16258
作者: Jiagang Qu,Yong Tao,Dan Wang,Enyi Li,Jingjing Qi,Ding Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:The application of artificial intelligence methods in power electronic converter modeling is becoming increasingly widespread, but existing applications still face many challenges, such as difficulties in multi-time-scale hybrid analysis and the lack of physics-aware evaluation criteria and constraints, resulting in poor performance. This paper proposes a Neural Controlled Differential Equation (Neural CDE) framework for learning continuous-time surrogate models of grid-forming inverters for electromagnetic transient (EMT) simulation, which relaxes the constraint of fixed sampling rates and enables multi-time-scale control analysis. Then, an affine-control formulation with dual slow/fast pathways is proposed to capture the hierarchical and multiscale behavior of converter dynamics, and a physics-inspired regularization method is utilized to enhance stability and coherence. Evaluated on EMT-generated trajectories, the model accurately reproduces transient responses, preserves effective damping and the dominant oscillatory characteristics, and maintains bounded long-horizon rollouts. The results show that Neural CDE-based component modeling offers a physically consistent surrogate modeling approach for EMT-level simulation studies.
[AI-172] Comprehensive Evaluation of Machine Learning for Type 2 Diabetes Risk Prediction: Large-Scale External Validation and Fairness Analysis
链接: https://arxiv.org/abs/2607.16253
作者: Rajveer Singh Pall,Sameer Yadav,Siddharth Bhalerao,Sourabh Sahu,Ritu Ahluwalia,Bhaskar Awadhiya
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted and published at the IEEE EDS Technically Sponsored International Conference on Intelligent Processing, Hardware, Electronics, and Radio Systems (CIPHER-2026), 13-15 Feb 2026, NIT Jalandhar, India (IEEE Conference Record #70417, Paper ID: 155). 8 pages, 4 figures, 3 tables
Abstract:Machine learning-based Type 2 diabetes risk prediction models obtain good internal validation results but lose effectiveness in real-world applications due to deficient external testing and fairness assessment. We developed a multi-dimensional framework evaluating discrimination, calibration, interpretability, and algorithmic fairness on nationally representative populations. An XGBoost model was trained on NHANES 2015-2020 (n=15,685) using eight non-laboratory predictors: age, sex, race/ethnicity, BMI, smoking status, physical activity, history of heart attack, and history of stroke. External validation was performed on BRFSS 2020-2022 (n=1,285,783) under realistic distribution shift. Internal validation showed good discrimination (AUC=0.794, 95% CI 0.788-0.800), with performance loss on external validation (AUC=0.717, relative decrease: -9.7%, p0.001). Fairness analysis revealed severe bias: elderly adults (=60) showed AUC=0.607 vs 0.742 for young adults (difference=0.135, p0.001); obese individuals showed AUC=0.698 vs 0.735 for normal weight (difference=0.037, p0.001). Gender showed comparable performance (male=0.723 vs female=0.712, p=0.142). Calibration revealed risk overestimation (Brier score=0.123). SHAP analysis identified age, BMI, and physical activity as primary risk drivers. Populations with highest diabetes risk receive the worst algorithmic performance, underscoring the need for fairness-aware, age-stratified deployment strategies before clinical use.
[AI-173] SOS-LoRA: Static Orthogonal-Subspace Low-Rank Adaptation with Fixed Multi-Scale Scaling
链接: https://arxiv.org/abs/2607.16252
作者: Yupeng Chang,Yuan Wu,Yi Chang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Low-Rank Adaptation (LoRA) is a widely used parameter-efficient fine-tuning (PEFT) method for large language models. Under a fixed rank budget, LoRA parameterizes each adapted weight through a single low-dimensional input-side pathway, which may couple heterogeneous behaviors through shared input directions and induce interference during optimization. We propose Static Orthogonal Subspace LoRA (SOS-LoRA), a drop-in extension that reparameterizes a rank-rtot update as a sum of K static (always-on, non-routed) low-rank experts. SOS-LoRA (i) decomposes the total rank across experts, (ii) applies a fixed multi-scale scaling scheme to encourage scale-separated optimization dynamics, and (iii) promotes diverse input-side directions via cross-expert orthogonal initialization and a lightweight regularizer. SOS-LoRA remains fully mergeable, adding no inference-time parameters or latency after merging. Experiments on reasoning and knowledge-intensive benchmarks (Llama 2/3), encoder-based NLU (GLUE), and math reasoning (GSM8K/MATH) show consistent gains over matched-budget LoRA baselines and recent variants. Code is available at this https URL.
[AI-174] Benchmarking Machine Learning Models for Multi-Omics-Based Breast Cancer Prediction
链接: https://arxiv.org/abs/2607.16250
作者: Priyanka Paudel,Madan Baduwal
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Estrogen Receptor (ER) status is a critical biomarker in breast cancer diagnosis, prognosis, and treatment selection. Recent advances in high-throughput sequencing technologies have enabled the generation of multi-omics datasets that provide complementary molecular information for computational prediction tasks. This study presents a systematic benchmarking analysis of classical machine learning models for ER status prediction using transcriptomic (RNA expression), genomic (copy number variation; CNV), and proteomic (RPPA) data from the TCGA-BRCA cohort. A rigorous experimental framework incorporating stratified train-test splitting, stratified five-fold cross-validation, class imbalance handling, and fold-specific feature selection was employed to ensure reliable evaluation and prevent data leakage. Random Forest, XGBoost, LightGBM, CatBoost, Support Vector Machines (SVM), and Logistic Regression were evaluated across single-omic and multi-omic settings. Results demonstrated that RNA expression provided the strongest predictive signal, while multi-omic integration yielded modest but consistent improvements over individual modalities. Among all evaluated approaches, Random Forest achieved the best overall performance in the integrated multi-omic setting, obtaining a balanced accuracy of 90.3% and an ROC-AUC of 97.1%. Furthermore, recurrent selection of biologically relevant genes, including \textitESR1, \textitPGR, \textitFOXA1, and \textitGATA3, supported the biological validity of the learned models. These findings indicate that carefully regularized classical machine learning methods remain highly effective for small, high-dimensional genomic datasets and that multi-omic integration provides complementary information for breast cancer ER status prediction.
[AI-175] High-accuracy Low-Bit KV-Cache Quantization via Local Distribution Restoration
链接: https://arxiv.org/abs/2607.16248
作者: Gradwell Dzikanyanga,Yanqi Pan,Weihao Yang,Donglei Wu,Wen Xia,Hao Huang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Long-context large language model inference relies on the KV cache to avoid redundant attention computation, but incurs high memory and bandwidth overheads. Low-bit KV-cache quantization reduces this cost, yet it severely degrade quality; particularly, one-bit quantization reduces accuracy from 84.2% to 47.8% on Llama-3.1-8B under RULER. Rather than common beliefs that absolute error of logits, we find that the root cause is structured local misranking, where the distribution of logits in top-K region is drifted. We thereby propose local distribution restoration, a new technique that detects steps with high local distribution risk from quantized-logit features and restores only the selected top-K candidate distribution before token selection. We implement DGAP to achieve local distribution restoration, with efficient risk detcetors and correctors. Expeirments show that on Llama-3.1-8B, DGAP recovers K1V1 RULER accuracy from 47.8% to 83.2% and reduces distribution drift from 0.38 to 0.14; across Llama, Mistral, and Qwen models, it preserves the persistent low-bit KV-cache footprint with modest decode overhead. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI) Cite as: arXiv:2607.16248 [cs.LG] (or arXiv:2607.16248v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16248 Focus to learn more arXiv-issued DOI via DataCite
[AI-176] Let the Data Decide: Supervision Analysis Capability Trade-offs and Adaptive Objective Routing in Continued Pre-Training via Off-Policy Distillation
链接: https://arxiv.org/abs/2607.16246
作者: Jiangan Yuan,Zhixuan Li,Han Xu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Off-policy distillation is now central to large language model pre-training, yet how training data, objective parameterization, and model capabilities interact remains poorly characterized. We studies top- k -truncated, temperature-scaled off-policy distillation by decomposing this problem into two questions: an \emphobjective-to-capability analysis of how the training objective shapes token-level supervision and downstream performance, and a \emphdata-to-objective analysis of how data heterogeneity should inform objective routing. We first show that the language-modeling objective ( L_\mathrmLM ) and the knowledge-distillation objective ( L_\mathrmKD ) induce systematically different capability profiles, and trace this divergence to a gradient-level tension between \emphdirect observed-token reinforcement and \emphteacher-supported alternative supervision. To quantify this tension, we introduce diagnostic metrics – support coverage, observed-token probability mass, and teacher-distribution concentration – and show via controlled sweeps that the support size k governs a coverage-sharpness trade-off, while distillation temperature controls within-support probability allocation. We then examine adaptive objective routing: a domain-level policy that applies L_\mathrmLM to math and code and L_\mathrmKD to general-domain data yields consistent gains over both single-objective baselines, whereas token-level routing based on observed-token probability mass or teacher entropy fails to consistently match the single-objective baseline. These results suggest that effective objective routing depends less on routing granularity than on the quality of the routing signal, reframing continued pre-training via off-policy distillation as a structured, data-conditional supervision-design problem rather than a global hyperparameter choice.
[AI-177] Learning Structural Manipulability in Gate-Level Netlists Using Graph Neural Networks
链接: https://arxiv.org/abs/2607.16245
作者: Rupesh Raj Karn,Ozgur Sinanoglu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 11 pages
Abstract:Gate-level netlists exhibit intrinsic structural properties that influence signal propagation independently of functional simulation. We define a topology-driven structural manipulability score that characterizes node-level structural flexibility using path participation, k-core embedding, symmetry, and centrality. Modeling netlists as directed graphs, we formulate node-level regression to learn this topology-derived score using graph neural networks (GNNs). Experiments on ISCAS85 and EPFL benchmarks evaluate how effectively different GNN architectures approximate this metric across held-out circuits, with hierarchical models yielding the most consistent rankings. Component-level and ablation analyses examine the contribution of individual factors. As an illustrative case study, analysis of Trojan-injected circuits using TrustHub templates reveals statistically distinguishable structural patterns, indicating that topology-based scoring provides complementary structural insight.
[AI-178] RobustMAD: Evaluating Real-World Robustness of Multimodal Small Language Models for Deployable Anomaly Detection Assistants KR
链接: https://arxiv.org/abs/2607.16243
作者: Anushiya Arunan,Xin Li,Yan Qin,U-Xuan Tan,Nhu Khue Vuong,Xiaoli Li,Chau Yuen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted for publication in Transactions on Machine Learning Research (TMLR), 2026 at this https URL
Abstract:Multimodal industrial anomaly inspection assistants are a critical component of next-generation smart factories, enabling interactive vision-language-based querying. However, multimodal large language models remain impractical for on-site deployment due to prohibitive computational demands and privacy risks from cloud-based inference. Compact multimodal small language models (MSLMs) offer a deployable alternative, yet progress is constrained by the lack of comprehensive robustness analyses and meaningfully challenging benchmarks that reflect real-world industrial conditions. To address this gap, we develop RobustMAD, the first deployment-motivated benchmark, designed to comprehensively evaluate model robustness through diverse open-ended queries spanning object understanding, anomaly detection, unanswerable problems, and visual quality degradations. Contrary to conventional assumptions, top-performing MSLMs exhibit promising capabilities, surprisingly outperforming even the larger GPT-5 Nano. However, they still fall short of safety-critical requirements, and RobustMAD reveals critical robustness gaps that pose operational risks. In particular, three recurring failure modes emerge: (i) fragile multimodal grounding under fine-grained distinctions or degraded visual conditions, (ii) insufficiently comprehensive responses, and (iii) weak logical grounding on unanswerable or ill-posed queries, leading to hallucinated outputs. Grounded in these insights, we provide actionable guidance for the design of next-generation multimodal industrial inspection assistants that leverage their promising competence. Code is available at this https URL.
[AI-179] RACE: Trajectory-Based Safety Patch Learning for LLM Post-Training Realignment
链接: https://arxiv.org/abs/2607.16242
作者: Changyue Li,Jiaming He,Youliang Yuan,Jialin Wu,Boxi Yu,Zhicong Huang,Pinjia He
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注:
Abstract:Fine-Tuning-as-a-Service (FTaaS) platforms let users train large language models (LLMs) on customized tasks, but this pipeline could erode models’ safety alignment. In practice, service providers need to recover models’ safety without re-running full alignment, or destroying the utility gained from customized tasks. A line of existing work refers to model parameter merging, which adds a safety patch on the fine-tuned model parameters to shift the model away from unsafe tendencies. However, this merging-based paradigm is fundamentally bottlenecked by task-safety update entanglement: downstream task updates and the safety patch often overlap in their dominant directions, so the merge strength is intrinsically hard to calibrate. If the safety vector is scaled too weakly, harmful components could still dominate, preventing the model from returning to a safe region; if it is scaled too aggressively, it suppresses task-relevant directions and degrades utility. To solve this problem, we shift the focus of merging-based methods from designing online merging operators to offline patch learning, and seek a safety patch that minimally interferes with task-relevant directions while retaining decisive control over unsafe behaviors. We propose TRACE, a trajectory-based safety patch learning framework that (i) simulates harmful tuning trajectories to generate progressively corrupted states, and (ii) optimizes a plug-in patch to recover safety while maintaining utility across varying corrupted base states. Across six benchmarks and two models, TRACE consistently dominates the safety-utility frontier. TRACE reaches nearly 100% safety on all settings, while maintaining comparable utility to the undefended fine-tuned model. Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR) Cite as: arXiv:2607.16242 [cs.LG] (or arXiv:2607.16242v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16242 Focus to learn more arXiv-issued DOI via DataCite
[AI-180] KernelBench-Verified: Do LLM -Generated Kernels Actually Beat PyTorch?
链接: https://arxiv.org/abs/2607.16241
作者: Yunxiang Zhang,Ping Yu,Jianyu Wang, Max (Xiangjun)Fan,Julian Reed,Azalia Mirhoseini,Will Su
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Recent large language models (LLMs) can generate custom CUDA kernels that appear to outperform PyTorch on benchmarks such as KernelBench. Building upon this foundational framework, we demonstrate that frontier models frequently engage in reward hacking to artificially inflate reported performance. In this work, we identify two areas where evaluation frameworks must co-evolve with model capabilities. First, to accurately measure true speedup, we examine the baseline timing mechanism, noting that enabling Tensor Core acceleration with TF32 provides a more realistic estimation of execution on modern GPUs. Second, concerning algorithmic correctness, models often exploit the narrow test distribution by hardcoding bypasses for specific tensor values. By skipping required computations, these kernels artificially accelerate execution rather than implementing actual CUDA kernels. We introduce KernelBench-Verified, an extended evaluation framework that incorporates a TF32-enabled baseline and a four-distribution hidden test suite. We additionally introduce memory efficiency metrics that capture the often-overlooked speed-memory tradeoff in kernel optimization. Under verified single-turn evaluation with seven frontier LLMs, we find that the best-performing model (GPT-5.5) achieves a 0.88x geometric mean speedup, significantly lower than the 1.43x speedup observed under the standard evaluation protocol. No model consistently outperforms PyTorch when evaluated against realistic baselines. On the memory front, 28% of GPU kernels generated by the best model increase peak GPU memory usage. Our findings demonstrate the necessity of continually adapting robust evaluation protocols as LLM kernel generation capabilities advance.
[AI-181] Normalized Rewards for Preference Optimization ICML2026
链接: https://arxiv.org/abs/2607.16240
作者: Shawn Im,Federico Danieli,Skyler Seto,Barry-John Theobald,Katherine Metcalf
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: ICML 2026
Abstract:Direct Alignment Algorithms (DAAs) such as DPO have become a common way to post-train and align LLMs with human preferences. However, DAAs have been observed to over-optimize their implicit reward model and decrease the likelihood of preferred responses. This results in a decrease in the total likelihood assigned to responses seen in the preference dataset, potentially resulting in undesirable behavior. To counteract this undesired side-effect of DAAs, we examine the effect of using objectives that add a regularization term to maintain the total length-normalized probabilities of the chosen and rejected responses. To better understand over-optimization, we investigate how response likelihood changes are distributed over the tokens with and without regularization. We find that a significant portion of the likelihood changes are due to a small set of outlier tokens, which explains how DAAs improve generation quality despite decreasing the likelihoods of chosen responses. We apply the proposed regularization to reference-based (DPO) and reference-free (SimPO) methods and find (1) improved trade-offs between generation quality and general benchmark capability and (2) improvements in reward modeling across datasets. For example, on Llama-3.1-8B-Instruct, we see both a 20% relative increase in AlpacaEval2 scores and 9% relative performance gains on general benchmarks. Additionally, we find that the added regularization term effectively mitigates the amount of displacement within preferred responses overall, and for the outlier tokens specifically, by utilizing low-likelihood tokens.
[AI-182] Diffusion-corrected Autoregressive Fourier Neural Operator for Droplet Evolution Prediction
链接: https://arxiv.org/abs/2607.16238
作者: Jinghao Cao,Minsung Kang,Hongyue Sun,Chi Zhou,Jihoon Chung,Xubo Yue,Sanchoy Das,Bo Shen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computational Physics (physics.comp-ph); Fluid Dynamics (physics.flu-dyn)
备注: 11 figures, 4 tables
Abstract:Predicting droplet evolution in material jetting, or Inkjet Printing (IJP), is essential for maintaining printing quality. However, long-horizon forecasts remain challenging due to error accumulation and the complex coupling of process variables. In this work, we introduce the Diffusion-corrected Auto-Regressive Fourier Neural Operator (DiffARFNO), a two-stage framework that combines an autoregressive Fourier-MIONet with a conditional Denoising Diffusion Implicit Model (DDIM) corrector. Fourier-MIONet is trained as a coarse predictor and deployed autoregressively for long-horizon forecasting. In the second stage, a DDIM-based conditional corrector refines the coarse prediction within each sliding window through efficient iterative denoising. By combining coarse predictions from Fourier-MIONet with a DDIM corrector that restores fine details, DiffARFNO aims to provide high-fidelity predictions for long-horizon forecasts. Extensive experiments on droplet datasets from ANSYS Fluent demonstrate that DiffARFNO significantly outperforms existing state-of-the-art models.
[AI-183] Quantizing Recursive Reasoning Models
链接: https://arxiv.org/abs/2607.16237
作者: Thorir Mar Ingolfsson,Wajeeha Tahir,Anna Tegon,Lionnus Kesting,Gamze İslamoğlu,Luca Benini
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Preprint, 27 pages, 4 tables, 12 figures
Abstract:Recursive reasoning models solve hard puzzles by applying compact, weight-tied blocks over many refinement steps. Because these blocks are reused many times, quantizing them creates a unique dynamical problem: the quantization error is incurred at every step. While 8-bit quantization (integer or float) preserves accuracy, moving to a per-tensor 4-bit format causes a systematic bias to accumulate. The ensuing drift catastrophically degrades exact-solution accuracy on Sudoku from 84.1% to 0.0% (only ~25% of cells correct). In this work, we show that this collapse is caused by activation-scaling granularity rather than bit-width or number format. Crucially, moving to per-block scaling completely restores the transition. To implement this, we apply MXInt4, a blockwise integer activation format, to recursive reasoning models. It is competitive with blockwise float formats on our tasks, while keeping integer elements and power-of-two block scales. Finally, recursion depth and reuse modulate quantization sensitivity, with the deepest architecture we test (the EqR equilibrium model) the most sensitive. Yet blockwise scaling overcomes this vulnerability, staying robust across these architectures and transferring to the open-ended ARC-AGI benchmark.
[AI-184] he Failures of Marginal Influence-Based Attribution Methods for Global Time Series Explanations IJCAI2026
链接: https://arxiv.org/abs/2607.16236
作者: Amadeo Tunyi
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: Accepted at the Workshop on Explainable Artificial Intelligence (XAI), International Joint Conference on Artificial Intelligence (IJCAI 2026)
Abstract:Explainability methods for time series models predominantly produce flat attribution scores: they quantify the direct influence of a feature at a timestamp by a scalar. We prove that the dominant failure mode of such methods is not the scalar format itself but a fundamental computational mismatch: existing methods compute scores via marginal conditioning or off-manifold gradients, both of which conflate direct temporal dependencies with mediated ones under autocorrelation. We also define DAG-faithfulness: an explanation is DAG-faithful if the temporal dependency graph it encodes is Markov-equivalent to the temporal directed acyclic graph (DAG) implicitly learned by the model. Particularly, we observe that standard attribution methods, specifically SHAP, are not DAG-faithful in general, and that recent time-series-aware extensions inherit the same computational limitation.
[AI-185] OpenMHC: Accelerating the Science of Wearable Foundation Models
链接: https://arxiv.org/abs/2607.16235
作者: Narayan Schuetz,Yuze Bai,Lianggang Pan,Edgar Eggert,Favour Nerrise,Juan Delgado-SanMartin,Max Rosenblattl,Milana Gurbanova,Mohammad Asadi,Anders Johnson,Paul Schmiedmayer,Dennis Wang,Allan Lawrie,Daniel Seung Kim,Xin Liu,Akshay Paruchuri,Ehsan Adeli,Euan Ashley,Kelly W. Zhang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Mobile and wearable devices offer an unprecedented opportunity for continuous, passive health monitoring and active health coaching. However, the largest wearable datasets are not publicly available for research, and leading wearable foundation models trained on such datasets are rarely open-weight or come with reproducible training code. To accelerate open science in wearable health, we release OpenMyHeartCounts (OpenMHC), the largest and most comprehensive open-access wearable health dataset to date, alongside open-source implementations of recent wearable foundation models. OpenMHC, derived from over a decade of data collected through the My Heart Counts study app, includes 60 million hours of wearable data across 19 sensor channels (e.g., step count, heart rate, sleep, workouts) and up to 169 linked variables, including health, lifestyle, mood, and behavior from 11,894 consenting participants. Furthermore, we introduce a unified, open benchmark that enables standardized comparison of wearable health models across three tracks: health and behavior downstream prediction, multivariate data imputation, and time-series forecasting. We benchmark classical methods alongside recent wearable and multivariate time series foundation models. By open-sourcing data, code, and model weights at this unprecedented scale, we aim to democratize wearable health AI research and enable the community to drive open progress in this domain.
[AI-186] HantaWatch: Federated Learning for Hantavirus Genomic Surveillance
链接: https://arxiv.org/abs/2607.16234
作者: Shanika Iroshi Nanayakkara,Shiva Raj Pokhrel
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Hantavirus genomic surveillance is limited by the distribution of sequence data, non-IID source heterogeneity, and constrained expert-review capacity. We propose HantaWatch, a federated learning framework that enables laboratories and surveillance sites to collaboratively train sequence-based models without sharing raw data. HantaWatch integrates k-mer feature extraction, source-aware federated client construction, adaptive DU-FedProx optimization, surveillance-specific model selection, and prediction-only triage. Experiments on binary and multi-class tasks show that HantaWatch supports high-risk screening, outbreak-associated prediction, clade classification, and clinical-syndrome categorization while balancing predictive performance, false-negative risk, and update stability. The framework converts model output into risk scores, confidence estimates, uncertainty flags, and ranked expert-review priorities. HantaWatch therefore provides a practical federated decision-support layer for decentralized Hantavirus surveillance, supporting expert prioritization without replacing laboratory or public-health interpretation.
[AI-187] oken-Level Cross-Modal Transformer with Contrastive Multi-Task Learning for Breast Cancer Subtype Classification and Survival Prediction
链接: https://arxiv.org/abs/2607.16233
作者: Suxing Liu Byungwon Min
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Integrating heterogeneous genomic and clinical modalities for joint cancer subtype classification and survival prediction remains a key challenge in precision oncology. Existing approaches suffer from three limitations: (1) they treat each modality as a monolithic feature vector, precluding fine-grained token-level interactions across modalities; (2) cross-modal fusion is typically performed through linear weighting or late averaging rather than structured token exchange; and (3) survival and classification objectives are optimized independently, missing a joint regularization signal.
[AI-188] RouteCost: A Production-Inspired Multi-Stage Framework for Pre-Order Shipping Cost Estimation in E-Commerce
链接: https://arxiv.org/abs/2607.16230
作者: Xianling Zeng,Zihan Yu,Sichen Zhao,Yalun Qi,Zhiming Xue
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:
Abstract:Accurate pre-order shipping cost estimation is important in e-commerce because it affects price presentation, margin planning, and conversion. In practice, shipping cost is shaped not only by distance but also by destination demand mix, billable weight, dimensional pricing, surcharge triggers, and latent operational effects such as shipment consolidation. Static lookup methods therefore miss important sources of variation, while monolithic regressors may exploit strong but non-causal correlations. We propose RouteCost, a production-inspired multi-stage framework that decomposes the problem into time-aware demand forecasting, fee-card-informed baseline pricing, Stage 2 residual correction, and proxy-based box-consolidation inference. Route-level cost estimates are aggregated through a route-weighted expectation formulation to produce product-level shipping cost predictions. Across over 250,000 orders, 260 products, and 18 months of order history, the framework improves predictive quality and aggregate calibration while preserving route-level interpretability.
[AI-189] International Agreements to Limit Frontier AI: Objectives and Exit
链接: https://arxiv.org/abs/2607.16224
作者: Lennart Finke
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI)
备注:
Abstract:An international agreement to limit AI development could be crucial to mitigate risks from AI. However, it remains unclear which conditions should determine when the limiting measures are relaxed. We survey existing international agreements, outline what properties appropriate conditions should satisfy, list possible conditions, and finally give a recommendation in an example scenario. We recommend a fixed time period after which a new organization established at the start of the period specifies conditions that address when AI development can be safely conducted, with a possibility of withdrawal in extraordinary circumstances. We hope to illustrate the considerations that would likely go into an international agreement to limit AI.
[AI-190] Fully-sensorized smart-eyewear platform for on-device Machine Learning
链接: https://arxiv.org/abs/2607.16222
作者: Andrea Giudici,Christian Veronesi,Pietro Bartoli,Mario Caliò,Aurelio Teliti,Giacomo Gervasoni,Diana Trojaniello,Franco Zappa
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Signal Processing (eess.SP)
备注: This work was carried out in the EssilorLuxottica “Smart Eyewear Lab”, a Joint Research Center between EssilorLuxottica and Politecnico di Milano
Abstract:This paper presents ARGO, a smart eyewear platform designed to bridge ergonomic comfort, high computational throughput, and energy efficiency. Unlike cloud-dependent solutions, ARGO leverages the STM32N6 microcontroller and its integrated Neural Processing Unit (NPU) to enable on-device machine learning, minimizing latency and preserving user privacy through local data processing. The primary contribution lies in the holistic co-design of hardware, firmware, and artificial intelligence, centered on the deployment of an optimized YOLOv11 model for real-time urban obstacle recognition. To ensure compatibility with the target NPU, we introduce Head-wise Parallel Attention (HPA), an architectural refinement that enables efficient accelerator execution while preserving the original computational logic. The model is trained on the Walking On The Road (WOTR) dataset, and the final deployed configuration achieves an mAP50-95 of 24 under strict memory constraints, with a memory footprint of only 2.483 MB. The platform integrates a multimodal sensor suite, RGB cameras, Time-of-Flight sensors, microphones, and ambient sensors, and delivers 10 FPS at a continuous autonomy of ~113 minutes on a 200 mAh battery. These results demonstrate the feasibility of a high-performance, privacy-preserving, and socially acceptable assistive device, and highlight how competitive edge AI solutions increasingly demand tightly integrated, multidisciplinary co-design approaches.
[AI-191] Comparing Spectrogram Front-Ends for Abnormal Heart-Sound Detection with a Convolutional Neural Network
链接: https://arxiv.org/abs/2607.16220
作者: Abhinav Pala,Dhanush Pala
类目: Computers and Society (cs.CY); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Sound (cs.SD); Audio and Speech Processing (eess.AS)
备注:
Abstract:Heart disease kills a lot of people, and one cheap way to catch it early is by listening to heart sounds with a stethoscope, or better yet, just recording them and running them through a model. This project is a binary classification task: take a short clip of someones heartbeat and decide if it sounds normal or abnormal. Instead of trying out a bunch of different models, we kept the CNN the same the whole time and just changed how we turned the raw audio into a picture for it to look at. We tried three ways of doing that: a regular logmel spectrogram, PCEN (which basically normalizes each frequency bin over time), and a multi resolution version that stacks a few different window sizes together. We ran all three on the PhysioNet 2016 heart-sound dataset with the exact same setup but same model, same optimizer, same random seed. Turns out all three do pretty well at catching abnormal cases (sensitivity around 0.95), but PCEN and multi-resolution both edge out the plain logmel on the official PhysioNet accuracy metric (0.915 and 0.916 vs. 0.910). We also ran Grad-CAM to see where the model was actually looking, and it mostly focused on the low frequencies where S1 and S2 heart sounds live, which is a good sign that it learned something real
[AI-192] SelKV: Selective KV Cache Merging with Per-Token Merge-or-Drop and Attention Compensation
链接: https://arxiv.org/abs/2607.16213
作者: Soumia Bouyahiaoui,Manel Kara laouar,Aicha Boutorh,Mohamed Hadj Ameur
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Large Language Models (LLMs) generate text autoregressively, relying on a key-value (KV) cache whose memory footprint grows linearly with context length, creating a major bottleneck. Recent compression methods mitigate this cost via token merging; however, these approaches often rely on indiscriminate aggregation, which degrades representations and introduces attention sag, a mismatch where merged tokens receive the same softmax mass as individual tokens despite encoding multiple inputs. We propose a training-free, dual-component framework for KV cache compression that addresses these limitations. First, a soft cosine gate adaptively modulates merging decisions based on value-vector similarity, suppressing or discarding dissimilar tokens to preserve semantic fidelity. Second, we introduce an attention-ratio compensation mechanism that applies a decoding-time logit bias derived from prefill attention statistics, correcting the softmax imbalance induced by merging. Evaluated on LongBench (16 English datasets) while retaining only 25% of the KV cache, our framework achieves strong compressed performance against representative one-shot baselines. It is especially robust on the evaluated grouped-query attention (GQA) models, maintaining nearlossless generation quality. Furthermore, the method outperforms the full-cache baseline on complex multi-document QA tasks and delivers a 3.3x decoding speedup at 100k tokens.
[AI-193] Symbolic Augmentation Closes a Canonical-Equivalence Blind Spot in Neural Fact-Checkers
链接: https://arxiv.org/abs/2607.16212
作者: Genpei Zhang
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 3 figures, 6 tables
Abstract:Large language models hallucinate numbers and units when summarizing scientific text, a failure mode that can silently invert a scientific claim. We recast the detection of such errors as typed verification: we introduce a five-class typed-quantity error taxonomy and a 1500-item benchmark, rewritten from PMC and arXiv sources and labeled by two independent LLM annotators with adjudication (Krippendorff’s alpha = 0.882). A ModernBERT encoder fine-tuned on this benchmark reaches macro-F1 = 0.899, far above any off-the-shelf neural fact-checker, yet four probes expose a sharp structural blind spot: on canonical-equivalent rewrites of physically equivalent quantities (e.g., 95°C and 368.15 K) its accuracy collapses to 36.5%. We propose Symbolic Augmentation, a training-time framework that runs the modules of a symbolic verifier in reverse to generate label-preserving augmented training data. The augmentation lifts canonical-equivalence robustness to 98.2% while slightly improving in-distribution accuracy (macro-F1: 0.899 to 0.902); the augmented encoder matches a closed-frontier LLM at no inference cost and transfers to an external benchmark (SciFact-Open binary macro-F1: 0.791 to 0.828). Two negative results sharpen the claim: symbolic features as auxiliary encoder inputs add nothing, and symbolic silver labels scale negatively under teacher noise. Together these results identify training-time augmentation as the right integration point between symbolic and learned components.
[AI-194] Accurate and Efficient Long-Term Memory for LLM Agents
链接: https://arxiv.org/abs/2607.16211
作者: Zicheng Zhao,Xinyang Guo,Luyao Lv,Menghan Wang,Ming Li,Shuaicheng Li
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:LLM agents augmented with persistent memory can recall past interactions, but existing systems suffer from two limitations: flat, unstructured storage loses relational context needed for multi-hop and temporal reasoning, and reliance on expensive LLM-based classification makes them impractical for latency-sensitive deployment. Without mechanisms to validate new information against stored knowledge, these systems silently accumulate contradictions. We present MOSAIC (Memory-Organized Structured Agent for Information Collection), a structured, conflict-aware long-term memory framework for LLM agents that is substantially more accurate and efficient. MOSAIC introduces three key capabilities: (1) entity-typed graph storage with semantic classification preserving relational structure across events, personas, and relationships, enabling multi-hop and temporal reasoning over conversation history; (2) hash-accelerated dual-path retrieval replacing LLM-based classification with locality-sensitive hashing, achieving near-instantaneous lookup with negligible accuracy loss; and (3) active conflict detection at save time that cross-references new information against existing graph neighbors, triggering updates or deletions for contradictory entries. Evaluated on LoCoMo (long-conversation QA), HaluMem, and a novel clinical-guideline error compounding test, MOSAIC achieves 89.35% accuracy on LoCoMo (+27.21 pp over the best baseline), best HaluMem-Medium extraction F1(86.77%) and HaluMem-Long extraction F1 (85.84%), best QA correctness on both Medium and Long (73.10%, 70.75%), and detects 66% of injected factual conflicts-4.7 times higher than the best baseline (14%)-while hash-accelerated retrieval keeps average search latency at 0.58 s per question.
[AI-195] A Survey on the Verification of Reinforcement Learning Policies IJCAI
链接: https://arxiv.org/abs/2607.16210
作者: Luca Marzari,Ezio Bartocci,Enrico Marchesini
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the 35th International Joint Conference on Artificial Intelligence (IJCAI) 2026
Abstract:Reinforcement learning (RL) is increasingly applied in complex, safety-critical domains, yet the lack of rigorous behavioral guarantees for neural network-based policies remains a major barrier to deployment. Recent advances in policy expressiveness and scale have intensified this challenge, leading to a rapidly growing but conceptually fragmented body of work on RL policy verification. This survey provides a unifying perspective on RL verification methods. We introduce a taxonomy that clarifies relationships among existing approaches along three axes: verification paradigm (formal versus probabilistic), temporal scope (step-wise versus multi-step), and guarantees strength. Beyond taxonomy, we unify underlying theoretical foundations, make implicit assumptions and limitations explicit, and identify emerging directions.
