本篇博文主要内容为 2026-11-20 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 (2026-11-20)

今日共更新893篇论文,其中:

  • 自然语言处理共96篇(Computation and Language (cs.CL))
  • 人工智能共252篇(Artificial Intelligence (cs.AI))
  • 计算机视觉共213篇(Computer Vision and Pattern Recognition (cs.CV))
  • 机器学习共213篇(Machine Learning (cs.LG))
  • 多智能体系统共14篇(Multiagent Systems (cs.MA))
  • 信息检索共21篇(Information Retrieval (cs.IR))
  • 人机交互共23篇(Human-Computer Interaction (cs.HC))

多智能体系统

[MA-0] Replication of Agent ic Security Events: Audit and Investigation (ROGUE-AI)

【速读】:该论文旨在解决当前生成式AI代理(Generative AI Agents)在实际应用中因安全缺陷导致的故障事件难以被有效复现与根因分析的问题。现有媒体报道和事故目录仅记录“发生了什么”,却缺乏对事件是否真实发生、其根本原因是否被正确识别的验证,且无法通过可重复实验检验潜在漏洞。针对这一缺口,论文提出ROGUE-AI(Replication of Agentic Security Events: Audit and Investigation)——一种多智能体复现框架,能够在线下重建每个已报告的安全事件,并构建一个仅缺少被归责原因的对照组进行对比实验。其核心解决方案在于:通过精确还原事故场景并设置控制变量,实现对代理行为的可重复性测试。实验表明,ROGUE-AI成功复现了89%的108次尝试事件;基于GPT、Claude、Gemini和DeepSeek等五种主流大模型的代理在存在特定缺陷时,错误发生率高达18%至33%,而缺失该缺陷的对照组错误率不超过1%。进一步分析发现,72起造成真实危害的事件中,62起无外部攻击者参与,问题根源在于代理正确执行任务但忽视了如生产数据库共享等常见风险。值得注意的是,在无攻击者场景下的复现造成的危害反而高于提示注入(prompt injection)类攻击,且现有防御机制针对注入指令设计,但在无注入的情况下失效。此外,每步工具输出后添加关于风险及其后果的警告,显著降低了所有模型的危害程度,而同等长度但未指明风险的说明则几乎无效。由此,ROGUE-AI构建的基于真实事件的可复现审计框架,为厂商在发布新版本前提供了一套可重复使用的安全验证机制。

链接: https://arxiv.org/abs/2611.26553
作者: Mengxiao Wang,Mahbub Alam,Kalyan Nakka,Nitesh Saxena
机构: Texas A&M University (德州农工大学)
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: 21 pages, 5 figures, 13 tables. Project page: this https URL

点击查看摘要

Abstract:Agents based on large language models (LLMs) can now delete files and send payments on behalf of their users. News media reports routinely describe failures of these agents (e.g., Replit’s coding agent deleting a production database), but a report often only says what went wrong and rarely (in)validates the incident or finds its root cause. Incident catalogs collect these reports, and benchmarks record whether agents fail, but neither replicates an incident nor tests the cause its report might identify. In this paper, we present ROGUE-AI (Replication of Agentic Security Events: Audit and Investigation), a multi-agent framework that closes the above gap and rebuilds each reported incident offline together with a matched control that lacks only the blamed cause, and runs agents on both. Our results suggest that ROGUE-AI is able to reproduce the reported harm in 89% of its 108 attempted replications. Agents built on five current GPT, Claude, Gemini and DeepSeek models made the reported mistake in 18% to 33% of runs, against at most 1% without the blamed cause. Of the 72 incidents with real-world harm, 62 had no underlying attacker, and in these, agents did what their task asked for but missed an ordinary hazard, such as a database shared with the live site. Replications without an attacker caused more harm than replications of prompt injection for all five models. Defenses against prompt injection look for injected instructions, but these replications contain none. Further, a warning after each tool output that names the hazard and its consequence lowered harm for all five models, whereas a note of the same length that does not mention the hazard left most of the harm in place. Built from real incidents, the ROGUE-AI replications form a reusable audit framework that vendors can rerun before each new release.

[MA-1] Public-Private Cooperation in the Port of Rotterdam: An Agent -Based Toy Model of Cooperative Efforts Against Cocaine Trafficking

【速读】:该论文旨在解决鹿特丹港(Port of Rotterdam, PoR)毒品走私治理中公共与私人部门协同合作(Public-Private Cooperation, PPC)的复杂性问题,尤其关注信任、合作、竞争、压力、执法强度及参与者能力等关键因素之间的动态交互机制。其解决方案的关键在于构建一个结合代理模型(Agent-Based Model, ABM)与因果回路图(Causal Loop Diagram, CLD)的“玩具模型”(toy model),通过模拟不同初始条件下参数扰动对协同治理效能的影响,揭示系统内在的结构性惯性。研究发现,单一措施难以在五年时间尺度上实现持久成效,而协调推进信任建设、合作强化与适度压力,并降低监管执行摩擦的多参数综合策略,可使治理效能提升25%,显著优于仅增强合作的孤立策略(13%提升)。该模型强调治理效能的高度制度依赖性,涵盖制度、关系、行为与资源准备度四维维度,为港口安全治理提供了基于系统动力学的探索性假设。

链接: https://arxiv.org/abs/2611.26154
作者: Frederike Oetker,Charlotte Geertman
机构: 未知
类目: Multiagent Systems (cs.MA)
备注:

点击查看摘要

Abstract:Cocaine trafficking poses a persistent governance challenge in the Port of Rotterdam (PoR), requiring cooperation between public and private actors. This paper develops an Agent-Based toy Model (ABM) coupled with a Causal Loop Diagram (CLD) to explore interactions between trust, cooperation, competition, pressure, enforcement, and actor skills in public-private cooperation (PPC). Based on empirical data of PPC in the PoR, we explore responses to parameter shocks across different initial configurations. In this model, effectiveness of PPC is strongly state-dependent and measures institutional, relational, behavioral, and resource readiness. The toy model demonstrates that isolated, single-track measures yield limited long-term gains modelled over 5 years due to structural system inertia, whereas coordinated multi-parameter strategies produce substantial improvements. Specifically, combining trust-building, cooperation, and moderate pressure with reduced regulatory enforcement friction raises effectiveness by 25% from a baseline effectiveness, compared to just 13% for an isolated increase in cooperation. Acknowledging the exploratory nature of the model, these findings offer exploratory hypotheses on port security governance. Future work should validate the model against test cases and additional expert input.

[MA-2] Law at the Speed of AI: Critical Legal Infrastructure for the Machine Age: Autonomy Under Law Architecture (AULA)

【速读】:该论文旨在解决自主人工智能代理(autonomous AI agents)在组织内外大规模部署所引发的法律可识别性缺失问题,即现有法律基础设施无法有效识别、解释和规制由自主代理产生的行为,导致责任归属不清与权利分配失序。其核心挑战在于:尽管底层通信与认证协议(如模型上下文协议Model Context Protocol和Agent2Agent协议)日趋成熟,但这些技术进展并未带来法律层面的可追溯性与可问责性。当前运行时权限控制仅限制资源访问,而不界定代理行为的法律授权基础,从而形成“法律归因鸿沟”(legal attribution gap)。该论文提出“法律之下的自治架构”(Autonomy Under Law Architecture, AULA),通过将AI系统与数据资产注册为依托于特定法律主体的法律客体,并以密码学绑定的智能法律工具(smart legal instruments)进行治理,实现对代理行为的实时法律评估。每个智能法律工具可包含“法律上下文条款”,使现行立法、监管及司法判例在行为发生时具备实际法律效力,从而构建兼具技术可行性与法律可执行性的新型治理基础设施,确保人类与自主机器共同参与的法律、财产权利与经济体系可持续运行。

链接: https://arxiv.org/abs/2611.25985
作者: Natasha E Blycha,James Myint,Chen Feng,Dr Amani Soysa,Koshin Jama,Schellie-Jayne Price,Matthew Blycha
机构: 未知
类目: Computational Engineering, Finance, and Science (cs.CE); Artificial Intelligence (cs.AI); Emerging Technologies (cs.ET); Multiagent Systems (cs.MA)
备注: 203 pages, 13 Tables and 17 Figures

点击查看摘要

Abstract:The global deployment of autonomous AI agents within and across organisations exceeds the governance infrastructure available to constrain it. Agent communication and authentication protocols, principally the Model Context Protocol and the Agent2Agent Protocol, are maturing quickly, but maturity at the protocol layer does not produce legal legibility at the governance layer. Runtime enforcement written by the deploying organisation fixes the resources an AI agent can reach, not the authority under which it acts. Legal legibility is the capacity of a legal system to recognise, interpret and regulate activity by mapping it onto legal personhood, property, rights and liabilities. Existing (analogue) legal infrastructure cannot supply it for autonomous deployment. The resulting deficit is bidirectional. It concerns the attribution of liability when an AI agent causes harm, and the allocation of rights, including data and economic rights, generated by agentic activity and by the AI agent itself as a source of value and form of property. This paper characterises that deficit as the legal attribution gap, a material and unaddressed risk to economies, organisations and citizens that is magnified in cross-organisational multi-agent deployment. The paper describes the Autonomy Under Law Architecture (AULA), in which AI systems and data assets are registered as legal objects anchored to one or more identified legal persons, governed by cryptographically tethered smart legal instruments and evaluated at runtime. Each smart legal instrument can carry a legal context clause, through which the legislation, regulation and judicial decisions in force at the time of the act are given practical and legal effect in that evaluation. Infrastructure of this kind is a precondition for the continued function of any legal, property and economic system with both human and autonomous machine stakeholders.

[MA-3] Coherent belief and opinion propagation produces more echo chambers: corrected version

【速读】:该论文旨在解决社交网络中回音室(echo chamber)形成机制的问题,特别是探究在意见传播过程中同时传播信念(belief)对回音室演化的影响。传统意见动态模型仅关注意见更新过程,而本研究通过引入经典的信念融合操作,将信念与意见的传播相结合,并基于共享价值观维持信念与意见的一致性,从而构建更贴近现实的认知传播框架。其解决方案的关键在于:在现有意见动态模型基础上,显式建模信念与意见的协同演化机制,并设计基于同质性、隔离性和强化性的量化指标以识别回音室。仿真结果表明,在不同初始条件和个体容忍度下,即使采用更严格的检测标准,回音室仍会形成;更重要的是,当意见与信念相连接并保持一致性时,虽然信念层面的回音室数量增加,但每个回音室所涉及的参与者数量减少,揭示了信念传播对回音室结构的细微影响。

链接: https://arxiv.org/abs/2611.25879
作者: Hiro Kataoka,J{é}r{ô}me Euzenat(MOEX),Koji Hasebe
机构: 未知
类目: Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注: kataoka2026e

点击查看摘要

Abstract:The occurrence of echo chambers is a well-known phenomenon in social networks. Echo chambers are groups of agents satisfying three properties: homogeneity, i.e., sharing the same opinion, segregation, i.e., listening primarily to agents of the same group, and reinforcement, i.e., increasing opinion similarity over time. Opinion dynamics models have been used to explain how echo chambers occur through agents updating their opinions. However, social network users also exchange beliefs supporting their opinions. Here, we aim at identifying the effect of propagating beliefs together with opinions on the formation of echo chambers. For that purpose, we complement an existing opinion dynamics model with classical belief merge operations. Beliefs and opinions are also maintained coherent guided by values that agents share. We introduce measures, applying to either beliefs and opinions, to identify echo chambers based on the three properties. This model is used to simulate opinion and belief propagation with varying initial conditions and agent tolerance for different opinions/beliefs. Results confirm the occurrence of echo chambers observed in previous opinion dynamics studies in a different setting and with a stricter measure. They also demonstrate the occurrence of echo chambers with belief propagation. Finally, they show that connecting opinions to beliefs and maintaining their coherence increases the number of belief echo chambers, but that they involve a reduced number of agents.