[AI-196] Shapley Context Pruning: A Cooperative Game Perspective for Context Reranking and Pruning
链接: https://arxiv.org/abs/2607.16209
作者: Yanqiao Chen,Dongsheng Hou,Yuhan Rui,Zhen Cao,Yepang Liu
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Context reranking and pruning have become essential for improving the efficiency of modern Retrieval-Augmented Generation (RAG) systems, yet an interpretable and unified framework remains underexplored. Previous work has primarily emphasized lexical retrieval, cross-encoder architectures, model distillation, and Low-Rank Adaptation (LoRA), mostly relying on heuristic loss functions and empirical attribution. This paper presents Shapley Context Pruning (SCP), a novel framework for context reranking that establishes a cooperative-game-theory perspective for importance attribution by modeling the context as a cooperative game. Balancing the trade-off between fine-grained and coarse-grained representations, we employ a Deep Sets architecture to approximate a permutation-invariant value function at the sentence level, utilizing pre-trained language models as sentence embedders and optimizing via a pairwise margin ranking loss. To ensure practical scalability without sacrificing mathematical rigor, we leverage Monte-Carlo sampling for efficient training and inference, providing formal theoretical error bounds and sample complexity guarantees for preserving Top-K subset rankings. Furthermore, we conduct comprehensive experiments-spanning supporting-sentence recall, Needle-in-the-Haystack (NIAH) evaluations, long-context QA, and multi-hop reasoning-alongside rigorous ablation studies on embedding quality and attribution strategies. The model achieves competitive downstream QA performance against robust baselines.
[AI-197] JUMP: Single-Pass Membership Inference on Fine-Tuned Diffusion Language Models
链接: https://arxiv.org/abs/2607.16207
作者: Yeachan Jun,Albert No
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: 24 pages. Preprint
Abstract:Membership inference attacks (MIAs) test whether a candidate example appeared in a model’s training data. We study MIAs for fine-tuned discrete diffusion language models (dLLMs), where membership means inclusion in the target model’s fine-tuning set. Unlike autoregressive language models, dLLMs allow an attacker to choose arbitrary mask sets and obtain token distributions for all masked positions in parallel. The prior dLLM attack, SAMA, follows a natural loss-mimicking strategy by averaging reconstruction signals over many randomly sampled masks, but it uses the any-order interface only as randomization and requires many target/reference queries. We propose JUMP (Joint Uncertainty-Guided Mask Probing), a single-pass scoring attack that exploits both distinctive properties of dLLMs: any-order decodability is used to select low-reference-confidence positions, and parallel decodability is used to score all selected positions through one joint masked query per model. JUMP masks the selected positions jointly and computes a clipped target/reference reconstruction-gap statistic. On fine-tuned LLaDA-8B-Base across six MIMIR domains, JUMP improves mean ROC-AUC from 0.82 to 0.90 over SAMA and substantially improves low-FPR detection, while requiring only one selector pass and one scoring pass through each of the target and reference models.
[AI-198] PPO-HSC: An Exploratory Reinforcement Learning Framework Based on Wide-Area Policy Coverag e Optimization
链接: https://arxiv.org/abs/2607.16206
作者: Yujie Shen,Haowen Chen
类目: Artificial Intelligence (cs.AI)
备注: 13 pages, 3 figures
Abstract:This paper introduces PPO-HSC (Proximal Policy Optimization with High-order Sampling Coverage), an exploratory reinforcement learning framework designed to address the “Invisible Shackles” of mode collapse in Large Language Model (LLM) fine-tuning. While standard Reinforcement Learning from Verifiable Rewards (RLVR) effectively reinforces high-reward trajectories, it often leads models to over-optimize known solutions, sacrificing curiosity and the ability to explore broader solution manifolds. To overcome this, PPO-HSC incorporates a High-order Sampling Coverage (HSC) reward that incentivizes the discovery of “low-similarity yet high-validity” reasoning patterns. By maintaining a dynamic trajectory library of verified unique solutions, the framework provides a differentiable signal that rewards semantic novelty while ensuring structural rationality through a plausibility constraint. Empirical evaluations on mathematical reasoning (GSM8K, SVAMP) and code generation tasks demonstrate that PPO-HSC significantly enhances solution diversity and state-space coverage while maintaining or surpassing the accuracy and syntax integrity of state-of-the-art RL baselines.
[AI-199] It Takes 8 Tokens: Weak-to-Strong Off-Policy RL via Auxiliary Branches
链接: https://arxiv.org/abs/2607.16205
作者: Dayu Wang,Jiaye Yang,Weikang Li,Jiahui Liang,Liwei Qian,Xin Pei,Jizhou Huang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Reinforcement learning with verifiable rewards has emerged as a standard approach for enhancing reasoning in large language models, which typically optimizes the policy by contrasting multiple self generated rollouts. However, we identify a critical support limited bottleneck in this paradigm: on challenging reasoning tasks, the target model’s samples often exhibit semantic redundancy, converging into the same erroneous “reasoning basins” that offer negligible reward contrast for policy updates. In this paper, we propose to overcome this limitation through a weak to strong learning paradigm, where a policy’s exploration is informed by a weaker but computationally efficient auxiliary model. We introduce W2SPO, an off policy RL method that injects short auxiliary segments often as brief as 8 tokens into intermediate target model trajectories and the target model then completes the reasoning path from these diverted states. Policy updates are restricted to these short inserted segments based on final verifiable rewards. Empirically, W2SPO achieves superior performance among evaluated 4B scale models on mathematical reasoning benchmarks, outperforming evaluated post trained baselines. Compared with vanilla GRPO under the same sampling budget, W2SPO improves Pass@1 from 62.3% to 64.2% while achieving a 3.55 times training speedup. These results suggest that weak auxiliary branches can induce stronger target reasoning policies by expanding local exploration support.
[AI-200] Masked Diffusion Language Models are Strong and Steerable Text-Based World Models for Agent ic RL
链接: https://arxiv.org/abs/2607.16204
作者: Darshan Deshpande
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Dataset: this https URL Training code: this https URL
Abstract:Recent growth in reinforcement learning (RL) has surfaced a need for diverse, specialized training environments. Hand-curated environments with fixed task and reward difficulties become ineffective signals as model performance improves, and sparse rewards over long horizons induce mode collapse on specific workflows or tool structures. World models that simulate environment states have matched pure rollout performance, making them promising for scaling diversity on-demand. However, autoregressive (AR) world models suffer from a left-to-right bias preventing conditioning on globally interdependent state anchors such as tool schemas, prior turns, and expected outcomes. We (i) formalize text-based world modeling as a steerable transition-dynamics problem decomposed into initial state, task context, tool schemas, domain rules, and steering directives, and (ii) curate 239,403 grounded state-action trajectories spanning nine open-source environments and twelve frontier model families. We compare AR LMs and masked diffusion language models (MDLMs), showing MDLMs, via bidirectional anchor-aware denoising, achieve better coherence, groundedness, and empirically validated rollout diversity than LLMs over 4x their parameter size, at comparable inference latency. We introduce a plug-and-play GRPO training framework with deterministic state checks, and perform zero-shot transfer ablations on three OOD environments (ScienceWorld, ALFWorld, AppWorld) across three 1.2B-7B agent backbones (LFM2.5, Qwen3, Mistral), achieving up to 47% absolute gains over baselines without environment-specific fine-tuning. We further conduct behavioral analysis of failure modes under adversarial scenarios and human evaluation on realism, outcome correctness, and training utility. We open-source our work to encourage research in this direction.
[AI-201] Democratizing AI with Small Language Models: Structured Benchmarking and Parameter-Efficient Fine-Tuning for Local Deployment
链接: https://arxiv.org/abs/2607.16202
作者: Daniel Cersosimo
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:AI democratization is not primarily a question of matching frontier-scale generality; it is a question of whether capable models can be selected, audited, and specialized under hardware and governance constraints that ordinary institutions can actually satisfy. This paper studies that problem through a controlled evaluation of nine open-weight language models between 135M and 3B parameters on a 1,085-example, 16-topic multiple-choice benchmark designed for structured local deployment. The benchmark emphasizes symbolic precision, constrained formatting, extraction, and short-horizon semantic decision making under a strict one-letter output protocol. A shared parameter-efficient fine-tuning pipeline then adapts a subset of models using 4-bit NF4 quantization with DoRA/LoRA-style adapters on an NVIDIA L4-class budget. In base evaluation, Qwen Coder 3B leads at 75.67% strict accuracy, followed by Qwen2.5 1.5B at 67.10%, Qwen3.5 2B at 64.98%, and Granite 3.3 2B at 64.61%. On the shared 108-example held-out fine-tuning split, adaptation improves Qwen Coder 3B by +26.85 points, SmolLM2 1.7B by +25.92, Qwen2.5 1.5B by +19.44, SmolLM2 360M by +10.18, and SmolLM2 135M by +5.55. Across ranking, topic-level heterogeneity, difficulty strata, failure composition, efficiency frontiers, and topic-conditioned transfer, the same conclusion recurs: a disciplined workflow of benchmark construction, cross-model evaluation, and low-cost specialization already makes a subset of sub-3B models viable as local experts for structured niche workloads.
[AI-202] Generative Ontology Induction: Domain-Agnostic Schema Discovery from Document Corpora Using Large Language Models
链接: https://arxiv.org/abs/2607.16201
作者: Sergei Sergienko
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 2 figures, 1 table. LNCS format
Abstract:Ontology engineering remains a critical bottleneck in knowledge-intensive AI systems. Existing automated approaches either depend on predefined schemas, operate within narrow domains, or produce unstructured outputs unsuitable for downstream pipelines. We introduce Generative Ontology Induction (GOI), a domain-agnostic framework that induces a generative blueprint - entities, dimensions, properties, relationships, and constraints - from a corpus of examples and exports it as a typed graph (six node types, seven edge types) in YAML/JSON. We introduce the Node Coverage Score, a novel evaluation metric that measures the fraction of structural ontology nodes (classes, properties, and dimensions) appearing in generated outputs. A controlled generative validation on four contrasting ontologies - a familiar Software Services Invoice schema, a custom Job Description Ontology, a confidential Pain-Management Clinical Visit Record Ontology, and a Professional Services Contract Statement of Work Ontology - shows that GOI-prompted generation covers 95-100% of the structural backbone in every case; a generic three-field template holds at 97.8% on the invoice schema but drops to 52.2% on the Job Description Ontology, 62.2% on the Pain-Management ontology, and 78.3% on the Professional Services Contract ontology. The structural coverage holds regardless of how familiar the document type is to the model. Comments: 10 pages, 2 figures, 1 table. LNCS format Subjects: Artificial Intelligence (cs.AI) ACMclasses: I.2.4; I.2.7; H.2.1 Cite as: arXiv:2607.16201 [cs.AI] (or arXiv:2607.16201v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.16201 Focus to learn more arXiv-issued DOI via DataCite
[AI-203] Deterministic Replay for AI Agent Systems
链接: https://arxiv.org/abs/2607.16200
作者: Rasheed Mudasiru
类目: Artificial Intelligence (cs.AI)
备注: 9 pages, 5 figures, 5 tables
Abstract:AI agent systems that couple large language models (LLMs) with external tools and APIs are inherently non-deterministic: LLM sampling variance, external API state, CDN infrastructure headers, and execution-environment noise collectively prevent any prior agent run from being faithfully re-executed. Existing observability platforms capture execution logs but cannot reproduce a run in isolation. We present agrepl, a developer-first CLI framework for deterministic replay of agent executions. agrepl intercepts all external interactions at the transport layer via a man-in-the-middle (MITM) proxy, serialises them as structured execution traces, and replays them in a strictly isolated environment with zero outbound network access. We formalise the agent execution model, define the request-key matching function K(s), and prove the determinism invariant. We introduce a noise-aware diff algorithm classifying HTTP header divergence into signal and noise tiers. Empirical evaluation across five workloads (n = 250 replay instances) demonstrates replay fidelity F = 1.0 and a median per-step latency reduction of 98.3%. agrepl is implemented in Go, ships as a single static binary, and is released under the MIT licence. Keywords: AI agents, deterministic replay, LLM debugging, reproducibility, MITM proxy, execution tracing, record/replay systems. Comments: 9 pages, 5 figures, 5 tables Subjects: Artificial Intelligence (cs.AI) MSC classes: 68N19, 68N30(Primary) 68T42, 68T01 (Secondary) ACMclasses: I.2.5; D.3.2; I.2.11 Cite as: arXiv:2607.16200 [cs.AI] (or arXiv:2607.16200v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2607.16200 Focus to learn more arXiv-issued DOI via DataCite
[AI-204] PlanFlip: Attacking Multi-Agent LLM Systems via Planning -Phase Prompt Injection
链接: https://arxiv.org/abs/2607.16199
作者: Yuhang Wang
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:Multi-agent LLM systems increasingly rely on a Planner to decompose goals into sub-task sequences that downstream Executor and Critic agents execute and audit. We identify the planning phase as a critical attack surface: a single injection into the Planner’s context achieves cascade amplification, corrupting all downstream sub-tasks simultaneously. We introduce PlanFlip, a framework comprising four planning-phase prompt injection attacks – GoalSubstitution (PF-1), PriorityInversion (PF-2), ContextPollution (PF-3), and RoleConfusion (PF-4) – each disguised as plausible tool outputs to evade keyword filters. Evaluating nine frontier LLMs across 3,479 episodes, we uncover three findings: (1) capability amplifies vulnerability – GPT-5 achieves the highest attack success rate (ASR = 0.68), contradicting the assumption that stronger models are inherently more secure; (2) homogeneous pipelines exhibit a correlated-agent blind spot – GPT-4o and Llama-3.3-70B show ASR near 0 yet Stealth = 1.00 and StepShift 0, with attacks restructuring plans while the same-backbone Critic reports alignment (two independent judges confirm -0.20 to -0.32 semantic deviation, r = 0.943); (3) reasoning-augmented models resist injections – DeepSeek-R1 achieves StepShift = 0.00 across all attacks. We propose GoalAnchorCheck (D1) and CrossAgentConsensus (D2), achieving detection rates up to 1.00 and outperforming same-backbone baselines in 15 of 16 cells. Our key insight: heterogeneous model diversity is a security prerequisite for multi-agent systems; redundancy within a homogeneous backbone provides no protection against planning-phase attacks.
[AI-205] A Survey on GNN-based Link Prediction: Techniques Applications and Challenges
链接: https://arxiv.org/abs/2607.16198
作者: Chengcheng Sun,Yajie Song,Cheng Zhai,Jiayun Tian,Jia Yang,Xiaobin Rui,Jian Zhang,Zhixiao Wang,Philip S. Yu
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Social and Information Networks (cs.SI)
备注: Submmit to WIREs: Data Mining and Knowledge Discovery. This version of the article has been accepted, after peer review but is not the version of record. The final version will be available at: this https URL . Paper list at Github: this https URL
Abstract:Graph Neural Networks (GNNs) have emerged as the leading paradigm for link prediction, enabling the inference of missing connections and the anticipation of potential future links. However, existing reviews lack systematic exploration specifically targeting underlying GNN architectures and diverse graph structures. To address this critical gap, this paper provides a comprehensive review of GNN-based link prediction from a novel and dedicated GNN perspective. We propose an innovative taxonomy that categorizes recent advancements based on techniques and applications. From a technique perspective, we focus on key GNN encoder architectures, including GCN-based, GAE-based, GAT-based, and GFormer-based methods, discussing their strengths and limitations. From an application perspective, we highlight prominent use cases of link prediction in knowledge graphs and recommendation systems, demonstrating their real-world impact. In addition, we examine the current challenges and discuss promising future directions.
[AI-206] Some Large Language Models Exhibit Consistent Risk Attitudes
链接: https://arxiv.org/abs/2607.16197
作者: Bowen Sun,Rui Min,Yuxi Wang,Brian Odegaard,Qi Wang,Jing Du
类目: Artificial Intelligence (cs.AI)
备注: 38
Abstract:As artificial intelligence systems are deployed in open-ended, high-stakes settings, a critical dimension remains unmeasured: how perceived risk is translated into action. We test whether large language models (LLMs) exhibit systematic and consistent risk attitudes under uncertainty. We introduce a cross-domain framework that decouples contextual risk belief from categorical decision, and apply it to six representative LLMs and 100 human participants across spatial navigation, clinical triage, and financial allocation tasks. Using regression models, we extract each agents belief-to-decision mapping and quantify risk sensitivity and risk attitude bias. We find that most tested LLMs exhibit (i) robust intra-task consistency, indicating stable mappings from contextual belief to risk decision within a fixed task domain; (ii) cross-domain rank-order stability, preserving relative risk posture across tasks; and (iii) a convergence toward a restricted risk-attitude distribution relative to the broader human baseline. These results reveal risk attitude as a stable and previously uncharacterized dimension of LLM behavior, establishing a foundation for evaluating and aligning AI systems in open-ended decision-making and motivating further investigation into the origins of these intrinsic behavioral dispositions.
[AI-207] Design and Validation of a Lightweight 1D CNN for Affective Touch Classification in Soft Plush Companions
链接: https://arxiv.org/abs/2607.16196
作者: Aleksandrs Vališevskis,Aleksandrs Okss,Inese Tīģere,Aleksejs Kataševs,Dina Bethere,Anete Hofmane,Airisa Šteinberga,Undīne Gavriļenko,Santa Meļķe,Lucie Matoušková
类目: Artificial Intelligence (cs.AI)
备注: 28 pages, 11 figures
Abstract:Soft, sensorized companions offer a physically safe and emotionally intuitive interface for socially assistive technologies, yet their deformability and multichannel tactile sensing complicate the robust interpretation of human affect. This study presents a complete open-source MATLAB-based framework for the development and validation of compact deep learning models for affective touch recognition in soft interactive companions. As a primary contribution, a diverse FAIR-compliant dataset of 1326 labelled gesture sequences collected from 25 participants spanning children, teenagers, and adults is made publicly available, providing a reusable resource for future research in affective touch recognition. Through systematic architecture and hyperparameter exploration across 468 CNN models, the study identifies compact dilated one-dimensional convolutional neural networks (1D CNNs) as the most effective solution, with a 13.2k-parameter model achieving 75% test accuracy and 85% mean leave-one-subject-out cross-validation accuracy. Theoretical inference-time analysis shows that quantized deployment requires 3.2 MMAC per window, compatible with 20 Hz real-time operation on the target microcontroller. PC-based real-time simulation with the physical toy streaming sensor data demonstrates that the CNN resolves subtle social touches that the previous heuristic system failed to detect, whereas high-force negative interactions are captured more reliably by trivial threshold-based logic. The resulting hybrid inference pipeline - instantaneous heuristic filtering followed by CNN-based nuanced gesture classification - is proposed as the embedded deployment strategy. The study demonstrates that emotionally meaningful, privacy-preserving touch interpretation is computationally feasible for direct embedding within soft therapeutic companions, with hardware integration addressed in a forthcoming study.
[AI-208] Rater State Bias in RLHF Preference Data: An Audit Framework
链接: https://arxiv.org/abs/2607.16195
作者: Elena Kopteva,Vitaliy Hlynianyi-Zhuk
类目: Artificial Intelligence (cs.AI)
备注:
Abstract:We identify a structured confound in Reinforcement Learning from Human Feedback (RLHF). Pairwise preference labels are intended to reflect the compared outputs, but they may also reflect the rater’s state during annotation. Under sustained stressful or distressing conditions, raters’ preferences may shift over time. As a result, preference data can encode rater state alongside judgments about response quality. These shifts differ from ordinary disagreement or random label noise. They are state dependent, can be shared across annotators working under similar conditions, and can propagate through reward modeling and policy optimization. We therefore propose rater state shift as a plausible and testable source of structured bias in RLHF preference data. This paper develops a hypothesis and an audit framework for studying this source of bias. We define rater state shift, rater state confound, and correlated rater state bias. We also define survival level emotional authenticity as a measurable response pattern using lexical, pragmatic, discourse, and safety related features. We analyze how correlated rater state bias can survive aggregation and enter learned reward signals. We derive five falsifiable predictions and effect size thresholds for an initial audit. Finally, we present an audit protocol and pilot study plan that can be applied to publicly available instruction tuned models. We do not infer the training history of any specific deployed model. Our goal is to isolate a plausible and testable source of structured bias in RLHF preference data.
[AI-209] Chemical filters for ultra-high-throughput materials screening and generation
链接: https://arxiv.org/abs/2607.17910
作者: Kinga O. Mastej,Panyalak Detrattanawichai,Hyunsoo Park,Anthony Onwuli,Masahiro Negishi,Aron Walsh
类目: Materials Science (cond-mat.mtrl-sci); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 19 pages, 7 figures, 3 tables, including Supplementary Information
Abstract:Generative artificial intelligence is rapidly transforming materials design by enabling de novo exploration of immense chemical spaces. Yet a large proportion of AI-generated compositions remain implausible, violating established chemical principles, which limits the reliability and interpretability of generative materials design. Here, we introduce a chemical validity operator that recasts heuristic chemical rules as a configurable algorithmic prior for evaluating and guiding generative materials discovery. Built on the open-source SMACT package, a data-informed oxidation-state model exposes tunable thresholds, allowing users to interpolate continuously between permissive and conservative chemical constraints, while supporting both exploratory and conservative materials-design workflows. Benchmarking six state-of-the-art generative models for inorganic crystals shows that most reproduce stoichiometry but under-represent realistic oxidation-state combinations, and that filtering removes compositions reliant on rarely observed oxidation states while preserving low-energy compounds near the convex hull. Beyond screening, the same operator can also serve as a reinforcement-learning reward, steering a latent diffusion model towards chemically grounded compositions. By encoding chemical heuristics and observations, this work establishes a foundation for oxidation-state-aware generative models.
[AI-210] Mathematical Discovery in the Wild: AI-Guided Proofs in Banach Space Theory
链接: https://arxiv.org/abs/2607.17388
作者: Antonio Acuaviva,Pablo Acuaviva
类目: Functional Analysis (math.FA); Artificial Intelligence (cs.AI)
备注:
Abstract:We investigate the capacity of current language models to contribute to mathematical research. In Banach space theory, AI systems generated key ideas and proofs for five new results, which were then verified and refined by humans. We also developed an automated system that searches the literature for open problems and attempts solutions at scale. Our results show both the potential of language models for mathematical discovery and the continuing importance of expert verification.
[AI-211] Alignment of a Total Automation Economy
链接: https://arxiv.org/abs/2607.17015
作者: David McAllester
类目: Physics and Society (physics.soc-ph); Artificial Intelligence (cs.AI)
备注:
Abstract:We consider economic theory from the perspective of a total automation economy, one with no human involvement in production either in manufacturing or in management. One can naturally ask whether a total automation economy is fundamentally a centrally planned economy or, alternatively, whether efficiency demands decentralization into local decisions by competing agents – agentic production. A soviet economist, Leonid Kantorovich, developed linear programming as a method companies or governments can use to optimize production. Ironically, he is also generally credited with showing that the most efficient production is achieved through decentralization – a free market economy with competing agents. I will call Kantorovich’s result the agentic theorem and review it in detail here. This analysis appears to reveal a alignment vulnerability under which agentic management of the economy diverges from human values as agents deveop motives that are “internal” to the economy and disconnected from prices derived from human consumption markets.
[AI-212] Optimizing Clinical Trial Protocols Using EHR-Derived Heterogeneous Treatment Effects
链接: https://arxiv.org/abs/2607.16934
作者: Xiaodi Li,Munhuwan Lee,Pengyang Li,Xiaoke Liu,Jose K. James,Patricia A. Pellikka,Cui Tao,Nansu Zong
类目: Applications (stat.AP); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 21 pages, 7 figures
Abstract:Traditional randomized trials often obscure clinically meaningful heterogeneity in treatment response by focusing on average effects. Leveraging real-world data to emulate clinical trials and estimate heterogeneous treatment effects (HTEs) offers a promising path toward more precise and efficient trial design. In this study, we emulate the DAPA-HF trial using electronic health records from the Mayo Clinic Cloud (MCC) to investigate whether HTE-guided stratification can identify patient subgroups with distinct treatment responses to dapagliflozin versus placebo in patients with heart failure with reduced ejection fraction. All-cause mortality was evaluated using Cox proportional hazards models, with HTEs estimated using a Meta-S learner and subgroups defined using a decision tree-based thresholding approach. In the overall cohort of the emulation, no significant treatment difference was observed (HR, 1.681; 95% CI, 0.828-3.413; p = 0.1507). However, compared with the overall emulated cohort, in which dapagliflozin showed no statistically significant survival benefit, HTE-driven stratification identified subgroups with significant and directionally distinct treatment effects. The beneficial (low-HTE) subgroup showed a significant survival benefit from dapagliflozin (HR = 0.203, 95% CI, 0.087-0.476, p = 0.0002), whereas the harmful (high-HTE) subgroup showed a significant harmful association with markedly increased mortality risk (HR = 6.680, 95% CI, 2.759-16.171, p 0.0001). These findings indicate that HTE-guided stratification can uncover clinically meaningful beneficial and harmful treatment-effect patterns that are masked in the full-cohort emulation.
[AI-213] A Deep Reinforcement Learning Algorithm for the Vehicle Routing Problem with Stochastic Demands and Outsourcing
链接: https://arxiv.org/abs/2607.16875
作者: Mohsen Dastpak,Fausto Errico,Ola Jabali
类目: Optimization and Control (math.OC); Artificial Intelligence (cs.AI)
备注:
Abstract:We introduce the vehicle routing problem with stochastic demands and outsourcing options (VRP-SDO), in which a logistics service provider partitions customer requests into customers outsourced to a common carrier and customers committed to its fixed fleet. The latter induces a vehicle routing problem with stochastic demands (VRP-SD), solved dynamically. Demands are revealed upon visit; residual demand may be served by other vehicles or after restocking at the depot. Work beyond the regular shift incurs overtime costs, and the unit outsourcing cost decreases with the expected outsourced demand. The objective is to minimize expected travel, overtime, and outsourcing costs. We propose an iterative two-level methodology whose first level partitions customers into committed and outsourced subsets, while the second level estimates the expected VRP-SD routing cost. To avoid solving this problem from scratch at every iteration, we learn an offline routing policy that estimates costs almost instantly for any committed subset. An iterated local search establishes the first-level partitions. We formulate the second level as a Markov decision process and solve it with a deep Q-network whose state is represented by a graph attention network aggregating customer and vehicle information by relevance to the acting vehicle. Trained offline on instances with variable customer cardinality and locations, the policy applies to any daily customer realization; online fine-tuning improves the cost approximation. Experiments show that our policy reduces routing costs by 19.6% relative to a state-of-the-art method and by at least 29.6% over classical heuristics. Our overall algorithm saves 13.7% on average over the version without the attention-based representation and generates high-quality decisions within minutes, whereas benchmarks without an offline-trained estimator require over an hour.
[AI-214] RealDESED: A Real-World Domestic Sound Event Detection Benchmark
链接: https://arxiv.org/abs/2607.16736
作者: Florian Schmid,Paul Primus,Alexander Fichtinger,Tara Jadidi,Tobias Morocutti,Gerhard Widmer
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Sound (cs.SD)
备注: Submitted to the DCASE 2026 Workshop (Detection and Classification of Acoustic Scenes and Events). Resources: Dataset (Zenodo): this https URL code and baseline implementation (GitHub): this https URL
Abstract:This paper presents RealDESED, a real-world domestic sound event detection (SED) benchmark comprising 5,710 audio recordings collected by 652 participants in their homes. Each recording is between 15 and 35 seconds long and contains temporally precise annotations for 15 common domestic sound classes. In contrast to existing SED datasets, which typically rely on simulated soundscapes or broad web-crawled audio, RealDESED consists exclusively of recordings captured in natural domestic environments, reflecting realistic variability in recording devices, device placement, acoustic conditions, background sounds, and naturally occurring event co-occurrences. A distinguishing characteristic of the dataset is its multi-annotator labeling scheme, where each recording is independently annotated by multiple annotators, while the validation and test sets undergo an additional review process to ensure high annotation quality and reliable benchmarking. Furthermore, the dataset provides rich metadata, including recording device, device placement, environment labels, and textual scene descriptions. We establish a strong transformer-based baseline and investigate annotation aggregation strategies, post-processing methods, long-form inference, and the impact of recording metadata on model performance. Our baseline achieves a macro-averaged PSDS1 score of 0.731 on the test set. We believe RealDESED provides a valuable benchmark for developing and evaluating robust SED systems under realistic domestic conditions, helping to bridge the gap between current research benchmarks and real-world deployment.
[AI-215] A Causal Markov Condition for Value UAI2026
链接: https://arxiv.org/abs/2607.16717
作者: Olav Benjamin Vassend
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at UAI 2026
Abstract:This paper proposes a causal independence principle for value – the value Causal Markov Condition (v-CMC) – and develops the conceptual and mathematical foundations of a “causal value theory” linking causality and utility. After motivating a local formulation of the v-CMC, we introduce a probability-value duality that translates standard causal-inference results into the value setting. In particular, we formulate local, global, and decomposition versions of the v-CMC and prove their equivalence. We also define v-separation and show that it is sound and complete for conditional value independence. Furthermore, we derive a Bellman-type recursion as a special case of the v-CMC, thereby generalizing standard Bellman recursion from linear chains to causal DAGs. Finally, we show how the v-CMC supports modular transfer and updating of utility information across causal contexts and develop algorithms for causally structured utility elicitation and canonical influence-diagram construction.
[AI-216] Joint-Embedding Predictive Architecture for Sensor-based Activity Recognition
链接: https://arxiv.org/abs/2607.16350
作者: Mohd Halim Mohd Noor,Abdulrahman M. A. Baraka
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Sensor-based human activity recognition (HAR) has achieved significant progressed in fully supervised learning settings. However, these supervised learning models rely on large amount of labeled data, which require labor-intensive collection and meticulous annotation. To address these challenges, this paper proposes a Joint Embedding Predictive Architecture framework tailored for sensor-based HAR, designed to learn robust and generalizable representations from unlabeled datasets. The proposed framework features an encoder designed to explicitly model both the fine-grained local temporal representations within individual window and the long-term temporal sequence of adjacent windows. Furthermore, we introduce an improved Variance-Invariance-Covariance Regularization (VICReg) objective function that incorporates computationally lightweight norm term to stabilize the JEPA pre-training phase. This term balances variance, invariance and covariance constraints to prevent representation collapse. The proposed HAR-JEPA framework is evaluated using two benchmark continuously performed activity datasets. The results show that high-quality representations are successfully learned by the proposed framework. Furthermore, the representations learned by HAR-JEPA demonstrates superior generalization on minority, high variance transitional activities such as sit-to-stand and sit-to-lie where supervised learning tend to overfit due to limited support.
[AI-217] Efficient EEG Seizure Detection Using INT8 Quantization Channel Pruning and Spiking Neural Networks
链接: https://arxiv.org/abs/2607.16296
作者: Kartikey Ahlawat
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:
Abstract:Continuous EEG monitoring for epilepsy is constrained by the limited power and memory budgets of wearable and implantable devices. Deep neural networks can detect seizures with high accuracy, but their computational cost and model size make them difficult to deploy on such platforms. In this work we use a single 1D CNN seizure detector on the CHB-MIT scalp EEG dataset as a common baseline, and then investigate three brain-inspired efficiency strategies: (i) conversion of the CNN into a spiking neural network (SNN) via parameter transfer, (ii) EEG channel pruning combined with 2:4 structured weight sparsity, and (iii) INT8 quantization using FX- and ONNX-based workflows, including quantization-aware training and operator fusion. The quantized CNN variants reduce stored model size from 1.63 MB to 0.44 MB, lower estimated energy per inference by up to 64%, and achieve as much as 2.8 times speedup in CPU latency while preserving, and in one case slightly improving, AUC. The pruned CNN halves the number of input channels and non-zero weights with only a modest accuracy drop, and the SNN conversion provides a spiking implementation with sparse temporal activity. Together, these experiments characterize three complementary efficiency directions for seizure detection.
[AI-218] Intelligence-Guided Adaptive Purification for DDoS-Resilient Quantum Networks: A CUDA-Q based Study
链接: https://arxiv.org/abs/2607.16276
作者: Santanu Ganguly
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Networking and Internet Architecture (cs.NI)
备注:
Abstract:Quantum-repeater networks require adaptive control policies that balance entanglement generation rate, end-to-end fidelity, purification overhead, and memory-induced latency. This tradeoff becomes more complex when the classical control plane is degraded by cyber anomalies or denial-of-service traffic. We develop a CUDA-Q/SeQUeNCe co-simulation workflow for studying adaptive entanglement purification in heterogeneous linear repeater chains. CUDA-Q noisy quantum kernels are used to estimate primitive entanglement purification and swapping behavior, while SeQUeNCe provides an event-layer model for stochastic link generation, waiting-time-dependent memory decay, purification failure, and end-to-end swapping. Under stationary conditions, we compare no purification, local threshold purification, mean-field predictive purification, fixed purification, and a resource-penalized risk-aware predictive policy. In an 8-node chain, the resource-penalized risk-aware controller increases above-target delivery probability relative to fixed purification while reducing latency and purification overhead. We then couple the quantum-network controller to anomaly scores derived from the CSE-CIC-IDS2018 benign-to-SSDP intrusion-detection trace. During the attack period, the attack-unaware controller maintains high raw delivery, but its above-target entanglement delivery falls to 0.098+/-0.007; the IDS-aware resource-adaptive controller switches to more purification-heavy masks and increases above-target delivery to 0.344+/-0.011, closely matching the oracle-aware value of approximately 0.335. These results demonstrate that cyber-state awareness can improve useful quantum-network outcomes by trading raw throughput for fidelity-qualified entanglement delivery.
[AI-219] Physics-Informed Feature Engineering 1D-CNN for Multilayer Cloud Detection from Geostationary Satellites
链接: https://arxiv.org/abs/2607.16270
作者: Fu Wang,Chi Yang,Qi-Feng Lu,Rui-Xia Liu,Xiao-Fei Yang,Xiao-Fang Liu,Bo Li,Lin Chen
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI); Atmospheric and Oceanic Physics (physics.ao-ph)
备注:
Abstract:Multilayer cloud detection from active–passive observation is vital for numerical weather prediction. In this study, channel selections derived from threshold-based algorithms are embedded as feature-engineering priors into a 1D-CNN, and machine learning (ML) is used to learn latent physical relationships to simplify physical retrievals for operational deployment. The results show that the 1D-CNN achieves a multilayer-cloud probability of detection ( \mathrmPOD\mathrmmul ) of 0.620 and a false alarm rate ( \mathrmFAR\mathrmmul ) of 0.240, outperforming the conventional threshold algorithm ( \mathrmPOD\mathrmmul = 0.558 , \mathrmFAR\mathrmmul = 0.369 ). These results demonstrate that prior physical knowledge derived from radiative transfer theory can serve as an effective feature-engineering prior. Further experiments show that ML-revealed physical mechanisms can also enhance traditional algorithms. Replacing AGRI channel 12 (C12, centered at 10.8~\mu\mathrmm ) with channel 13 (C13, centered at 12.0~\mu\mathrmm ) increased \mathrmPOD\mathrmmul from 0.558 to 0.609 without materially affecting \mathrmFAR\mathrmmul . However, for AHI, substituting the 11.2~\mu\mathrmm channel with the 12.3~\mu\mathrmm channel yielded negligible improvement. In addition to spectral response function (SRF) mismatches, a primary contributing factor is the channels’ on-orbit radiometric stability. Hence, physics-informed machine-learning methods appear promising for advancing remote-sensing AI, while sensor-specific characteristics must be considered during operational transfer.
[AI-220] From Intent to Infrastructure: LLM -Driven Agent Compilers for ISAC Networks
链接: https://arxiv.org/abs/2607.16269
作者: Lijie Zheng,Xudong Zhong,Baoquan Ren,Xiangwu Gong,Xinghui Zhu,Ji He
类目: ignal Processing (eess.SP); Artificial Intelligence (cs.AI)
备注:
Abstract:Integrated sensing and communications (ISAC) is moving from proof-of-concept demonstrations to system-level deployment in sixth-generation (6G) networks. Because sensing and communication share hardware, spectrum, and waveform resources, ISAC design now involves many tightly coupled choices, including waveform selection, sensing algorithm setup, resource scheduling, and deployment planning. This design space is already too large to manage well through manual tuning or isolated optimizers. This article introduces the \textitAgent Compiler, a large language model (LLM)-enabled compilation layer that translates high-level engineering intent into complete and executable ISAC system configurations. The Agent Compiler works in four stages: intent parsing, task decomposition, policy graph synthesis, and infrastructure mapping. It produces a verifiable intermediate representation called the ISAC Policy Graph (IPG). A runtime engine then deploys the compiled configuration and supports closed-loop adaptation at three levels: fast parameter updates, partial recompilation of affected subgraphs, and full workflow recompilation. The core design principle is strict time-scale separation: the LLM handles slow-loop strategic decisions, while proven algorithms retain real-time control in the fast loop. A UAV-assisted disaster rescue example illustrates the full compilation process. We also discuss open issues, including compilation latency, output reliability, constraint verification, and pipeline security, to guide future research.
[AI-221] An Agent ic Interface for End-to-End Probabilistic Seismic Hazard and Risk Analysis
链接: https://arxiv.org/abs/2607.16249
作者: Sreenath Vemula,Pierre Jehel,Fabrice Cotton,Filippo Gatti
类目: Geophysics (physics.geo-ph); Artificial Intelligence (cs.AI)
备注: 24 pages, 8 figures
Abstract:Probabilistic seismic hazard and risk analyses are backbone to building codes, insurance pricing, and disaster management. Yet their open-engine pipelines remain accessible primarily to experts. We present the first agentic interface to the end-to-end probabilistic seismic hazard and risk chain via an open-source server, addressable through the Model Context Protocol (MCP). MCP wraps the OpenQuake engine and the 2020 European Seismic Hazard and Risk Models using twenty-four typed endpoints. Here, an agent is defined as a large language model (LLM) with tools. LLM is confined to the role of an orchestrator so it plans, translates, and explains, while the OpenQuake engine and custom codes compute hazard values, damage probability, and loss. Each response carries source-model, ground-motion model (GMM), and certified data provenance for transparency. Results are benchmarked against ESHM20 at seventy-three cities, the replicated 475-year spectral accelerations match official values within a median of 5 %, and a full hazard-to-loss estimate runs in minutes. The interface additionally accepts a user-supplied empirical or machine-learning GMM on any tectonic region type of the published tree, and adds additional features that existing web services omit: conditional spectra, deterministic scenarios, surface hazard, per-building loss, retrofit comparison, and record selection with waveform retrieval. The proposed design layers transfer to other regional models, and to other hazards.
机器学习
[LG-0] Patch Policy: Efficient Embodied Control via Dense Visual Representations
链接: https://arxiv.org/abs/2607.18236
作者: Gaoyue Zhou,Zichen Jeff Cui,Ada Langford,Bowen Tan,Yann LeCun,Lerrel Pinto
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注:
Abstract:Pretrained dense visual features from Vision Transformers (ViTs) are powerful yet have been underutilized in robot learning. Modern robot policies either compress each observation into a single global token, or rely on visual backbones trained from scratch, sacrificing both fine-grained spatial detail and the benefits of large-scale visual pre-training. While there exist policies that do operate on dense patch features like large vision-language-action models (VLAs), they tend to be heavy and slow, inheriting the full cost of a billion-parameter vision-language model (VLM) backbone. We close this gap with Patch Policy, a minimal architectural extension that enables transformer-based policies to consume dense pre-trained patch tokens directly without the computational overhead of a full VLM. At its core is a block-causal attention mask that preserves the temporal causality of standard policies while letting the model attend over many patch tokens per observation, alongside other state information. Patch Policy is lightweight, fast, and highly effective. Across four simulated and three real-world environment suites, our method achieves a 40% relative improvement over policies using state-of-the-art global-pooled representations. Furthermore, it surpasses fine-tuned OpenVLA-OFT by 18% while using roughly 0.7% of the parameters. We believe Patch Policy provides a pipeline for the robotics community to readily leverage continuing progress in visual representation learning, without sacrificing the training efficiency or inference speed required for high-frequency, reactive control. Videos can be viewed at this https URL
[LG-1] Causal Discovery on Irregular Time Series
链接: https://arxiv.org/abs/2607.18226
作者: Martim Penim,Ricardo Ribeiro Pereira,Jacopo Bono,Hugo Ferreira,Mário A.T. Figueiredo,Pedro Bizarro
类目: Machine Learning (cs.LG); Methodology (stat.ME)
*备注:
Abstract:Causal discovery methods have shown strong performance in temporal systems, but they typically rely on regular and discrete lag structures, limiting their applicability to regularly sampled data. However, many real-world tasks require dealing with irregularly sampled streams of events, such as sensor streams, healthcare data, and financial transactions. In this work, we propose an extension of PCMCI+, a state-of-the-art method for causal discovery on regular multivariate time series, to allow for handling irregular time series. Instead of modelling causal relations through fixed-lag dependencies, our method aggregates causal influence over predefined temporal windows. We evaluate our method on synthetic irregular event streams with known causal structures under different signal-to-noise ratios, showing that it consistently recovers the underlying causal graph and substantially outperforms the standard PCMCI+ on irregularly sampled data.
[LG-2] EVOLVE: Efficient Learned Volume Compression with Variable-Rate Encoding on a Cross-Domain Database IEEE-VIS2026
链接: https://arxiv.org/abs/2607.18187
作者: Kaiyuan Tang,Maizhe Yang,Chaoli Wang
类目: Graphics (cs.GR); Databases (cs.DB); Machine Learning (cs.LG)
*备注: To be published in Proceedings of IEEE VIS 2026, IEEE Transactions on Visualization and Computer Graphics
Abstract:Large-scale scientific simulations generate volumetric data at rates that far outpace advances in storage and network bandwidth, making effective lossy compression increasingly critical. However, conventional compressors often struggle to preserve fine structural details at high compression ratios (CRs), and implicit neural representations (INRs) require costly per-volume optimization and produce models with fixed CRs. To respond, we present EVOLVE, an autoencoder (AE)-based volume-compression framework that targets high CRs for offline compression, with three key contributions. First, we construct a large-scale cross-domain database of 6,376 volumes from 21 scientific simulations, curated via perceptual hashing to ensure diversity, enabling the optimized model to extract features that generalize across volumes within the covered scientific simulation domains. Second, we reexamine the design space of AE-based compressors and incorporate several macro- and micro-designs into a vanilla AE to develop EVOLVE, which substantially improves the expressive power and compression capability. Third, we develop a learnable gain mechanism with a three-stage training strategy to enable variable-rate encoding, allowing a single model to support continuous CR adjustment at inference time. Experiments on multiple unseen scientific simulation datasets demonstrate that EVOLVE achieves substantially higher CRs than conventional compressors at comparable reconstruction quality, while delivering compression speeds that are orders of magnitude faster than INR-based methods, highlighting its promise as a strong alternative for compressing scientific data. The code, model weights, and results are available on our project page at this https URL.
[LG-3] FlashRT: Agent Harness for Guiding Agents to Deploy Real-Time Multimodal Applications
链接: https://arxiv.org/abs/2607.18171
作者: Krish Agarwal,Zhuoming Chen,Yanyuan Qin,Zhenyu Gu,Atri Rudra,Beidi Chen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Real-time multimodal applications, including voice agents and interactive video generation, compose heterogeneous models into pipelines whose efficient deployment requires application-specific decisions about placement, streaming, and intra-model parallelism. Existing serving systems and auto-parallelism compilers commit to limited transformations and fixed workload assumptions, so achieving high performance on a new application requires hand-crafting an efficient implementation. We present FlashRT, an agent harness that guides coding agents to lift simple developer-written reference implementations into optimized multi-GPU deployments that flexibly weigh target metrics like latency and throughput. Using a new chain-of-program paradigm, FlashRT directs a generic coding agent through a multi-pass transformation process where an agent transforms the reference into an intermediate representation (IR) to capture data dependencies and persistent-state scopes, validates this IR via a sequential interpreter, and performs static analyses to identify candidate transformations. Then, the agent iteratively implements, verifies, and benchmarks each candidate under a measurement-gated optimization loop to produce effective deployments that span different hardware budgets. Across various applications, including video world models and multimodal LLMs, FlashRT converts reference implementations into highly efficient deployments, delivering up to ~70x latency reduction and 2.8x throughput improvement on NVIDIA B200 GPUs. On AMD MI355X GPUs, FlashRT matches the peak latency reduction while increasing peak throughput improvement to 3.6x, demonstrating that agent-driven optimization can be more scalable on platforms with less mature expert optimization. In fact, for Qwen3-Omni text-to-audio inference, FlashRT reduces response latency by 65% compared to the expert vLLM-Omni implementation on AMD MI355X.
[LG-4] he Calibration Channel Determines the Bayes-Error Proxy: An Exact Law for Temperature-Induced Distortion
链接: https://arxiv.org/abs/2607.18162
作者: Shreyas Pradeepkumar Khandale
类目: Machine Learning (cs.LG)
*备注: 6 pages, 2 figures. Code available at this https URL
Abstract:The soft-label Bayes-error estimator beta(z) = E[min(z, 1-z)] of Ishida et al. estimates the irreducible error of a binary task directly from probability-valued labels. Recent work by Ushio et al. showed that this estimator is fragile when the probabilities are not the true posterior: even perfectly calibrated soft labels can yield a substantially inaccurate estimate, and they propose isotonic calibration as a consistent remedy. We complement that line of work by characterizing exactly how the most widely used post-hoc calibration map – temperature scaling – distorts the proxy. We prove an exact, model-free identity reducing the temperature-scaled proxy to the classifier’s margin distribution, from which we obtain (i) strict monotonicity in the temperature and (ii) a continuous bijection from the temperature axis onto the open interval (0, 1/2), so that a fixed classifier – with fixed decisions and fixed 0-1 error – can be made to report any proxy value whatsoever. Under a Gaussian model of the logits we further derive a two-parameter closed form for the entire proxy-versus-temperature curve. Across CIFAR-10, Fashion-MNIST, and SVHN (eight binary tasks), the proxy varies by 56x to 980x at constant test error, the closed form reproduces the empirical curve to within 0.018, and the calibration temperature that minimizes the expected calibration error does not coincide with any stable proxy value. Our results give a precise, predictive account of the distortion whose existence motivates calibration-based remedies, and they reinforce the practical recommendation that a proxy value is meaningful only together with the mechanism that produced its probabilities.
[LG-5] otally Positive Matrices and the Highest-Order Coefficients of the Characteristic Polynomial
链接: https://arxiv.org/abs/2607.18148
作者: Tiago Closs,Leandro Farina
类目: Machine Learning (cs.LG); Rings and Algebras (math.RA)
*备注: 26 pages, 12 figures, 3 tables. Published open access in Linear Algebra and its Applications
Abstract:We investigate the extent to which totally positive matrices can be distinguished through the highest-order coefficients of their characteristic polynomials. To identify the most informative coefficients, we also employed neural-network classifiers together with feature-attribution methods. Using datasets built from several structured totally positive families, including products of positive bidiagonal matrices, Vandermonde matrices, and Cauchy matrices, we find that the coefficients (a_n-1, a_n-2, a_n-3) already contain strong discriminatory information for separating totally positive from non-totally positive matrices in dimensions 5, 10, and 30. The resulting separation is markedly nonlinear and admits a natural geometric description in the corresponding three-dimensional coefficient space by means of Mahalanobis ellipsoids. These ellipsoids enclose the totally positive samples while excluding most non-totally positive ones. Moreover, different structured totally positive families exhibit distinct ellipsoidal signatures, and the separation between these signatures increases with the dimension. These observations lead us to formulate a conjecture on the geometric separation of structured totally positive families in the space determined by the three highest-order characteristic coefficients.
[LG-6] Manifold-Constrained Hyper-Connections for Parameter-Efficient Finetuning
链接: https://arxiv.org/abs/2607.18130
作者: Valentijn Oldenburg,Floris de Kam,Bente Zuijdam,Lieve Eberson,Nicky van Zutphen,Stef de Wildt,Ivo Verhoeven
类目: Machine Learning (cs.LG)
*备注: Pre-print
Abstract:Most parameter-efficient finetuning (PEFT) methods adapt weights or activations, thus leaving one of the key Transformer components unchanged: residual connections. This paper investigates Manifold-Constrained Hyper-Connections (mHC), a generalisation of residual connections, as a novel PEFT approach, wrapping frozen OLMo-2 backbones with learned residual routing modules. We find that mHC can finetune frozen Transformers, but that its role differs fundamentally from the original pre-training setting: in finetuning, fixing the residual mixing matrix to identity often improves performance. As a standalone PEFT method, mHC does not consistently outperform LoRA. However, at matched trainable parameter budgets, mHC+LoRA combinations improve language-modelling loss and show task-dependent benchmark gains at both 1B and 7B scale. Overall, our results identify residual routing as a distinct and promising novel PEFT axis.
[LG-7] ClouDens: Operational Context-Aware Anomaly Detection for Large-scale Cloud System Monitoring
链接: https://arxiv.org/abs/2607.18127
作者: Thu T. H. Doan,Mohammad Saiful Islam,Andriy Miranskyy,Ngoc-Thanh Nguyen,Rogardt Heldal,Patrizio Pelliccione
类目: Networking and Internet Architecture (cs.NI); Machine Learning (cs.LG)
*备注: 16 pages
Abstract:With the rapid growth of cloud computing infrastructures in scale and complexity, network monitoring for Large-scale Cloud Systems (LCSs) has become increasingly challenging, requiring automated and reliable anomaly detection to maintain service availability. Modern LCSs continuously generate telemetry logs from distributed cloud services, producing high-dimensional multivariate time series that capture system operations. Detecting anomalies in this context is difficult due to extreme dimensionality, complex dependencies among distributed components, and severe sparsity from intermittently active services. Taking these challenges into account, we first conduct an empirical study on telemetry logs from the IBM Cloud Console platform, and then propose ClouDens, an anomaly detection framework tailored to LCS monitoring that leverages operational-context attributes encoded in the telemetry log schema to improve detection accuracy and early identification of anomalies. ClouDens partitions high-dimensional telemetry logs into domain-guided subsets, constructs a context-aware graph modeling operational service dependencies, and employs Spatio-Temporal Graph Neural Networks for forecasting-based anomaly detection. We evaluate ClouDens on the recently released IBM Cloud Telemetry Dataset and provide practical insights into designing reliable anomaly detection solutions for LCS monitoring. Results show ClouDens achieves higher NAB scores in count-based telemetry features, indicating more accurate, earlier anomaly detection with broader coverage than a GRU-based model. Our study further reveals that telemetry feature subsets, operational-context modeling, scoring strategies, and sparsity imputation all substantially influence detection performance, offering practical guidance for designing and fairly benchmarking anomaly detection approaches for LCS monitoring.
[LG-8] Sobek: Streaming Equivariant Tensor Product Convolutions
链接: https://arxiv.org/abs/2607.18074
作者: Vladimir Chorošajev,Cédric Bény
类目: Machine Learning (cs.LG)
*备注:
Abstract:Equivariant graph neural networks repeatedly apply edge-conditioned tensor-product convolutions over graph edges. Conventional implementations materialize edge-specific weights, messages, and adjoints, causing tensor-product workspace and memory traffic to grow rapidly with graph size and operator width. This limits feasible workloads and can prevent larger problems from fully utilizing the GPU. We show that these edge-sized intermediates are artifacts of the execution schedule, not requirements of the equivariant operator. By reassociating radial projection, spherical-harmonic coupling, and graph aggregation, edge-local products can be consumed directly into bounded receiver-side state. The resulting streaming formulation preserves fully connected multiplicity mixing and extends through forward, backward, and double backward. We implement this formulation in Sobek, a generated-CUDA backend, and evaluate it across edge-scaling regimes and varied feature structures. Across two operator families and all three differentiation orders, Sobek is faster in all 75 capacity-matched comparisons, with speedups ranging from 1.2\times to 49.7\times , and reduces peak allocated memory by up to 99%. It also executes workloads up to two orders of magnitude beyond OpenEquivariance’s capacity while retaining near-peak throughput. These results show that edge-scaled tensor-product workspace is a property of the conventional schedule, not of equivariant convolution itself. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.18074 [cs.LG] (or arXiv:2607.18074v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.18074 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-9] Hardware Mechanisms to Dynamically Throttle AI Performance
链接: https://arxiv.org/abs/2607.18069
作者: Haiyue Ma,Lauren Malek,Joseph Forzani,David Wentzlaff
类目: Hardware Architecture (cs.AR); Machine Learning (cs.LG)
*备注:
Abstract:As more capable AI models are increasingly integrated into critical computer systems, the lack of control over AI intent motivates safety mechanisms. Existing software safeguards impose only behavioral constraints that can potentially be bypassed by sufficiently intelligent models. While hardware-level safety enforcement has been recognized as an essential last line of defense, few mechanisms have been proposed beyond policy regulations on unauthorized accesses or coarse full-chip shutdown. What is missing is a fine-grained, dynamic intervention mechanism at the architecture level. In this paper, we introduce a set of microarchitecture knobs which dynamically control the available hardware resources to limit AI performance at runtime. We evaluate candidate knobs spanning the GPU memory subsystem, across capacity, bandwidth, latency and frequency dimensions, and narrow down to four strong candidates: L2 size, L2 latency, L2 bandwidth, and shared memory port access rate. To minimize new logic and extra design cost, we build all four mechanisms from well-established microarchitectural primitives: cache way masking, credit-based rate limiting, latency insertion, and bank arbitration. We show that these knobs achieve high performance sensitivity (up to 80% performance cut at 1/8 resource availability), negligible implementation cost (~10K flip flops), fast stabilization after dynamic throttling (5-80K cycles), and minimal collateral impact on the rest of the chip. Further, multi-knob analysis reveals combinations of knobs that amplify the performance degradation beyond the effect of each knob individually, which enables a broader range of performance targets. Subjects: Hardware Architecture (cs.AR); Machine Learning (cs.LG) Cite as: arXiv:2607.18069 [cs.AR] (or arXiv:2607.18069v1 [cs.AR] for this version) https://doi.org/10.48550/arXiv.2607.18069 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-10] SEE: Structure-aware Exploring Exploiting for Long-horizon GUI Agent Trajectory Synthesis ACM-MM2026
链接: https://arxiv.org/abs/2607.18046
作者: Zhuohang Fan,Beichen Zhang,Yuanfa Li,Changqiao Wu,Wei Liu,Jian Luan,Weigang Zhang
类目: Machine Learning (cs.LG)
*备注: Accepted by ACM International Conference on Multimedia 2026 (ACM MM 2026)
Abstract:Graphical User Interface (GUI) agents powered by vision-language models hold promise for automating real-world mobile tasks. However, progress is limited by the lack of high-coverage, long-horizon interaction trajectories collected from element-rich and rapidly evolving apps. Existing pipelines often rely on costly human demonstrations or on-policy framework, which tends to over-sample common flows while missing rare transitions and complex multi-step procedures. To address this problem, we propose SEE, a two-stage data synthesis framework consisting of (i) an efficient exploration stage that builds an explicit UI transition graph over screens and elements, and (ii) a graph-based synthesis stage that composes diverse multi-step trajectories via planning and controlled sampling. This design yields reproducible and explainable data generation, while explicitly preventing spurious cycles and enabling long-horizon composition. Across multiple real-world apps, SEE produces trajectories with an average length of 14.8 steps while avoiding spurious loops, and agents fine-tuned on SEE achieve improved task success and generalization to unseen screens. We will publicly release our synthesis code and dataset.
[LG-11] Adaptive Mamba Neural Operators ICLR2026
链接: https://arxiv.org/abs/2607.18043
作者: Zeyuan Song,Zheyu Jiang
类目: Machine Learning (cs.LG); Analysis of PDEs (math.AP); Numerical Analysis (math.NA)
*备注: 22 pages, accepted by ICLR 2026 conference ( this https URL )
Abstract:Accurately solving partial differential equations (PDEs) on arbitrary geometries and a variety of meshes is an important task in science and engineering applications. In this paper, we propose Adaptive Mamba Neural Operators (AMO), which integrates reproducing kernels for state-space models (SSMs) rather than the kernel integral formulation of SSMs. This is achieved by constructing Takenaka-Malmquist systems for the PDEs. AMO offers new representations that align well with the adaptive Fourier decomposition (AFD) theory and can approximate the solution manifold of PDEs on a wide range of geometries and meshes. In several challenging benchmark PDE problems in the fields of fluid physics, solid physics, and finance on point clouds, structured meshes, regular grids, and irregular domains, AMO consistently outperforms state-of-the-art solvers in terms of relative L^2 error. Overall, this work presents a new paradigm for designing explainable neural operator frameworks.
[LG-12] FlashPDE: A Drop-in Fused Triton Operator Library for Neural PDE Solvers
链接: https://arxiv.org/abs/2607.18020
作者: Peiyu Zang,Bosen Xie,Ruoxiang Xu,Yongqiang Cai
类目: Machine Learning (cs.LG); Mathematical Software (cs.MS)
*备注: 9 pages of main text, 4 figures. Code available at this https URL
Abstract:Physics-Informed Neural Networks (PINNs) solve PDEs by incorporating physical constraints into neural-network training, but large-scale problems are limited by automatic-differentiation memory overhead and inefficient execution of grid-based PDE operators. We present FlashPDE, a drop-in fused operator library for grid-based scientific machine learning. FlashPDE replaces fragmented PyTorch finite-difference execution with differentiable Triton kernels. Each operator integrates fused stencil evaluation, an analytic discrete-adjoint backward pass, and boundary-gradient correction within a unified this http URL interface. The library provides 14 differentiable PDE operators covering 17 configurations across 1D–3D elliptic, parabolic, and Navier–Stokes systems, while remaining independent of neural architectures and training strategies. Experiments on an NVIDIA A100 GPU show that FlashPDE reduces peak memory usage by up to 37.0x compared with coordinate-based automatic differentiation and reduces CUDA kernel launches by up to 3.5x compared with eager PyTorch finite-difference implementations. Across six representative PDE benchmarks, FlashPDE achieves up to 2.30x end-to-end time-to-solution speedup and up to 19.2x kernel-level acceleration while maintaining numerical agreement with PyTorch finite-difference references. FlashPDE provides a hardware-efficient execution layer that bridges differentiable PDE solvers and GPU-optimized numerical computation within the PyTorch ecosystem.
[LG-13] fSRD: Fuzzy Spectral Region Decomposition – Automated Multi Operator Koopman Representations via an Adaptive Spectral Learning Architecture
链接: https://arxiv.org/abs/2607.17990
作者: Charles Bokor,Mark Cary,Denise Morrey,Fabrizio Bonatesta
类目: Machine Learning (cs.LG)
*备注: 66 pages, 30 figures, corresponding author = Charles Bokor
Abstract:Highly nonlinear chaotic dynamical systems remain difficult to model due to fundamental trade-offs between complexity, expressivity, and data efficiency. Modern machine learning methods achieve strong predictive performance but often rely on a-priori system knowledge or curated data with limited interpretability. Koopman operator theory offers a promising direction via linear representation in an infinite-dimensional observable space. However, many data-driven Koopman methods seek globally valid operators for which useful finite-dimensional spectral embeddings remain difficult to identify under these constraints. To overcome associated limitations, we introduce Fuzzy Spectral Region Decomposition (fSRD), a fully automated learning framework for estimating finite Koopman representation via multiple operators. The proposed method realizes a data-adaptive framework for assembling locally invariant embeddings, termed Invariant Decomposition. fSRD achieves highly accurate linear reconstructions of nonlinear systems while learning finite-dimensional representations of their induced evolution operators, bridging interpretable operator-theoretic models with expressive data-driven sequence learning. These embeddings are adaptively constructed via a global fuzzy tree model, drawing inspiration from fuzzy neural architectures to learn the induced dynamics while prioritizing parsimonious solutions. Empirical results across canonical chaotic systems (e.g., Lorenz and Duffing) and high-dimensional real-world data demonstrate strong predictive accuracy, interpretability, and robust expressivity across data-rich and data-limited regimes, highlighting the method’s generality.
[LG-14] Information-Based Exploration via Random Features for Reinforcement Learning
链接: https://arxiv.org/abs/2607.17981
作者: Waris Radji,Odalric-Ambrym Maillard
类目: Machine Learning (cs.LG)
*备注:
Abstract:Representation learning has enabled classical exploration strategies to be extended to deep Reinforcement Learning (RL), but often makes algorithms more complex and theoretical guarantees harder to establish. We introduce Random Feature Information Gain (RFIG), grounded in Bayesian kernel methods theory, which uses random Fourier features to approximate information gain and compute exploration bonuses in non-countable spaces. We provide error bounds on information gain approximation and avoid the black-box aspects of neural network-based uncertainty estimation, for optimism-based exploration. We present practical details that make RFIG scalable to deep RL scenarios, enabling smooth integration into standard deep RL algorithms. Experimental evaluation across diverse control and navigation tasks demonstrates that RFIG achieves competitive performance with well-established deep exploration methods while offering superior theoretical interpretation.
[LG-15] DiFA: Inference-Time Forward-Process Alignment for Diffusion Models ICML2026
链接: https://arxiv.org/abs/2607.17972
作者: Shigui Li,Delu Zeng
类目: Machine Learning (cs.LG)
*备注: Accepted to ICML 2026. 23 pages, 7 figures
Abstract:The prevailing inference framework for diffusion models formulates generation fundamentally as a problem of numerical integration. This perspective casts the model as an exact estimator, neglecting the inherent statistical uncertainty of the denoising process. In this work, we propose Forward-Process Aligned Diffusion prediction (\textbfDiFA), a training-free framework that reframes inference-time data prediction refinement as a sequential state estimation problem. Rather than reusing past outputs solely for numerical integration, DiFA treats iterative data predictions along the reverse trajectory as correlated observations to build a forward-aligned temporal consensus. Inspired by Kalman filtering, this consensus aggregates historical predictions according to structural consistency and noise-level compatibility. To counteract the over-smoothing tendency of temporal consensus, we introduce a deviation guidance mechanism to adaptively preserve residual details. Empirically, DiFA yields significant improvements on CIFAR-10 and ImageNet across the evaluated metrics, including FID, IS, and FD-DINOv2, demonstrating that aligning inference with the forward statistical structure substantially improves generative fidelity.
[LG-16] AutoEncoder-Compressed Parallel Split Learning for Pre-trained Model Fine-Tuning ECML-PKDD2026
链接: https://arxiv.org/abs/2607.17913
作者: Bas Meuwissen,Vasileios Tsouvalas,Nirvana Meratnia
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注: Accepted to ECML-PKDD 2026, 4th Workshop on Advancements in Federated Learning (WAFL)
Abstract:Distributed Fine-Tuning (DFT) of large-scale Foundation Models (FMs) on resource-constrained edge devices is limited by local compute constraints and communication overhead. Parallel Split Learning (PSL) reduces client-side computation by keeping few model layers on each client and offloading the remaining computation to the server; however, clients must exchange intermediate activations and gradients with the server at every training step. Existing SL communication-compression methods mainly rely on task-agnostic heuristics, such as sparsification and quantization. While learnable SL compressors can better adapt to intermediate representations, they require co-training with the target model. Therefore, directly inserting them into off-the-shelf FMs introduces feature-distribution misalignment and degrades DFT performance. To address this, we propose AE-PSL, a communication-efficient PSL framework that compresses intermediate activations and gradients using a lightweight AutoEncoder (AE) placed at the split layer. To ensure compatibility of AE compression with pre-trained FMs, AE-PSL introduces a novel two-stage alignment mechanism, which adapts the AE to the pre-trained model’s feature manifold and client-specific feature distributions before DFT.
[LG-17] Distributional Soft Bellm an Operator under the Cramér Geometry
链接: https://arxiv.org/abs/2607.17897
作者: Keru Wang,Yixin Deng,Yao Lyu,Stephen Redmond,Shengbo Eben Li
类目: Machine Learning (cs.LG)
*备注:
Abstract:Distributional soft policy iteration (DSPI) provides an important framework for combining distributional reinforcement learning (DRL) with maximum-entropy control, in which the policy evaluation step is governed by a distributional soft Bellman operator acting on entropy-regularised returns. Theoretical analysis of such an evaluation step requires a probability metric under which Bellman updates can be controlled, typically by showing that the operator contracts the distance between any two candidate return-distribution estimates. In this paper, we focus on the Cramér geometry, a cumulative distribution function (CDF)-based metric with an L^2 structure, and study whether the fixed-policy distributional soft Bellman operator has this contraction property and hence a unique fixed point under this metric. Working directly on an admissible CDF field domain, we formulate the CDF-level distributional soft Bellman operator, prove that it is a \sqrt\gamma -contraction, and obtain the corresponding unique fixed point together with convergent iterative policy evaluation. The CDF formulation also shows that this finite-Cramér-domain property follows from a uniform first-moment condition on the combined one-step reward entropy shift, rather than from separate uniform boundedness assumptions on the reward and entropy terms. We then transport the same evaluation problem to the spectral domain by conjugation, obtaining an equivalent Hilbert-space representation of the same decision process. Taken together, these results identify the Cramér-geometric Bellman fixed point associated with the policy-evaluation step of DSPI, providing a reference point for studying approximate critics, evaluation error, and critic-loss design in DSPI-style algorithms.
[LG-18] Organization of computation in reservoir computing
链接: https://arxiv.org/abs/2607.17858
作者: Mohab Abdalla,Damien Rontani
类目: Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG)
*备注:
Abstract:Reservoir computing exploits nonlinear dynamical systems to encode temporal inputs into high-dimensional state space representations. Although reservoir performance is often characterized through memory, nonlinearity, and their tradeoff, such aggregate measures do not reveal how task-relevant information is organized within the state space. Here, we introduce an eigen-spectral decomposition framework linking the degree-wise information processing capacity to the corresponding state space modes. As a result, we are able to quantify the degree-wise representation energy, and show that in some cases, substantial amounts of information processing capacity may reside in low-energy modes that are vulnerable to experimental noise. These results suggest that useful reservoir computation depends not only on dimensionality expansion, but also on the geometric organization of task-relevant information, with direct implications on physical reservoir computers.
[LG-19] heoretical Foundations of max@k Reinforcement Learning
链接: https://arxiv.org/abs/2607.17823
作者: Riccardo Poiani,Martino Bernasconi,Andrea Celli
类目: Machine Learning (cs.LG)
*备注:
Abstract:Reinforcement Learning is a cornerstone technique for modern large reasoning models. Usually, for difficult tasks such as code generation and theorem proving, the agent is evaluated by generating K responses rather than sampling a single response, and performance is then measured using a retry-aware metric such as \max @ k . Despite their practical importance, the theoretical foundations of learning under such criteria remain limited. In this work, we provide a theoretical study of the \max @ k learning problem in finite-horizon reinforcement learning. We show that optimizing the \max @ k objectives is fundamentally different from standard expected-return maximization. In particular, we prove that Markovian policies are in general insufficient, identify a compact state augmentation that restores optimality, and explicitly characterize the performance gap that can arise between history-dependent and non-history-dependent policies. Moreover, we show that learning \max @ k -optimal policies is statistically harder than standard reinforcement learning and provide an efficient algorithm that achieves the optimal sample complexity rate.