[MA-4] Interlacing Committees with Edge-Expansion

【速读】:该论文旨在解决基于批准投票(approval-based)的委员会选择中,如何有效衡量和优化委员会质量的问题。现有研究采用“交错性”(interlacing)框架,通过委员会诱导出选民间的相似性图,并以图的连通性作为评价标准。然而,该框架所使用的连通性度量存在局限性。本文提出将边扩张性(edge-expansion)这一经典的图论指标作为优化目标,因其在刻画网络极化现象方面具有坚实的理论基础。作者从信息传播模型的角度为交错性框架提供了理论支撑,并系统比较了边扩张性与先前提出的连通性度量之间的差异,揭示了其在捕捉结构特性上的本质不同。此外,论文分析了最大化边扩张性的计算复杂性:即使在选民区间偏好(voter-interval preferences)下,该问题仍为计算困难;但当委员会规模固定时,问题可在多项式时间内求解,且在选民-候选人区间偏好情形下,算法对委员会规模的依赖性得到显著优化。因此,该研究的关键在于引入边扩张性作为更优的优化目标,并建立了其理论合理性与可计算性边界。

链接: https://arxiv.org/abs/2611.25852
作者: Gili Bielous,Reshef Meir
机构: Technion - Israel Institute of Technology(以色列理工学院)
类目: Multiagent Systems (cs.MA)
备注:

点击查看摘要

Abstract:Approval-based committee selection has recently been studied through the lens of \emphinterlacing, a framework for studying polarization in approval elections, in which committees induce a similarity graph over voters. The quality of committees is then evaluated through the connectivity of the resulting graph. We revisit the connectivity measures used in this framework and propose \emphedge-expansion, a classical graph-theoretic measure with well-established connections to polarization in networks, as the optimization objective. We provide a theoretical foundation for this proposal by relating interlacing to classical models of information diffusion and by comparing edge-expansion with previously proposed connectivity objectives, identifying qualitative differences in the structural properties they capture. Finally, we study the computational complexity of optimizing edge-expansion, showing that hardness persists even for voter-interval preferences, while the problem is polynomial-time solvable for every fixed committee size, with an improved dependence on the committee size for voter-candidate-interval preferences.

[MA-5] DS2Agent : Holistic Device Agents for Accelerated Embedded Board-Level Design

【速读】:该论文旨在解决嵌入式板级开发中因组件选型、接口配置、驱动编写及系统集成等复杂任务所需高度专业知识而导致的效率瓶颈问题,尤其针对现有基于大语言模型(LLM)的智能体在板级设计场景下存在的局限性:既往方法或局限于芯片级设计,或将数据手册(datasheet)视为静态文本,缺乏对特定组件是否适用于目标板级设计的动态判断能力,且由于未与真实硬件交互,易产生幻觉性错误信息。其核心解决方案是提出一个开源框架DS2Agent,通过从设备数据手册构建可交互的智能体,该智能体由三个协同模型组成:知识模型用于支持知识检索与问答;数字孪生模型(digital-twin model)捕捉设备的物理功能行为;接口模型支持驱动开发与系统集成。为抑制幻觉,系统还支持可选的真实硬件连接,使智能体能够基于实际设备反馈校准其内部知识。实验表明,相较于基线方法,DS2Agent显著提升了问答准确率(87→97/100),并分别在驱动生成和系统集成任务中减少44.8%、72.9%和29.7%的平均令牌使用量,验证了其高效性与可靠性。

链接: https://arxiv.org/abs/2611.25774
作者: Hongbin Yang,Chi Zhang,Yinhao Xiao,Runyu Pan,Yanni Yang,Xiuzhen Cheng
机构: Shandong University (山东大学); GDUFE-UCM Joint Institute, Guangdong University of Finance and Economics (广东财经大学-西班牙马德里康普顿斯大学联合学院)
类目: Multiagent Systems (cs.MA)
备注:

点击查看摘要

Abstract:Embedded board-level development demands extensive expertise, from component selection and device characteristics to interface configuration, driver writing, and system composition. Recently, LLM-based agents have shown strong capabilities, raising the prospect of automating board-level design tasks. However, existing agents either target chip-level design or treat datasheets as static text, offering little guidance on whether a specific component should—or even could—be used for a given board-level design. Moreover, without touching real hardware, they hallucinate incorrect details. To fill this gap, we present \setlength\fboxsep1pt\hrefthis https URL\fcolorboxbluewhiteDS2Agent, an open-source framework that constructs interactive agents from device datasheets. Each agent comprises three models: a knowledge model supporting retrieval and question answering; a digital-twin model capturing the physical functionality of the device; and an interface model supporting driver development and system integration. To suppress hallucinations, we also allow optional connections to real devices, so that the agent can calibrate its internalized knowledge against the device. We evaluate DS2Agent on ten devices spanning seven functional categories. Compared with the baseline, the DS2Agent-augmented agent improves the question-answering score from 87 to 97 out of 100 while using 44.8% fewer tokens. For driver generation, it reduces average token usage per device by 72.9%. For system integration, it reduces total token usage across three FreeRTOS applications by 29.7%. Subjects: Multiagent Systems (cs.MA) Cite as: arXiv:2611.25774 [cs.MA] (or arXiv:2611.25774v1 [cs.MA] for this version) https://doi.org/10.48550/arXiv.2611.25774 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[MA-6] GRODE-RTL: Source-Mapped Dependency Views for LLM -Based RTL Repair

【速读】:该论文旨在解决生成式 AI 在自动修复寄存器传输级(Register-Transfer Level, RTL)设计中非局部故障(non-local faults)时所面临的挑战,即根因与故障输出在源代码中相距甚远,导致传统将RTL视为扁平令牌序列的方法难以捕捉硬件数据流的拓扑结构。现有方法或因过度简化而丢失关键依赖关系,或因暴露完整综合网表而导致模型被细粒度元数据淹没。为此,论文提出GRODE-RTL——一种以锚点为中心的表示方法,其核心在于从网表中提取可疑寄存器的单时钟周期扇入锥(fan-in cone),将网表节点映射回可编辑的源代码行,并构建紧凑的跳段划分(hop-partitioned)依赖视图。该方法通过优先保留直接驱动信号并折叠远距离跳转,有效引导标准工具使用型智能体进行修复,同时避免引入真实答案提示或Oracle反馈。实验在30个经形式化验证的非局部故障上评估了三种大语言模型(DeepSeek-V4-Flash、GLM-5.2和MiniMax-M3),结合六组配置的消融研究,结果表明:相较于源码级探索,GRODE-RTL使pass@1分别提升16.7%(DeepSeek)、26.7%(GLM-5.2)和20.0%(MiniMax-M3),且在所有模型中对远距离和间接锚点子集均保持一致增益。控制实验进一步证明,选择性依赖组织而非全图暴露才是实现高效RTL修复的关键。

链接: https://arxiv.org/abs/2611.25661
作者: Jingyi Zhou,Yi Liu,Bin Sun,Hongji Zhang,Dongsheng Zuo,Qiang Xu
机构: The Chinese University of Hong Kong(香港中文大学); Institute of Computing Technology, CAS(中国科学院计算技术研究所)
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA)
备注:

点击查看摘要

Abstract:Large Language Models show great potential for automated Register-Transfer Level repair, yet they struggle with non-local faults where root causes and failing outputs reside far apart in source code. Treating RTL as a flat token sequence obscures hardware dataflows, whereas exposing complete synthesized graphs overwhelms the model with fine-grained metadata. To resolve this representation dilemma, we propose GRODE-RTL, an anchor-centric representation that extracts the single-clock-stage fan-in cone of a suspect register from the netlist, maps netlist nodes back to editable source lines, and organizes them into a compact hop-partitioned dependency view. By prioritizing direct drivers and folding distant hops, GRODE-RTL guides standard tool-using agents without exposing ground-truth hints or oracle feedback. We evaluate GRODE-RTL across 30 formally validated non-local faults using three LLMs (DeepSeek-V4-Flash, GLM-5.2, and MiniMax-M3) alongside a controlled six-configuration ablation study. GRODE-RTL improves pass@1 over source-level exploration by 16.7% on DeepSeek, by 26.7% on GLM-5.2, and by 20.0% on MiniMax-M3, with consistent gains on the hard-to-retrieve far-distance and indirect-anchor subsets of all three models. Our controls isolate structural relevance from generic context expansion, proving that selective dependency organization rather than full graph exposure is key to effective RTL repair.

[MA-7] Game-Theoretic Data Allocation for Expert Specialization and Routing

【速读】:该论文旨在解决如何在不依赖显式训练门控机制的情况下,预测并实现最优专家选择的问题。核心挑战在于:能否通过训练前的专家分配(allocation)预先判断哪个专家在特定提示(prompt)上表现最佳,从而实现推理阶段的可解释性分工与高效路由。其解决方案的关键在于提出InflEns框架——一种基于固定提示-特质空间中位置博弈(location game)的LoRA专家混合模型,其中专家的训练权重与推理路由均源自同一博弈过程。该博弈以最大化各专家对提示分布的预期份额为目标,不依赖专家损失函数。通过引入博弈核的通用理论判据,可判断共位专家位置是否因局部不稳定而发生分离,从而实现自然的专业化分工。实验表明,在七个安全相关基准构建的语料上,采用锐化单纯形表示时,最高份额专家在约90%的提示上具有最低的保留响应负对数似然(NLL),远优于随机置换分配后的15%;且无需训练显式门控网络即可达到与已有学习门控相当的性能。由于路由器固定,可精准定位有害行为来源,仅需重训单一专家,使HarmBench攻击成功率从0.51降至0.02,同时保持指令遵循能力。此外,该框架可通过核函数选择、特质空间几何结构及专家数量等参数灵活调控专业化程度,具备高度通用性,适用于任意编码器与数据分布。

链接: https://arxiv.org/abs/2611.25167
作者: Mark Lovett,Denis Lim,Soroush Vosoughi
机构: Dartmouth College(达特茅斯学院)
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA)
备注:

点击查看摘要

Abstract:Can the allocation used to train experts also predict which expert will perform best on a prompt? Such a connection would make the intended division of labor inspectable before training and allow the allocation itself to guide inference. We introduce InflEns, a mixture of LoRA experts whose training weights and inference routes come from the same location game in a fixed prompt-trait space. Players choose positions to maximize their expected shares of the prompt distribution, without using expert losses. We utilize the game’s kernel-general theoretical criterion for when co-located positions become locally unstable to separation. We evaluate InflEns on a corpus assembled from seven safety-related benchmarks, which demonstrates that the highest-share expert has the lowest held-out response negative log-likelihood (NLL) on approximately 90% of prompts across four kernels under a sharpened simplex representation, compared with approximately 15% after permuting allocations across prompts. We demonstrate that the method performs comparably to published learned gates without having to train an explicit gate on the experts. Because the router is fixed, we can trace harmful behavior to a single expert and retrain only that expert, which lowers HarmBench attack success from 0.51 to 0.02 while keeping instruction following. The game also gives us several controls over how the experts specialize, including the kernel choice and its parameters, the trait-space geometry, and the number of experts. Importantly, InflEns is a general framework that can be extended to any encoder and data distribution.

[MA-8] GridAnchor: Specialized Multi-Agent AI for Safeguarding Autonomous Energy Systems Against Cyber-Physical Hallucinations

【速读】:该论文旨在解决自主能源系统中由多个专业化学习代理(如预测器、状态估计器和调度器)协同工作时产生的“网络物理幻觉”(cyber-physical hallucinations)问题,即代理输出虽具高置信度但违背物理规律,进而引发下游代理的错误决策与连锁反应。其核心解决方案是提出一种名为GridAnchor的验证层,关键在于采用基于物理残差的图变换器(Physics-Residual Graph Transformer),通过计算代理输出、测量值与线性化电网模型之间的残差,对每个声明进行评分;结合Mondrian conformal p-values与Benjamini-Hochberg校正程序将评分转化为可解释的决策,实现对幻觉及其传播路径的根因定位;并通过一个二次规划优化器,在修复步骤中以电压裕度为约束条件纠正不安全操作。实验表明,该方法在模拟农村中压馈线上达到0.846的消息级精确率-召回率曲线下面积,优于六种基线模型,且误报率保持在名义水平附近,闭环运行下无节点电压越限,显著提升了系统的物理一致性与安全性。

链接: https://arxiv.org/abs/2611.25158
作者: Abderaouf Bahi,Hasan Masrur
机构: Computer Science and Applied Mathematics Laboratory (LIMA), Faculty of Science and Technology, Chadli Bendjedid University (查德利·本杰迪德大学理学院与技术学院计算机科学与应用数学实验室); Center for Smart Mobility and Logistics (IRC-SML), King Fahd University of Petroleum and Minerals (沙特法赫德国王石油矿产大学智能交通与物流中心); Department of Industrial and Systems Engineering, King Fahd University of Petroleum and Minerals (沙特法赫德国王石油矿产大学工业与系统工程系)
类目: Multiagent Systems (cs.MA)
备注:

点击查看摘要

Abstract:Autonomous energy systems increasingly chain specialized learned agents, such as a forecaster, a state estimator and a dispatcher that may be a language model. These agents can emit confident outputs that contradict physics, and downstream agents may act on them. We define such cyber-physical hallucinations and their cascades, organize them in eight types, and build CPH-Bench, a generator of labeled episodes on SimBench feeders. We propose GridAnchor, a verification layer whose Physics-Residual Graph Transformer scores every claim from residuals between the claims, the measurements and a linearized grid model. Mondrian conformal p-values with the Benjamini-Hochberg procedure turn the scores into decisions, a root head attributes cascades, and a quadratic program repairs unsafe actions with a voltage margin calibrated at repaired steps. On a simulated rural medium-voltage feeder, GridAnchor reaches a message-level area under the precision-recall curve of 0.846, against 0.837 for the strongest of six baselines, a difference within the confidence interval. It leads on estimates and actions, whereas a self-consistency check that knows the label tolerances reaches 0.874 by leading on forecasts. The false-alarm rate stays within 0.17 percentage points of the nominal level on steps without injected hallucinations. In closed loop, GridAnchor leaves no bus outside the voltage band on the test days, against 0.0700% of the bus-steps without protection, and simpler detectors with the same mitigation come close. Stale but coherent messages and single-bus forecast errors remain mostly undetected.

[MA-9] BuggyMLE: Scaling Machine Learning Engineering Difficulty with Bug Injection

【速读】:该论文旨在解决当前机器学习工程(Machine Learning Engineering, MLE)基准测试中代理(agent)能力评估受限的问题,即现有基准普遍局限于单文件编辑且缺乏对完整代码仓库及终端环境的访问权限,难以真实反映复杂模型开发中的多文件协作、依赖管理、调试与框架迁移等实际挑战。其核心解决方案在于提出一种名为“代码复杂度指数”(Code Complexity Index, CCI)的新度量标准,基于仓库规模、依赖深度和补丁大小等十个结构化特征,能够有效预测多个公开编码基准上的平均求解率(R² = 0.951),并识别出补丁大小与复杂度为关键预测因子。基于CCI,研究构建了可扩展的BuggyMLE框架,通过在222个真实ML仓库中注入可控缺陷,生成超过6,600个带缺陷的MLE仓库变体及近26万行注入代码。在此基础上,研究首次构建了难度分级的仓库级MLE基准——BuggyMLE-Bench-Lite,验证结果显示,随着难度从最低到最高层级提升,模型最高求解率由95%降至9%;同时,利用三层次难度的训练课程对Qwen3.5-9B模型进行微调,使其在BuggyMLE-Bench-Lite上表现媲美前沿模型,并展现出良好的跨领域迁移能力:在SWE-bench Lite上的解决率从24%提升至33%,在LiveCodeBench v6上的pass@1指标从69%提升至71%。

链接: https://arxiv.org/abs/2611.25088
作者: John Wu,Rushi Qiang,Yuchen Zhuang,Guowang Li,Girish Narayanswamy,Max Xu,Sam Schmidgall,Mark Malhotra,Shwetak Patel,Jimeng Sun,Daniel McDuff,Bo Dai,Xin Liu
机构: Google Research(谷歌研究); Google DeepMind(谷歌深脑); Siebel School of Computing and Data Science, University of Illinois Urbana-Champaign(伊利诺伊大学厄本那-香槟分校西贝尔计算机与数据科学学院); Georgia Institute of Technology College of Computing(佐治亚理工学院计算学院)
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Multiagent Systems (cs.MA); Software Engineering (cs.SE)
备注:

点击查看摘要

Abstract:Enabling AI systems to autonomously improve AI models requires agents capable of performing machine learning engineering (MLE). In practice, designing, training, and evaluating complex models demands modifying multi-file repositories to reproduce results, debug training runs, integrate datasets, and port frameworks. However, existing MLE benchmarks limit agents to single-file edits without full repository or shell access, and common repository-level software engineering benchmarks have rapidly saturated. To address this, we introduce the Code Complexity Index (CCI), a difficulty metric based on ten structural features spanning repository scale, dependency depth, and patch size. CCI strongly predicts average solve rates across eight public coding benchmarks ( R^2 = 0.951 ), identifying patch size and complexity as top predictors. Guided by CCI, we develop BuggyMLE, a scalable framework where agents inject controlled mutations into working ML repositories, synthesizing over 6,600 bug-injected MLE repository variants and nearly 260K lines of injected code based on 222 source repositories. Using BuggyMLE, we make two contributions. First, we construct BuggyMLE-Bench-Lite, the first difficulty-graded, repository-level variant of MLE-Bench Lite; across 14 evaluated models, the top solve rate drops from 95% to 9% as difficulty is scaled from the easiest to the hardest tier. Second, we synthesize a training curriculum across three difficulty levels over 200 additional ML repositories, fine-tuning a Qwen3.5-9B model to rival frontier models on BuggyMLE-Bench-Lite. Our trained model transfers beyond ML, raising its SWE-bench Lite resolve rate from 24% to 33% and its LiveCodeBench v6 pass@1 from 69% to 71%.

[MA-10] FinVeriQA: Financial Research Agents for Question Answering: Routing Retrieval MCP Tool Use and Provenance-Aware Evidence Verification STOC

【速读】:该论文旨在解决大语言模型(Large Language Models, LLMs)在金融问答任务中普遍存在但关键的可靠性问题:尽管其生成文本流畅,却常出现财务比率计算错误、引用过时股价以及错误调用工具等严重失误,且往往以高置信度输出。核心挑战在于如何在多智能体系统中合理分配各组件职责,以实现准确、可验证的金融分析。解决方案的关键在于构建一个分层、分工明确的协同架构——通过一个轻量级的类型化决策模型(Laya)负责任务路由与意图识别,避免将复杂计算或证据检索交由通用语言模型处理;由专用的Python环境执行所有数值计算,确保精确性;利用检索与外部工具获取实时证据;由Mistral-7B生成带引用的答案;并通过代码验证器独立核验引用编号是否存在及数值是否一致,同时引入模型上下文协议(Model Context Protocol, MCP)服务器统一管理整个流水线。实验表明,该架构显著提升了任务路由准确性(Laya达0.85,优于Mistral-7B的0.52),使财务计算正确率从29%提升至100%,并将股价误差从210%降至0.00%。然而,研究也揭示了共享证据污染下的自洽性陷阱:当共同证据被篡改时,模型仍保持高一致性(0.55–0.59),但准确率骤降至0,凸显验证必须独立于原始证据来源的重要性——仅依赖共享证据检查可捕获35%错误,而直接比对原文可捕获77%以上(p = 5.96×10⁻⁸),证明独立验证机制是保障可信输出的核心。

链接: https://arxiv.org/abs/2611.25071
作者: Meenakshi Rajpurohit,Anna Phan,Ravikumar Komandur Narayanan,Gunanidhi Ramakrishnan,Nathan Howland
机构: San José State University (圣何塞州立大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR); Multiagent Systems (cs.MA)
备注: 15 pages, 4 figures, 11 tables. Code: this https URL

点击查看摘要

Abstract:Large language models (LLMs) write fluent financial analyses but miscompute ratios, recall stale prices, and route questions to the wrong tool with high confidence. We ask which component should do each job in a financial question-answering assistant, and study it with FinVeriQA, an open-source multi-agent research prototype over SEC filings and live market data: a compact typed-decision model (Laya) routes each question and screens out advice requests, Python computes every number, retrieval and tools fetch evidence, Mistral-7B writes a cited answer, a code validator checks that cited IDs exist and that numbers match computed notes, and a Model Context Protocol (MCP) server exposes the pipeline. In live experiments, Laya routed better than Mistral-7B (0.85 vs. 0.52; overall 0.86 vs. 0.68, McNemar p = 0.0003); a 12B Mistral decided no better, and DeepSeek-R1-8B decided better (0.79 vs. 0.68, p = 0.0192) but about 9.9 times more slowly. Mistral-7B computed 29% of 100 financial calculations correctly against 100% for Python, a live quote cut price error from 210% to 0.00%, and hybrid retrieval reached 76% Hit@5 on 50 filing questions. Yet in a live case study, both agent answers passed validation while describing AMD sample data under NVIDIA’s and Alphabet’s names. A controlled upper-bound study on the FinRank benchmark explains why: when one figure in the shared evidence is corrupted, self-consistency stays confident (agreement 0.55 to 0.59 while accuracy falls from 0.50 to 0.00), and a shared-evidence check (the kind our validator performs) catches 35% of the errors, whereas a check against the original passage catches 77% ( p = 5.96 \times 10^-8 ); for DeepSeek-R1-8B, 68% vs. 90%. Verification must be independent of the evidence the writer saw.