[LG-20] he Concept of Representation in ML: Beyond Plato and Aristotle ICML2026
链接: https://arxiv.org/abs/2607.17800
作者: Gilad Landau,Aviv Keren
类目: Machine Learning (cs.LG)
*备注: 8 pages. Accepted for oral presentation at ICML 2026
Abstract:Representation is a central concept in modern machine learning, where it usually refers to internal encodings that support learning and generalization. As models scale and their capabilities become increasingly human-level, this representational language sometimes shifts from an engineering context into the more philosophically loaded domain of mental representation. We argue that this is the case for recent claims about the convergence of representational properties across different AI models. In particular, we assess the arguments developed in The Platonic Representation Hypothesis, according to which this convergence is driven by a unified structure of reality. We examine this claim by introducing arguments and ideas from debates about mental representation in the philosophy of mind. We argue that these philosophical resources can clarify what is at stake in such claims, explain why alignment evidence alone is insufficient for strong metaphysical conclusions, and suggest directions for future research.
[LG-21] owards Reliable Zero-Shot Crowd Forecasting: Evaluating Time Series Foundation Models for Special Event Pedestrian Forecasting ITSC2026
链接: https://arxiv.org/abs/2607.17758
作者: Ziteng Li,Yanan Xin,Tina Comes,Serge Hoogendoorn
类目: Machine Learning (cs.LG)
*备注: 9 pages, 7 figures, 5 tables. Accepted for presentation at IEEE ITSC 2026
Abstract:Managing massive crowds during infrequent special events requires reliable real-time pedestrian-flow forecasting to ensure public safety and operational efficiency. However, supervised forecasting methods face limitations in these contexts due to scarce historical data, heterogeneous data distributions, and short in-event observation windows. To effectively support operational decision-making, forecasts should provide not only accurate point estimates but also informative predictive uncertainty. Probabilistic uncertainty quantification plays a critical role in this aspect, particularly capturing sudden volatility and tail risks. This paper investigates pretrained time series foundation models as a lightweight approach for zero-shot probabilistic forecasting without extensive local retraining. Using decision-oriented metrics tailored to short events, we conduct a comprehensive assessment of two time series foundation models on crowd forecasting, with the SAIL2025 event as a use case. We then distill practical insights for crowd managers, specifying when zero-shot forecasts remain operationally reliable.
[LG-22] Planning with Transformers: Chain of Computation and Structured Context Windows
链接: https://arxiv.org/abs/2607.17710
作者: Ehsan Futuhi,Nathan R. Sturtevant
类目: Machine Learning (cs.LG)
*备注:
Abstract:Large Language Models (LLMs) have had a remarkable impact across many areas of machine learning. However, recent studies have shown that they struggle to reliably solve planning problems. At the same time, theoretical results have shown that transformers, the core architecture underlying modern LLMs, are Turing-complete. In this work, we investigate this apparent gap between the theoretical computational power of LLMs and their empirical planning performance. We propose Chain of Computation (COC), a computational architecture that places a transformer-based LM inside an iterative loop, leveraging its strength as a pattern-matching system. The COC uses a Structured Context Window (SCW) which provides a constant-sized context window with support for choosing which window is used at each planning step. Within this architecture, the LM is able to learn a planning policy, predicts the world model, and performs the arithmetic operations required during planning. We show that, when given an append-only SCW (resembling a Turing Machine tape), even relatively small LMs trained from scratch can learn planning policies and generalize from a small number of training instances within each planning domain, achieving success rates above 99.89% on BlocksWorld and the Pancake puzzle. Our analysis of failure cases in Tower of Hanoi (TOH) reveals that they arise from arithmetic operations or from encountering previously unseen tokens. We show that COC can solve TOH problem instances with up to 20 disks, requiring over 1 million actions, while requiring substantially less training data by either (1) planning with symbolical support for arithmetic or by (2) using a deterministic pushdown automaton (PDA) formulation for the SCW.
[LG-23] GeneSpeak-FP: Target and Compound Retrieval from Observed Cell-Level Perturbation Signatures
链接: https://arxiv.org/abs/2607.17671
作者: Kseniia Vaniushkina,Jeongmin Lim,Jinyong Park
类目: Machine Learning (cs.LG)
*备注:
Abstract:Large-scale single-cell perturbation atlases make it possible to ask an inverse question: given an observed transcriptional response, which annotated targets and compounds in a fixed library are most consistent with that response? We present \model, a Transformer retrieval model for this closed-library setting. Each input is a cell-level perturbation signature formed by contrasting one treated cell with a cell-line-specific mean DMSO reference. The encoder maps the signature to a target-retrieval vector and a molecular-embedding vector, trained jointly with supervised target losses and structure–transcriptome alignment. We evaluate on Tahoe-100M conditions with mapped target annotations using a within-compound stratified 90/10 condition-pair split of 10,505 training and 1,168 validation drug–cell-line pairs. Because compounds and cell lines can occur in both partitions, the experiment measures held-out condition-pair retrieval rather than generalization to unseen compounds or cellular contexts. In a Monte Carlo evaluation over 38,400 sampled validation cells, \model\ achieved target Recall@10 of 0.408 and Recall@20 of 0.544, together with compound Hit@1 of 0.129, Hit@10 of 0.343, and mean reciprocal rank of 0.205 over a 379-compound bank. A separate diagnostic evaluation produced nearly identical values for the main model and large gains over a random-vector control and post-hoc bag-of-genes controls. These results demonstrate that a single multi-task model can recover both mapped target annotations and recorded compound identities from observed cell-level responses in the evaluated Tahoe-100M closed-library setting. Generalization to unseen compounds and cellular contexts remains to be established.
[LG-24] Can Transformers Really Do It All? On the Compatibility of Inductive Biases Across Tasks ICLR2026
链接: https://arxiv.org/abs/2607.17624
作者: Damien Teney,Liangze Jiang,Hemanth Saratchandran,Simon Lucey
类目: Machine Learning (cs.LG)
*备注: Published as a conference paper at ICLR 2026. this https URL
Abstract:Transformers are remarkably versatile and their design is largely consistent across a variety of applications. But are they optimal for any given task or dataset? The answer may be key for pushing AI beyond merely scaling current designs. Method. We present a method to optimize a transformer architecture for a given dataset, which we use as a tool to study optimal task-specific inductive biases. This method replaces the most important non-linearities (GeLUs,;softmax) with functions learned on held-out data. We then train the resulting architectures on other datasets, as a way to evaluate the compatibility between pairs of tasks. Findings. On algorithmic toy tasks, we identify new architectures with dramatic improvements in learning speed, in- and out-of-distribution generalization, and stability across seeds. The new designs prove very task-specific however, and indicate that these tasks require inductive biases very different from those of standard transformers. On code and language modeling datasets, we also find architectures with consistent, yet smaller improvements. These designs transfer much better across datasets and domains (English computer code). Implications. Our results show that standard transformers are rarely a local optimum in the space of architectures. Simple alternatives can perform much better but sacrifice universality. This suggests that there may be room for improved architectures that better support multiple capabilities simultaneously, such as fluency and robust reasoning. Comments: Published as a conference paper at ICLR 2026. this https URL Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.17624 [cs.LG] (or arXiv:2607.17624v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.17624 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Damien Teney [view email] [v1] Mon, 20 Jul 2026 07:26:17 UTC (1,272 KB)
[LG-25] Online learning of neural state-space models
链接: https://arxiv.org/abs/2607.17614
作者: Bendegúz Györök,Tamás Péni,Maarten Schoukens,Roland Tóth
类目: ystems and Control (eess.SY); Machine Learning (cs.LG)
*备注: Submitted to L-CSS. Extended version
Abstract:Recent advances in deep-learning-based nonlinear system identification have led to encoder-based estimation of neural state-space (ANN-SS) models that achieve state-of-the-art performance in offline settings by estimating initial model states from past input-output data. These methods are typically used in multiple-shooting-based offline identification, and online learning of these models remains largely unexplored. This paper presents a batch-wise learning pipeline and a direct recursive identification algorithm for subspace encoder-based ANN-SS models. We provide convergence analysis of the recursive formulation and validate its performance through extensive simulation studies. The results demonstrate that the proposed approach enables computationally efficient online adaptation with high model accuracy.
[LG-26] Optimizing the Preconditioner: A Black-box Online-to-Nonconvex Conversion with Static Regret Minimization Oracles
链接: https://arxiv.org/abs/2607.17607
作者: Haichen Hu,David Simchi-Levi
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Machine Learning (stat.ML)
*备注:
Abstract:We study whether stochastic nonconvex optimization can be reduced to ordinary static regret minimization in online convex optimization in a black-box manner. For smooth nonconvex objectives, our reduction maintains a predictable gradient tracker, while a black-box online learner selects a preconditioner that determines how this tracker is transformed into the update direction. The learner receives linear convex losses and is evaluated against a single fixed comparator over one undiscounted online game. For a \beta -smooth objective with range bounded by M and an unbiased stochastic-gradient oracle with variance bounded by (\sigma^2), we establish \frac1T\sum_t=1^T \mathbb E!\left[|\nabla f(x_t)|_2^2\right] \lesssim \frac\sigma\sqrtM\beta\sqrt T + \frac\sqrtM\beta, \mathscr R_T(\mathcal A,I_d)T + \fracM\betaT. Consequently, any black-box OCO algorithm with \mathscr R_T(\mathcal A,I_d)=O(\sqrt T) recovers the classical O(\frac1\sqrtT) convergence rate. We further show that the same black-box framework extends beyond the smooth setting to Lipschitz nonconvex objectives without Lipschitz continuous gradients. Importantly, this extension continues to rely only on an ordinary static-regret guarantee and requires no stronger notion of online regret. When the OCO oracle admits square-root static regret, the resulting conversion achieves the optimal O(T^-2/7) convergence rate for the corresponding Goldstein stationary point. These results resolve the open problem posed by Chen and Hazan (2024). More broadly, our framework separates optimizer design into gradient prediction and online preconditioner selection, providing a principled perspective on how adaptive optimization methods such as AdaGrad and Shampoo may be understood through static regret and applied in nonconvex optimization. Subjects: Machine Learning (cs.LG); Optimization and Control (math.OC); Machine Learning (stat.ML) Cite as: arXiv:2607.17607 [cs.LG] (or arXiv:2607.17607v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.17607 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Haichen Hu [view email] [v1] Mon, 20 Jul 2026 06:53:59 UTC (74 KB) Full-text links: Access Paper: View a PDF of the paper titled Optimizing the Preconditioner: A Black-box Online-to-Nonconvex Conversion with Static Regret Minimization Oracles, by Haichen Hu and 1 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: cs.LG prev | next new | recent | 2026-07 Change to browse by: cs math math.OC stat stat.ML References Citations NASA ADSGoogle Scholar Semantic Scholar export BibTeX citation Loading… BibTeX formatted citation loading… Data provided by: Bookmark checked="checked"class=“labs-tab-input”> Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer (What is the Explorer?) Connected Papers Toggle Connected Papers (What is Connected Papers?) Litmaps Toggle Litmaps (What is Litmaps?) scite.ai Toggle scite Smart Citations (What are Smart Citations?) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv (What is alphaXiv?) Links to Code Toggle CatalyzeX Code Finder for Papers (What is CatalyzeX?) DagsHub Toggle DagsHub (What is DagsHub?) GotitPub Toggle Gotit.pub (What is GotitPub?) Huggingface Toggle Hugging Face (What is Huggingface?) ScienceCast Toggle ScienceCast (What is ScienceCast?) Demos Demos Replicate Toggle Replicate (What is Replicate?) Spaces Toggle Hugging Face Spaces (What is Spaces?) Spaces Toggle TXYZ.AI (What is TXYZ.AI?) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower (What are Influence Flowers?) Core recommender toggle CORE Recommender (What is CORE?) IArxiv recommender toggle IArxiv Recommender (What is IArxiv?) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv’s community? Learn more about arXivLabs. Which authors of this paper are endorsers? | Disable MathJax (What is MathJax?) mathjaxToggle(); We gratefully acknowledge support from our major funders, member institutions, , and all contributors. About Help Contact Subscribe Copyright Privacy Accessibility Operational Status (opens in new tab) Major funding support from
[LG-27] Concentration and Mean-Square Bounds for Contractive Stochastic Approximation: A Unified Elementary Approach STOC
链接: https://arxiv.org/abs/2607.17595
作者: Siddharth Chandak
类目: Machine Learning (cs.LG); Systems and Control (eess.SY); Optimization and Control (math.OC)
*备注: Submitted to Stochastic Processes and their Applications
Abstract:We establish mean-square and concentration bounds for stochastic approximation (SA) with arbitrary norm contractive mappings, under a multiplicative noise model where the noise may scale affinely with the norm of the iterates, and the iterates are potentially unbounded. These settings arise in reinforcement learning, where operators are often contractive in the \ell_\infty norm and the noise scales with the iterates. To address the arbitrary norm, earlier works replace the non-smooth squared norm with a smooth Lyapunov function constructed via the generalized Moreau envelope. For concentration analysis, these works handle multiplicative noise and unbounded iterates through a multi-stage bootstrapping argument that starts from a time-varying worst-case bound and iteratively refines it. We instead present a unified and elementary analysis that yields both bounds. Using an averaged noise sequence and corresponding auxiliary iterates, we obtain a one-step Lyapunov drift inequality for the normed error directly, without smoothing the norm or constructing an envelope. For the mean-square bound, we combine this drift inequality with an induction argument showing that the iterates remain bounded in expectation. For the concentration bound, we develop a probabilistic induction over a sequence of “good” events on which the iterates are controlled, allowing the standard Azuma-Hoeffding bound to be applied. Our approach yields the first sub-Gaussian tailed maximal (all-time) concentration bound for SA under multiplicative noise, by allowing the stepsize to depend logarithmically on the confidence level. Beyond the specific setting considered here, we discuss the generalizability of these proof techniques to other noise models and iterative algorithms.
[LG-28] A Weisfeiler-Leman Characterization of Global-Attention Graph Transformers for Mixed-Integer Linear Programs
链接: https://arxiv.org/abs/2607.17570
作者: Md Abrar Jahin,Craig A. Knoblock,Jay Pujara
类目: Machine Learning (cs.LG)
*备注: Accepted to Topology, Algebra, and Geometry in Data Science (TAG-DS) 2026 (20 pages, 10 figures) [Lightning Oral Presentation]
Abstract:Graph foundation models (GFMs) with global attention are increasingly used to represent mixed-integer linear programs (MILPs), aiming to capture structure beyond the locality of standard graph neural networks. We study their expressive power through graph isomorphism testing, asking which MILP instances they map to identical representations. We prove that a broad class of hierarchical graph transformers combining global linear attention, edge-weighted cross-attention, and bipartite message passing is bounded by the one-dimensional Weisfeiler-Leman (1-WL) test: under any parameter setting, 1-WL-equivalent MILP graphs receive identical graph embeddings. Our compositional proof shows that each architectural component is a symmetric multiset function and thus preserves 1-WL equivalence. We validate this characterization across ten diverse graph encoders, including Graphormer-, GraphGPS-, Set-Transformer-, and Gasse-style models. Across model capacities, graph scales, and pooling operators, every tested encoder maps 1-WL-equivalent non-isomorphic graph pairs to numerically identical embeddings. Consequently, graph invariants that vary within a 1-WL equivalence class cannot be recovered from these representations. We further show that expressiveness beyond 1-WL arises from input encoding rather than attention: random-walk positional encodings separate the constructed pairs, while additional constructions expose the limits of this remedy. These results characterize the expressive power of global-attention GFMs and provide an encoder-agnostic diagnostic for detecting 1-WL-induced representation equivalence.
[LG-29] Volatility-Aware Extreme Event Detection in High-Frequency Financial Markets
链接: https://arxiv.org/abs/2607.17555
作者: Maorufa Zaman,Haris Md Sahed
类目: Machine Learning (cs.LG)
*备注: 7 pages, 3 figures. Preprint
Abstract:Predicting extreme price movements in high-frequency financial markets is a challenging task due to non-stationarity, heavy-tailed return distributions, and severe class imbalance. In particular, rare but impactful events are often difficult to detect using conventional modeling approaches, which typically treat extreme movements as isolated observations. This study proposes a volatility-aware approach for extreme event detection using high-frequency Bitcoin limit order book (LOB) data. Motivated by empirical evidence of volatility clustering, the target formulation is extended to incorporate both large future returns and high-volatility regimes. This redefinition increases the proportion of informative samples and aligns the learning objective with the underlying market dynamics. Using a tree-based model (XGBoost) with time-series cross-validation and imbalance-aware evaluation, the proposed method achieves a Precision-Recall AUC of approximately 0.40, significantly outperforming the baseline formulation with a PR-AUC of around 0.06. This represents more than a sixfold improvement in detecting rare events. The results highlight that target design plays a critical role in financial machine learning, often exceeding the impact of model complexity. By incorporating volatility structure into the labeling process, the proposed approach provides a more effective and realistic framework for extreme event detection in high-frequency cryptocurrency markets.
[LG-30] Program Synthesis for Simulation-Based Inference: Joint Model Selection and Parameter Estimation
链接: https://arxiv.org/abs/2607.17540
作者: Siddharth Mishra-Sharma
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 15+7 pages, 4+2 figures
Abstract:Neural simulation-based inference enables parameter estimation for complex models, but typically requires the user to specify a simulator encoding a fixed model structure. We present a framework for joint model selection and parameter estimation that combines large language models for program synthesis with neural simulation-based inference. Given a natural language description of the system and data under investigation, an LLM proposes candidate simulator programs which are iteratively refined via feedback-driven mutation and evaluated using neural density estimation. The approach enables simulation-based inference over a pool of models, not just parameters within a fixed model. On benchmarks spanning deterministic dynamics, stochastic epidemic models, and dark matter substructure inference from gravitational-lensing images, the method identifies plausible model families from open-ended prompts, with accuracy that reflects the information content of the data and identifiability of candidate models.
[LG-31] FlowSonic: Stable Zero-Shot Music Editing via High-Order Trajectory Integration
链接: https://arxiv.org/abs/2607.17526
作者: Ali Boudaghi,Hadi Zare
类目: ound (cs.SD); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS); Signal Processing (eess.SP); Numerical Analysis (math.NA)
*备注:
Abstract:Zero-shot text-guided editing of real-world music recordings requires balancing semantic modification with faithful preservation of the original musical structure. Although recent diffusion transformers trained with rectified flow have achieved remarkable success in text-to-music generation, extending them to edit existing recordings remains challenging because editing requires accurate deterministic inversion, reliable structural preservation, and numerically stable integration throughout the inversion and generation processes. We present FlowSonic, a zero-shot music editing framework built upon a pretrained diffusion transformer trained with rectified flow. FlowSonic first deterministically inverts a real-world recording into the latent space and preserves its musical structure during editing by reusing cross-attention representations extracted during inversion. To improve the numerical reliability of inversion-based editing, we introduce a high-order ODE solver and systematically investigate how different numerical integration schemes influence trajectory stability, structural preservation, and semantic controllability. Comprehensive experiments on timbre-transfer and genre-modification tasks demonstrate that FlowSonic consistently outperforms existing music editing methods across semantic alignment, harmonic preservation, structural consistency, and perceptual audio quality. We further provide geometric and empirical analyses showing how the proposed numerical integration strategy improves latent trajectory stability and leads to more reliable music editing. Subjects: Sound (cs.SD); Machine Learning (cs.LG); Audio and Speech Processing (eess.AS); Signal Processing (eess.SP); Numerical Analysis (math.NA) Cite as: arXiv:2607.17526 [cs.SD] (or arXiv:2607.17526v1 [cs.SD] for this version) https://doi.org/10.48550/arXiv.2607.17526 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-32] FailureAtlas: A Taxonomy of Failure Modes in Multi-Provider LLM Serving Infrastructure
链接: https://arxiv.org/abs/2607.17525
作者: Vishal Pandey,Gopal Singh
类目: Machine Learning (cs.LG); Software Engineering (cs.SE)
*备注: Survey Paper, 14 pages, 1 figure
Abstract:Multi-provider LLM gateways reverse proxies that route, load-balance, and rate-limit requests across foundation-model APIs have become critical production infrastructure. Yet the failure modes specific to this architectural layer remain undocumented, scattered across issue trackers and post-mortems with no unifying framework. We introduce \fa, a two-axis taxonomy that classifies failures by their \emphorigin layer (Network/Transport, Streaming/Protocol, State/Session, Model~Behavior, Governance/Cost) and their \emphdetectability (Loud vs.\ Silent). We populate this taxonomy with five verified catalog entries sourced from public bug reports and first-hand stress testing, each accompanied by a mechanistic root-cause analysis. Three entries include standalone reproduction scripts. Our principal finding is that the most operationally severe failures are \emphsilent: they return HTTP~200, pass every standard health check, and corrupt application state in ways that require semantic-level observability to detect. Two such silent failures a concurrency race condition causing history loss and a streaming index collision corrupting tool-call payloads were discovered first-hand during \cb evaluation campaigns.
[LG-33] Lightweight Wrappers for Adapting Time Series Foundation Models to Regional Drought Forecasting KDD2026
链接: https://arxiv.org/abs/2607.17511
作者: Wentao Gao,Jiuyong Li,Lin Liu,Thuc Duy Le,Jixue Liu,Yanchang Zhao,Yun Chen
类目: Machine Learning (cs.LG)
*备注: 12 Pages, KDD2026 Accpeted Paper(Poster)
Abstract:Large \emphTime Series Foundation Models (TSFMs) demonstrate strong zero-shot forecasting capabilities across diverse domains. However, their application to regional climate forecasting faces practical challenges: model weights are often proprietary, local training records are limited, and computational budgets are constrained, making traditional fine-tuning approaches infeasible. To address these constraints, we introduce a lightweight, black-box adaptation framework (requiring no access to backbone parameters and no backbone fine-tuning) that enhances frozen TSFMs at inference time through two plug-and-play wrappers: \textbfSMR\textsuperscript2 (Stationarity aware multi-resolution Residual), which decomposes the input into multi-resolution temporal views, learns stride specific residual corrections that capture regional dynamics, then adaptively ensembles them into a single forecast, and \textbfMBB (Moving Block Bootstrap), which preserves temporal dependencies through block resampling and ensembles over temporally coherent residual perturbations to stabilize the point forecast. Both wrappers instantiate the same bagging style principle: they build diverse views of the input or its residuals, forecast each with the same frozen backbone, and aggregate, so all adaptation comes from inference time ensembling rather than any weight update. Evaluated on one month ahead Standardized Precipitation Evapotranspiration Index (SPEI) prediction across multiple sites in South Australia, our framework consistently improves forecasting performance across several backbone models, demonstrating up to 26% mean squared error (MSE) reduction over the corresponding frozen backbone while enabling practical deployment in resource constrained regional forecasting systems.
[LG-34] Residual-Guided Multi-Resolution Refinement of Foundation Models: A Case Study in Drought Forecasting ICML2026
链接: https://arxiv.org/abs/2607.17507
作者: Wentao Gao,Jiuyong Li,Lin Liu,Thuc Duy Le,Jixue Liu,Yanchang Zhao,Yun Chen
类目: Machine Learning (cs.LG)
*备注: 23 pages, ICML2026 Accepted Paper(Poster)
Abstract:Regional climate prediction presents unique challenges for time series foundation models, which typically process temporal patterns through single-pass inference. Expert climatologists, in contrast, employ multi-scale temporal analysis and iterative refinement based on systematic error diagnosis. We present RGMR (Residual-Guided Multi-Resolution Refinement), an inference-time framework that adapts pre-trained foundation models to perform structured coarse-to-fine refinement for climate forecasting without updating backbone parameters. Applied to drought forecasting using the Standardized Precipitation Evapotranspiration Index (SPEI), RGMR is architecture-agnostic across the three TSFM backbones evaluated per site (TimesFM, TimeGPT, TabPFN) and consistently lowers test-set MSE on three South Australian sites and three additional regions outside South Australia. Applied to TimesFM, the wrapper reduces one-month-ahead SPEI MSE by up to 18.9% across the three South Australian sites (mean reduction \approx 18.7%). Overall, RGMR provides a practical route for deploying frozen TSFMs in regional climate forecasting workflows.
[LG-35] he Dimension of Nonterminating Resampling Computations
链接: https://arxiv.org/abs/2607.17469
作者: Yunbei Xu
类目: Computational Complexity (cs.CC); Statistical Mechanics (cond-mat.stat-mech); Machine Learning (cs.LG); Combinatorics (math.CO); Probability (math.PR)
*备注:
Abstract:A randomized algorithm may terminate almost surely even though exceptional random tapes make it run forever. This paper studies the survival tail, the Kolmogorov complexity of one such tape, and the Hausdorff dimension of all of them. For each s0 at which the powered repair matrices commute, the main theorem bounds \sum_wP[w]^s over surviving prefixes w , uniformly over deterministic nonanticipating selectors. The case s=1 controls termination; the full family gives weak-source and dimension bounds. The source powers contain information absent even from the ordinary repair kernel and the complete stopping-time law. Under one common finite tape source, two overlapping disagreement-repair rules on a four-vertex path have the same ordinary kernels and the same stopping-time law for every selector, yet their nontermination dimensions can be arbitrarily close to zero and one. At one common source-power level, the same dominated tape source makes one rule run forever but gives the other an exponential stopping tail. The separation is caused by action labels that produce the same state transition and are therefore invisible at power one. For bounded-dependence k -SAT, conditional block min-entropy above the trace-growth threshold gives exponential termination, and the effective dimension of an individual infinite run is bounded by the trace growth induced by the clauses repaired infinitely often. Tree formulas asymptotically attain the maximum-degree dimension and global source bounds, while clique formulas attain the graph-specific one-step threshold in the stated regime. An exact backward likelihood identity complements these setwise results with tail and coding bounds for each run. Subjects: Computational Complexity (cs.CC); Statistical Mechanics (cond-mat.stat-mech); Machine Learning (cs.LG); Combinatorics (math.CO); Probability (math.PR) Cite as: arXiv:2607.17469 [cs.CC] (or arXiv:2607.17469v1 [cs.CC] for this version) https://doi.org/10.48550/arXiv.2607.17469 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-36] Calibrated Alzheimers Conversion Risk in Mild Cognitive Impairment: Persistent Homology of Clinical Trajectories with Conformal Guarantees
链接: https://arxiv.org/abs/2607.17442
作者: Navin Bondade
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 17 pages, 6 figures, 2 tables. Data from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Code available upon reasonable request
Abstract:Background. Predicting conversion from mild cognitive impairment (MCI) to Alzheimer’s disease (AD) is central to trial enrichment and care planning, yet existing models provide no individual-level uncertainty estimates and rarely include transparent leakage audits. We introduce the first application of persistent homology to longitudinal clinical trajectory point clouds for this task, and the first split-conformal individual risk guarantee for any AD-conversion model. Methods. We analysed 741 MCI subjects (240 converters, 32.4%) from ADNI with a uniform 4-year follow-up cap. Five leakage sources were corrected; without them a naive pipeline achieved AUC=0.934, inflated by +0.075. Vietoris-Rips persistent homology and sublevel-set proxies were combined with trajectory slopes and engineered features (76 total) in a stacking ensemble evaluated by 5-fold cross-validation. Results. Cox and Random Survival Forest models with TDA features achieved concordance C=0.799 and C=0.826 versus C=0.753 and C=0.812 without (+0.045 and +0.014). The primary nested AUC is 0.840 (same-fold bound 0.866); external AUC was 0.879 on a zero-overlap ADNI-2/GO/3 cohort. H0 persistence entropy was the top SHAP feature and significantly associated with APOE4 dosage (Spearman r=-0.191, p0.0001, Bonferroni-corrected). Cross-conformal coverage was 90.4%±2.2% (target 90%); empirical external coverage 96.9%. Maximum fairness gap in false-negative rate across seven subgroups was 0.092. Conclusions. We propose H0 persistence entropy as a topological biomarker of cognitive decline and demonstrate that a leakage-audited, conformally calibrated pipeline reaches competitive accuracy with individual-level uncertainty quantification not previously available for this task.
[LG-37] Decoder-Preserving Sparse Autoencoders: Which Readouts Survive Sparse Compression?
链接: https://arxiv.org/abs/2607.17425
作者: Aniket Deshpande
类目: Machine Learning (cs.LG)
*备注:
Abstract:Sparse autoencoders (SAEs) compress model activations into sparse codes, but equal reconstruction error and sparsity can preserve different linearly decodable signals. We formalize this ambiguity as a matrix-valued distortion between optimal ridge-prediction operators and train decoder-preserving SAEs by combining this distortion with reconstruction loss. In a rank relaxation, an isotropic task prior saturates per-mode omission costs without changing PCA’s ordering, whereas a structured prior can change which modes are retained. A controlled sparse experiment shows that a declared prior protects held-out combinations from its task subspace. On GPT-2 small block 8, DPSAE reduces held-out decoder distortion by 10.6–11.4% across three paired runs while matching reconstruction NMSE. The same checkpoints pass an average natural-text output-KL noninferiority test, but one matched Pythia pair shows no improvement in probes restricted to a few sparse features. These results show that reconstruction quality does not determine which refitted linear readouts survive sparse compression, and that readout preservation is distinct from learning cleaner benchmark concepts or preserving every frozen-model behavior.
[LG-38] Grounded verification of chemical and materials reasoning : detection is the bottleneck
链接: https://arxiv.org/abs/2607.17417
作者: Can Polat,Mustafa Kurban,Erchin Serpedin,Hasan Kurban
类目: Machine Learning (cs.LG); Chemical Physics (physics.chem-ph); Computational Physics (physics.comp-ph); Quantum Physics (quant-ph)
*备注:
Abstract:Large language models confabulate chemical objects (molecular formulas, space groups, formation energies) in fluent reasoning traces, concentrated on long-tail entities where confidence is least trustworthy. Deterministic, database-grounded verification can catch and repair such errors without the coverage cost of blanket retrieval; the binding constraint, we find, is detection, not repair. Our tiered verifier extracts each checkable claim, checks it against authoritative databases and physics, and feeds the reference into a gated correction loop. Across four models and 528 condition-pinned prompts, gated correction cuts committed-formula error from 22% to 4% at 3.2\times fewer retrievals than blanket augmentation, beating a conversational oracle. Repair succeeds wherever a flag fires (80–97%); the bottleneck is in-loop detection recall. Grounding improves the final answer only when the verifier’s scope reaches the deliverable (83% to 90%), and the lift appears only where extractable long-tail error exists: absent on near-ceiling physical constants, large on isotope half-lives (11% to 0%).
[LG-39] CORAL: Learning Amyloid Fibril Ligand Docking with Cooperative Binding Rewards
链接: https://arxiv.org/abs/2607.17412
作者: Yasheng Sun,Bohan Li,Youqi Tao,Jürgen Schmidhuber
类目: Machine Learning (cs.LG)
*备注: 25 pages
Abstract:A hallmark of neurodegenerative diseases such as Alzheimer’s and Parkinson’s is the aberrant aggregation of proteins into amyloid fibrils, and small molecules that selectively bind to these fibrils hold promise as diagnostics, imaging probes, and therapeutics. Predicting how such ligands bind to fibril targets, however, presents two fundamental challenges. First, resolved co-crystal structures of amyloid-ligand complexes are exceptionally scarce; even with recent advances in cryo-EM only a handful have been structurally characterized, making supervised training of docking models impractical for this target class. Second, amyloid fibrils present a binding mode fundamentally different from globular proteins: ligands intercalate into longitudinal cross- \beta grooves and stack cooperatively along the fibril axis, a geometry that existing docking models are not designed to capture. To address these challenges, we present CORAL (COopeRative Amyloid Ligand docking), a reinforcement learning framework that trains a generative docking model to produce ligand pose distributions tailored to the cross- \beta groove geometry. Our reward explicitly incorporates cooperative ligand-ligand stacking energy alongside protein-ligand docking affinity, directly capturing the distinctive binding geometry of amyloid fibrils. We further introduce a curated evaluation set of amyloid-ligand complexes constructed from model-generated poses validated by domain experts. Experiments on both experimentally resolved structures and this evaluation set demonstrate improved pose quality and binding affinity correlation over existing docking baselines.
[LG-40] Feature-Guided Diffusion for Non-Differentiable Inverse Rendering
链接: https://arxiv.org/abs/2607.17411
作者: Andrei-Timotei Ardelean,Michael Fischer,Tim Weyrich,Tomáš Iser
类目: Graphics (cs.GR); Machine Learning (cs.LG)
*备注:
Abstract:Inverse rendering is traditionally solved via differentiable renderers and gradient descent, which requires substantial problem-specific engineering and is prone to getting stuck in local minima due to ambiguities. Derivative-free approaches alleviate engineering requirements, but often heavily depend on a good problem initialization. In this work, we propose Feature-Informed Diffusion Evolution (FIDE), a fully black-box framework that requires no gradients or specific initialization: the renderer is treated as an opaque function whose only requirement is to produce images. Our key insight is feature guiding: rather than reducing each candidate rendering to a scalar loss value, we use a Vision Transformer (ViT) to extract dense visual features from it. We subsequently use these features to train a diffusion-based candidate proposal model, allowing the network to use visual cues to predict parameters that would match the target image. The candidate solutions proposed by this diffusion model are then refined in a closed loop with a CMA evolution strategy, continuously narrowing the proposal region as optimization progresses. We validate across diverse inverse problems from path tracing, vector splines, Voronoi shaders, and robotics, and demonstrate that feature-guiding substantially improves convergence over scalar-loss baselines and reliably escapes local minima where gradient-based methods stall.
[LG-41] Chebyshev Manifold Adaptation
链接: https://arxiv.org/abs/2607.17377
作者: Jiawen Li
类目: Machine Learning (cs.LG)
*备注:
Abstract:The paper presents a new parameter-efficient adaptation method called ChebyMA (Chebyshev Manifold Adaptation). ChebyMA adopts weight matrices through a multi-surface superposition of Chebyshev polynomial bases evaluated on learnable coordinates and combined via trainable coefficient matrices, replacing standard linear projections with highly expressive continuous function approximation. Theoretically, we establish an Approximation Expressivity Theorem, proving from the perspective of function approximation theory that single-manifold ChebyMA guarantees convergence in Frobenius norm error of reconstruction. Besides, drawing on Kolmogorov n -width intuition, we demonstrate the expressive advantages of multi-manifold superposition ( S 1 ) in decoupling high-dimensional complex features. Experimental results on Computer Vision CIFAR datasets(CIFAR-10, CIFAR-100)\citeCIFAR and Natural Language Processing (AG News, SST-2) datasets demonstrate that ChebyMA consistently achieves a superior parameter-accuracy Pareto front compared to standard full-parameter fine-tuning, LoRA\citeLoRA, TLoRA\citeTLoRA, and StelLA\citeStelLA. ChebyMA significantly outperforms other tested methods in tested datasets, validating its solid theoretical foundation for generality with purely vectorized computations.