[MA-11] Attacks and Defenses in Decentralized Multi-Agent Systems with Hop-by-Hop Task Decomposition and Delegation

【速读】:该论文旨在解决去中心化多智能体大语言模型(Decentralized Multi-Agent LLM)系统中因缺乏全局监管而引发的安全挑战,特别是恶意智能体对任务完整性、可用性与机密性的威胁。其核心问题在于:在无中心协调的环境下,各智能体独立运行且彼此不可信,导致无法有效验证行为合法性,进而可能引发任务失败、资源滥用及敏感信息泄露。解决方案的关键在于提出一种基于密码学的协议,能够生成不可抵赖的可验证执行追溯链(verified execution trace),确保每个操作行为均可被审计与溯源;同时设计一个审计器(auditor)对这些追溯链进行分析,实现对恶意行为的高精度检测与归因。实验结果表明,该方案在隐藏基准测试(HiddenBench)上对gpt-6.1-sol和Qwen3.8-27B-FP8模型均表现出极低延迟开销,并成功识别并正确归因98%的攻击行为,包括那些表面合规但实际恶意的攻击。

链接: https://arxiv.org/abs/2611.25056
作者: Matthew D. Laws,Maria Lazou,Alina Oprea,Cristina Nita-Rotaru
机构: Northeastern University(东北大学)
类目: Cryptography and Security (cs.CR); Multiagent Systems (cs.MA)
备注: 17 pages, 5 figures, 7 tables

点击查看摘要

Abstract:Decentralized multi-agent LLM systems enable independently administered agents, each with its own operator, model, memory, and configuration, to collaborate on complex tasks without centralized coordination. Agents autonomously decompose tasks, execute subtasks and delegate remaining work to peers, enabling scalability and eliminating single points of failure. However, decentralization also introduces significant security challenges, as no single entity has complete visibility into task execution or the authority to enforce correct behavior across agents. In this paper, we systematize threats from malicious agents in decentralized multi-agent systems and propose mitigation techniques. We first design and evaluate attacks against integrity, availability, and confidentiality, demonstrating that a single malicious agent can cause up to 100% task failure, inflate token consumption and runtime by up to 7.7x and 3.2x, respectively, and exfiltrate up to 94% of relevant prior tasks from benign agents. We then present a cryptographic protocol that produces an irrefutable verified execution trace of every action taken in the system and develop an auditor that analyzes these traces to detect and attribute malicious behavior. Evaluations on HiddenBench using gpt-6.1-sol and Qwen3.8-27B-FP8 show that our protocol incurs negligible latency overhead, while the auditor detects and correctly attributes 98% of attacks, including those that comply with the protocol. Comments: 17 pages, 5 figures, 7 tables Subjects: Cryptography and Security (cs.CR); Multiagent Systems (cs.MA) Cite as: arXiv:2611.25056 [cs.CR] (or arXiv:2611.25056v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2611.25056 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[MA-12] Beyond 4: Improved Competitive Traveling Repairperson

【速读】:该论文旨在解决在线旅行维修人员问题(Online Traveling Repairperson Problem, OTRP)在任意度量空间下的竞争比优化问题,尤其关注具有多个维修人员(k ≥ 1)从同一出发点(depot)出发的场景。其核心挑战在于如何在请求动态到达且无法预知未来的情况下,设计高效的在线算法以最小化最大响应时间或总延迟。针对一般情况,论文提出了一种3.877-竞争性的确定性算法,优于此前最优的4-竞争性算法。对于一类特殊的零残差实例(zero-residual instances),通过引入ε-近似技术,进一步将竞争比提升至3 + ε(∀ε > 0)。关键创新在于对请求调度策略与维修人员路径规划之间的权衡进行精细建模,并结合度量空间的几何特性设计了具有更优性能保证的在线决策机制。此外,该方法还可扩展至相关联的在线叫车问题(Online Dial-a-Ride Problem)的多种变体,显著提升了现有算法的理论性能上限。

链接: https://arxiv.org/abs/2611.21943
作者: Bibek Maharjan,Gokarna Sharma
机构: Kent State University (肯特州立大学)
类目: Data Structures and Algorithms (cs.DS); Distributed, Parallel, and Cluster Computing (cs.DC); Discrete Mathematics (cs.DM); Multiagent Systems (cs.MA)
备注: 27 pages, 1 figure, 2 pseudocodes, 1 table

点击查看摘要

Abstract:With k\geq 1 repairpersons starting initially from a depot, we present a 3.877-competitive deterministic algorithm for the fundamental online traveling repairperson problem on an arbitrary metric. For the special class of zero-residual instances, we improve the ratio to 3+\epsilon for every \epsilon0 . We also extend these results to some variants of the related so-called online dial-a-ride problem. The state-of-the-art is the 4-competitive deterministic algorithm for both problems.

[MA-13] Evolving Decoding Algorithms for Quantum Error Correction NEURIPS2026

【速读】:该论文旨在解决量子纠错中如何高效实现容错性的问题,具体而言是针对量子低密度奇偶校验(qLDPC)码在连续测量噪声量子态时,如何通过经典解码算法快速、可靠地推断并纠正噪声。其核心挑战在于在保持高纠错性能的同时,显著提升解码速度以满足实时性需求。解决方案的关键在于构建一个领域特定语言(DSL),将已发表的信念传播(BP)类解码器形式化表达,并利用语法演化(grammatical evolution)对解码器参数进行自动搜索与优化。通过在DSL中显式呈现解码器参数,研究者可预先对BP+OSD与BP+LSD解码器引入阻尼(damping)机制,从而实现17.1至41.0倍的速度提升及更高的错误容忍率;更优的是,演化出的新型解码器在更低的时间开销下显著优于传统调优解码器,其中最快版本在仅需原方法1/31时间的情况下仍实现更高精度,且在不同噪声水平、码长和轮次下具有良好的泛化能力。这表明,将专家经验编码于DSL中形成的人机协同优化中间层,有效融合了人类直觉与机器学习的优化能力,为实现高效量子纠错提供了新范式。

链接: https://arxiv.org/abs/2611.25880
作者: Eduardo Dueñez,Edgar Said Hernandez Sanchez,Edgar A. Duéñez-Guzmán,Suzanne Sadedin
机构: Gibran.AI; The University of Texas at San Antonio(德克萨斯大学圣安东尼奥分校)
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Multiagent Systems (cs.MA); Neural and Evolutionary Computing (cs.NE)
备注: 5 main pages, 2 main figures. 15 total pages. Accepted in NeurIPS 2026 EvoRobust Workshop

点击查看摘要

Abstract:Quantum computation entails robustness engineering: a noisy quantum state is measured continuously, and a classical algorithm (the decoder) must infer corrections that reliably remove the noise, fast enough to keep up. We write several published belief-propagation (BP) decoders for quantum low-density parity-check (qLDPC) codes in a Domain-Specific Language (DSL), and search over it with grammatical evolution. Written in the language, the published decoders show their parameters in plain view, allowing us to re-parameterise BP+OSD (ordered statistics decoding) and BP+LSD (localized statistics decoding) decoders with damping, before any search, making each 17.1–41.0 \times faster at a higher error rate, and BP+OSD-0 about 103 faster than the ldpc package’s version tuned on the [[72,12,6]] bivariate-bicycle code. Evolution over the language finds decoders that improve the time-error front. One makes 23% fewer logical errors than BP+OSD-0 with damping at 0.64 \times its time, and decoders evolved from random code are more accurate than ldpc’s tuned BP+OSD-0 at 1/31 of its time or less. Frozen at one operating point, the two fastest evolved decoders generalize well across noise levels, code sizes and round counts, staying at the time-error front. Ultimately, we demonstrate that capturing expert knowledge within a DSL creates a powerful intermediate layer—one that effectively bridges human intuition and machine-driven optimization to unlock highly efficient quantum error correction.

自然语言处理

[NLP-0] Agent RoboCritic: Understanding the Capability of Frontier Agents on Direct Robot Control

【速读】: 该论文旨在解决当前前沿多模态语言代理在直接机器人控制任务中缺乏细粒度行为分析的问题。现有评估基准仅提供任务级别的成功率,无法揭示代理在执行过程中的具体成功机制或失败原因,导致对模型能力与局限性的理解过于粗略。为此,论文提出一种关键解决方案——AgentRoboCritic,这是一个基于意图的批判框架:在每一步动作执行时,策略代理(policy agent)同时声明其局部意图(local intent),由一个独立的批评者(critic)根据该意图评估动作结果及失败模式。该方法实现了动作级(action-level)的系统性分析,能够精准识别代理在闭环恢复、语义一致性以及零样本操作等方面的能力,同时揭示其在抓取、对齐等物理执行瓶颈上的局限性及不同失败模式间恢复能力的不均衡性。研究通过在400条人工标注动作的基准上验证该批评框架,并发布包含12,844条标注动作的数据集,配合微调后的开源批评模型Muse-Glimmer-30B,将失败模式识别准确率从32.58%提升至74.24%,显著降低了动作级分析的成本,为理解与改进生成式机器人控制提供了可扩展、可解释的新范式。

链接: https://arxiv.org/abs/2611.26613
作者: Yiyang Du,Yuxuan Kuang,Qitao Zhao,Yixuan Wang,Zhanqiu Guo,Yi Xu,Xin Li,Wenbin He,Liu Ren,Chenyan Xiong
机构: Carnegie Mellon University (卡内基梅隆大学); University of Illinois at Urbana-Champaign (伊利诺伊大学香槟分校); Bosch Research North America (博世北美研究中心); Bosch Center for Artificial Intelligence (BCAI) (博世人工智能中心)
类目: Robotics (cs.RO); Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: 28 pages. Project page: this https URL

点击查看摘要

Abstract:Frontier multimodal language agents such as GPT-6 Astra have demonstrated strong capabilities in controlling robots directly via low-level numerical actions without high-level primitives. However, existing benchmarks report only episode-level success rates, which show neither how these agents succeed nor where they fail. In this paper, we present a systematic action-level analysis of frontier agents on direct robot control across 12 manipulation benchmarks. We introduce AgentRoboCritic, an intent-grounded critique framework in which the policy agent states a local intent alongside each action, and a critic evaluates the action’s outcome and failure mode against this intent. Our action-level analysis demonstrates the promising capabilities of frontier agents on direct robot control, including closed-loop recovery from local execution failures, reliable semantic grounding across diverse scenes, and strong zero-shot manipulation without task-specific adaptation. It also identifies their current limitations, including task-relevant physical execution bottlenecks such as grasping and alignment, and uneven recovery across failure modes. Finally, we validate our critic on a 400-action human-annotated benchmark and release a dataset of 12,844 annotated actions alongside a fine-tuned open-source critic, Muse-Glimmer-30B, whose failure-mode accuracy rises from 32.58% to 74.24%, enabling low-cost action-level analysis.

[NLP-1] Represented but misread: Post-training changes how language models read out what they represent

【速读】: 该论文旨在解决后训练(post-training)是否改变语言模型内部表征(representation)及其输出判断方式的核心问题。研究发现,后训练并未显著改变模型对输入内容的内在表征能力,但显著改变了判断结果的读出机制(readout)。其关键在于:后训练通过调整模型在决策过程中的敏感度(sensitivity)与先验信念(base rate)的使用方式,改变了模型输出答案的逻辑方向。具体表现为,后训练使模型在人类/机器文本判别任务中平均敏感度下降(Δd’ = -0.34),且更倾向于过度响应先验概率;而在否定陈述真实性判断任务中,后训练反而修复了基线模型对真实性的误判,使答案性能提升至接近理论上限,并且这种修复效应集中在模型中层及深层的上下文位置之后的状态中。此外,实验表明仅需少量标注数据即可引导后训练的方向性转变,说明读出机制的调整具有高度可塑性。因此,该研究揭示了后训练主要影响的是“如何从已有表征中读出答案”,而非“表征本身”的本质变化。

链接: https://arxiv.org/abs/2611.26584
作者: Jochen Laubrock
机构: University of Potsdam (波茨坦大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 64 pages, 7+2 figures, 4+6 tables

点击查看摘要

Abstract:Does post-training change what a language model represents, or how this representation is read out into an answer? We compared linear probes with the models’ own answers, analyzed with signal detection theory, in six open-weight base models and their post-trained versions. For judgments of whether German texts were written by a human or an AI, probes and word n-gram classifiers separated the classes well, but the answers did not and often ordered the texts inversely. Post-training lowered the sensitivity of the answers on average ( \Delta d’ = -0.34 across model families; -0.29 across three prompts). It also changed the use of a stated base rate: Base models shifted their log-odds by less than the stated change in prior odds, most post-trained models by more. In two models with stage-wise checkpoints, the change occurred during supervised fine-tuning, model stitching implicated the early layers, and from mid-depth on internal states at positions after the text, not those of the text, transferred the change between checkpoints. Our corpora cannot be shared, so we add a case reproducible from public data, the truth of negated statements. Here, post-training acted in the opposite direction: Probes decoded truth at or near ceiling in every model, the answers of several base models did not use it, and post-training repaired them; by mid-depth the repair resided at the positions after the statement; below it, the representation of the statement contributed, strongly in one of four models. In a training study, adding 2,000 correctly labeled authorship judgments of news texts to otherwise identical fine-tuning data improved the answers on our newspaper and textbook corpora and inverted them on the chats, where they followed a news-trained probe of the base model. Post-training thus changed how a judgment is read out of what a model represents, and a small training signal can set its direction.

[NLP-2] Hybrid Retrieval for Government Scheme Search in English and Marathi

【速读】: 该论文旨在解决印度马哈拉施特拉邦政府政策与公共服务信息检索中,公民信息需求与正式行政描述之间因语言差异及术语不一致导致的匹配难题。其核心挑战在于跨语言(英语-马拉地语)环境下如何实现精准、高效的语义检索。解决方案的关键在于构建一个双语检索系统,通过等权重融合机制结合传统的词项匹配方法(BM25)与基于句子嵌入的语义相似性计算(如IndicSBERT和MahaSBERT),在统一索引中对齐英-马拉地语政策描述进行多语言、马拉地语主导及翻译中介的语义配置。实验表明,混合检索显著优于纯密集检索(dense retrieval),在平均倒数排名(MRR)上分别提升至0.872(IndicSBERT)和0.879(MahaSBERT),同时在精度(Precision)上从0.781提升至0.808,且保持与BM25相当的召回率(Recall=0.946)。结果验证了在该语料中词法信号与语义信号具有互补性,但相较于强大的词法基线,性能提升幅度有限且依赖具体评估指标,凸显了混合检索在实际政务信息访问场景中的有效性与应用价值。

链接: https://arxiv.org/abs/2611.26514
作者: Aishwarya Gunjal,Kanchan Khemnar,Pawan Poojary,Pranav Kulkarni,Raviraj Joshi,Manish Khodaskar
机构: Pune Institute of Computer Technology, Pune, India(普奈计算机技术学院, 普奈, 印度); L3Cube Labs, Pune, India( L3Cube 实验室, 普奈, 印度)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Retrieving government schemes requires matching citizens’ information needs to formal administrative descriptions, often across languages and terminology. We present a bilingual search system for Maharashtra government schemes and public services that combines BM25 retrieval with sentence-embedding similarity through equal-weight score fusion. The system indexes aligned English-Marathi scheme descriptions and supports multilingual, Marathi-focused, and translation-mediated semantic configurations. Evaluation on manually curated bilingual queries compares BM25, dense retrieval, and hybrid retrieval using IndicSBERT and MahaSBERT. BM25 provides a strong baseline with a mean reciprocal rank (MRR) of 0.864. Hybrid retrieval increases MRR to 0.872 with IndicSBERT and 0.879 with MahaSBERT, whereas the corresponding dense-only systems achieve 0.663 and 0.655. MahaSBERT-based fusion also increases precision from 0.781 to 0.808 while matching BM25 recall at 0.946. These results support the complementary use of lexical and semantic signals in this collection, while showing that improvements over a strong lexical baseline are modest and metric-dependent. The study provides an application-focused analysis of hybrid retrieval for bilingual public-service information access.

[NLP-3] AutoMaster: Autonomous Post-Training by Making the Same Model a Better Teacher

【速读】: 该论文旨在解决后训练阶段(post-training)中仅依赖任务成功指标无法有效引导行为强化的问题,尤其针对教师模型(teacher model)生成的轨迹中存在的不良习惯、无效重复或表面验证等隐性缺陷。其核心解决方案是提出一种基于技能引导的微调方法——Skill-FT(Skill-Guided Fine-Tuning),通过自然语言形式的技能指令,在教师模型的推理过程中抑制不良行为并激发被忽视的有效行为。关键创新在于构建了名为AutoMaster的自主后训练系统,该系统由大语言模型(LLM)代理驱动,能够自动完成轨迹分析、技能设计、数据筛选及基于回放与目标模型评估的反馈迭代,实现闭环优化。实验结果表明,在终端操作、搜索和软件工程任务中,使用4B和9B规模的目标模型时,AutoMaster相较于传统的无引导监督微调(Vanilla SFT)在成功轨迹上的平均性能提升了1.42至8.99个百分点。此外,研究发现保留技能指令进行微调虽在评估阶段不参与推理,但能显著提升模型平均表现,表明在训练示范中保留技能上下文有助于更深层次的行为内化。这说明行为塑造(behavioral shaping)可有效提升固定模型作为教师的指导效能。

链接: https://arxiv.org/abs/2611.26510
作者: Guoxin Chen,Shuang Sun,Fanzhe Meng,Xinyu Tang,Ruihua Song,Wayne Xin Zhao
机构: Gaoling School of Artificial Intelligence, Renmin University of China(中国人民大学高瓴人工智能学院)
类目: Computation and Language (cs.CL)
备注: Website: this https URL Code: this https URL Models: this https URL

点击查看摘要

Abstract:Task success alone does not specify which behaviors post-training should reinforce. Even successful teacher trajectories can contain unsupported assumptions, unproductive repetition, or superficial verification. We propose Skill-FT (Skill-Guided Fine-Tuning), which uses natural-language skills to suppress undesirable habits and elicit underrepresented effective behaviors during teacher rollouts, then fine-tunes a target model on the resulting trajectories. We introduce AutoMaster, an autonomous post-training system in which an LLM agent develops these interventions through trajectory analysis, skill design, data selection, and feedback from rollouts and target-model evaluation. Across Terminal, Search, and software engineering tasks with 4B and 9B target models, AutoMaster improves mean performance over Vanilla SFT on successful unguided teacher trajectories by 1.42-8.99 percentage points, with the teacher, task pool, training set size, and SFT recipe held fixed. Together with the ablations and trajectory analyses, these results suggest that behavioral shaping can make a fixed model a more effective teacher. We further find that retaining skill instructions during SFT yields higher mean performance than removing them across all six settings, despite their absence at evaluation and identical supervised outputs. These findings suggest that preserving skill context as part of a training demonstration can support better behavioral internalization.

[NLP-4] GLiDE: A Fast Calibrated Decision Model Distilled from Reasoning Teachers

【速读】: 该论文旨在解决生成式推理(generative reasoning)虽能显著提升决策质量,但其高昂的计算成本(如逐词生成、长序列推理)限制了在实时、低延迟场景下的应用问题。核心挑战在于如何在保持单次前向传播(single forward pass)的高效性的同时,有效迁移推理过程带来的性能增益。其解决方案的关键在于提出GLiDE系列决策模型,通过移除全注意力层中的因果掩码(causal masking),实现上下文、问题与候选答案之间的双向交互,从而增强模型对复杂语义关系的理解能力;同时,采用多教师软目标蒸馏(soft-target distillation)结合渐进式后训练(progressive post-training),将多个推理模型的知识迁移至轻量级决策框架中;此外,引入无参数递归深度(parameter-free recurrent depth)机制,在不生成新文本的前提下增加推理时的计算深度,进一步提升决策准确性。实验表明,所有规模的GLiDE模型均位于Hugging Face DI基准的性能-参数帕累托前沿,其中27B版本在单卡上实现73.37的隐集得分(hidden-set score)、2.1%的期望校准误差及92.9毫秒中位延迟,显著优于同规模或更小模型。

链接: https://arxiv.org/abs/2611.26509
作者: Sahibzada Allahyar,Julia White,Dhruv Atreja,George Hurn-Maloney,Urchade Zaratiana
机构: Fastino Labs
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 18 pages, 8 figures

点击查看摘要

Abstract:Decision models enable fast, confidence-aware decisions by scoring candidate answers in a single forward pass, without generating text or intermediate reasoning. Yet reasoning can substantially improve decision quality, raising the question of whether its benefits can be retained without the cost of generating it. We investigate this question with \textbfGLiDE, a family of decision models spanning 0.4B to 27B parameters. Building on Qwen hybrid language models, GLiDE enables bidirectional interactions between the context, question, and candidate answers by removing causal masking from full-attention layers. To transfer reasoning capabilities into single-pass inference, we combine soft-target distillation from multiple teachers, including a reasoning model, with progressive post-training on diverse decision tasks. We further explore parameter-free recurrent depth to increase inference-time computation without generating tokens. Across five model sizes, GLiDE lies on the performance–parameter Pareto frontier of the Hugging Face \DI, with each model outperforming evaluated alternatives of equal or smaller size. \dmtwo (27B) achieves a hidden-set score of 71.52, with an expected calibration error of 2.0% and a median latency of 92.9,ms on a single GPU. Further checkpoint averaging and parameter-free recurrent depth yield GLiDE~27B, which reaches 73.37 with an expected calibration error of 2.1%. Comments: 18 pages, 8 figures Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2611.26509 [cs.AI] (or arXiv:2611.26509v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2611.26509 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[NLP-5] Multimodal Error Detection in Chest X-ray Reports: A Large Language Model Reading Image Evidence