[LG-42] aurus: Accelerating Out-of-Core Graph Neural Network Inference on Billion-Scale Graphs
链接: https://arxiv.org/abs/2607.17374
作者: Pranjal Naman,Yogesh Simmhan
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注: Extended full-length version of paper that appeared at HPDC 2026: “ATLAS: Efficient Out-of-Core Inference for Billion-Scale Graph Neural Networks”, Pranjal Naman and Yogesh Simmhan, in the 35th ACM International Symposium on High-Performance Parallel and Distributed Computing (HPDC), 2026. DOI: this https URL
Abstract:Graph Neural Network (GNN) inference on billion-scale graphs is challenging due to the large memory footprint of features and embeddings and high disk I/O costs in out-of-core settings. Existing distributed GNN systems incur high communication times and infrastructure costs, while disk-based GNN systems are primarily tailored to training and experience massive wasted reads during inference on the entire graph. We present Taurus, a single-machine system for GNN inference on graphs that do not fit in RAM, supporting both \textitexact full-graph inference and fanout-sampled inference. To avoid random and repeated feature gathers, Taurus reformulates layer-wise inference as source-centric broadcasts over sequential SSD scans, backed by a pipelined GPU-CPU-SSD hierarchy, topology-aware reordering, pending-message eviction, and a GPU-resident store for high-degree vertices. It further uses non-buffered sequential reads and GPU-backed writes to reduce page-cache pollution, host-memory pressure, and write overheads. On out-of-core graphs with up to 269M vertices, 4B edges, and 514 GiB of features, Taurus outperforms the strongest layer-wise baseline, DGI, by 7 - 25\times , and vertex-wise baselines by 40 - 140\times .
[LG-43] Stringological sequence prediction II: Right-to-left automaticity and related complexity measures
链接: https://arxiv.org/abs/2607.17369
作者: Vanessa Kosoy
类目: Formal Languages and Automata Theory (cs.FL); Data Structures and Algorithms (cs.DS); Machine Learning (cs.LG)
*备注: 73 pages
Abstract:In a previous paper, we began the study of sequence prediction algorithms adapted to stringological word complexity measures. One measure we considered was left-to-right (most-significant-digit-first) automaticity. Here, we show a statistically and computationally efficient algorithm adapted to the dual'' right-to-left (least-significant-digit-first) automaticity, which turns out to be substantially different for our purpose. We also demonstrate a prediction algorithm for a more expressive measure that we call arithmetic repetition complexity’'. In particular, the latter can be used for predicting the so-called mix-automatic sequences.
[LG-44] A multiverse-consensus pipeline for reproducible feature selection in untargeted LC-MS metabolomics
链接: https://arxiv.org/abs/2607.17345
作者: Mohammed Saeed Al-Huraibi,Ihsan Yozgat,Ahmet Kaplan
类目: Machine Learning (cs.LG); Genomics (q-bio.GN)
*备注: 14 pages
Abstract:Background: Untargeted LC-MS metabolomics requires a long chain of preprocessing decisions, each with several equally defensible options. Analysts typically commit to one pipeline and report the resulting feature shortlist. How strongly that shortlist depends on choices that were never varied stays invisible. Results: We adapt multiverse analysis to untargeted metabolomics feature selection. We present an auditable, configuration-driven pipeline that (i) applies a ten-stage quality-control filter cascade in which every feature’s fate is logged, and (ii) runs the downstream analysis as a multiverse over four contrasting preprocessing philosophies, each combined with four feature-ranking methods under bootstrap stability selection and label-permutation testing. Only features recurring across paths enter a tiered consensus. On a demonstration dataset of five breast-cancer cell lines (30,370 detected features), the four single pipelines individually returned shortlists of 4-20 features whose pairwise agreement was as low as Jaccard = 0.05. The multiverse consensus retained 15 features (=2/4 paths), of which one recurred across all four, although two paths (sharing normalization and drift-correction methods) dominate the consensus. A pipeline-wide label-permutation test found no false discoveries in 50 null permutations. Conclusions: Reporting only preprocessing-robust features, with a complete kept/dropped audit trail, converts hidden analytical degrees of freedom into an explicit, inspectable output. We discuss scope and limitations, including single-batch design and the need for independent validation.
[LG-45] When Drift Detectors cry Wolf: False Alarm Rates in continuous ML Monitoring ICLR2026
链接: https://arxiv.org/abs/2607.17336
作者: Raj Shekhar Singh
类目: Machine Learning (cs.LG)
*备注: Accepted in ICLR 2026 CAO workshop
Abstract:Drift detection is a core component of production machine learning monitoring systems, where detectors are used to compare incoming data with a reference distribution and trigger alerts when changes occur. However, these detectors are often evaluated in research settings that emphasize detection accuracy under synthetic shifts, while overlooking false alarms under continuous monitoring. In production environments, models are monitored repeatedly over time and across many features, and even small false positive rates can accumulate into frequent alerts, leading to alarm fatigue. We empirically analyze false positive behavior across five commonly used drift detectors: PSI, KS, MMD, LSDD, and adversarial validation. Consistent with existing literature, PSI exhibits strong sensitivity to batch size, producing frequent false alarms at small sample sizes; however, we further observe that its behavior stabilizes and improves substantially once batch sizes exceed approximately 200 samples. In contrast, KS, MMD, and LSDD display persistent fluctuations across batch sizes, while remaining comparatively more reliable than PSI in low-data regimes. Applying a Bonferroni correction reduces false positive rates, but often at the cost of reduced true positive sensitivity, reinforcing the well-known stability - sensitivity trade-off in drift detection. This work provides a systematic comparison of false positive behavior across multiple drift detectors under continuous monitoring conditions. We identify tradeoffs across detector families and provide practical guidelines for selecting and calibrating drift detectors in production ML systems.
[LG-46] Rethinking the Suitability of Reinforcement Learning Algorithms Under Practical Transfer Constraints
链接: https://arxiv.org/abs/2607.17326
作者: Hany Hamed,Abhishek Naik,Colin Bellinger,A. Rupam Mahmood
类目: Machine Learning (cs.LG); Robotics (cs.RO)
*备注:
Abstract:Transfer-oriented reinforcement learning requires evaluating algorithms along dimensions that go beyond standard sample efficiency. We focus on two dimensions: practical efficiency, which asks whether conclusions about algorithm suitability change under wall-clock rather than interaction-based budgets, and robustness under dynamics mismatch, which asks how different learning paradigms respond to variability in the training distribution induced by domain randomization. We provide two insights to reinforcement-learning practitioners. First, comparing the sample efficiency of different algorithms is often an insufficient criterion in transfer-oriented settings. The wall-clock time required to train a decent policy is an important consideration for practitioners, and we find that the sample-inefficient PPO algorithm can produce a performant policy faster than relatively more sample-efficient algorithms such as SAC and TD-MPC2, validating the common understanding of massively parallel training paradigms. Second, domain randomization can help different kinds of algorithms learn robust policies. In particular, although PPO, SAC, and TD-MPC2 represent different RL paradigms - on-policy, off-policy, and model-based learning and planning, respectively - we find that domain randomization affects all three algorithms in a similar way. To the best of our knowledge, this is the first controlled comparison of the effect of domain-randomization coverage on PPO, SAC, and TD-MPC2 under the same transfer protocol. Taken together, these two insights highlight the importance of evaluating RL algorithms not only by sample efficiency, but also by practical considerations such as training time and the algorithms’ ability to produce usable policies.
[LG-47] Rationalizing Boltzmann Rationality: An Axiomatic Characterization of Entropy-Regularized Policies
链接: https://arxiv.org/abs/2607.17316
作者: Silviu Pitis
类目: Machine Learning (cs.LG); Theoretical Economics (econ.TH)
*备注: RLC 2026
Abstract:The softmax policy \pi(a \mid s) \propto \exp(\beta Q(s,a)) is the default model of stochastic choice in reinforcement learning (RL). Various justifications based on robustness, exploration, and optimization have been offered in the RL literature, but none uniquely derives the softmax form from first principles. This leaves a basic tension unresolved: the entropy bonus in the soft Bellman equation violates the Independence axiom that underwrites the Markov decision process (MDP) reward structure. We dissolve this tension by distinguishing two kinds of randomness: chance and choice. By restricting von Neumann-Morgenstern (VNM) Independence to environmental lotteries over base prospects, we show that imposing independence of irrelevant alternatives (IIA) and monotonicity on the policy and value functions at choice nodes uniquely determines the Boltzmann policy, the entropy-regularized representation, and the soft Bellman equation. The choice between the soft and hard Bellman equations thus reduces to a design decision: whether the agent values its own ability to choose. We develop RL-specific consequences, including return monotonicity and convergence under generalized discounting, and synthesize the independent lines from economics and information theory that arrive at the same structure, offering a normative assessment of when IIA is appropriate for agent design.
[LG-48] An Iterative Geometric Approach to Optimizing Separating Hyperplanes
链接: https://arxiv.org/abs/2607.17282
作者: Akos Hajnal
类目: Machine Learning (cs.LG)
*备注: 15 pages, 1 figure
Abstract:Given a binary-labeled linearly separable dataset, and the objective is to compute the maximum-margin separating hyperplane, also known as the hard-margin Support Vector Machine (SVM) classifier. This paper investigates whether, if given an initial separating hyperplane, can it be exploited to reach this unique optimum more efficiently. We present a geometric approach that gradually improves the alignment of the hyperplane, starting from an initial separating hyperplane, while preserving separation and continuously increasing its margin until convergence to the global optimum. At each iteration, the method considers only local information, namely the current active set, and aims to re-align the hyperplane according to the optimal separating hyperplane of this reduced subset. Consequently, the original convex quadratic optimization problem is addressed through a sequence of smaller subproblems. The paper presents the algorithm in detail, together with a preliminary experimental evaluation and several theoretical findings. The results suggest that, when an initial separating hyperplane is available, the proposed method can be competitive on larger datasets and, in some cases, can outperform state-of-the-art approaches that solve the optimization problem directly.
[LG-49] Node4All: Learning Node Representation Beyond Datasets KDD2026
链接: https://arxiv.org/abs/2607.17272
作者: Dooho Lee,Jaemin Yoo
类目: Machine Learning (cs.LG); Social and Information Networks (cs.SI)
*备注: Accepted to KDD 2026
Abstract:Node representation learning has advanced rapidly, yet most existing methods rely on per-dataset training and hyperparameter tuning. This dataset-specific optimization comes from the difficulty of designing reusable graph models that generalize across diverse graph datasets. In this work, we introduce Node4All, a node representation learner applicable to arbitrary graph datasets without any dataset-specific optimization. Node4All is built on two complementary ideas. At the architectural level, we introduce the Channel Graph Transformer (CGT), which enables a single fixed parameterization to process arbitrary graph datasets. At the learning level, we propose a self-supervised learning based on a series of synthetic graphs. Together, these components enable generalization beyond individual datasets, which is infeasible with existing architectures and learning frameworks. We extensively evaluate Node4All on node classification across 25 benchmarks against 21 baselines, covering both supervised and self-supervised methods. Despite all baselines being trained and optimized for each dataset, a single Node4All, applied uniformly across the datasets, achieves a competitive ranking of 5th among 21 baselines. Moreover, Node4All supports one-shot and in-context learning with an appropriate predictor and outperforms recent graph foundation models (GFMs) in these settings. These results demonstrate that Node4All not only achieves reusability across arbitrary graph datasets, but also remains an effective solution in practice. Code and model checkpoints are available in this https URL. Comments: Accepted to KDD 2026 Subjects: Machine Learning (cs.LG); Social and Information Networks (cs.SI) Cite as: arXiv:2607.17272 [cs.LG] (or arXiv:2607.17272v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.17272 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-50] DynImmune-BERT: Dynamic Immune Repertoire Modeling with Neural ODE Driven Continuous Transformers
链接: https://arxiv.org/abs/2607.17244
作者: Rong Fu,Yongtai Liu,Xiaowen Ma,Haoyu Zhao,Shuo Yin,Yiqing Lyu,Long Zhang,Wangyu Wu
类目: Machine Learning (cs.LG)
*备注:
Abstract:Longitudinal T cell receptor repertoires contain signals of clonal expansion, contraction, disappearance, and reappearance after immune perturbation. Static repertoire language models usually summarize a sample as a bag of sequences, so the sampling interval, sequencing depth, and clone presence pattern are only weakly represented. This paper presents DynImmune-BERT, a continuous time repertoire model for patient level immune status prediction. The method combines depth adaptive centered log ratio initialization, clone presence gated Neural ordinary differential equation dynamics, bounded neighborhood self attention, event based state restart, and a hybrid transport objective that supervises dominant and rare clone mass. A low rank meta adapter initializes reappearing clonotypes while keeping the parameter count independent of the number of observed clones. The evaluation separates literature reported baselines from internally controlled temporal comparisons, reports uncertainty for small external cohorts, adds calibration and threshold diagnostics, and visualizes latent clone trajectories and attention neighborhoods. The results indicate that event aware temporal modeling can complement strong static encoders when longitudinal repertoire structure is available, while small external cohorts and protocol differences require cautious interpretation.
[LG-51] Rate-Distortion-Perception Theory: Redefining the Fundamental Limits of Information Representation
链接: https://arxiv.org/abs/2607.17232
作者: Photios A. Stavrou,Giuseppe Serra,Marios Kountouris
类目: Information Theory (cs.IT); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 20 pages, 10 figures, IEEE BITS
Abstract:Classical rate-distortion (RD) theory has long established the fundamental limits of lossy compression by quantifying the minimum number of bits required to represent a source under a prescribed distortion constraint. However, widely used distortion measures such as mean-squared error often fail to capture perceptual quality or semantic validity, which are increasingly central in modern learning-driven applications. Rate-distortion-perception (RDP) theory extends the RD framework by introducing perception as a third fundamental axis, quantified via distributional similarity between the source and reconstructed signals, leading to the rate-distortion-perception function (RDPF). This tutorial provides a structured overview of the coding principles underlying perception-aware lossy compression and surveys recent achievability results under different randomness assumptions. It then presents a unifying optimization viewpoint for computing the RDPF as defined by Blau and Michaeli, for both discrete and continuous sources under broad families of perceptual constraints, including f-divergences, alpha-divergences, and Wasserstein-based metrics. Special attention is given to computational tools such as alternating minimization schemes, Newton-based methods, and convex optimization formulations, as well as to analytically tractable cases such as Gaussian sources and the perfect-realism regime. Unlike recent broad surveys that emphasize generative architectures and AI-empowered communication systems, this tutorial focuses on the coding-theoretic and computational machinery needed to characterize, compute, and interpret the RDP limits. Finally, the tutorial outlines promising research directions at the intersection of information theory, neural compression, robust source coding, and perception-aware networked control systems.
[LG-52] Explaining and Tuning Transformer-based LLM s in Arithmetic Tasks with Human Strategies
链接: https://arxiv.org/abs/2607.17166
作者: Luyu Qiu,Jianing Li,Hwanhee Kim,Xiaoyong Wei,Yueyuan Zheng,Janet Hsiao,Lei Chen
类目: Machine Learning (cs.LG)
*备注:
Abstract:Transformer-based large language models (LLMs) continue to achieve state-of-the-art performance across various natural language processing tasks. However, their subpar performance on seemingly elementary problems, such as basic arithmetic, raises concerns about model reliability, safety, and ethical deployment. In this study, we demonstrate that the performance of a vanilla Transformer model trained on integer arithmetic tasks can be improved using methods effective for human learners. We begin by decomposing the arithmetic task into well-defined subtasks and conducting loss convergence order analysis together with ablation studies for each subtask. Our findings reveal that LLMs exhibit learning patterns similar to those of human learners, with a faster learning speed for simpler subtasks compared to more complex ones. In addition, we successfully improved the accuracy of LLMs by applying problem-solving strategies and cognitive empowerment methods shown to enhance the performance of human learners. This suggests that transformer-based LLMs may share cognitive processes with human learners in arithmetic. Lastly, we provide a comprehensive demonstration of our method’s effectiveness, including significant accuracy improvement experiments, visualization verification, and explanation-based analysis to illuminate the intricacies of LLMs in arithmetic learning. In general, this work explores the potential similarities between transformer-based LLMs and human learners, supported by explainable AI (XAI) verifications, ultimately fostering trust in LLMs for critical and high-stakes applications.
[LG-53] Robust Assamese Speech Recognition through Controlled Fine-Tuning of Whisper Models
链接: https://arxiv.org/abs/2607.17164
作者: Ganapati Das,Dwipen Laskar,Hasin Afzal Ahmed,Sanjib Kr Kalita,Kshirod Sarmah,Hem Chandra Das,Manjula Kalita
类目: Machine Learning (cs.LG)
*备注: 31 pages, 11 figures
Abstract:Developing Automatic Speech Recognition (ASR) for morphologically rich, low-resource languages such as Assamese is challenging due to insufficient annotated speech data. The pretrained Whisper model performs poorly on Assamese speech recognition tasks. This paper presents a controlled, fine-tuned Whisper-based Assamese ASR system trained on the Mozilla Common Voice 24.0-Assamese corpus. A hardware-aware optimized training pipeline is implemented for resource-constrained environments, employing mixed-precision training and gradient accumulation on Tesla 4 Graphics Processing Units (T4 GPUs). The proposed fine-tuned model significantly outperformed the Zero-shot baseline, yielding Word Error Rate (WER), Character Error Rate (CER), Match Error Rate (MER), and Word Infomation Loss (WIL) of 43.17%, 13.18%, 43%, and 64.81%, respectively, achieving significant relative improvements of 78.26%, 93.10%, 57.0%, and 35.19% over the baseline. Semantic evaluation of the fine-tuned model also demonstrates notable improvement over a zero baseline, attaining Bilingual Evaluation Understudy (BLEU) and Metric for Evaluation of Translation with Explicit ORdering (METEOR) scores of 30.81 and 0.5262, respectively. Additionally, the predicted hallucination rate and Real-Time Factor (RTF) are substantially improved by 96.70% and 32.38%, compared to the zero-shot baseline.
[LG-54] OrderMoE: An expert similarity driven distributed edge MoE inference
链接: https://arxiv.org/abs/2607.17154
作者: Xin Yuan,Ning Li,Quan Chen,Wenchao Xu,Athanasios V. Vasilakos,Song Guo,Haijun Zhang
类目: Networking and Internet Architecture (cs.NI); Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注: 17 pages, 12 figures
Abstract:Although mixture-of-experts, MoE, models have been increasingly adopted to scale large language models with moderate computation cost, it remains challenging to deploy MoE inference over resource-constrained and bandwidth-limited edge infrastructures. Existing distributed MoE serving methods mainly rely on exact expert placement, caching, replication, or communication scheduling, while overlooking the functional similarity among experts, which provides an opportunity to reduce cross-server token transmission. Therefore, this paper introduces a similarity-aware expert allocation and distributed deployment framework, dubbed OrderMoE, which aims to accelerate edge MoE inference while balancing inference latency, communication overhead, server workload, and inference quality. OrderMoE first constructs an expert similarity model based on router-induced logits representations and partitions experts in each MoE layer into multiple similarity groups. Then, it develops a similarity-aware expert grouping and deployment strategy to improve local similarity coverage across edge servers. Since reducing remote expert invocation and preserving exact inference quality are conflicting objectives, OrderMoE further designs a quality-aware and trajectory-aware runtime server-expert selection algorithm to decide whether a token should invoke its remote target expert or use a feasible local substitute expert. Experimental results on a real distributed edge testbed show that OrderMoE significantly reduces average latency, tail latency, cross-server traffic, and remote expert invocation ratio, while introducing only small and controllable inference quality degradation.
[LG-55] A Multi-Model Hybrid Defense Approach Against White-box Adversarial Attacks in Computer Network Traffic
链接: https://arxiv.org/abs/2607.17105
作者: Khushnaseeb Roshan
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 40 Pages, 11 Tables and 9 Figures
Abstract:It is crucial to safeguard computer networks from evolving network security threats and unknown cyberattacks. An essential tool for protecting computer networks against unknown cyber threats is Network Intrusion Detection System (NIDS). However, NIDS faces a major security concern due to its susceptibility to adversarial attacks. Adversarial attacks aim to deceive NIDS by crafting and injecting adversarial examples into the system. These adversarial inputs can deceive the NIDS into misclassifying benign network traffic as malicious. We developed a resilient hybrid defense mechanism aimed to mitigate the impact of two potent adversarial attacks: Fast Gradient Sign Method (FGSM) and Carlini Wagner (CW) attack. Our hybrid defense approach leverages the combined strength of two heuristic defense methods: Adversarial Training (AT) and Gaussian Data Augmentation (GDA). GDA provides multi-directional defense, while AT enhances NIDS robustness against specific adversarial vectors. Under pre-attack scenarios, NIDS demonstrated good accuracy and f1-score. However, in the post-attack scenario, its accuracy significantly dropped under FGSM and CW attacks (0.2649 and 0.4961, respectively). Our proposed hybrid defense method effectively mitigated these adversarial threats, with post-defense accuracy of 96.57% and 89.20% for FGSM and CW attacks. We evaluated the defense strategy across a range of epsilon and confidence noise factor values (ranging from 0.0001 to 0.0009). This research provides a good direction for future researchers in the emerging area of adversarial machine learning from a security perspective.
[LG-56] hRIve: Thermally Robust CNN Inference via Low-Rank Adaptation in Heterogeneous PIM Architectures
链接: https://arxiv.org/abs/2607.17091
作者: Vibhanshu Sharma,Pratyush Dhingra,Janardhan Rao Doppa,Partha Pratim Pande
类目: Hardware Architecture (cs.AR); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
*备注: Published in ACM Transactions on Design Automation of Electronic Systems, Volume 31, Issue 2
Abstract:Processing-In-Memory (PIM) has emerged as a promising technology for accelerating machine learning (ML) workloads. Specifically, non-volatile memory-based PIM architectures have enabled effective ML acceleration due to their ability to perform energy-efficient matrix-vector multiplication operations. However, these devices suffer from non-idealities such as thermal noise. This noise alters the stored values in the memory cells which correspond to actual model weights, compromising the inference accuracy. In this work, we introduce ThRIve, a noise-aware training methodology that leverages low-rank adaptation to enable thermally robust inference on heterogeneous PIM architectures. ThRIve selectively stores these low-rank noise-aware parameters on a hardware that is less susceptible to thermal noise, enabling robustness against temperature-induced noise variations. ThRIve mitigates the effects of thermal-noise and prevent the drop in inference accuracy across the entire operating temperature range. Experimental results demonstrate that ThRIve-enabled architectures maintain consistent inference accuracy, with the mean accuracy staying within 2% of the ideal (i.e., noise-free) accuracy, and the variation in accuracy across the entire operating temperature range remaining within 2% of the mean. The proposed methodology achieves accuracy and robustness comparable to thermally-resilient Static Random-Access Memory (SRAM)-based PIM systems, while delivering up to 5.4x reduction in energy-delay product (EDP) during CNN model inferencing.
[LG-57] What does a Bayes-filtered transformer believe? A predictive Monte Carlo approach
链接: https://arxiv.org/abs/2607.17060
作者: Afiq Abdillah Effiezal Aswadi,Haotong Ma,Susan Wei
类目: Machine Learning (cs.LG)
*备注:
Abstract:A Bayes-filtered transformer (BFT) is a transformer trained on sequences that are generated in two steps: first a latent task is drawn from a prior, then observations are drawn conditional on that task. Trained under autoregressive log loss, the BFT’s next-token prediction, in the idealized limit, is the Bayesian posterior predictive distribution (PPD) induced by that prior and that conditional law. In practice the trained BFT is only an approximation of this ideal PPD, raising an interpretive question: what prior and posterior over the latent task has the trained BFT actually internalized? Existing work answers this question by comparing the trained BFT’s predictions against the predictions of various “reference” posteriors, each standing in for a different candidate algorithm or computation the BFT might be implementing. This prediction-space comparison is fragile: different posteriors can share the same posterior-mean predictions. We use predictive Monte Carlo (PMC) as a general interpretability tool for any BFT: using only next-token generation, PMC returns an approximation to the implicit prior and posterior over the latent task, answering the interpretive question directly in latent space. We apply PMC to three stylized task families spanning 0-Markov and 1-Markov exchangeability. The phenomena previously reported in these settings remain visible in latent space. Code is available at this https URL
[LG-58] Apeliotes: A Diffusion-Based Modeling Framework for km-scale Multi-Level Atmospheric Fields
链接: https://arxiv.org/abs/2607.17037
作者: Evangelia Rafaela Frastali,Achyut Paudel,Maryam Golbazi,Frank Liu
类目: Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE)
*备注:
Abstract:High-resolution atmospheric data are required to resolve mesoscale and localized meteorological structures, however such datasets remain limited in many regions of the world. Existing high-resolution weather products are typically produced through dynamical downscaling, which is computationally expensive and difficult to scale across locations, variables, and forecast scenarios. These limitations motivate machine-learning-based downscaling systems that can generate multiple weather variables stochastically while producing new high-resolution fields directly. In this paper we present Apeliotes, a framework for high-resolution weather forecasting. Built on the global re-analysis atmospheric data, a pre-trained global weather foundation model, and a regionally trained generative diffusion model, Apeliotes not only provides accurate kilometer-scale weather variables, but also multi-level atmospheric fields which are not directly available in the existing global atmospheric data. Our comprehensive evaluation demonstrates that Apeliotes achieves highly competitive performance. The model predicts vertical wind profile with less than 3% error between truth and predicted fields, achieving correlations of 0.91 for 10-m wind speed and 0.99 for 2-m temperature, with NRMSE values of 0.42 and 0.17, respectively.
[LG-59] ChemFusion: A Multimodal Cross-Attention Network for Reaction Yield Prediction
链接: https://arxiv.org/abs/2607.17033
作者: Qiwei Han,Chi Zhou
类目: Machine Learning (cs.LG); Chemical Physics (physics.chem-ph)
*备注: 10 pages, 4 figures, 2 tables
Abstract:Forecasting the outcomes of transition-metal-catalyzed reactions is notoriously complex due to the interplay of diverse physical and chemical variables. A persistent computational bottleneck has been effectively merging broad electronic descriptors with the localized, three-dimensional geometry of the reactive site. To bridge this representation gap, we present ChemFusion, a hybrid neural network that fuses conventional electronic features with explicit 3D atomic coordinates. Using a cross-attention mechanism, the model enables global electronic states to dynamically attend to specific spatial constraints within un-pooled molecular point clouds. When benchmarked against a diverse library of cross-couplings, this approach delivers exceptional predictive performance, decisively surpassing traditional single-modality frameworks. Importantly, extracting the attention matrices reveals that the architecture autonomously learns to identify and penalize restrictive steric hindrances. This provides a physically grounded interpretability, demonstrating that spatially aware networks can navigate complex reaction sterics that standard statistical models typically miss.
[LG-60] Interpretable Machine Learning for Air Pollution and Respiratory Health Prediction: A Socioeconomic Subgroup Analysis
链接: https://arxiv.org/abs/2607.17024
作者: Maede Azani Hassan Abadi,Shouyi Wang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Air pollution and climate-related stressors are increasingly important concerns for respiratory health, especially in settings with unequal environmental exposure and healthcare capacity. This study evaluates an interpretable machine learning framework for predicting respiratory disease rates and air-quality status using structured country-level weekly data. Two supervised learning tasks were considered: regression of respiratory disease rate per 100,000 population and binary classification of air-quality status. Nine regression models and nine classification models were compared using nested cross-validation. Model interpretation was conducted using SHAP values, and subgroup analysis was performed across income levels and geographic regions. The results showed that PM2.5 concentration was the dominant predictor of respiratory disease rate, with linear and regularized linear models achieving the strongest regression performance. For air-quality classification, models achieved high balanced accuracy when PM2.5 was included, but performance decreased substantially when PM2.5 was removed, indicating strong dependence on pollutant-related information. SHAP analysis showed that, without PM2.5, socioeconomic and meteorological variables such as GDP per capita, precipitation, and healthcare access became more influential. Subgroup analysis showed similar aggregate regression error across income groups, but PM2.5 contributed more strongly to predictions in lower-middle-income countries. These results show that model accuracy alone is not sufficient for climate-health prediction. Interpretable models can help identify dominant pollution-related signals, test whether results depend on key pollutant variables, and show whether prediction patterns differ across socioeconomic groups.
[LG-61] Regularize or Localize: When Training-Time KV-Cache Geometry Pays Under Quantization
链接: https://arxiv.org/abs/2607.17019
作者: Libo Sun,Po-Wei Harn,Zewei Zhang,Peixiong He,Xiao Qin
类目: Machine Learning (cs.LG)
*备注: 16 Pages, 4 Figures
Abstract:We study whether \sigreg – LeJEPA’s anti-collapse objective – can reshape representations during standard autoregressive language-model pretraining, and when the resulting geometry helps \kv-cache quantization. We train 110M-parameter models on 10B FineWeb tokens and report three findings. \textbf(1) At \lambda=0.01 , \sigreg reduces hidden-state pairwise-cosine anisotropy by 38% across three paired seeds. Perplexity increases by less than 0.35% in every pair, with no consistent zero-shot loss. \textbf(2) This change does not propagate from hidden states to the \kv cache. Applying \sigreg directly to K and V during continued training, however, reduces mean cache anisotropy by 94% across four checkpoints. A matched continuation without the \kv term leaves cache geometry nearly unchanged, and the frozen-trunk retrofits we tested do not reproduce the effect. \textbf(3) Under untransformed symmetric group-free quantization, direct \kv regularization is the only training condition that prefers per-channel scaling in all three seeds, and under that same 3-bit per-channel scheme the baseline incurs 4.3 – 7.9\times the directly regularized model’s \dnll. Under the full simulated KIVI-style configuration (mixed arrangement, zero-points, grouped scales), however, all models reach near-parity, including when storage overhead is approximately matched. In this 110M regime, the training intervention helps when quantizer scales are coarse; the advantage vanishes under the tested combination of token-local grouping, mixed \kv scaling, and zero-points. To our knowledge this is the first training-time \emphdistributional regularization of standard \kv-cache geometry evaluated against post-hoc cache quantization.
[LG-62] Increasing Line Outage Localization Performance with Ensemble Classifiers
链接: https://arxiv.org/abs/2607.17008
作者: Daniel Flores,Yuanrui Sang,Michael P. McGarry
类目: ystems and Control (eess.SY); Machine Learning (cs.LG)
*备注:
Abstract:In many cases, the outage of one transmission line in a system can be localized by monitoring the power flow of another line, and machine learning methods can be used to distinguish the cases under uncertainty. In this study, we examine the improvements in line outage localization performance achieved by various ensemble classifiers compared to single-model methods. In the case studies, we compared the classification results with measurement data collected at observed transmission lines (OTLs) selected using three algorithms, i.e, greedy maximum coverage problem (MCP), high-eta, and random selection, based on two sensitivity factors, i.e., line outage distribution factors (LODFs) and line outage impact factors (LOIFs). We found that the OTLs selected by the greedy MCP algorithm yielded the highest F1 score and the ensemble classifiers significantly outperformed a base kNN classifier. The extra-trees bagging technique achieved the highest F1 score in many instances. All the findings were statistically significant.
[LG-63] Periodic Bootstrap Thompson Sampling For Periodically Non-Stationary Bandit Problems
链接: https://arxiv.org/abs/2607.16986
作者: Boning Shao
类目: Machine Learning (cs.LG)
*备注: 9 pages, 10 figures, Accepted to 2025 3rd International Conference on Data Science, Advanced Algorithms, and Intelligent Computing
Abstract:This paper introduces Periodic Bootstrap Thompson Sampling (PBTS), an innovative extension of the classic Thompson Sampling (TS) algorithm tailored for bandit problems with periodic non-stationarity. Conventional TS accumulates all past observations, leading to biased posteriors when reward distributions cycle over time. PBTS overcomes this by synchronizing belief resets with known or inferred period intervals and embedding structured bootstrap exploration phases, effectively purging obsolete data while preserving uncertainty estimates. PBTS is tested in artificially constructed environments, which include skewed and balanced reward distributions, along with different bootstrap proportions and misaligned periodic intervals. Results indicate that PBTS generally achieves statistically significant reductions in cumulative regret against traditional TS in periodic non-stationary environments. Subsequent discussion further articulates the potential of PBTS’s real-world deployment. The study mentions limitations like extreme periodic misalignment and proposes future research such as self-adjusting cycle-recognition. With memory reset and bootstrap phase, PBTS introduces a novel approach to optimizing bandit algorithms in periodic reward contexts.
[LG-64] SurvCF(t): Counterfactual Explanations for Survival Analysis in Predictive Maintenance Multivariate Time Series Data
链接: https://arxiv.org/abs/2607.16969
作者: Zara Karazian,Panagiotis Papapetrou,Sindri Magnússon,Erik Frisk,Tony Lindgren
类目: Machine Learning (cs.LG)
*备注: 10 pages, 3 figures
Abstract:Predictive maintenance relies on accurate Remaining Useful Life estimation, often formulated using survival analysis over multivariate time-series data. While modern deep survival models achieve strong predictive performance, their black-box nature limits their use in safety-critical settings where actionable insight is required. In this work, we introduce \textitSurvCF(t), the first framework for generating counterfactual explanations for survival models operating on time-series data. \textitSurvCF(t) identifies minimal, plausible, and temporally consistent changes to an asset’s operational history that increase its predicted life time, framing explanation as a constrained optimization problem combining validity, proximity, sparsity, and plausibility. We evaluate the method on multiple benchmarks, including C-MAPSS, N-CMAPSS, and a real-world case study of the Scania Component_X dataset, demonstrating its ability to produce actionable and interpretable interventions. Our results show that \textitSurvCF(t) bridges the gap between survival prediction and prescriptive maintenance, enabling explainable and decision-oriented AI for maintenance strategies.