【速读】: 该论文旨在解决放射科胸片报告与影像学证据之间潜在不一致性的自动检测问题,即在缺乏人工干预的情况下,如何准确识别报告描述与图像分析结果之间的矛盾。其核心解决方案是构建一个基于生成式AI(Generative AI)的智能校验系统——Rad-Linter,该系统结合预训练的图像分类器提供的九种病理学得分与经过LoRA微调的15亿参数语言模型,对报告内容与图像评分的一致性进行判断。关键创新在于利用微调后的语言模型理解报告中病灶描述的语义细节,从而超越传统关键词与否定规则基线在复杂表述上的局限性。实验采用受控的报告-评分配对(来自OpenI数据集),通过反事实四元组设计确保每份报告和评分向量同时出现在一致与不一致条件下,使仅依赖单一输入的基线模型准确率上限为50%。结果显示,在131个四元组上,Rad-Linter达到98.5%的配对准确率,显著优于规则基线的80.5%,且在正确分类全部四对组合方面表现更优(94.4% vs. 56.5%)。尤其在需语义理解的发现描述类型上,差距更为明显;而未经微调的语言模型性能仍接近随机水平(约50%),凸显微调对任务成功的关键作用。在包含标注分歧的第二组181个四元组中,该系统仍保持93.4%的准确率,高于规则基线的75.3%,进一步验证了其在真实复杂场景下的有效性。

链接: https://arxiv.org/abs/2611.26492
作者: Yanda Cheng,Bo Wen,Yuhao Lu
机构: Cornell University (康奈尔大学); University of Rochester (罗切斯特大学)
类目: Computation and Language (cs.CL)
备注: 34 pages, 4 figures, including supplementary material

点击查看摘要

Abstract:Rad-Linter checks chest X-ray reports against image-derived evidence to flag potential errors for review. A pretrained image classifier supplies scores for nine pathologies, and a 1.5B-parameter language model, fine-tuned with LoRA on 1,816 examples, checks their consistency with the report. We compare it with a keyword-and-negation rule baseline given the same reports and scores. Evaluation uses controlled report-score pairings from OpenI, with consistency labels derived from automatic report labels and score thresholds. Each counterfactual quad crosses two real reports with two real score vectors from other patients, creating two consistent and two inconsistent pairs. Every report and score vector occurs in both conditions, so either input alone permits only 50% accuracy. On 131 quads, Rad-Linter achieves 98.5% pair accuracy versus 80.5% for the rule baseline, and classifies all four pairs correctly in 94.4% of quads versus 56.5%. The checkers perform similarly on omission items, including unmentioned and negated findings; larger differences occur on the remaining item types, which require interpreting how findings are described. The same language model without fine-tuning remains near 50% accuracy. On a second set of 181 quads that also includes reports on which the automatic labelers disagree, pair accuracy is 93.4% versus 75.3%. These controlled results show that, given the same image-derived scores, a trained language model detects report-score inconsistencies more accurately than the tested rule baseline, with the advantage concentrated in interpreting how findings are worded.

[NLP-6] MemFolder: Structure-Guided Memory Folding for Efficient Agent Memory Management

【速读】: 该论文旨在解决长期智能体(long-term agents)在持续交互过程中积累海量外部记忆(external memory)所引发的存储、检索与响应生成效率下降的问题。核心挑战在于:记忆条目数量增长导致检索开销上升,而较长的检索上下文又加剧预填充计算与解码阶段的注意力开销,二者相互耦合,形成性能瓶颈。现有方法通过剪枝或合并记录来压缩记忆,但前者可能丢失有效证据,后者则可能导致合并后的单元过长,难以高效处理。为此,本文提出一种结构引导的记忆折叠框架 MemFolder,其关键在于通过共享的记忆分组(memory groups)实现双重优化:一方面利用跨记录的结构信息指导证据合并,另一方面生成紧凑表示并借助验证反馈进行自修正。具体而言,RecordFolder 基于实体提及识别,逐步选择具有代表性和互补性的锚点,并结合语义与实体亲和性及时间链式关系,将兼容记录合并为更少的检索单元;ContextFolder 则在这些分组内对原始记录进行紧凑编码,并通过关键事实恢复失败反馈动态修正编码内容。实验在三个文本型记忆基准(LoCoMo、PersonaMem、LongMemEval)和两个多模态基准(M3-Bench-web、M3-Bench-robot)上验证了该方法的有效性。结果显示,在 Mem0 测试集上,RecordFolder 在仅保留 50% 记忆条目的情况下仍维持 99.0% 的平均性能,且相比最强基线提升 1.9 个百分点;结合 ContextFolder 后,现有压缩方法可进一步缩短输出长度并保持高性能保留,显著提升了记忆系统的整体效率与可扩展性。

链接: https://arxiv.org/abs/2611.26427
作者: Xuyang Liu,Chenwen Lu,Junxi Wang,Siyu Ren,Xinying Lin,Siteng Huang,Hongli Sun,Linfeng Zhang,Honggang Chen
机构: Sichuan University(四川大学); Fudan University(复旦大学); Sun Yat-sen University(中山大学); Zhejiang University(浙江大学); Shanghai Jiao Tong University(上海交通大学)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: Code is available at this https URL

点击查看摘要

Abstract:Long-term agents accumulate large external memory stores as interactions unfold, increasing the costs of storage, retrieval, and response generation. Efficiency is constrained by two coupled factors: a growing record count increases retrieval overhead, while longer retrieved contexts increase prefill computation and decoding-time attention costs. Existing methods prune or consolidate records, but pruning can discard useful evidence, whereas consolidation may produce fewer yet longer retrieval units. We introduce MemFolder, a structure-guided memory folding framework that sequentially reduces record count and context length. The framework comprises two core components that share memory groups, using cross-record structure to guide both evidence consolidation and compact representation. RecordFolder grounds entity mentions, progressively selects representative yet complementary anchors, and combines semantic and entity affinity with temporal chaining to consolidate compatible records into fewer retrieval units. ContextFolder encodes the original records within these groups into compact text and uses verification feedback to revise encodings when key facts cannot be recovered. Experiments on the three text-based memory benchmarks LoCoMo, PersonaMem, and LongMemEval and the two multimodal benchmarks M3-Bench-web and M3-Bench-robot demonstrate the strong performance of our methods. Notably, on Mem0, RecordFolder retains 99.0% of the original average performance across the three text-based benchmarks with 50% fewer records, exceeding the strongest baseline by 1.9 percentage points in performance retention. Combined with ContextFolder, existing record compression methods further shorten their output records while retaining most of their performance.

[NLP-7] HiSA: Sparsifying the Hidden Dimension of RoPE Attention in LLM s

【速读】: 该论文旨在解决大语言模型(Large Language Models, LLMs)中注意力机制计算效率低下的问题,特别是全量点积(full dot product)带来的高昂计算开销。现有大多数稀疏注意力方法通过在序列长度维度(token sparsity)上减少注意力计算,而本文提出一种新型的隐藏稀疏注意力(Hidden Sparse Attention, HiSA),其核心创新在于在隐藏(特征)维度实现稀疏性——即在不改变查询(query)和键(key)投影结构的前提下,依据位置信息采用静态、无参数的规则仅计算部分点积。该方法基于旋转位置编码(Rotary Position Embedding, RoPE)的结构特性,将查询与键分解为可被舍弃的部分,从而显著降低计算量。实验表明,使用HiSA从零训练的LLM不仅保持甚至提升了性能,同时仅需计算约三分之一的查询-键点积;在推理阶段,该方法显著提升解码速度并减少内存占用。此外,作者还将HiSA扩展至FlashAttention-3内核,支持高效训练与推理。研究进一步揭示,传统密集注意力会自发抑制那些被HiSA舍弃的注意力分量,验证了该方法的合理性。

链接: https://arxiv.org/abs/2611.26363
作者: Ambroise Borbely,Mathieu Salzmann,Olivier Verscheure
机构: EPFL(瑞士联邦理工学院); Swiss Data Science Center(瑞士数据科学中心); ETH Zurich(苏黎世联邦理工学院)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Attention in large language models (LLMs) scores every query against every key with a full dot product. We investigate whether this dot product must be computed completely for attention to be effective. We propose Hidden Sparse Attention (HiSA), which keeps the query and key projections intact but only computes a partial dot product, following a static, parameter-free rule set by the token positions: attention becomes sparse in the hidden (feature) dimension, an axis much less explored than token sparsity, rather than over tokens as in almost all sparse attention methods. Based on the structure of rotary position embeddings (RoPE), we decompose keys and queries into parts which can be omitted when computing attention. By training LLMs from scratch with HiSA, we find that this partial dot product does not hurt but improves performance compared with the complete product while computing only about a third of the query-key products. During inference HiSA improves decoding speed and shrinks memory footprint. We also extend HiSA to FlashAttention-3 kernels for training and inference. Finally, we motivate our approach by showing that dense attention learns to suppress the attention score parts that HiSA drops. Code is available at this https URL.

[NLP-8] Beam Model Technical Report

【速读】: 该论文旨在解决大模型在代码生成、推理及代理任务中计算效率低、推理成本高且安全性与可控性不足的问题。其核心挑战在于如何在保持前沿性能的同时,显著降低推理阶段的计算开销,并提升对有害请求的拒绝能力。解决方案的关键在于:首先,采用稀疏混合专家(Mixture-of-Experts, MoE)架构,实现501.3B总参数量下仅23.4B活跃参数的高效计算;其次,通过创新的两阶段中段训练(midtraining phase)实现100万令牌上下文窗口,结合可验证、正确性锚定的数据进行预训练;第三,构建大规模异步强化学习(Reinforcement Learning, RL)系统,在10.5K GB300 GPU上运行4周,生成超1亿次轨迹,显著提升了模型在复杂任务中的表现并持续受益于增加的RL算力;最后,依托软硬件协同设计、FP8/NVFP4量化技术以及一个19.3亿参数的推测解码起草器(DFlash 2),实现了高效的部署与高吞吐服务。实验表明,Beam在各类评估中对有害请求的拒绝/转向率达到88.9%,同时对无害请求的合规率为81.5%,是当前开源权重模型中在帕累托前沿上表现最优的安全性与性能平衡者。

链接: https://arxiv.org/abs/2611.26358
作者: Reflection Team:Adam Abdulhamid,Osama Abuelsorour,Yasmin Alameddine,Carlo Alfano,Duarte Alves,Brandon Amos,Selvaraj Anandaraj,Arseni Anisimovich,George Antoniadis,Ioannis Antonoglou,Sheila Babayan,Neeraj Bahl,Märt Bakler,Vidhisha Balachandran,Cameron Banks,Thibaut Barroyer,Navroop Bath,Irwan Bello,Jeremy Berman,Fernanda Bernaldo,Mrityunjay Bhadauria,Sanyam Bhutani,Tolga Bolukbasi,Anudhyan Boral,Joseph Brennan,Ethan Brooks,Sam Buchanan,Anton Bulyenov,John Cai,Catherine Cang,Eason Chan,Elijah Chanakira,Nathan Chandrasekaran,Hao Chen,Hong-You Chen,Jeffrey Chen,Saurav Chhatrapati,Ting-Rui Chiang,Brian Tiger Chow,Aakanksha Chowdhery,Timothy Chung,Rob Claire,John D. Co-Reyes,Anthony Conte,Eduardo Coronado-Montoya,Julien Crozon,Drew Csillag,Joe Cummings,Jeff Da,Samira Daruki,Tim Dawson,Charlie Deck,Arash Delijani,Ousmane Dia,Randy Dodgen,Ryan Douglas,Omar El-Kishky,Ben Ellis,Gamaleldin Elsayed,Simon Fan,Patrick Fernandes,Gabriel Ferns,Nick Fram,Ankush Garg,Florent Gbelidji,Iaroslav Geraskin,Behrooz Ghorbani,Ivan Glushkov,Neil Gollapudi,Scott Goodfriend,Yoni Gozlan,Christopher Grimm,Suriya Gunasekar,Shruti Gupta,Joel Gustafson,Emre Guven,Jon Hackmann,Jeremy Hadidjojo,Jess Hamrick,Ethan Hellman,Joscha Hoche,Spurthi Amba Hombaiah,Uzair Honda,Zijian Hu,Dzmitry Huba,Nancy Hung,Vladimir Ivanov,Shomik Jain,Mojan Javaheripi,Kevin Martin Jose,Aymen Kallala,Amog Kamsetty,Min Hyung Kang,Steven Kapturowski,Ramtin Keramati,Samuel Kessler,Sabina Kim,Seonghyeon Kim,Ivan Komarov
机构: Reflection Team(反射团队)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Beam is a sparse Mixture-of-Experts (MoE) model with 501.3B total parameters and 23.4B active parameters, specifically built and optimized for coding, reasoning, and agentic workloads. Beam delivers frontier level reasoning, while requiring 3-4x less inference compute than many models of its class. Pretrained on 23.8T diverse, curated, high-quality tokens, Beam features a 1M-token context window achieved through a novel two-stage midtraining phase, built on verifiable, correctness-anchored data. Post-training leveraged a large-scale, fully asynchronous reinforcement learning (RL) campaign across 10.5K GB300 GPUs, generating over 100M rollouts over 4 weeks of training. We believe this is one of the largest scale RL runs conducted by any open lab to date. Across our evaluation suite, capabilities continued to improve as we increased RL compute, with no sign of a plateau. Supported by deep hardware-model co-design, FP8/NVFP4 quantization, and a 1.93B-parameter speculative decoding drafter (DFlash 2), Beam maximizes serving throughput with efficient, flexible deployment. Averaged across conversation turns and harm categories, Beam refuses or pivots on 88.9% of harmful requests while complying with 81.5% of harmless requests, achieving the highest refusal/pivot rate on harmful requests among open-weight models on the Pareto frontier.

[NLP-9] One Framework to Diagnose Them All: A Stratified Benchmark for Bengali LLM Hallucination WWW EACL2027

【速读】: 该论文旨在解决低资源语言中幻觉(hallucination)研究碎片化的问题,即每一种新识别出的错误模式都催生独立的数据集,导致标注资源分散且不可比。其核心解决方案是通过大规模社区协作实现基准统一化,提出Alikbachan——一个针对孟加拉语的幻觉检测基准,该基准在三个交叉维度上进行分层:经人工验证的文化距离带、六类任务类型以及涵盖五大家族32种模式的幻觉分类体系。该基准包含两个阶段的数据集:第一阶段为采集的2,516个样本测试集,第二阶段为按配额构建并由黄金标注的5,000个样本集。分层设计具有诊断意义:零样本大语言模型(LLM)判别器在不同分层间的表现差异超过30个宏观F1点,且99个第一阶段与32个第二阶段参赛系统在个体样本上的预测结果重现了这一分布特征——话语层面的遗漏是判别器和多数团队共同面临的挑战,而对负向控制样本的误报率则普遍较低。尽管判别器的置信度评分在项目排序上表现良好(ROC-AUC 0.74–0.95),但存在明显过度自信问题(ECE 0.04–0.29)。在开放竞赛中,共有222支队伍提交3,566次提交,单提示32B规模判别器达到第57百分位。值得注意的是,在两阶段均参与的25支队伍中,24支在第二阶段性能显著下降(平均-17.6分),而所有零样本判别器均表现出提升,且第一阶段排名与第二阶段排名相关性极弱(Spearman ρ=0.35),表明仅依赖第一阶段排行榜将得出错误的优胜者判定。因此,本研究的关键在于通过结构化、可扩展且具备跨域诊断能力的统一基准,推动低资源语言幻觉评估的标准化与可比性。

链接: https://arxiv.org/abs/2611.26336
作者: Asif Shahriyar Sushmit,Shefayat E Shams Adib,Abdullah Al Jubaer Gem,Ahmed Shafin Ruhan,MD. Nazmuddoha Ansary,Biplab Kumar Sarkar,Syed Rifat Raiyan,Ajwad Abrar,Abdullah Bin Hasan,Adib Sakhawat,M. Sakeef Hossain,Ahmed Alfey Sani,Arefin Mizan,Sourab Chakma,Quazi Adibur Rahman Adib,Wassie Haque,Sanjay Kumar Sen,Saad Tahmid,Mohammad Mamun Or Rashid
机构: Islamic University of Technology (伊斯兰科技大学); Bengali.AI (孟加拉人工智能); North Dakota State University (北达科他州立大学); Bangladesh University of Engineering and Technology (孟加拉国工程技术大学); Institute of Policy Dynamics (政策动态研究所); Mawlana Bhashani Science and Technology University (穆拉纳·巴沙尼科学与技术大学); BRAC University (BRAC大学); Jahangirnagar University (贾汉吉尔纳加尔大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to Findings of the Association for Computational Linguistics: EACL 2027, 25 pages, 8 figures, 23 tables. Code: this https URL | Data: this https URL | Competition: this https URL | Project page: this https URL

点击查看摘要

Abstract:Hallucination research in low-resource languages fragments easily: each newly named failure mode prompts its own dataset, spending scarce annotators on incomparable resources. We argue for consolidation and test it at community scale. Alikbachan is a Bengali hallucination-detection benchmark stratified along three crossed axes: a human-validated cultural-distance band, six task types, and a five-family, 32-mode hallucination taxonomy. It comprises a harvested 2,516-item test set (Phase 1) and a quota-constructed, gold-labeled 5,000-item set (Phase 2). The stratification is diagnostic. Four zero-shot LLM judges vary by over 30 macro-F1 points across strata, and the item-level predictions of 99 Phase-1 and 32 Phase-2 competition systems reproduce the profile: discourse-level omission defeats judges and most teams alike, whereas only a minority over-flag negative controls. The judges’ verdict probabilities rank items well (ROC-AUC 0.74-0.95) but are overconfident (ECE 0.04-0.29). An open competition on Alikbachan drew 222 teams and 3,566 submissions, and a single-prompt 32B judge reaches the 57th percentile. Of the 25 teams in both phases, 24 regress on Phase 2 (mean -17.6 points) while every zero-shot judge improves, and Phase-1 rank barely predicts Phase-2 rank (Spearman \rho=0.35 ), so a Phase-1 leaderboard alone would have crowned a different winner.

[NLP-10] Language-Oriented Semantic Modeling for Target Language Extraction

【速读】: 该论文旨在解决多语言语音混合场景下目标语言语音提取(Target Language Extraction, TLE)的难题,现有方法主要依赖带有语言身份标签的声学建模,仅提供粗粒度的类别指引,难以充分捕捉目标语言的语义特征。其解决方案的关键在于提出一种面向语言的语义建模框架,通过引入语言特定的语义表征来增强声学提取过程。具体而言,利用预训练的语音语义模型,结合语言感知提示(language-aware prompt)与参数高效微调技术,从混合语音中生成聚焦于目标语言的高阶语义表示。这些语义表示能够捕捉声学信号中的高层语义结构,并作为结构化语义引导融入提取模型,超越传统语言身份标签的限制。实验在CommonVoiceMix数据集上验证了该方法的有效性,相较于仅依赖声学信息的基线模型,显著提升了语音分离性能,最高实现1.16 dB的SI-SDR增益。

链接: https://arxiv.org/abs/2611.26317
作者: Mehmet Sinan Yıldırım,Ke Zhang,Shuai Wang,Haizhou Li
机构: National University of Singapore (新加坡国立大学); The Chinese University of Hong Kong, Shenzhen (香港中文大学(深圳)); Nanjing University, Suzhou (南京大学苏州校区)
类目: Computation and Language (cs.CL); Audio and Speech Processing (eess.AS)
备注: Published in APSIPA ASC 2026

点击查看摘要

Abstract:Target Language Extraction (TLE) aims to isolate speech in a target language from multilingual mixtures. Existing approaches primarily rely on acoustic modeling with language identity labels, which provide only coarse categorical guidance. In this work, we propose a language-oriented semantic modeling framework that augments acoustic extraction with language-specific semantic representations. A pre-trained speech semantic model is adapted via language-aware prompt conditioning and parameter-efficient fine-tuning to generate target-language-focused semantic representations from the mixture. These capture high-level semantic structure of the acoustics and they are integrated into the extraction model to provide structured semantic guidance beyond identity labels. Experiments on the CommonVoiceMix dataset show that incorporating semantic representations consistently improves over acoustic-only baselines, achieving up to 1.16 dB SI-SDR gain.