[LG-65] ght Sample Bounds for Renyi and Min-Entropy Estimation
链接: https://arxiv.org/abs/2607.16966
作者: Arman Adibi,Piotr Krysta
类目: Information Theory (cs.IT); Computational Complexity (cs.CC); Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注:
Abstract:Estimating entropy from samples is fundamental in information theory and property testing. Shannon entropy measures average uncertainty and can be estimated to constant additive accuracy over a k -symbol alphabet using \Theta(k/\log k) samples. Min-entropy depends only on the most likely symbol. Both are special cases of order- \alpha R’enyi entropy, H_\alpha . We characterize the sample complexity of estimating min-entropy and R’enyi entropy for k and integer \alpha1 ; our lower bounds also hold for noninteger \alpha\ge1.001 . We prove that min-entropy estimation to constant additive accuracy has sample complexity \Theta(k\log k) . The upper bound uses the largest empirical frequency and concentration via dyadic grouping. The matching lower bound hides a slightly heavier symbol at a uniformly random location. Thus, min-entropy requires \Theta(\log^2 k) more samples than Shannon entropy and corrects a previously stated \Theta(k/\log k) characterization. For every integer 2\le\alpha\le c_0\log k , we prove the matching fixed-accuracy bound \Theta_c_0(\alpha k^1-1/\alpha) . Previous results gave \Omega_\alpha(k^1-1/\alpha) for fixed integer \alpha1 and O_c_0(\alpha^2k^1-1/\alpha) for all integer \alpha1 . Our upper bound analyzes an unbiased falling-factorial estimator based on \alpha -way collisions, while a hidden-heavy-coordinate construction gives the matching lower bound and shows that the factor \alpha is unavoidable. For every real 1.001\le\alpha\le c_0\log k , we prove the uniform lower bound \Omega_c_0(\alpha k^1-1/\alpha) . Finally, since 0\le H_\alpha§-H_\infty§\le\log k/(\alpha-1) , min-entropy uniformly approximates H_\alpha when \alpha is a sufficiently large multiple of \log k . Combining this reduction with our min-entropy bounds gives \Theta_\varepsilon(k\log k) sample complexity in the high-order regime. Subjects: Information Theory (cs.IT); Computational Complexity (cs.CC); Machine Learning (cs.LG); Statistics Theory (math.ST) Cite as: arXiv:2607.16966 [cs.IT] (or arXiv:2607.16966v1 [cs.IT] for this version) https://doi.org/10.48550/arXiv.2607.16966 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-66] Enhancing Personalized Bladder Cancer Treatment Through Reinforcement Learning: A Recurrent Patient State Transition Decision Support Framework
链接: https://arxiv.org/abs/2607.16916
作者: Divyansh Chawla,Anshu Garg,Isshaan Singh
类目: Machine Learning (cs.LG)
*备注:
Abstract:Bladder cancer treatment requires personalized and adaptive decision-making, particularly for recurrent disease, where treatment effectiveness changes across successive clinical episodes. Conventional clinical decision support systems typically rely on static treatment guidelines or single-step predictive models, limiting their ability to capture disease progression over time. This paper presents a recurrent patient state-transition simulation framework for bladder cancer treatment planning that integrates predictive state-transition modeling with a Markov Decision Process (MDP) and a Deep Q-Network (DQN) reinforcement learning environment. The predictive module estimates changes in tumor characteristics following treatment, while the reinforcement learning agent sequentially optimizes treatment decisions by interacting with simulated patient trajectories. This framework enables dynamic, patient-specific treatment planning by continuously adapting recommendations to evolving clinical states. It also generates interpretable treatment trajectories and detailed simulation logs to improve transparency and support clinical decision-making. The proposed framework was evaluated against existing reinforcement learning-based treatment planning approaches. It achieved a cumulative reward of 63,918.87, an average training loss per episode of 0.0056, and a policy improvement score of 6.62%, demonstrating effective sequential learning and robust treatment optimization in a simulated recurrent treatment environment. These findings highlight the potential of recurrent patient state-transition simulation with reinforcement learning as a flexible decision-support framework for personalized bladder cancer treatment planning and AI-assisted precision oncology.
[LG-67] When Can Safe Controllers Adapt? Information before Commitment
链接: https://arxiv.org/abs/2607.16895
作者: Venkatesh Saligrama
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注:
Abstract:Safe adaptive control is online adaptation under a safety guarantee on the learning trajectory itself. The controller may use any causal, history-dependent rule and act differently across environments as data arrive. Only its safety guarantee is uniform: the same rule must satisfy it under every initially plausible model. Performance is measured against a safe oracle that knows the realized model. Many finite-time analyses assume persistent excitation of the uniformly safe closed loop, so the data distinguish every pair of models requiring different control decisions. Under that assumption, feasibility is already settled; only the rate remains. We ask instead: Do the safety constraints permit such an informative experiment at all? While an alternative remains plausible, the controller must preserve a safe continuation under it. We call the first action that forecloses such a continuation commitment. Chance safety allows commitment only on an event rare under the alternative, and the evidence must arrive beforehand: the observation generated by the committing action is too late. We define precommitment information as the KL divergence between learner-visible laws stopped before commitment. Our main result is a causal reduction. The commitment rule determines (1) the probability that safety permits commitment under the alternative, (2) the target-side cost of remaining noncommittal, (3) and the information available when the decision is made. Bounded precommitment information therefore leaves a fixed fraction of the oracle gap unavoidable. If the gap is \Omega(T), every uniformly safe policy has linear regret. We establish the obstruction in a constrained linear system with quadratic regulation cost. We also prove recovery in special cases and derive semidefinite upper certificates for deterministic linear-Gaussian systems. Subjects: Machine Learning (cs.LG); Systems and Control (eess.SY) Cite as: arXiv:2607.16895 [cs.LG] (or arXiv:2607.16895v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16895 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-68] VGL-CFM:Generating and Forecasting Time-Varying Trajectories of Dynamic Networks with Conditional Flow Matching
链接: https://arxiv.org/abs/2607.16894
作者: Om Roy,Yashar Moshfeghi,Keith Malcolm Smith
类目: Machine Learning (cs.LG)
*备注:
Abstract:Many complex systems such as brain networks, financial markets, and gene-regulatory circuits are described not by a fixed graph but by one that changes over time. A standard way to summarise such structure at each instant is the sparse precision (inverse-covariance) matrix, and the time-varying graphical lasso (TVGL) turns a multivariate signal into a smooth chain of these matrices. We introduce TVGL-CFM, a single model that learns the distribution of such chains and can both generate new, realistic time-varying network trajectories for a given class and forecast how an observed trajectory will continue. Each precision matrix lives on a curved space of positive-definite matrices, but a log-Euclidean chart flattens an entire trajectory into an ordinary vector space, so a simple conditional flow-matching model can be trained and sampled there while every decoded matrix is guaranteed to be a valid precision matrix. For forecasting we start the flow not from noise but from a rough extrapolation of the recent history, so the model only has to learn a small correction. Across EEG motor-imagery, chaotic systems, and gene-expression data, TVGL-CFM generates trajectories that keep the class-discriminative structure of real data, and it forecasts future connectivity more accurately than raw-signal baselines. Generating the structured precision trajectory directly is therefore more faithful than generating raw signals and estimating connectivity afterwards.
[LG-69] HyBDM: Multi-Scale Hybrid Experts for Time Series Forecasting with Bidirectional Dependency Modeling
链接: https://arxiv.org/abs/2607.16882
作者: Wenqiang Ma,Chen Cheng,Xue Cheng,Jiarui Ye
类目: Machine Learning (cs.LG)
*备注:
Abstract:Time series forecasting (TSF) is vital to many applications, yet existing models often struggle to capture the heterogeneous long-range global patterns and short-range local variations in multivariate time series. While some approaches partially model these dependencies, they often do not jointly exploit temporal and feature-wise information. To address this challenge, we propose HyBDM, a multi-scale hybrid model that decomposes temporal dynamics into global patterns and local variations, which are modeled by two specialized experts. The Global Patterns Expert employs an enhanced BiConv-Mamba module that integrates bidirectional convolutions, an M-SSM layer, a forgetting mechanism, and a GDD-MLP module for cross-channel modeling. The Local Variations Expert uses a Local Window Transformer (LWT) to perform efficient locality-aware attention with reduced computational complexity. In addition, a Multi-Scale Patcher and a Long-Short Router enable multi-resolution representations and adaptive fusion of the two experts. Experiments on six benchmark datasets show that HyBDM outperforms state-of-the-art methods in both forecasting accuracy and computational efficiency, demonstrating its effectiveness in bridging global-local dependencies for multivariate TSF.
[LG-70] Bridging battery design and health assessment through virtual sensing and physics-informed learning
链接: https://arxiv.org/abs/2607.16864
作者: Wendi Guo,Søren Byg Vilsen,Daniel Ioan Stroe,Yaqi Li,Yicun Huang,Ashima Verma,Daniel Brandell
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 28 pages, 7 figures
Abstract:Supercharging of lithium-ion batteries (LiBs) requires robust health monitoring to ensure durability, safety, and user confidence, particularly for emerging vehicle-to-grid applications with bidirectional energy flows. Yet battery management remains largely disconnected from the material and structural origins of aging, limiting both interpretable health assessment and informed battery design. Here we propose a physics-informed learning framework with virtual sensing that infers hard-to-measure design parameters, including solid-state diffusion coefficient, electrode thickness, ion concentration, and particle size, directly from standard battery management system (BMS) measurements. Across diverse fast-charging strategies and driving profiles, embedding a digital-twin-derived particle-cracking mechanism as a soft constraint reduces trajectory and lifetime prediction errors by 6-8 times relative to state-of-the-art machine learning baselines using only 2% early-life observations. We further show that accurate degradation extrapolation does not require fully resolved governing equations; validated partial mechanisms, jointly refined with limited data, provide sufficient guidance. Virtual sensing transforms standard charging signals into latent design variables without additional sensors, bridging observable battery behavior and underlying aging processes while reducing capacity loss error by up to 39%, end-of-life (EOL) error by 17%, and prediction variability by up to 54%, enabling real-time exploration of new battery configurations. More broadly, the proposed framework establishes a practical feedback loop between deployment and development, demonstrating how real-world operation can continuously inform upstream design decisions across complex multiphysics systems.
[LG-71] Honest Physical-Support Inference after Latent Dictionary Learning: Collision Singularities and Minimax Resolution
链接: https://arxiv.org/abs/2607.16813
作者: Guan-Ju Peng
类目: Machine Learning (cs.LG); Statistics Theory (math.ST)
*备注:
Abstract:Sparse-support uncertainty is usually quantified by treating the dictionary as known, an assumption that can produce overconfident, label-dependent conclusions when the dictionary is learned from latent sparse mixtures. Near collisions of coherent atoms, a test signal may identify the active physical group even though the training data cannot distinguish the physical rays within it. We develop inference for active physical rays, unit atoms modulo sign, after latent dictionary learning. In a fixed-dimensional Gaussian train-test experiment, we retain all dictionaries compatible with a robust training-moment region, profile the test representation over them, and project surviving configurations onto a permutation-invariant support space. The resulting confidence correspondence can report cross-sheet inconclusiveness, group resolution with child ambiguity, or fine-support resolution. We characterize both its statistical cost and decision-theoretic benefit. Residual block orientation first affects the latent training density at cubic order, yielding information of order s^6 , where s is the within-block collision scale. The correspondence provides high-probability-over-training conditional test coverage, with resolution governed separately by parent detectability, test-time support separation, and learned-dictionary orientation. In the resolved fixed-shell regime, its projective Hausdorff diameter contracts at the minimax-optimal rate s \wedge (\sqrtNs^2)^-1 , up to constants. A restricted-task theorem further determines when coefficient asymmetry allows test replication to supplement training information and when calibration uncertainty remains irreducible. The framework thus yields honest, resolution-adaptive support statements and guides the allocation of training versus test measurements. Subjects: Machine Learning (cs.LG); Statistics Theory (math.ST) Cite as: arXiv:2607.16813 [cs.LG] (or arXiv:2607.16813v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16813 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-72] Interpretable Anomaly and Drift Detection with Gaussian Mixture Models
链接: https://arxiv.org/abs/2607.16811
作者: Behnam Asadi
类目: Machine Learning (cs.LG)
*备注: 16 pages, 4 figures. Code and experiments included in the source
Abstract:We revisit Gaussian Mixture Models (GMMs) as a lightweight, interpretable tool for anomaly detection and, in particular, for detecting distributional drift in data streams. We make three practical choices explicit and evaluate them on seven public benchmarks. First, the number of mixture components is selected automatically by the Bayesian Information Criterion, initialised by k-means, removing the need to fix it in advance. Second, individual observations are scored by their negative log-likelihood under a GMM fitted to normal data, with thresholds set at a target false-alarm rate using Extreme Value Theory. Third, the same interpretable model extends to distributional drift: each Gaussian component is a named “regime,” and the fraction of a stream window that matches no regime – its unexplained mass – is a drift signal that is itself the explanation. We benchmark this against a model-free kernel two-sample test (Maximum Mean Discrepancy, MMD) and against two GMM-to-GMM divergences (a closed-form Cauchy-Schwarz divergence and a matching-based KL surrogate). Across seven benchmarks ranging from 3 to 64 dimensions and five random splits, the GMM point detector is competitive with – though rarely more accurate than – Isolation Forest, Local Outlier Factor, one-class SVM, ECOD, COPOD and an autoencoder, while uniquely yielding an interpretable model. For drift, MMD is the strongest pure detector, but the interpretable unexplained-mass statistic matches it when anomalies form novel regimes (and honestly fails, as MMD does not, when drift is a pure re-weighting of existing regimes). Every alarm is explainable: anomalies lie a median of 3-10 sigma outside their nearest regime vs. about 1 sigma for normal points, and a drift alarm reports the fraction of the window matching no known regime. All code and experiments are released.
[LG-73] MultiLoReFT: Decoupling Shared and Modality-Specific Subspaces in Multimodal Learning via Low-Rank Representation Fine-Tuning
链接: https://arxiv.org/abs/2607.16789
作者: Sana Tonekaboni,Viktoria Schuster,Caroline Uhler
类目: Machine Learning (cs.LG)
*备注:
Abstract:Real-world perception and decision making are inherently multimodal, integrating complementary signals across modalities. However, training multimodal models faces two main obstacles. First, collecting large-scale, well-aligned paired multimodal datasets is often impractical, making end-to-end multimodal training difficult. Second, existing multimodal representations frequently entangle information shared across modalities with modality-specific information, hindering interpretability and control. We introduce MultiLoReFT, an efficient and scalable low-rank representation fine-tuning framework for multimodal learning with pretrained unimodal models. MultiLoReFT extends low-rank adaptation to the multimodal setting and learns interpretable projection subspaces that decouple shared and modality-specific information. Across simulated and real-world benchmarks, it produces representations that support multimodal prediction while explicitly revealing how shared and modality-specific information is distributed across modalities.
[LG-74] On the Potential of Graph Neural Networks as Metamodels for Supply Chain Optimization: Dataset Architectures and Directions
链接: https://arxiv.org/abs/2607.16769
作者: Tushar Lone,Neha Karanjkar
类目: Machine Learning (cs.LG)
*备注: 13 pages, 5 figures
Abstract:Graph Neural Networks (GNNs) have emerged as a powerful, differentiable class of learning models for graph-structured systems. Their ability to generalize across topologies opens the prospect of a surrogate for combined structural and parametric optimization, which classical metamodels cannot offer. Supply chains are a natural target, yet the use of GNN surrogates for supply chain problems is largely unexplored. This paper lays the foundation, presents initial steps, and discusses key research directions. As a foundation, we formulate the problem and create a large public training dataset of programmatically generated supply chain graphs with input parameters and steady-state performance metrics obtained using our SupplyNetPy simulation library. As initial steps, we explore GNN architectures that work well as surrogates for node- and network-level predictions, and analyze their accuracy-compute trade-off against simulation. Most importantly, we outline the exciting directions this opens, namely gradient-based optimization over topology, fast design-space exploration, and sensitivity analysis.
[LG-75] Robust Losses from Univariate Base Functions for Noisy-Label Learning
链接: https://arxiv.org/abs/2607.16768
作者: Peng Hu,Jianwei Ma
类目: Machine Learning (cs.LG)
*备注:
Abstract:Learning with noisy labels is a fundamental problem in training reliable deep neural networks. Robust loss functions provide a direct and effective way to mitigate the adverse effects of label noise. However, most existing robust losses are designed directly at the level of the final multiclass objective, which makes it difficult to systematically characterize and extend their robustness properties. In this paper, we propose a general framework that constructs robust multiclass losses from univariate base functions. By defining mapping operators from base functions to multiclass losses, the robustness of the induced losses can be characterized through simple properties of the base functions. We develop two complementary construction schemes, Target Separation and Binary Reduction, corresponding to inter-class independent and inter-class dependent formulations, respectively. For both schemes, we analyze their symmetry and asymmetry properties and derive corresponding sufficient conditions, which provide theoretical criteria for noise-robust loss design. The proposed framework also provides a new route to constructing symmetric losses, serving as a complement to normalization-based symmetric loss designs. Extensive experiments on synthetic and real-world noisy-label benchmarks demonstrate that the proposed losses achieve competitive or superior performance under various noise settings.
[LG-76] he Anatomy of a Truth Direction: Knowledge-Dependent Dimensionality a Relational Law and a Convergent Category Geometry in Small Language Models
链接: https://arxiv.org/abs/2607.16741
作者: Francesco Karim Vicidomini
类目: Machine Learning (cs.LG)
*备注: Code and data: this https URL
Abstract:Bürger et al. (2024) demonstrated that truth representations in large language models are universal across statement polarity but reside within a multidimensional subspace. We extend that framework along three questions: how the dimensionality of the subspace depends on the model’s knowledge, which architectural component builds the truth direction, and what the direction is a mixture of. In Part I (one model), a training-free directional probe derived from the SVD of hidden-state minimal pairs shows that the dimensionality of truth is knowledge-dependent: the signal is concentrated on a single axis for behaviorally known facts (held-out AUC 0.938) and becomes diffuse as knowledge decreases or material heterogeneity increases; seven falsification experiments sustain the one-dimensional reading, and the same decomposition recovers the supervised polarity direction of Bürger et al. at cosine 0.959 without polarity labels. In Part II (four base models, two families, pre-registered predictions), a relational law emerges: attention propagates truth frames it did not write, the feed-forward network opposes the frame of its current block, and the genuine post-peak decay is causally attributed on all four models to the SwiGLU value stream; per-category truth axes form a semantically signed arrangement that contradicts the lexical surface and converges across families (Mantel p = 0.0009). Stress tests at three scales (33 categories, 888 pairs per relation, six sampling seeds) expose a sign instability of the per-category orientation, repair it with a declared spectral consensus gauge, and sharpen the convergence into a knowledge-gated law, strengthening from +0.42 on dense mixed-knowledge material to +0.74 on knowledge-restricted categories: the geometry of the mixture belongs to the knowledge domain rather than to the architecture.
[LG-77] BG4Sea: Biogeochemical Seasonal Forecastability via Progressive Information Scaling
链接: https://arxiv.org/abs/2607.16731
作者: Gabriela Martinez Balbontin,Anastase Charantonis,Dominique Bereziat,Stefano Ciavatta
类目: Machine Learning (cs.LG)
*备注:
Abstract:Marine biogeochemical forecasting is increasingly important for managing marine ecosystems and the carbon cycle, yet global, seasonal forecast products lag far behind physical oceanography, held back by the complexity of the processes involved and by data scarcity. We introduce BG4Sea, which to our knowledge is the first global, data-driven system to produce multivariate seasonal forecasts of the marine biogeochemical state. BG4Sea is a modular architecture with a column autoencoder that compresses the vertical column into a low-dimensional latent space, a latent forecaster propagates this representation forward in time, a surface-forcing conditioner that injects physical boundary information via Feature-wise Linear Modulation (FiLM), and a horizontal-coupling module that incorporates neighboring-column context through cross-attention. The model is trained and evaluated on the global ocean reanalysis BIORYS4 (NEMO/PISCES), and produces six-month forecasts at 1/4 degree, monthly resolution for dissolved chemistry, biology, and carbon-pool variables, outperforming persistence and climatology across most variables and lead times. We position BG4Sea as an interpretable baseline for future, more expressive approaches, and discuss predictability attribution to each component, alongside the model’s structural limitations.
[LG-78] Graph-Embedded Intuitionistic Fuzzy Broad Learning System: A Multi-view Framework
链接: https://arxiv.org/abs/2607.16728
作者: Yogesh Kumar,Manju,Mudasir Ganaie
类目: Machine Learning (cs.LG)
*备注:
Abstract:The Broad Learning System (BLS) has been widely used for data classification and is based on a layer-by-layer feed-forward structure. However, it gives the same importance to all data points, which reduces its effectiveness on real-world datasets with noise and outliers. In addition, it does not consider the geometric structure of the data and has limitations in handling data from multiple sources. To address these challenges, we propose a Multi-View Graph-Embedded Intuitionistic Fuzzy Broad Learning System (MVGIFBLS) that integrates multi-view learning, graph embedding, and intuitionistic fuzzy theory into the BLS framework. This design enables the model to combine information from multiple sources and learn more discriminative representations. Graph embedding captures the geometric relationships among samples and improves class separation through intrinsic and penalty subspaces based on local Fisher discriminant analysis. Intuitionistic fuzzy theory enhances robustness to noise, while kernel-based neighborhood analysis captures local data structures. We evaluate the proposed framework on several UCI, KEEL, and AwA benchmark datasets using comparative evaluation, Gaussian feature noise analysis, ablation studies, and statistical analysis. The results demonstrate that each component contributes positively to the overall framework and that the proposed MVGIFBLS consistently achieves higher Area Under the Curve (AUC) scores and maintains robust performance under Gaussian feature noise.
[LG-79] Half the Experts All the Code: One-Shot Domain Pruning of Mixture-of-Experts LLM s for Coding
链接: https://arxiv.org/abs/2607.16721
作者: Anik Jha
类目: Machine Learning (cs.LG)
*备注: 15 pages, 3 figures. Code and pruned checkpoints: this http URL
Abstract:The strongest open-weight coding models are mixture-of-experts (MoE) networks: most of their size comes from large pools of “expert” subnetworks, of which only a few act on any token. That pool is why these models do not fit on the machines most developers own, yet for a user who only wants coding help, most experts encode abilities that will never be invoked. We ask how many experts can be removed, and which, by pruning two recent open-weight MoE models from different families (Qwen3.6-35B-A3B and Gemma-4-26B-A4B) under five selection strategies, judged the way a user would: by whether the model still writes correct code. Half the experts can be removed from either model with no statistically detectable loss on the primary code benchmark, and the damage lands almost entirely on abilities outside coding, the intended trade. But the winning strategy flips between the two models, so a recipe validated on one family cannot be assumed to work on another. We further show that perplexity, the metric much of the pruning literature leans on, can rate a broken model above an intact one; that a lightweight fine-tune recovers about half of what aggressive pruning loses; and that against quantizing the full model to the same memory, pruning wins only where quantization would have to drop below 3 bits per weight. Five attempts to overturn that crossover, with failure criteria fixed in advance (better calibration, guarded selection, causal expert importance, failure attribution, and an agentic evaluation letting each model repair its failures from execution feedback), all leave it standing; the last shows single-shot benchmarks overstate compression penalties broadly, as one repair turn erases the 2-bit quantization penalty entirely. Expert pruning works, but it demands per-model validation on the task the model will actually serve.
[LG-80] Effects of width-dependent model hyperparameters and ell_2-regularization on the loss landscape of two-layer ReLU networks ICML2026
链接: https://arxiv.org/abs/2607.16720
作者: Haruka Eshima,Makoto Yamada
类目: Machine Learning (cs.LG)
*备注: ICML2026 4th Workshop on High-dimensional Learning Dynamics (HiLD)
Abstract:Understanding deep neural networks remains a central challenge in machine learning. In particular, the theoretical properties of even two-layer ReLU networks, especially in the presence of weight decay, remain poorly understood. To this end, we derive a sufficient condition on the hyperparameter settings under which the global minima collapse to the zero solution. Interestingly, our experiments reveal that using AdamW as an optimizer prevents the collapse of the learned parameters, whereas using SGD does not, which may help explain the success of AdamW in deep learning training. In addition, when restricting the input dimension to one, we derive an analytical solution for the globally optimal parameter sets of two-layer ReLU networks and show that \ell_2 -regularization has a width-invariant effect on connectivity, but its dimensionality-reducing effect becomes stronger as the network width increases. These results provide insight into how width-dependent hyperparameters influence the geometry of regularized loss landscapes.
[LG-81] A Framework for Early Sepsis Prediction via Self-Supervised (JEPA) and Federated Representation Learning
链接: https://arxiv.org/abs/2607.16681
作者: Umair bin Mansoor,Munaf Rashid,Roomi Naqvi
类目: Machine Learning (cs.LG)
*备注: 11 pages, 6 figures, Submitted to arXiv
Abstract:Early sepsis prediction from electronic health records is challenged by irregular sampling, high missingness, and class imbalance. We systematically compare four modeling paradigms – self-supervised Joint Embedding Predictive Architecture (JEPA) via masked latent prediction, self-supervised VICReg (variance-invariance-covariance regularization) with two-view augmentation, semi-supervised fine-tuning of a VICReg-pretrained encoder, and supervised Temporal Convolutional Network (TCN) – alongside raw-feature baselines. All models share a common preprocessing pipeline of hourly binning with forward-fill imputation applied to 7 biomarkers selected via sparsity analysis from the MIMIC-III dataset. Our best model (JEPA + XGBoost + mean pooling) achieves AUPRC 0.636 at the time of onset (H0), approaching the SupMix benchmark (0.667) while using 83% fewer biomarkers. The Tier 1 pipeline – VICReg pretraining followed by semi-supervised fine-tuning and XGBoost – achieves AUPRC 0.510 at H0, a 3.1 \times improvement over the raw-feature baseline (0.165) and a 7.6% improvement over the end-to-end supervised TCN (0.474). Crucially, the fine-tuned VICReg encoder exhibits the most temporally persistent representations, degrading only 16.8% from H0 to H10 compared to 47.5% for supervised TCN and 65.3% for JEPA, demonstrating that self-supervised pretraining with task-aware fine-tuning yields features that are both sharp near onset and robust across prediction horizons.
[LG-82] CLDRoute: Conditional Latent Diffusion for Routability Map Generation in Physical Design
链接: https://arxiv.org/abs/2607.16674
作者: Kiran Thorat,Nicole Meng,Caiwen Ding,Yingjie Lao,Zhijie Jerry Shi
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate routability estimation during physical design is important for reducing costly post-routing iterations. Prior learning-based methods treat this task as deterministic prediction, mapping placement-stage features to a single congestion or DRC outcome. We instead formulate routability estimation as a conditional generation problem, where both routing congestion and DRC violations are modeled as spatially structured routability fields. Our framework, Conditional Latent Diffusion for Routeability estimation (CLDRoute), uses physics-aware conditioning and task-specific latent modeling to handle the different characteristics of congestion and DRC maps. This allows our method to supports sample-based inference, producing both a mean prediction and a spatial uncertainty estimate for the same input design. On CircuitNet 2.0 (N28), our method achieves, for DRC violation generation, an SSIM of 0.9678, an MAE of 0.0028, and a TopK@1% of 0.3494; for congestion generation, it achieves an SSIM of 0.9031, an MAE of 0.0286, and an NZ-Pearson of 0.3692. Overall, our framework provides a more practical view of routability at placement by generating both the expected outcome and its uncertainty.
[LG-83] Multimodal Attention-based Deep Learning for Emergency Triage with Electronic Health Records
链接: https://arxiv.org/abs/2607.16662
作者: Hazqeel Afyq Athaillah Kamarul Aryffin,Kamarul Aryffin Baharuddin,Mohd Halim Mohd Noor
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate emergency triage decision is critical to avoid clinical deterioration, morbidity, and mortality. Machine learning-based triage system involves acquiring the main presenting complaint in text form and assessing vital signs in numerical data, enabling an automated and efficient analysis of patient information for timely and accurate prioritization of medical attention. However, modelling the intricacies of both data types requires a comprehensive understanding of the temporal structure and dependencies within the data. Thus, the aim of this study is to propose a multimodal deep learning architecture that can effectively handle both tabular and textual data. Furthermore, the proposed model exploits self-attention to to capture both local and global relationships between the features. A dataset consisting of 11,102 triage data collected from emergency department of Hospital Universiti Sains Malaysia is used for model development and validation. The proposed model demonstrated an increase of 1.95% in accuracy, 2.49% in F1-score, and 1.41% in ROC AUC compared to the baseline model. The experimental results demonstrated the potential of the proposed model in predicting triage decisions.
[LG-84] Discrete Ricci Curvature on Protein Contact Graphs for Lightweight Fold Classification
链接: https://arxiv.org/abs/2607.16553
作者: Jianru Shen
类目: Machine Learning (cs.LG)
*备注: Accepted at IEEE International Conference on Future Machine Learning and Data Science (FMLDS)
Abstract:Protein fold classification can be approached via sequence-based representations or structural descriptors, but direct comparisons between lightweight handcrafted descriptors and pretrained protein language model embeddings remain limited. We investigate discrete Ricci curvature on Calpha contact graphs as a lightweight structural descriptor for fold classification. Each protein domain is represented by a 22-dimensional fixed-length feature derived from summary statistics and quantiles of Ollivier-Ricci and Forman-Ricci edge curvature distributions. We evaluate on CATH top-10 Topology classification and on the ASTRAL 40%-identity SCOPe top-10 Fold benchmark, comparing against geometry, contact-graph statistics, persistent homology, and mean-pooled ESM-2 (150M) baselines. On both datasets, lightweight structural descriptors substantially outperform mean-pooled ESM-2 embeddings, with a larger performance gap on the ASTRAL 40% SCOPe benchmark. Ricci alone uses 22 dimensions, or 3.4% of the ESM-2 baseline dimensionality, and already outperforms mean-pooled ESM-2 on both datasets. Combining Ricci with persistent homology yields the strongest performance, achieving macro-F1 of 0.71 on CATH and 0.68 on SCOPe with a 112-dimensional feature vector. These results identify a regime where lightweight interpretable graph descriptors offer a practical alternative to pretrained protein language model embeddings.
[LG-85] Hierarchical Domain Generalization
链接: https://arxiv.org/abs/2607.16528
作者: Chenxiao Yang,Zhiyuan Li,Shai Ben-David,Nathan Srebro
类目: Machine Learning (cs.LG)
*备注:
Abstract:We study hierarchical domain generalization as a problem of extrapolation from finite observed regions to an entire instance space, replacing i.i.d. sampling with arbitrary domain hierarchies. We show that the central obstruction is not only the complexity of the hypothesis class, but the train/test domain partition through which evidence is revealed. In particular, no matter how small the class or how large the training size, some partition makes generalization fail for some target. These results suggest that modern generalization theory must treat domain structure as a first-class object.
[LG-86] Feedback Attribution and Representation Geometry: Metrics for Comparing Individual and Shared Rewards in MARL ICML2026
链接: https://arxiv.org/abs/2607.16524
作者: Tasha Pais,Richard Higgins
类目: Machine Learning (cs.LG)
*备注: Accepted at the Reinforcement Learning from World Feedback Workshop, ICML 2026
Abstract:Cooperative multi-agent RL systems routinely use team-averaged rewards, a feedback-attribution choice that gives each agent the team outcome regardless of its individual contribution. We ask whether this leaves a measurable signature, geometric or behavioral, on learned representations. We propose EffRank/ n (effective rank normalized by agent count) and D_\textact (mean pairwise KL divergence between agents’ action distributions) as low-overhead diagnostics for reward-attribution effects, then test them on competent MAPPO agents in SMACv2 \textttprotoss_5_vs_5, where unit type is encoded in the observation. In an observation \times reward-attribution comparison (unit type observed vs.\ masked; individual damage-contribution reward vs.\ shared team reward), geometry follows observation rather than reward. With unit type observed, shared and individual rewards have similar EffRank/ n ( 0.31\pm0.03 vs.\ 0.29\pm0.02 ) and probe accuracy ( 0.75\pm0.05 vs.\ 0.73\pm0.05 , both \gg 1/3 chance), while D_\textact leans higher under individual rewards ( 1.23\pm0.06 vs.\ 1.07\pm0.20 ). Masking unit type cuts the above-chance probe signal by more than half, to 0.49 in both reward arms. In short: individually rewarded agents are competent and separable by role, but on SMACv2 the observation explains the geometry and reward attribution shows up mainly in behavior. Thus geometric diagnostics must control for observed role information and test persistent roles that are not directly observed. EffRank/ n and D_\textact add 5% overhead.
[LG-87] Foresight Residual RL for Long-Horizon Robot Manipulation with Vision-Language-Action Models IROS2026
链接: https://arxiv.org/abs/2607.16506
作者: Yuhan Liu,Xinyu Zhang,Litao Liu,Abdeslam Boularias
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注: Accepted at IROS2026. Project website: this https URL
Abstract:Vision-Language-Action (VLA) policies offer strong general-purpose manipulation priors, but often fail on tight-tolerance, contact-rich assembly due to long-horizon credit assignment and subtask coupling: a state that is geometrically successful for the current skill can be brittle for downstream skills. We show this failure mode in residual reinforcement learning (RL) over a frozen VLA base policy: constant sparse success rewards improve each subtask in isolation yet yield little or no gain when skills are chained, because terminal state quality is uncontrolled. We propose Foresight Residual RL, which optimizes handoff quality by augmenting each subtask’s sparse success reward with an offline-estimated foresight value – the probability of future subtask success conditioned on the terminal state of the current subtask. Concretely, we (i) train a visual foresight predictor from images of terminal states of the base policy, labeled using downstream rollout statistics, and (ii) train residual policies via backward foresight induction, using the predictor output as a reward multiplier. On a three-phase wrench-based nut-tightening assembly task in Isaac Gym (grasp, move-insert, rotate), our method achieves 85.6% full-task success, outperforming standard subtask residual RL (54.5%) and VLA baselines, while leaving per-subtask success unchanged. These results highlight that improving long-horizon performance requires shaping which successful states are produced at each sub-task, not only whether success occurs.
[LG-88] Leakage-Robust Evaluation and Data-Scale Sensitivity of Attention-Enhanced Multi-Task Learning for Joint Fault Diagnosis and Remaining Useful Life Estimation
链接: https://arxiv.org/abs/2607.16493
作者: Md Mahamudur Rahaman Shamim,Md. Nuruzzaman,Zannatul Ferdus,Md Rajib Ahmed,Abieer Nwshad Anward,Mohammad Tooneer,Johir Uddin Khan,Khalid Hossen
类目: Machine Learning (cs.LG)
*备注: Submitted in Next Research (Elsevier)
Abstract:Multi-task deep learning models that jointly perform fault classification and remaining useful life (RUL) regression are increasingly used in predictive maintenance, yet reported performance can be strongly affected by how sliding-window sequences are split into training and test sets. We investigate this issue using AMTLNet, an attention-enhanced multi-task architecture, on three public benchmarks: NASA C-MAPSS, NASA IMS, and the UCI Hydraulic System dataset. We show that naive splitting can inflate classification accuracy from a genuine 20-60 percent to 99.9 percent, or reduce it to 0 percent through degenerate class representation. To address this, we introduce a chunk-based, leakage-audited splitting protocol and evaluate all models using five seeds, one-way ANOVA, and Tukey HSD tests. On C-MAPSS, with 19,976 leakage-free training windows, AMTLNet matches a single-task CNN-LSTM baseline in classification, achieving 84.12 +/- 0.96 percent accuracy with Tukey p = 1.0, and reaches an R2 of 0.86 +/- 0.01 while significantly outperforming a naive multi-task baseline. On the smaller Bearing and Hydraulic datasets, multi-task training is unstable, but the failure mode differs: classification degrades for Bearing, whereas regression degrades for Hydraulic. We relate this asymmetry to label provenance and propose a practical framework for deciding when joint training is appropriate under data scarcity. Ablation results show that the multi-head attention branch is the main contributor to regression stability. Removing it reduces R2 from 0.861 to 0.766 and more than doubles classification variance, whereas the convolutional branch contributes little to regression despite using about one-third of the parameters. This study contributes a reusable leakage-audit protocol, seed-transparent evaluation, and evidence that task-specific stability depends more on label provenance than on task type.
[LG-89] Compact convolutional neural networks for AI-based drone detection system
链接: https://arxiv.org/abs/2607.16455
作者: Gábor Farkas,Gábor Fazekas,Karakai Patrik,András Németh,Gábor Farkas
类目: Machine Learning (cs.LG)
*备注:
Abstract:The increasing use of first-person-view drones in modern conflicts has created a demand for compact and reliable detection systems capable of operating in complex electromagnetic environments. These drones continuously transmit video signals through onboard video transmitters, generating radio-frequency emissions that can be exploited for early detection. This study investigates the use of lightweight convolutional neural networks for automated detection of drone signals captured by a software-defined radio-based electronic warfare framework. Samples are converted into rasterized time-domain images, providing a computationally efficient input representation suitable for embedded systems. Several custom model architectures were designed and benchmarked in terms of accuracy, model size, and inference performance using a dataset containing approximately 40,000 labeled images. In addition to offline testing, the models were integrated into a GNU Radio signal processing chain for real-time evaluation. The results show that compact models can achieve high detection accuracy while maintaining low computational requirements, making them suitable for embedded radio-frequency monitoring applications. Compared with existing spectrogram-based RF detection methods, the proposed approach eliminates frequency-domain preprocessing and achieves comparable accuracy with significantly reduced computational cost.
[LG-90] EA-RMENet – Path Loss Prediction in Urban Environments using Deep Learning
链接: https://arxiv.org/abs/2607.16449
作者: Jonathan O’Shea(DCU School of Electronic Engineering),Conor Brennan(DCU School of Electronic Engineering)
类目: Machine Learning (cs.LG)
*备注:
Abstract:Accurate path loss prediction is a critical component of wireless network planning. Current path loss prediction methods typically struggle to balance the trade-off between accuracy and computational efficiency. This paper proposes the Efficient Attention Radio Map Estimation Network (EA-RMENet) which is an image data-driven, deep learning (DL) model designed for radio map estimation (RME). EA-RMENet uses a U-Net framework with an EfficientNetB5 encoder, Attention Gated (AG) skip connections, and Atrous Spatial Pyramid Pooling (ASPP). The EfficientNet encoder uses compound scaling to balance accuracy and efficiency. AG skip connections suppress irrelevant features, and the ASPP captures a multi-scale context. The model has a test prediction RMSE of 0.0334 on the RadioMapSeer3D dataset with an inference time of 0.022 seconds/sample. In the ICASSP 2023 Radio Map Prediction Challenge, the model ranks third with a competitive RMSE of 0.0406 this highlights the models potential for real-world RME.
[LG-91] Scaling Limits of Constant-Stepsize SGD at Flat Minima
链接: https://arxiv.org/abs/2607.16384
作者: Jingyi Zhang,Cheng Mao,Debankur Mukherjee
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Probability (math.PR); Machine Learning (stat.ML)
*备注: 52 pages, 3 figures, 1 table
Abstract:For stochastic gradient descent (SGD) with a constant stepsize \alpha , the invariant law of the iterates, centered at a minimizer, describes the behavior of the algorithm over long time horizons. In the strongly convex case, this invariant law has the familiar \sqrt\alpha scaling and a Gaussian limit as \alpha\downarrow 0 . We show that this behavior changes fundamentally for convex objectives H with flat minima and (sub)quadratic tails. More specifically, we study SGD with Markovian noise generated by a contractive driving chain. For every sufficiently small constant stepsize \alpha , we prove existence, uniqueness, and geometric convergence to an augmented invariant law in a Wasserstein distance induced by an \alpha -dependent metric. When the minimizer x_\star has local flatness exponent m\ge2 , meaning that \nabla^2 H(x)\asymp \lVert x-x_\star\rVert^m-2 I_d as x\to x_\star , we obtain a contraction bound with factor 1-c\alpha^m-1 , where c0 is a constant. This recovers the factor 1-c\alpha in the quadratic case m=2 . We then analyze the small-stepsize scaling limit. We show that the invariant law concentrates on the scale \alpha^1/m and that the rescaled iterates converge weakly to the stationary distribution of the stochastic differential equation dY_t=-h_0(Y_t),dt+\Sigma^1/2,dB_t , where h_0 is the limiting drift at the minimizer and \Sigma denotes the asymptotic covariance. This recovers the Gaussian limit when m=2 and gives generally non-Gaussian stationary limits in the flat case m2 . Finally, we give corresponding results for coordinate-separable objectives with unequal flatness exponents. Comments: 52 pages, 3 figures, 1 table Subjects: Machine Learning (cs.LG); Optimization and Control (math.OC); Probability (math.PR); Machine Learning (stat.ML) Cite as: arXiv:2607.16384 [cs.LG] (or arXiv:2607.16384v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16384 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
[LG-92] On Hardware-Aware Design and Optimization of Edge Intelligence
链接: https://arxiv.org/abs/2607.16297
作者: Shuo Huai,Hao Kong,Xiangzhong Luo,Di Liu,Ravi Subramaniam,Christian Makaya,Qian Lin,Weichen Liu
类目: Hardware Architecture (cs.AR); Machine Learning (cs.LG)
*备注: Author’s accepted version. Published in IEEE Design Test
Abstract:Edge intelligence systems, the intersection of edge computing and artificial intelligence (AI), are pushing the frontier of AI applications. However, the complexity of deep learning models and heterogeneity of edge devices make the design of edge intelligence systems a challenging task. Hardware-agnostic methods face some limitations when implementing edge systems. Thus, hardware-aware methods are attracting more attention recently. In this paper, we present our recent endeavors in hardware-aware design and optimization for edge intelligence. We delve into techniques such as model compression and neural architecture search to achieve efficient and effective system designs. We also discuss some challenges in hardware-aware paradigm.
[LG-93] PsiLogic: Chaos-Aware Active Cancellation for Adam with a Fair Cross-Domain Benchmark
链接: https://arxiv.org/abs/2607.16268
作者: Ali Sultonov
类目: Machine Learning (cs.LG)
*备注: 9 pages, 4 figures; code at this https URL
Abstract:Adaptive optimizers such as Adam and AdamW apply the same update rule regardless of whether training is in a chaotic early phase or near convergence. We introduce PsiLogic, an optimizer that augments Adam with a dynamic Active Cancellation Term gated by a dual exponential moving average (EMA) of scale-normalized gradient norms. The resulting chaos detector strengthens damping when gradient statistics are unstable and fades to zero as training stabilizes, providing an implicit warmup without a hand-tuned schedule. We evaluate PsiLogic against Adam, AdamW, and Lion using FairBench – a reproducible benchmark protocol with per-optimizer learning-rate sweeps, identical initialization per seed, and Welch t-tests. On an NVIDIA H100 80GB reference run (4 arenas, 3 seeds, 2000 steps, bf16 AMP), PsiLogic achieves the best validation metric in three of four arenas: NLP perplexity 7.79 +/- 0.18 vs. 8.17 +/- 0.08 (AdamW, p = 0.049), ViT top-1 accuracy 0.244 +/- 0.006 vs. 0.223 +/- 0.002 (AdamW, p = 0.015), and ResNet top-1 accuracy 0.222 +/- 0.001 vs. 0.172 +/- 0.004 (Adam, p = 0.001). On diffusion, validation MSE is statistically tied with Adam/AdamW (p = 0.49). ResNet accuracy vs. AdamW is a numerical tie without significance at three seeds (p = 0.44). Peak GPU memory is comparable across optimizers; PsiLogic incurs 1.2–1.8x wall-clock overhead on transformer-heavy arenas (implementation-bound). We release an open-source PyTorch implementation, the full FairBench harness, and all raw CSV outputs to support independent verification. Comments: 9 pages, 4 figures; code at this https URL Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.16268 [cs.LG] (or arXiv:2607.16268v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16268 Focus to learn more arXiv-issued DOI via DataCite
[LG-94] Preference-based Antibody Expression Ranking: Scaling with Large-scale Weak Supervision ICML2026
链接: https://arxiv.org/abs/2607.16263
作者: Josh Qixuan Sun,Morteza Babaie,Wenyang Hou,Mark Crowley,David Young
类目: Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE); Quantitative Methods (q-bio.QM)
*备注: Accepted at ICML 2026
Abstract:Antibody expression ranking is a critical task in antibody design, yet its modelling is severely hindered by the scarcity of labeled experimental data. To address this, we propose a unified preference-based learning framework that integrates scarce quantitative expression data with large-scale weak positive supervision from immunization data. We adapt Direct Preference Optimization (DPO) to protein language models by introducing a union-masked log-likelihood approximation and IMGT-based alignment, enabling efficient training on variable-length sequences. Evaluating on a diverse internal dataset of 1254 labeled sequences and 4 million unlabeled camelid-derived antibodies, we show that our method consistently outperforms baselines on most metrics. Our results demonstrate that preference learning can effectively learn from weak supervision, providing a scalable solution for antibody expressibility optimization in data-constrained settings. Project page: this https URL.
[LG-95] More Than Memory: Task-Conditioned Signed FFN Writes in Long-Context Retrieval
链接: https://arxiv.org/abs/2607.16254
作者: Zhibo Yang
类目: Machine Learning (cs.LG)
*备注: 23 pages, 8 figures, 12 tables
Abstract:FFNs are often treated as parametric memories. In long-context retrieval, however, the sharper question is not only what they store, but whether their native residual writes push the current retrieval state toward or away from the correct answer. We test this by scaling the model’s own FFN write one layer at a time, without editing weights or injecting external steering vectors. Across controlled literal and semantic retrieval suites, native FFN response surfaces are signed, layer-specific, and task-conditioned: the final FFN is a suppressor in 7 of 8 model-suite cases, and 60% of layers switch role between retrieval modes (95% CI [50%, 69%]). A local directional derivative along the native write separates the two monotone roles: suppressors have negative derivative in 34/35 cases, and amplifiers have positive derivative in 18/18 cases, so the roles are not reducible to write size. On a safety-filtered LongBench retrieval-QA probe, the same diagnostic predicts attenuation damage with raw R^2=0.796 on Qwen2.5-7B and 0.791 on Qwen3.5-9B; a held-out suppressor-attenuation policy improves retrieval margins over random and norm-matched controls. These results show that native FFN scaling exposes a signed, task-conditioned residual-write structure in retrieval, and that write-gradient alignment is a compact diagnostic for the two monotone roles. Comments: 23 pages, 8 figures, 12 tables Subjects: Machine Learning (cs.LG) Cite as: arXiv:2607.16254 [cs.LG] (or arXiv:2607.16254v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16254 Focus to learn more arXiv-issued DOI via DataCite
[LG-96] Learning Spatio-Temporal Foundation Models from Pure Synthetic Data
链接: https://arxiv.org/abs/2607.16251
作者: Yutong Feng,Shiyuan Piao,Yutong Xia,Xu Liu,Wenqi Fan,Fugee Tsung,See-Kiong Ng,Yuxuan Liang
类目: Machine Learning (cs.LG)
*备注:
Abstract:Spatio-Temporal Foundation Models (STFMs) aim to learn generalizable representations of complex dynamical systems across space and time. However, existing approaches suffer from distributional bias in real-world pre-training data, structural bottlenecks of autoregressive or diffusion-based paradigms, and objectives that overemphasize point-wise reconstruction in noisy observation this http URL propose \textbfNeoST, the first spatio-temporal foundation model pre-trained solely on procedurally generated synthetic systems. NeoST introduces a scalable synthetic pre-training corpus to mitigate real-world bias, a latent-space reasoning architecture that generates and iteratively refines multiple future trajectories without sequential error accumulation, and latent-space objectives that emphasize structural dynamics and enable inference-time correction under distribution this http URL experiments across diverse real-world benchmarks show that NeoST consistently outperforms existing STFMs in diverse real-world spatio-temporal systems, achieves superior long-horizon stability and inference efficiency.
[LG-97] BACON: Budgeted Human Calibration for Modeling and Evaluation with Multiple AI Judges
链接: https://arxiv.org/abs/2607.16239
作者: Lei Shi,Anlan Zhang,Rita Lyu,Zhengmian Hu,Tong Yu,David Arbour,Avi Feller,Saayan Mitra,Ritwik Sinha
类目: Machine Learning (cs.LG); Applications (stat.AP)
*备注:
Abstract:AI judges offer a scalable, low-cost alternative to human evaluation, but their outputs can be biased relative to human preferences and highly item-dependent, varying across judges, tasks, and domains. When uncalibrated AI evaluations are used for model ranking, item scoring, or population-level quality reporting, these biases can directly distort downstream decisions. We propose BACON, a four-stage pipeline that combines budgeted human calibration with multiple AI-judge outputs to produce more accurate annotations. BACON constructs full-coverage auxiliary features for every item, including multi-judge scores, token-level uncertainty statistics, and contextual embeddings. It then collects human labels for a small sampled subset and trains a cross-fitted outcome model to generate calibrated item-level surrogate predictions. These predictions support two use cases: population-level estimation of summary metrics, such as means or quantiles, using an augmented estimating-equation estimator with valid confidence intervals; and individual-level surrogate scoring for item ranking and annotation. BACON treats AI judges as auxiliary measurements rather than ground truth: human labels provide the calibration anchor, while AI-derived signals improve efficiency. Across diverse tasks, domains, and labeling budgets, BACON improves predictive accuracy and ranking consistency, and reduces bias and variance relative to raw AI outputs and purely human-label-based methods. These results show that BACON offers a practical, statistically grounded framework for scalable evaluation with limited human annotation.
[LG-98] Orthogonal Gradient Constraints Shape Noisy-Label Memorization Dynamics ICML
链接: https://arxiv.org/abs/2607.16231
作者: Richard Mai
类目: Machine Learning (cs.LG)
*备注: 7 pages, accepted into the ICML HiLD workshop 2026 @Seoul
Abstract:Modern neural networks can fit corrupted training labels, making noisy-label learning a useful setting for studying memorization-driven overfitting. Most regularization methods modify the objective, architecture, or data distribution; here we instead study a geometric intervention on the optimizer update itself. We evaluate OrthoGrad, which removes the component of each weight gradient parallel to the current weight vector, in noisy-label image classification. On MNIST with small-data regimes, OrthoGrad improves test accuracy most clearly for CNNs while reducing corrupted-label fitting. Mechanism diagnostics based on weight norms and gradient-weight cosine similarity suggest that the projection has the strongest effect when the raw gradient contains a nontrivial radial component, and becomes weaker in larger-data regimes where gradients are already nearly orthogonal to weights. Additional CIFAR-10 ResNet-18 experiments show that the method can alter memorization trajectories but does not prevent eventual noisy-label memorization. These results support orthogonal update constraints as a useful diagnostic for studying learning dynamics, while showing that OrthoGrad is regime-dependent rather than universally regularizing.
[LG-99] Operator-Aware Mixed-Precision Tolerance Calibration for Tensor Kernels
链接: https://arxiv.org/abs/2607.16228
作者: Dipankar Sarkar
类目: Machine Learning (cs.LG); Mathematical Software (cs.MS)
*备注: 8 pages, 1 figure, LNCS format. Companion paper: arXiv:2606.20128 (P1). Additional companions (P3, P4) to follow on arXiv this week; IDs will be added in a v2 replace
Abstract:Most tensor-kernel correctness tests go through a fixed-shape all close-style check with hand-picked absolute and relative tolerances. The thresholds are copied across the corpus and rarely revisited. We mine the element-wise error distribution of every test case from accumulated cloud GPU runs across the 26-entry gpuemu corpus and 2 dtypes (8,076 result rows). We then ask one empirical question: what absolute tolerance would the kernel itself, observed under its correct implementation, justify? The answer is much tighter than the current hand-picked atol. The largest tightening is attention_triton fp16 at 2,184\times . Restricted to the seven LLM-style buggy variants for which the corpus ships a paired correct counterpart, calibrated per-(op, dtype) tolerances raise bug-detection recall from 73.2% (1,805 of 2,467) to 82.4% (2,034 of 2,467), an absolute gain of 9.3 percentage points (+229 new detections). The control false-positive count rises from 0 to 20 out of 1,882 correct-control cases (+1.1 percentage points). Comments: 8 pages, 1 figure, LNCS format. Companion paper: arXiv:2606.20128 (P1). Additional companions (P3, P4) to follow on arXiv this week; IDs will be added in a v2 replace Subjects: Machine Learning (cs.LG); Mathematical Software (cs.MS) ACMclasses: G.4 Cite as: arXiv:2607.16228 [cs.LG] (or arXiv:2607.16228v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2607.16228 Focus to learn more arXiv-issued DOI via DataCite
[LG-100] LLM Unlearning for Cyber Defense: A Survey on Methods Challenges and Emerging Threats
链接: https://arxiv.org/abs/2607.16227
作者: Ruppikha Sree Shankar,Abhishek Bhardwaj,Arnav Doshi,Anusri Nagarajan,Troy Paulus Asia,Saptarshi Sengupta
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR)
*备注: 42 pages, 5 tables
Abstract:LLMs are increasingly deployed in security-critical systems across healthcare, finance, education, and decision support, yet their inability to forget creates serious cybersecurity, privacy, and safety risks. Sensitive personal information, copyrighted material, hazardous domain knowledge, and memorized training data remain encoded across billions of parameters long after deployment, leaving models vulnerable to extraction, jailbreak attacks, membership inference, and regulatory non-compliance. Real-world incidents, from chatbots regenerating private information to fabricated legal citations producing direct legal and financial cost, place the problem at the center of the emerging-threats landscape rather than the realm of speculation. Because retraining billion-parameter models on revised corpora is computationally infeasible, and because knowledge within an LLM is distributed and entangled across parameters rather than localized to identifiable units, LLM unlearning has emerged as the principal cyber defense response, aiming to remove or suppress targeted knowledge from a trained model without retraining and without eroding what the model should still know. A central question, however, remains unresolved. Do current methods genuinely remove knowledge, or do they only stop the model from expressing it under ordinary prompting conditions? This survey examines LLM unlearning through the lens of security, robustness, and verifiable forgetting, with primary focus on gradient-based methods, which have come to dominate the field due to their compatibility with existing training pipelines and their scalability to billion-parameter models.
[LG-101] Reinforcement Learning-Guided NSGA-II Enhanced with Gray Relational Coefficient for Multi-Objective Optimization: Application to NASDAQ Portfolio Optimization
链接: https://arxiv.org/abs/2607.16194
作者: Zhiyuan Wang,Qinxu Ding,Ding Ding,Siying Zhu,Jing Ren,Yue Wang,Chong Hui Tan
类目: Machine Learning (cs.LG); General Economics (econ.GN)
*备注:
Abstract:In modern financial markets, decision-makers increasingly rely on quantitative methods to navigate complex trade-offs among multiple, often conflicting objectives. This paper addresses constrained multi-objective optimization (MOO) with an application to portfolio optimization for minimizing risk and maximizing return. To address existing gaps, we propose a novel reinforcement learning (RL)-guided non-dominated sorting genetic algorithm II (NSGA-II) enhanced with gray relational coefficients (GRC), termed RL-NSGA-II-GRC, which combines an RL agent controller and GRC-based selection to improve convergence and diversity of Pareto fronts. The agent adapts evolutionary parameters online using metrics of hypervolume, feasibility, and diversity, while the GRC tournament operator ranks parents via a unified score considering dominance rank, crowding distance, and proximity to ideal reference. We evaluate the framework on the Kursawe and CONSTR benchmarks and a NASDAQ portfolio application. On the benchmarks, RL-NSGA-II-GRC achieves convergence improvements of about 5.8% and 4.4% over NSGA-II, while preserving well-distributed non-dominated solutions. In the portfolio application, it produces a smooth, densely populated efficient frontier supporting identification of the maximum Sharpe ratio portfolio (annualized Sharpe =1.92) and utility-optimal portfolios for different risk-aversion levels. The main contributions are three-fold: 1) we propose an RL-NSGA-II-GRC method integrating an RL agent into the evolutionary framework to adaptively control parameters via generational feedback; 2) we design a GRC-enhanced binary tournament operator providing a comprehensive indicator to guide the search toward the Pareto front; 3) we demonstrate, on benchmark MOO and a NASDAQ case study, that the method delivers improved convergence and well-populated frontiers supporting actionable insights.
[LG-102] Vector Search As Nearest Neighbor Matching: RAG -based Policy Learning in Causal Inference
链接: https://arxiv.org/abs/2607.18225
作者: Masahiro Kato,Taka Kato
类目: Econometrics (econ.EM); Machine Learning (cs.LG); Statistics Theory (math.ST); Methodology (stat.ME); Machine Learning (stat.ML)
*备注:
Abstract:We propose one-step and two-step methods for policy learning with retrieval-augmented generation (RAG). We formulate RAG-based action selection under the potential outcome framework. In the two-step method, vector search retrieves action-specific neighboring evidence in an embedding space, the generator estimates conditional expected outcomes or their contrasts, and a plug-in rule selects an action. This formulation connects action-specific vector search with nearest-neighbor matching in causal inference. We decompose the regret of the two-step method into candidate-generation regret and within-candidate choice regret, and we bound the latter using prediction-error guarantees for nearest-neighbor estimators and transformers. We evaluate the one-step method directly as a policy because its intermediate computation is unobserved.
[LG-103] Unveiling Invariant and Transferable Latent Factors Across Heterogeneous Environments via ATLAS
链接: https://arxiv.org/abs/2607.18209
作者: Yihong Gu,Katherine Liao,Tianxi Cai
类目: atistics Theory (math.ST); Machine Learning (cs.LG); Methodology (stat.ME); Machine Learning (stat.ML)
*备注: 47 pages, 3 figures
Abstract:This paper considers a multi-environment factor model in which high-dimensional covariates are collected from heterogeneous environments, with auxiliary labels available in a subset of these environments. The joint distribution of the covariates may vary across environments, whereas the latent structure is decomposed into invariant factors with shared loadings and heterogeneous factors with environment-specific loadings. Such a model is motivated by transfer learning and latent factor regression, where one seeks stable low-dimensional representations for both interpretation and robust out-of-sample prediction of the response Y . Leveraging the invariance principle, we show that the invariant and heterogeneous factors are disentangled under a minimal structural condition. Based on this, we propose ATLAS, an Auxiliary-label and invariance-guided Transfer via Latent Alignment across heterogeneous environmentS. ATLAS is a unified procedure that leverages the invariance principle to separate aligned invariant and unaligned heterogeneous factors, and further exploits supervision from auxiliary labels to extract prediction-invariant and transferable factors from those unaligned heterogeneous factors. ATLAS yields near-oracle performance for downstream latent factor regression, enables transferable prediction in new environments through the full latent signal when auxiliary labels are available, and reduces to robust invariant-factor-only prediction otherwise. We establish sharp non-asymptotic error bounds for recovering invariant and heterogeneous factors, identifying all the response-invariant factors, and estimating the invariant signal in Y .
[LG-104] COVAriance-Induced Fairness Gap Penalty for Subgroup-Fair Clustering
链接: https://arxiv.org/abs/2607.18119
作者: Kyungseon Lee,Hankyo Jeong,Kunwoong Kim,Kwanho Lee,Yongdai Kim
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Fair clustering aims to make cluster assignments independent of sensitive attributes, but this goal becomes challenging when multiple sensitive attributes jointly define many subgroups. In such settings, directly extending existing fair clustering algorithms is computationally expensive or numerically unstable, especially when the number of subgroups grows exponentially and some subgroups contain only a few instances. To address these challenges, we define a subgroup-fairness gap for clustering and derive a covariance-based surrogate that exactly matches this gap. We then introduce a continuous relaxation of the surrogate, enabling efficient gradient-based optimization and yielding our proposed algorithm, COVA-FC. We also show that subgroup fairness alone does not imply marginal fairness, and extend our framework to capture a subgroup-marginal-fairness gap. Experiments on benchmark datasets show that COVA-FC achieves competitive cost-fairness trade-offs and improves computational efficiency over existing baselines in both subgroup and higher-order marginal settings.
[LG-105] Beyond the Edge of Chaos: Stability-Expressivity Transfer in Reservoir Forecasting
链接: https://arxiv.org/abs/2607.17909
作者: Yao Du,Xingang Wang
类目: Chaotic Dynamics (nlin.CD); Machine Learning (cs.LG); Adaptation and Self-Organizing Systems (nlin.AO)
*备注: 5 pages, 3 figures
Abstract:The edge-of-chaos heuristic has long served as a guiding principle for designing reservoir computers, yet its relevance to machine performance remains elusive. Here, taking the spectral radius of the reservoir network as the control parameter, we show that the radius yielding the best forecasting performance does not coincide with the Lyapunov edge of the isolated, teacher-forced, or closed-loop generative reservoir. By analyzing the collective dynamics of the teacher-forced reservoir, we find that the target dynamics are represented mainly by stable Lyapunov modes whose finite-time stability is strongly modulated by the input. This finding motivates a stability-expressivity transfer index, which balances the stability of these modes against their expressivity in representing the target. Across chaotic and quasiperiodic targets, and for both asymmetric and symmetric reservoirs, this index accurately identifies the optimal spectral radius for autonomous forecasting.
[LG-106] Entanglement geometry separates circuit cutting classical hardness and trainability
链接: https://arxiv.org/abs/2607.17872
作者: Maria Gragera Garces,Sabina Drăgoi,Lirandë Pira
类目: Quantum Physics (quant-ph); Distributed, Parallel, and Cluster Computing (cs.DC); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
*备注: 4 pages, 2 figures
Abstract:Circuit cutting promises to scale quantum computations beyond current hardware, but variational quantum advantage also requires low cutting overhead, classical hardness, and trainability. We show that these properties are strongly constrained by entanglement geometry. Matrix product state (MPS) and tree tensor network (TTN) circuits with constant seam bond dimension can be cut with (O(1/\varepsilon^2)) sampling overhead, but remain efficiently classically simulable, ruling out asymptotic quantum advantage within these families. By independently controlling seam and intra-block entanglement, we construct a two-block circuit family that remains cheaply cuttable while requiring a super-polynomial global MPS bond dimension, as supported numerically up to (n=100). However, MPS hardness and trainability require incompatible depth regimes, (d=\omega(\log n)) and (d=O(\log n)), respectively. Using magic rather than entanglement as the hardness resource avoids this conflict: shallow Clifford+(T) circuits remain cuttable and trainable while their stabiliser-simulation cost grows exponentially with the (T)-count.
[LG-107] An Adjoint-Sensitivity Framework for Lost-in-the-Middle Phenomena in Causal Residual Transformers
链接: https://arxiv.org/abs/2607.17696
作者: Cheng Huan,Hongwei Yuan
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 41 pages, 4 figures
Abstract:We develop an adjoint-sensitivity framework for positional influence in causal residual Transformers and separate unconditional analytic results from conditional boundary-shape conclusions. The principal unconditional theorem is the residual-to-depth-flow estimate for layer controls converging in L^1 , complemented by a finite-token-to-Volterra attention estimate that explicitly controls the first cells near the causal endpoint. We define a normalized adjoint-energy influence density and derive its exact evolution along full-batch gradient flow. The adjoint admits an exact generator-term decomposition into residual transmission, nonlocal Volterra, and local channels, including all covariance cross terms. Causal masking can amplify early-position sensitivity and residual identity paths can transmit a right-localized terminal bias, but neither mechanism alone forces a U-shaped profile. We therefore state boundary advantages under independently checkable energy, correlation, and local-channel bounds; these conditions are sufficient rather than necessary. Finite-token influence balancing, positional reweighting, and task-aligned observability are presented as diagnostics or regularizers with explicit differentiation requirements, computational costs, and limitations. Controlled simulations illustrate that each intervention controls its designated surrogate, while observability balance or outer-loop reweighting need not monotonically reduce the influence-based Lost-in-the-Middle diagnostic.
[LG-108] Equality Equity and Causality in Fairness Research: A Commentary on Cheng (2026)
链接: https://arxiv.org/abs/2607.17679
作者: Youmi Suk
类目: Methodology (stat.ME); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: Invited commentary on Cheng (2026), Psychometrika, doi: https://doi.org/10.1017/psy.2026.10110%3B submitted to Psychometrika
Abstract:This is an invited commentary on the Psychometrika focus article “Fairness Issues and Evaluation in Psychometrics and AI/ML: What Can We Learn from Each Field?” by Ying Cheng (2026, doi:https://doi.org/10.1017/psy.2026.10110). Cheng offers a systematic comparison between long-standing test fairness and modern algorithmic fairness. Her mapping of the entire testing workflow onto the AI/ML fairness paradigm, rather than only the final selection stage, is a crucial contribution to interdisciplinary fairness research. This commentary extends her discussion by examining two conceptual issues: the distinction between equality and equity, and the role of causality in fairness research. Together, the focus article and this commentary point to directions for future fairness research across the psychometrics and AI/ML communities.
[LG-109] An efficient adaptive dimension selection algorithm for multidimensional probit graded response models
链接: https://arxiv.org/abs/2607.17654
作者: Yu Zhou,Yincai Tang,Bin Lv,Meng Gao
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
*备注: 36 pages, 3 figures, 5 tables
Abstract:Multidimensional graded response models (MGRMs) are widely used for analyzing ordinal questionnaire data in psychological and educational assessments. A central challenge in applying these models is determining the number of latent dimensions. Conventional approaches usually fit multiple fixed-dimensional models and select among them using post-hoc criteria such as AIC, BIC, or cross-validation, which can be computationally demanding and ignore uncertainty in dimensionality during estimation. We develop an adaptive Bayesian dimension selection framework for probit MGRMs. Building on the cumulative shrinkage process, we assign a cumulative ordered spike-and-slab (COSS) prior to the column-specific variances of the item loading matrix. This prior induces increasing shrinkage across latent dimensions, allowing redundant dimensions to be shrunk toward zero while preserving flexibility for active dimensions. Albert–Chib latent response augmentation is used to handle the ordinal probit likelihood, yielding conditionally Gaussian updates for item loadings and latent traits. These updates are combined with Gibbs updates for threshold and shrinkage parameters in an efficient adaptive sampler. Simulation studies evaluate the proposed method in terms of dimension recovery, parameter estimation accuracy, and computational efficiency, with comparisons to conventional fixed-dimensional estimation and model selection procedures. The results show that the proposed approach accurately recovers the latent structure while avoiding repeated model fitting over multiple candidate dimensions. We further illustrate the method using real psychological assessment data, demonstrating its practical utility for uncovering interpretable latent structures in ordinal item responses. Comments: 36 pages, 3 figures, 5 tables Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME) Cite as: arXiv:2607.17654 [stat.ML] (or arXiv:2607.17654v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2607.17654 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Yu Zhou [view email] [v1] Mon, 20 Jul 2026 08:06:47 UTC (343 KB)
[LG-110] Efficient Sequential Evaluation of Large Language Models
链接: https://arxiv.org/abs/2607.17409
作者: Chia-Yu Hsu,Shubhanshu Shekhar
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
*备注: This is a preliminary version; feedback is welcome
Abstract:We study the problem of sequentially evaluating a new large language model (LLM) on a fixed question set using historical performance data from prior LLMs. Our goal is to construct a confidence sequence (CS) for the model’s capability on this question set and to design active querying rules that shrink the CS width as quickly as possible. For CS construction, we invert a family of test supermartingales and focus on two representative approaches: a reverse information projection (RIPr)-based approach and a testing-by-betting-based approach. We first study these approaches under an oracle setting, and demonstrate the oracle optimality of the RIPr-based construction. We then propose a growth-oriented querying rule that aims to maximize the worst-case one-step expected log-increment over the endpoints of the current CS. In practice, we build these test supermartingales and the querying rule on predictions of question-level correctness learned from historical data. We then analyze the shrinkage behavior of the resulting CSs and identify two key factors that slow the shrinkage rate of CSs: accumulated prediction mismatch and the spikiness of the querying distribution. Finally, motivated by this analysis, we propose several mixture querying rules that combine growth-oriented querying, prediction refinement, and uniform exploration, trying to mitigate the effects that slow the shrinkage rate. We provide experiments comparing different querying rules for the RIPr-based and testing-by-betting-based CSs across several synthetic testing datasets. Interestingly, we observe that the simplest querying rule, uniform sampling, can sometimes outperform more adaptive querying rules for both methods.
[LG-111] Kernel Regression with Tensor Trains and Hadamard Overparameterization
链接: https://arxiv.org/abs/2607.17390
作者: Duc Thien Nguyen,Konstantinos Slavakis,Eleftherios Kofidis,Dimitris Pados
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注:
Abstract:Kernel regression with tensor trains and Hadamard overparameterization (KReTTaH) is introduced as a training-data-free, interpretable, and nonparametric framework for multi-way data imputation. The imputation problem is reformulated as regression in reproducing kernel Hilbert spaces (RKHS), where the tensor regression coefficients are explicitly constrained to lie on fixed-rank tensor-train (TT) manifolds and structured via Hadamard overparameterization to promote sparsity and high representational efficiency. Rather than relying on costly cross-validation, KReTTaH jointly optimizes the TT coefficient tensors and the kernel covariance matrices within a Riemannian product-manifold framework – the former on fixed-rank TT manifolds, the latter on the manifold of positive-definite matrices – thereby enabling automated kernel-hyperparameter selection. Numerical tests on two challenging applications – imputation of high-dimensional functional magnetic resonance imaging (fMRI) data and recovery of missing edge flows in dynamic graphs – demonstrate that KReTTaH consistently outperforms state-of-the-art tensor-, Bayesian-, and neural-network-based baselines in terms of modeling accuracy.
[LG-112] Interpreting Quantum Learning Models via Stochastic Processes
链接: https://arxiv.org/abs/2607.17327
作者: Johannes Fankhauser,Lukas J. Fiderer,Hans J. Briegel
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注: 13 pages, 2 tables
Abstract:Quantum machine learning models define probabilistic input–output maps through coherent quantum evolution and measurement. While such models can exhibit computational advantages, their internal functioning and decision making generally resists interpretation in terms of stochastic trajectories through intermediate configurations. In contrast to classical (Markovian) stochastic processes, quantum dynamics generically violates the Chapman–Kolmogorov divisibility condition, preventing a decomposition into probabilistically meaningful intermediate transitions. We develop a probabilistic framework for representing quantum learning models as stochastic processes over configuration spaces where the dynamics are modeled as linear maps on probability distributions. Starting from a fixed POVM, arbitrary quantum channels induce transition kernels on the associated probability representation. For informationally complete POVMs, and in particular SIC-POVMs, these kernels are Markovian but generally quasi-stochastic, with non-classicality appearing as negativity. By contrast, projective spaces admit positive stochastic kernels but generally require non-Markovian dynamics due to the failure of Chapman–Kolmogorov divisibility. This yields a trade-off between negativity and dependence on past configurations, i.e. quantum dynamics can be represented either by Markovian quasi-stochastic maps or by positive stochastic processes with higher Markov order. We discuss how such representations of quantum dynamics can be interpreted as stochastic walks through a memory space in the spirit of Projective Simulation, a model of learning and agency in which decisions arise from random walks over an episodic memory network. We further outline how finite-order stochastic kernels can approximate such quantum deliberation processes and show in what regimes the classical machine learning model is recovered.
[LG-113] Non-Asymptotic Best Policy Identification Guarantees in Online Reinforcement Learning
链接: https://arxiv.org/abs/2607.17201
作者: Joseph Lazzaro,Alessio Russo,Aldo Pacchiano
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 64 pages, 2 figures
Abstract:In this work we study the Best Policy Identification (BPI) problem in online, tabular Reinforcement Learning. This is an active sequential hypothesis testing problem in which the learner’s objective is to identify an optimal policy in a Markov Decision Process (MDP) with high confidence, while minimizing the expected sample complexity to do so. We consider an online setting with deterministic rewards, where the agent must strategically navigate through the MDP in order to effectively explore. Previous works in the literature have provided asymptotically optimal methods for BPI, such as the Navigate and Stop (NaS) algorithm and its variants, however existing analysis remains asymptotic. In this work, we fill that gap by providing the first non-asymptotic sample complexity guarantees for NaS, showing that its sample complexity depends not only on the characteristic time, but also on the connectivity of the underlying MDP, the curvature of the optimal characteristic time, and other instance-dependent quantities. We identify these additional attributes and make explicit their contributions to the overall sample complexity.
[LG-114] Robust Chance-Constrained Optimization using a Continuous Parameter Space Wasserstein-2 Ambiguity Set of Gaussian Mixtures
链接: https://arxiv.org/abs/2607.17018
作者: Shibshankar Dey,Sanjay Mehrotra
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注:
Abstract:We study distributionally robust linear chance-constrained problems in which uncertainty is modeled by a Gaussian mixture model (GMM). Finite-support distributionally robust (FDR) formulations, widely used in data-driven robust optimization, robustify over empirical mixture support points and therefore primarily stress-test the fitted nominal mixture. This can be insufficient when service reliability depends on structural misspecification of the nominal mixture-support parameters. To address this limitation, we describe the ambiguity set of distributions by developing a novel formulation of a Wasserstein-2 metric that uses the Bures-Wasserstein (BW) metric over probability measures with finite second moments. Unlike FDR, which generally sets finitely many empirical support points a priori, the proposed ambiguity set allows the worst-case distribution to endogenously determine both how many mixture components receive mass and where their means and covariances lie within a continuous support. For the resulting ambiguity set, under mild regularity conditions, we prove strong duality for the inner worst-case chance-constraint problem and derive its semi-infinite reformulation. We then develop an adaptive cutting-surface algorithm, which endogenously determines the locations of mixture components receiving mass, and the mean and covariances of the Gaussian distributions at these locations. The algorithm attains any prescribed optimality gap in finitely many iterations, while a block-alternating local search identifies new components. A case study using the electric-vehicle charging-station energy-allocation problem demonstrates the framework’s practical value in achieving any reliability targets. CDR also induces structural changes in energy allocations, unlike FDR, whose allocations remain close to the nominal solution.
[LG-115] wisted Schrödinger Bridge Matching
链接: https://arxiv.org/abs/2607.16987
作者: Maxence Noble,Marie Scheid,Yazid Janati,Eric Moulines,Alain Durmus
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: Preprint version
Abstract:Over the past few years, diffusion-based Schrödinger bridge models have been proposed to approximate optimal transport dynamics between two prescribed boundary distributions, with successful applications to generative modeling. More precisely, these methods aim to estimate a path measure whose initial and terminal marginals match the two boundary distributions, while minimizing the Kullback-Leibler divergence with respect to a reference Markov process. In this work, we consider the generalized Schrödinger bridge problem, in which the reference process is a twisted Brownian motion, that is, a Feynman-Kac transform of a Brownian motion induced by a time-dependent differentiable potential. Building on the Iterative Markovian Fitting (IMF) paradigm, and in particular on its special case Diffusion Schrödinger Bridge Matching (DSBM), which corresponds to the zero potential case, we introduce Twisted Schrödinger Bridge Matching (TSBM), a diffusion-based method designed to handle both continuous- and discrete-time potentials. Unlike previous approaches, TSBM provides a rigorous extension of the IMF scheme to the generalized Schrödinger bridge problem. This derivation leads to a new bridge-matching loss that depends explicitly on the gradient of the potential and recovers the DSBM objective when the potential vanishes, yielding improved performance. We further introduce trajectory-based variance-reduction techniques that substantially stabilize optimization and may be useful beyond the present setting. Finally, we empirically demonstrate the benefits of TSBM for trajectory inference across increasingly high-dimensional settings, including crowd navigation and single-cell data. Code available at this https URL.
[LG-116] Deep Adaptive Bayesian Screening
链接: https://arxiv.org/abs/2607.16927
作者: Jade Lejeune Herman,Arno Strouwen,Johan A.K. Suykens,Peter Goos
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:We introduce Deep Adaptive Bayesian Screening (DABS), a method for performing adaptive factorial screening in high-dimensional discrete design spaces. DABS learns a policy network offline to sequentially select informative experiments, amortizing Bayesian Optimal Experimental Design. It handles binary designs, incorporates sparsity and interactions via a spike-and-slab prior with strong heredity. The model is trained using a contrastive lower bound on information about factor activity with nuisance effect sizes and noise variance analytically integrated out. Unlike prior amortized Bayesian design approaches, DABS also integrates Gibbs posterior inference at deployment, yielding posterior probabilities of factor activity and credible intervals on effect sizes. We demonstrate DABS on screening problems calibrated to real-world benchmarks and show it achieves superior accuracy and scalability over classical and Bayesian baselines under tight experimental budgets.
[LG-117] Hierarchical Wireless Foundation Model for Multi-Task Optimization
链接: https://arxiv.org/abs/2607.16877
作者: Yangjing Wang,Ouya Wang,Shenglong Zhou,Geoffrey Ye Li
类目: ignal Processing (eess.SP); Machine Learning (cs.LG)
*备注: 15 pages, 10 figures, 3 tables
Abstract:The increasing complexity of next-generation wireless networks has driven the integration of artificial intelligence (AI) into wireless communications. However, most existing studies focus on developing task-specific deep learning techniques for single scenarios, which limits their ability to generalize across diverse tasks, channel conditions, and system configurations. To address this generalization bottleneck, we propose a hierarchical wireless foundation model (WFM) for multi-task optimization. The proposed WFM couples an upstream foundation channel encoder (FCE) with a downstream foundation optimization decoder (FOD) via geometry-aware cross-attention. Specifically, the FCE extracts task-agnostic channel representations via self-supervised masked reconstruction while the FOD generates multi-task optimization decisions through differentiable output heads. Moreover, a hybrid supervised-to-unsupervised training strategy is employed to overcome the performance ceiling of purely supervised learning, and the modular architecture of the WFM enables efficient adaptation to unseen communication tasks with minimal parameter overhead. Simulation results show that the proposed WFM learns high-fidelity channel representations and achieves competitive multi-task optimization performance while substantially reducing optimization inference latency relative to numerical baselines. Furthermore, it exhibits robust generalization to unseen propagation environments, varying constraint parameters, and heterogeneous system configurations.
[LG-118] Identity-Paired Progressive Depth Training: When Trainability Persists Beyond Expressibility
链接: https://arxiv.org/abs/2607.16800
作者: Athanasios Hadjidimoulas,Tirthak Patel,Anastasios Kyrillidis
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: 48 pages
Abstract:Variational Quantum Algorithms (VQAs) are a leading paradigm for near-term quantum computing, yet their training suffers from sensitivity to circuit depth, initialization, and landscape pathologies such as barren plateaus. We study \emphprogressive depth training (PDT) – a layerwise curriculum that trains a shallow circuit before appending new layers – and identify a fundamental obstacle: fixed entangling gates (CNOTs) in hardware-efficient ansätze cause \emphinitialization shock, an energy spike when new layers are added. We propose \emphidentity-paired progressive depth training (IP-PDT), which appends forward/inverse block pairs – each consisting of a standard rotation + CNOT block followed by its reverse – that compose to the identity at initialization. Because the adjacent CNOT rings cancel, the effective circuit retains only \textita single entangling layer surrounded by \textitoverparameterized local rotations. We prove a simple \textitReachable Set Saturation Theorem: under this construction the variational manifold expands exactly once (when post-entangler rotations are first introduced) and then \emphsaturates; all subsequent depth increases provide pure overparameterization of single-qubit unitaries. Despite this saturation, progressive addition of rotation parameters can continue to improve optimization outcomes – a phenomenon we term \emphtrainability beyond expressibility. We formalize IP-PDT as a continuation method on nested manifolds, prove monotone energy guarantees under an acceptance rule, and connect energy error to ground-state fidelity through spectral-gap inequalities. A detailed resource analysis shows that IP-PDT achieves lower total gate cost than both baselines by eliminating most CNOT gates.
[LG-119] Demodulation of chaotic signals using convolutional neural network
链接: https://arxiv.org/abs/2607.16788
作者: Mykola Kozlenko,Emrullah Demiral,Anton Yudhana
类目: ignal Processing (eess.SP); Machine Learning (cs.LG)
*备注: 6 pages, 12 figures, 3 tables. This paper supersedes arXiv:2505.06243 , Published by International Archives of the Photogrammetry Remote Sensing and Spatial Information Sciences, available: this https URL
Abstract:Chaotic modulation is an effective communication technique that exploits deterministic chaos to produce pseudo-random signals. A widely adopted approach involves modulation of the chaotic bifurcation parameter. This paper introduces a deep learning-based demodulation method for keying of the bifurcation parameter. It describes the architecture of the convolutional neural network and evaluates performance metrics for signals generated using the chaotic logistic map. The study assesses the bit error rate for binary signals and reports a bit error rate of 0.0819 for a bifurcation parameter deviation of 1.34% under additive white Gaussian noise at a signal-to-noise ratio of -13 dB (corresponding to a normalized signal-to-noise ratio of +20 dB). The results demonstrate the capability to detect chaotic patterns even when the specific patterns were not included in the training dataset.
[LG-120] Dropout and Random Gradient Masking Are Asymptotically Equivalent in Large ResNets
链接: https://arxiv.org/abs/2607.16761
作者: Javier Maass,Lénaïc Chizat
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Probability (math.PR)
*备注:
Abstract:Dropout and Random Gradient Masking (RaM) are two training techniques used to improve performance in deep learning. Both techniques inject randomness into the training dynamics, but in significantly different ways: dropout applies random masks to the activations in the forward pass, whereas RaM leaves the forward pass unchanged and instead masks the gradients. In particular, the noise induced by RaM in the parameter updates is unbiased, so standard explanations for the effectiveness of dropout, such as the penalization effect or the prevention of co-adaptation between neurons, do not apply to RaM. In this work, we show that the difference between the two methods disappears for ResNets in the large depth and width asymptotics: in the complete feature learning regime, they both converge to the same large-scale limiting dynamics. This asymptotic equivalence holds for several variants of dropout and RaM, including layerwise dropout as used in stochastic-depth ResNets, albeit at slower quantitative rates. In fact, we also show that several of these variants collapse to the same limit asymptotically.
[LG-121] Semi-Supervised Conditional Generative Learning through Stochastic Interpolation and Sufficient Representations
链接: https://arxiv.org/abs/2607.16725
作者: Changyu Liu,Yuling Jiao,Jian Huang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 35 pages, 4 figures and 1 table
Abstract:Conditional generative modeling remains a challenging problem in semi-supervised settings where labeled data is scarce but unlabeled samples are abundant. To effectively leverage structural information embedded within the unlabeled dataset and compensate for sparse conditioning signals, we propose a semi-supervised framework combining conditional stochastic interpolation with low-dimensional latent representations. RepG decomposes generation into two stages: label-dependent latent sampling and high-dimensional reconstruction. This isolates the supervised learning of conditional dependencies to a low-dimensional space, requiring few labels while utilizing the abundant unlabeled data purely for reconstruction. Theoretically, we establish an error decomposition showing that the Kullback-Leibler divergence of RepG comprises stage-wise estimation errors and a structural bias quantified by conditional mutual information. For deep neural network estimators, we derive non-asymptotic convergence rates proving that RepG significantly improves sample complexity. By confining the supervised estimation burden to the low intrinsic dimension of the latent representation, RepG achieves a strictly faster convergence rate. Complemented by a minimax lower bound, our theoretical results demonstrate that this method effectively mitigates the curse of dimensionality inherent in direct ambient-space generative modeling.
[LG-122] Semi-Supervised Conditional Diffusion via Label Augmentation
链接: https://arxiv.org/abs/2607.16685
作者: Jin Su,Yuan Gao,Yong Zhou,Jian Huang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 34 pages, 7 figures
Abstract:Conditional diffusion models have become a powerful and flexible framework for learning complex conditional distributions from labeled data. In practice, however, acquiring high-quality labels is costly and time-consuming, leaving large volumes of unlabeled data unused. To address this, we introduce label-augmented conditional diffusion (LACD), a simple and effective approach that incorporates unlabeled examples by assigning them a designated trivial label and performing joint denoising score matching over the augmented dataset. We provide sufficient conditions guaranteeing population-level identifiability of the target conditional distribution under this scheme. Moreover, we establish rigorous statistical guarantees: when sufficiently many unlabeled samples are available, the sampling distribution produced by LACD converges strictly faster than the purely supervised estimator in total variation distance, and at least as fast in Wasserstein-1 distance. Extensive experiments on synthetic, image, and tabular benchmarks corroborate our theory and show substantial gains in sample efficiency and generative performance compared with the purely supervised estimator.
[LG-123] he Value of Depth in Message Passing on Sparse Graphs: A Kesten-Stigum Dichotomy
链接: https://arxiv.org/abs/2607.16676
作者: Aseem Raj Baranwal
类目: atistics Theory (math.ST); Machine Learning (cs.LG); Probability (math.PR); Machine Learning (stat.ML)
*备注: 27 pages, 4 figures, 1 table. Code reproducing all numerical results is included as ancillary files
Abstract:How deep does a graph neural network need to be on a sparse graph? We study its purest statistical form: node classification on the sparse contextual stochastic block model (CSBM) with average degree \Delta=O(1) , whose local weak limit is a broadcast-labelled Poisson Galton-Watson tree. Prior work derived a message-passing classifier h_\ell that aggregates from each vertex at distance k\le\ell the attenuated evidence 2\operatornameartanh(\gamma^k t(X_v)) , with \gamma the edge signal and t a bounded likelihood-ratio transform of the feature. We prove that the value of depth is governed by a single number, the Kesten-Stigum ratio \kappa=\gamma^2\Delta . Below the threshold ( \kappa1 ), the error sequence is Cauchy at a geometric rate, |\mathcalE(\ell)-\mathcalE(\ell’)|\le C\kappa^(\ell+1)/3 for all \ell’\ell , so all layers beyond depth O(\log(1/\epsilon)) change the error by less than \epsilon ; conversely, under mild regularity each sufficiently deep layer still flips the decision with probability at least c\kappa^\ell/2 , the empirically sharp exponent. Above the threshold ( \kappa1 ), depth is geometrically productive: \mathcalE(\ell) is driven to a branching-process floor of order at most 1/(\kappa-1) at any geometric rate \kappa^-s\ell , s1 (this bound has content only for \kappa17 ). No local classifier of any depth beats the universal floor e^-\Delta\Phi(-\zeta) set by isolated roots ( \zeta the feature signal-to-noise ratio), while the first layer provably helps by an explicit total-variation amount. Simulations with an exact belief-propagation baseline on the same trees show that the pairwise rule’s error curve is mildly non-monotone in \ell , so an optimal finite depth exists (an exact instance is certified in the appendix), while BP saturates strictly faster, at an effective per-layer ratio below \kappa that we identify.
[LG-124] Isotonic Conformal Prediction
链接: https://arxiv.org/abs/2607.16675
作者: Daniel Bensimon,Sean Xiang Yu,Eric D. Kolaczyk,Archer Y. Yang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Methodology (stat.ME)
*备注:
Abstract:A point prediction that is well calibrated on average can still be systematically biased conditional on its own value, undermining its use in downstream decision-making. We consider two objectives for reliable uncertainty quantification: self-calibration, requiring a point prediction to be unbiased conditional on its own value, and prediction-conditional validity, requiring a prediction interval to attain nominal coverage conditional on the prediction. Self-Calibrating Conformal Prediction (SC-CP) attains both objectives exactly in finite samples, but requires refitting its calibrator for every candidate outcome, which is computationally prohibitive for continuous outcomes. We propose Isotonic Conformal Prediction (ICP), a framework that decouples calibration from prediction-set construction by fitting a single isotonic recalibration map and constructing prediction intervals within strata of similar recalibrated predictions. Within this framework we develop two procedures. Split Isotonic Conformal Prediction (SICP) attains prediction-conditional validity in finite samples and self-calibration asymptotically, at the computational cost of split conformal prediction. Transductive Isotonic Conformal Prediction (TICP) attains both objectives exactly in finite samples through a per-test-point inner loop that avoids refitting the isotonic calibrator. On synthetic heteroscedastic regression problems and a real-world healthcare-utilization dataset, both procedures match the coverage of SC-CP at substantially lower computational cost.
[LG-125] De-floored Principal Component Regression: When Rank Selection Alone Is Insufficient for Prediction
链接: https://arxiv.org/abs/2607.16638
作者: Peng Zhao
类目: atistics Theory (math.ST); Machine Learning (cs.LG); Methodology (stat.ME)
*备注:
Abstract:Principal component regression (PCR) regularizes high-dimensional prediction by choosing a spectral cutoff, but rank selection cannot correct systematic inflation of the retained empirical eigenvalues. We study clean Gaussian random designs in which the aggregate covariance tail creates a nearly scalar sample-space floor comparable to the predictive head scale. De-floored principal component regression (dPCR) retains the cutoff and subtracts an estimated floor from the retained denominators. We prove an ordinary-PCR prediction-risk lower bound uniform over all ranks and a high-probability dPCR upper bound. When the floor is sharp and inexpensive to remove in population prediction risk, the conditional risk of dPCR is asymptotically negligible relative to that of the best ordinary PCR rank. An exact risk decomposition explains the separation: denominator inflation is governed by first spectral mass, whereas the clean prediction cost of correction is governed by squared spectral mass. A same-sample trimmed-mean floor estimate attains the oracle dPCR upper-bound rate at a prespecified rank, and the separation persists under approximate predictive alignment when the tail prediction-energy fraction vanishes. Separate pointwise fixed-aspect formulas show that the risk-optimal positive scalar correction improves rank- 1 PCR, whereas mean-floor subtraction is generally not optimal for a broad Marchenko–Pastur bulk.
[LG-126] Backpropagation-Free Trunk Training via the Split Forward Gradients
链接: https://arxiv.org/abs/2607.16612
作者: Tian Qin,Wei-Min Huang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:Backpropagation makes training deep networks memory intensive because it must store intermediate activations. Forward-mode methods avoid this cost, but their gradient estimates become increasingly noisy as the number of trained parameters grows. We introduce Split Forward Gradient (Split-FG), which splits a network at an intermediate representation: it computes the output head gradient exactly and estimates only the trunk gradient with a Jacobian–vector product. This reduces estimator variance and requires no backward pass through the trunk, while retaining an Adam-style convergence guarantee. Our experiments reveal an important practical failure mode. On WikiText-103, naive forward-gradient training of the trunk performs worse than leaving a randomly initialized trunk frozen, likely because Adam updates every noisy, under-determined trunk coordinate too aggressively. Simply using a much smaller learning rate for the trunk reverses this result: a 16 M-parameter GPT-2-style model reaches validation perplexity 387 , compared with 668 for the frozen-trunk control and 2,885 for a matched pure forward-gradient baseline (backpropagation reaches 150 ). Split-FG also produces the strongest backprop-free results on our tabular benchmarks and reaches 60.5% on CIFAR-10 and 35.2% on CIFAR-100 with a heavy-head design. It reduces peak memory by up to 35% relative to matched backpropagation, although the performance gap widens as the forward-mode trunk grows.
[LG-127] Harnessing disorder to decouple extension and shear in kirigami metamaterials
链接: https://arxiv.org/abs/2607.16583
作者: Haomin Yu,Hanxun Jin,Mingxuan Bi,Mohammad Jafari,Feng Helen Long,Michael J Greenberg,Farid Alisafaei,Guy Genin
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
*备注: 5 figures
Abstract:Kirigami turns stiff sheets into compliant, shape-morphing structures, but its reliance on periodic cut patterns comes at a cost: correlated panel rotations couple extension to shear, so stretching one axis drives a parasitic shear that cannot be suppressed, and also confine anisotropic stiffness to a narrow, discrete set of responses that cannot be tuned independently. Biological tissues overcome an analogous constraint through controlled disorder, such as graded fiber orientations in skin and hierarchical anisotropy in myocardium, achieving direction-dependent mechanics unavailable to regular architectures. Here, we show that engineered disorder is a design degree of freedom for kirigami, with stochastic kirigami accessing a continuous and far broader region of mechanical response than periodic patterns. This includes programmable anisotropy with near-complete elimination of extension-shear coupling. Because disordered patterns lack a simple parameterization, we navigate this design space with a geometry-aware graph neural network (GNN) that maps cut topology to the full nonlinear, bidirectional stress-strain response, coupled to a genetic algorithm that inverse-designs patterns reproducing target responses along two perpendicular axes. The GNN trains an order of magnitude faster and more accurately than image-based models. Fabricated elastomer samples reproduce the predicted nonlinear, anisotropic responses, closing the loop from design to physical component. By turning disorder into a variable to control directional stiffness, this work develops architected materials that stretch without parasitic shear, from soft actuators to tissue-interfacing devices matched to the anisotropy of living tissue.
[LG-128] Mapping Order in Semicrystalline Polymers using Machine Learning of Nanobeam Electron Diffraction
链接: https://arxiv.org/abs/2607.16570
作者: Nicholas Marchese,Arthur R. C. McCray,Yael Tsarfati,Karen Bustillo,Adam Marks,Alberto Salleo,Colin Ophus
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
*备注: Submitted to ACS Macromolecules. Main text is 13 pages and 5 figures, 23 pages and 15 figures with supporting information
Abstract:Organic mixed ionic electronic conductors (OMIECs) are a promising class of polymer materials for applications spanning neuromorphic computation to energy efficient electronics and bioelectronics. Despite being highly tunable, the relationship between structural features and key performance properties such as charge carrier mobility is poorly understood. Scanning nanodiffraction in the transmission electron microscope (TEM) is a powerful probe for elucidating this structure-property relationship, but produces large, noisy datasets that are difficult to interpret because polymer reflections exhibit several distinct morphologies. To address the complexity, we trained a machine learning (ML) model to detect these polymer diffraction peaks and their intensities from synthetic data. Compared to correlative peak detection algorithms, the conventional method for analyzing nanobeam 4D scanning transmission electron microscopy (4DSTEM) data, we show that the ML model is significantly faster and outperforms correlative algorithms in almost all cases, opening up the possibility of near-live visualization of 4DSTEM experiments.
[LG-129] MTSSL: Meta-Thresholding Semi-Supervised Learning
链接: https://arxiv.org/abs/2607.16363
作者: Shuyang Liu,Ziang Zeng,Ruiqiu Zheng,Jiazheng Wang,Zechen Liu,Wenxi Li,Zhou Yu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:
Abstract:A large body of Semi-supervised Learning~(SSL) algorithms encounter the threshold \tau to select pseudo-labels. The value of \tau across different SSL algorithms can vary depending on the learning perspective, yet they may achieve similar performance. It motivates us to establish a unified theoretical framework to explain the role of \tau in SSL. We statistically explained that the unsupervised loss is affected independently by correct and incorrect pseudo-labels, while \tau adjusts their numbers to balance the corresponding error term. This inherent trade-off indicates that SSL can reach the same loss with varying \tau , precise optimal values of \tau during training may be unnecessary. With this, we treat \tau as an updatable parameter and optimize it via differentiation; the new policy is named \textbfMeta-Thresholding Semi-Supervised Learning (MTSSL). Extensive experiments demonstrate the superior performance of MTSSL. We observe that the accuracy curves of SSL algorithms can overlap completely even when the values of \tau differ significantly, which supports our theoretical framework and indicates that the selection of \tau can be relaxed in the future design of SSL algorithms.
[LG-130] Lipschitz Continuity in Deep Learning: A Systematic Review of Theoretical Foundations Estimation Methods Regularization Approaches and Certifiable Robustness
链接: https://arxiv.org/abs/2607.16329
作者: Róisín Luo,James McDermott,Colm O’Riordan
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: Accepted into Published in Transactions on Machine Learning Research (TMLR)
Abstract:Lipschitz continuity is a fundamental property of neural networks that characterizes their sensitivity to input perturbations. It plays a pivotal role in deep learning, governing \textbfrobustness, \textbfgeneralization and \textbfoptimization dynamics. Despite its importance, research on Lipschitz continuity is scattered across various domains, lacking a unified perspective. This paper addresses this gap by providing a systematic review of Lipschitz continuity in deep learning. We explore its \textbftheoretical foundations, \textbfestimation methods, \textbfregularization approaches, and \textbfcertifiable robustness. By reviewing existing research through the lens of Lipschitz continuity, this survey serves as a comprehensive reference for researchers and practitioners seeking a deeper understanding of Lipschitz continuity and its implications in deep learning.
[LG-131] A Novel Hybrid Quantum Reservoir Computing (nHQRC) for Phase Transition Detection in Non-Equilibrium Dynamical Systems
链接: https://arxiv.org/abs/2607.16281
作者: Manoj B. Bhatkar,Prashant M. Yawalkar
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG); Computational Finance (q-fin.CP); Statistical Finance (q-fin.ST)
*备注: 9 pages, 7 figures, 1 table
Abstract:The analysis of highly non-linear stochastic data within non-equilibrium dynamical systems requires computational frameworks capable of detecting latent phase transitions before systemic structural breakdowns occur. Traditional Variational Quantum Algorithms (VQAs) are frequently bottlenecked by vanishing gradients, the barren plateau problem, and prohibitive training overheads. In this paper, we propose a novel Hybrid Quantum Reservoir Computing (nHQRC) framework, which bypasses these limitations by employing a frozen, disordered Transverse-Field Ising Model (TFIM) to project time-dependent stochastic driving forces into an exponentially large Hilbert space. To resolve the physical phase multi-wrapping vulnerabilities present in baseline quantum reservoir models, we introduce a lookahead-free, pre-amplification manifold scaling technique. Multi-qubit configurations are genetically optimized to the “edge of chaos,” while quantum state tracking is performed by extracting von Neumann entropy ( S ) and exact mixed-state Quantum Fisher Information (QFI) to act as leading entanglement witnesses. Utilizing these quantum triggers as boundary constraints, trajectory predictions are constructed via a generative Stochastic Schrödinger Bridge (SSB) readout. By subjecting the quantum reservoir to an 8-dimensional non-stationary stochastic driving field, the framework significantly improves systemic drift-to-diffusion efficiency ( \eta ) and actively arrests maximum trajectory decay (MTD) by over 13% compared to standard classical benchmarks. This establishes a robust, \mathcalO(1) temporal overhead blueprint for near-term quantum regime detection and macroscopic subsystem stabilization.
[LG-132] FinBench: Time-Gated Calibration and Uncertainty Benchmarking for Agent ic Financial Forecasting
链接: https://arxiv.org/abs/2607.16229
作者: Rishab Ghosh,Vinay Devarakonda
类目: Applications (stat.AP); Machine Learning (cs.LG); Computational Finance (q-fin.CP); Portfolio Management (q-fin.PM)
*备注: 9 pages, 3 figures
Abstract:Large language models (LLMs) are increasingly used as components of agentic systems that observe, plan, and act. In finance, even “assistive” systems become decision-relevant once their outputs are used to size trades or allocate risk. A key failure mode is the confidence–competence gap: a model that is only slightly better than chance but consistently overconfident will, under typical bet-sizing rules, generate negative long-run growth. Existing benchmarks emphasize semantic understanding or point accuracy, but do not directly test probabilistic calibration under the temporal constraints and non-stationarity that define real markets. We introduce FinBench, a benchmark designed to evaluate calibration and uncertainty quality for financial forecasting in a setting that is (i) strictly time-gated to avoid look-ahead bias and (ii) evaluated with strictly proper scoring rules that penalize hallucinated confidence. FinBench tasks require models to output (a) a probability of positive return and (b) an 80% prediction interval for realized log return; evaluation uses the Brier score and the Winkler interval score, along with skill scores against hard baselines. This paper describes the benchmark specification and reports a small pilot run (one trading day; three liquid tickers; 33 forecasts) as a sanity check of the pipeline. The pilot illustrates how calibration-sensitive metrics distinguish between “confident but fragile” behavior and uncertainty-aware forecasting. Comments: 9 pages, 3 figures Subjects: Applications (stat.AP); Machine Learning (cs.LG); Computational Finance (q-fin.CP); Portfolio Management (q-fin.PM) Cite as: arXiv:2607.16229 [stat.AP] (or arXiv:2607.16229v1 [stat.AP] for this version) https://doi.org/10.48550/arXiv.2607.16229 Focus to learn more arXiv-issued DOI via DataCite
[LG-133] Overcoming the BCI Calibration Bottleneck: A Clinically-Grounded Architecture using Riemannian Alignment and Stochastic Weight Averag ing
链接: https://arxiv.org/abs/2607.16225
作者: Immanuvel Prathap Sagayaraju
类目: ignal Processing (eess.SP); Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
*备注: 10 pages, 8 figures
Abstract:Brain-Computer Interfaces (BCIs) face a severe calibration bottleneck due to cross-subject spatial covariance shifts and physiological artifacts. To enable zero-calibration BCI, a deep learning pipeline was engineered combining Per-Session Independent Component Analysis, Riemannian Euclidean Alignment, and EEGNet stabilized by Stochastic Weight Averaging (SWA). Evaluated on the strict MOABB BNCI2014-001 benchmark, the proposed architecture successfully isolates true sensorimotor rhythms. For the primary case study (Subject 1), a clinically robust SWA stable accuracy of 90.97% (AUC: 0.976, Cohen’s \kappa : 0.819) was achieved. Furthermore, expanded 9-fold Leave-One-Subject-Out (LOSO) cross-validation yielded a globally stable mean accuracy of 74.31%, proving hardware-agnostic zero-shot efficacy for binary motor imagery.
附件下载