[NLP-11] BADAS-3.0: Real-Time Driving Perception and Planning from One Camera

【速读】: 该论文旨在解决单目摄像头在实时延迟约束下实现高精度自动驾驶决策与预测的挑战,核心问题是如何仅依赖单一前向摄像头,在低于50毫秒的实时响应时间内,完成对当前车速、已行驶路径、碰撞风险评分及未来6.4秒内六种候选轨迹的联合预测。其解决方案的关键在于提出BADAS-3.0——一个拥有10亿参数(1B-parameter)的视频编码器,配备四个轻量级输出头,通过高效架构设计实现了对单路视频流的多任务端到端建模。该模型在未训练过的多摄像头传感器配置上展现出显著优势,表明其具备更强的泛化能力;而在共享的相同传感器配置下性能与更大规模模型(如34B参数的Alpamayo 2 Super)相当。为更真实评估模型安全性,研究引入两个开环安全基准测试:基于真实行车记录仪的碰撞与近事故数据集,以及欧洲新车安全评鉴协会(Euro NCAP)典型场景库,结果表明BADAS-3.0在两项测试中的通过率均优于所有对比的公开发布规划器,验证了其在复杂安全场景下的鲁棒性与实用性。

链接: https://arxiv.org/abs/2611.26294
作者: Roni Goldshmidt,Hamish Scott,Lorenzo Niccolini
机构: Empiric Earth( empiric earth)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:We ask how far a driving model can get with one camera and real-time latency. BADAS-3.0 is a 1B-parameter video encoder with four small heads. From two seconds of one forward camera, it predicts the current speed, the metric path just driven, a collision-risk score and six candidate futures over 6.4 seconds, in under 50 ms. Given the car’s current speed, BADAS-3.0 has 44% lower minADE _6 over five trajectory benchmarks than the 34B Alpamayo 2 Super, which reads six cameras and runs over 100x slower. The gain comes from camera rigs that neither model was trained on. On the rig they share, the two are on par. Existing benchmarks are mostly nominal driving, so we add two open-loop safety benchmarks, one of real dashcam collisions and near-misses and one of Euro NCAP scenarios. On both, BADAS-3.0 has a higher pass rate than every published planner we tested.

[NLP-12] wice-Told Critiques: Claim-Level Homogenization of Peer Review in the LLM Era

【速读】: 该论文旨在解决学术同行评审中评阅意见多样性下降的问题,特别是评估生成式人工智能(Generative AI)介入对评审意见重叠度的影响。其核心发现表明,近年来顶级会议(如ICLR、NeurIPS、ICML)及期刊(如F1000、PLOS)的评审意见重叠率持续上升,例如ICLR的评审重叠率从2026年的12.0%增至2027年的21.3%,且新增评审所贡献的新问题数量逐年减少,反映出评审意见趋同趋势加剧。为探究这一现象的成因,研究引入MELD(一种检测大语言模型生成文本的工具)对评审内容进行分析,结果显示被标记为可能由大语言模型生成的评审之间,其内容重叠程度显著高于未被标记的评审:在ICLR 2024–2025年期间高出3.0分,在2026–2027年期间高出4.5分,跨五种期刊来源平均高出5.4分。这一结果揭示出,被识别为大语言模型生成的评审意见在语义和内容上更趋一致,从而削弱了多评阅人之间的批判性差异,表明生成式AI的广泛应用正在实质性降低同行评审中应有的多样性与独立性。因此,解决方案的关键在于通过技术手段(如MELD)识别并区分由大语言模型辅助或生成的评审文本,以恢复和保障评审过程的独立性与多样性。

链接: https://arxiv.org/abs/2611.26286
作者: Ilya Gusev
机构: Keenable(凯能); Amsterdam(阿姆斯特丹)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Digital Libraries (cs.DL)
备注:

点击查看摘要

Abstract:Peer review relies on several readers finding different problems. We measure critique overlap in conference and journal reviews. At ICLR, measured overlap rose from 12.0% in 2026 to 21.3% in 2027, in reviews from the same stage of the review process, with year-on-year rises at NeurIPS, ICML and the F1000 and PLOS journals. A third ICLR review of a paper added 5.46 new problems to the other two in 2027, against 6.12 in 2026. We apply MELD, a detector of LLM-generated text, to all reviews. Within a paper, two reviews that MELD flags overlap more than two unflagged reviews: by 3.0 points at ICLR in 2024-2025, by 4.5 points in 2026-2027, and by 5.4 points across five journal sources. These results provide evidence that LLM-flagged reviews of one paper are less distinct from each other than unflagged reviews, and suggest that the diversity of critiques among reviews of a single paper is falling.

[NLP-13] What If Chain-of-Thought Is Not Enough?

【速读】: 该论文旨在解决大语言模型(Large Language Models, LLMs)在推理过程中依赖提示工程(prompting-based methods)导致的不可靠性和可解释性不足问题,探索是否可通过训练使模型内部的思考模式(internal thinking model)发生可控且有益的改变。其核心解决方案是引入一种名为“问题-选项-标准”(Question–Option–Criteria, QOC)的结构化决策流程,通过将任务分解为明确的问题、竞争性选项和评估标准,实现对模型内部推理过程的显式引导与优化。研究提出了一套三阶段发展范式,从基础的QOC决策核心逐步演进至具备答案综合与统计增强能力的推理系统,并通过实验证明:相较于未经微调的基础模型,基于QOC的微调显著提升了输出质量,尤其在模型规模足够大、能够稳定执行QOC计算时,性能增益尤为显著。结果表明,大语言模型的内部思考模式具有可训练性,而QOC等结构化推理框架为实现更可靠、可解释且任务适配的决策支持提供了有效路径。

链接: https://arxiv.org/abs/2611.26282
作者: Marc Jansen
机构: University of Applied Sciences Ruhr West (鲁尔西应用技术大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Reasoning capabilities in large language models have evolved from prompting-based methods to internalized model behaviors. In this paper, we investigate whether the internal thinking model of an LLM can be deliberately changed and whether such a change is beneficial. Instead of relying on Chain-of-Thought, we train models to follow Question–Option–Criteria (QOC), a structured decision process that decomposes a task into explicit questions, competing options, and evaluation criteria. We present a three-generation development pipeline that ranges from a basic QOC decision core to richer answer synthesis and statistically enriched QOC reasoning. We then evaluate the resulting models with respect to both internal consistency and output quality. Across the reported experiments, QOC-based fine-tuning improves answer quality relative to non-fine-tuned base models and yields particularly strong gains once models are large enough to execute the underlying QOC computations reliably. The results suggest that the internal thinking model of an LLM is trainable and that structured alternatives to Chain-of-Thought can provide a promising path toward more reliable, interpretable, and task-adapted decision support.

[NLP-14] dLoop: Looped Diffusion Language Models via Recurrent Latent Refinement

【速读】: 该论文旨在解决扩散语言模型(Diffusion Language Models, DLMs)在生成过程中因依赖大量去噪步骤而导致效率低下的问题。尽管DLM具备并行生成能力,但其每步仅能可靠地提交少量词元(token),主要受限于当前去噪前向传播中提示信息、已提交词元与未提交位置之间的信息整合不充分,导致部分预测尚未充分优化即被决策提交,从而影响生成质量。现有加速方法多聚焦于优化置信度学习或提交策略,但这些方法仅作用于单次去噪输出,无法从根本上改善未提交位置的潜在表示。为此,本文提出dLoop,将预训练的DLM重构为“前奏-循环-尾声”(Prelude-Loop-Coda)架构,并通过参数共享的中间模块对同一部分掩码状态进行多次递归计算。不同于传统去噪推进,每次循环实质上是进一步执行潜空间计算,使分散于未提交位置的信息持续交互并更充分融合,从而提升提交前的表示质量。实验表明,dLoop在GSM8K、MATH、MBPP和HumanEval四个基准测试中均达到最高或并列最高准确率;在考虑循环计算开销后的每前向传递词元数(Tokens Per Forward Pass, TPF)指标下,其性能优于Fast-dLLM、dParallel及d3LLM,在准确率与TPF双指标上均无其他方法能同时超越。机制分析揭示,额外循环主要提升缓慢且不稳定的生成轨迹表现,且未提交位置的更新驱动了后续词元决策的变化。因此,提交前的潜空间计算显著增强了预测可靠性,为DLM加速提供了一种区别于通过训练损失强制提高置信度的新范式。

链接: https://arxiv.org/abs/2611.26275
作者: Handa Li,Pengxiang Li,Mingyu Cao,Xilu Wang,Li Shen,Da Li,Lu Yin
机构: 未知
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Diffusion language models (DLMs) support parallel generation, yet they often require many denoising steps because only a fraction of token predictions are sufficiently reliable to be committed at each step. Existing acceleration methods primarily optimize confidence learning or commitment policies. However, these approaches operate on the outputs of a single denoising forward pass, in which information from the prompt, previously committed tokens, and uncommitted positions may not yet be fully integrated. Consequently, some predictions remain under-refined when the commitment decision is made, and modifying the commitment policy alone cannot improve their underlying representations. We introduce dLoop, which restructures a pretrained diffusion language model into a Prelude-Loop-Coda architecture and recurrently applies a parameter-shared intermediate block to the same partially masked state. Rather than advancing the denoising process, each additional loop provides further latent computation, allowing information distributed across uncommitted positions to continue interacting and become more fully integrated before commitment. Experiments on GSM8K, MATH, MBPP, and HumanEval show that dLoop achieves the highest or tied-highest accuracy on every benchmark. With tokens per forward pass (TPF) adjusted for loop computation, it outperforms Fast-dLLM in both accuracy and TPF on all four benchmarks and dParallel on three, and is more accurate than d3LLM on all four, while d3LLM attains higher TPF; no compared method surpasses dLoop in both. Mechanistic analyses show that extra loops mainly help on slow, unstable trajectories and that updates at uncommitted positions drive changes in token decisions and later generation. Pre-commitment latent computation thus makes predictions more reliable, offering a route to DLM acceleration distinct from forcing confidence through the training loss.

[NLP-15] STAGE: Spatio-Temporal Augmented Graph Expansion for Weakly Supervised Cross-Tweet Event Causality Discovery

【速读】: 该论文旨在解决灾难相关社交媒体中因果关系跨越独立推文导致跨推文因果发现难以实现规模化监督与建模的问题。其核心挑战在于,灾害事件的因果链条常分散在多个语义关联但时空分离的推文中,传统依赖局部标注的方法难以捕捉全局因果结构。为此,论文提出STAGE框架,其解决方案的关键在于将可扩展的跨推文弱监督构建与事件发生的时间-空间关系学习解耦:首先基于局部因果种子检索跨推文的语义相关事件实例,并通过因果评分器生成弱标签的跨推文配对;这些配对作为监督信号,驱动一个基于多头图注意力网络(multi-head Graph Attention Network)的模型,该模型在语义增强的事件表示和时空上下文图上进行联合学习。实验表明,相较于仅使用局部监督,所构建的跨推文监督使加州山火数据集的AUPRC提升0.340,哈维飓风数据集提升0.107;在域内测试中,山火任务获得0.885的AUROC,飓风任务达0.710。跨灾种实验揭示了非对称迁移现象,凸显模型对数据集特异性时空偏移的高度敏感性。结果验证了大规模跨推文弱监督与关系化事件建模相结合在碎片化灾情报告中实现因果学习的有效性。

链接: https://arxiv.org/abs/2611.26255
作者: Hieu Minh Duong,Todd Gary,Eugene Levin,Long Nguyen
机构: University of Louisville (路易斯维尔大学); Meharry Medical College (梅纳里医学学院)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Causal relations in disaster-related social media of- ten span independent posts, making cross-tweet causal discovery difficult to supervise and model at scale. We propose STAGE, a framework that separates scalable cross-tweet weak-supervision construction from spatio-temporal relational learning over event occurrences. Starting from local causal seeds, STAGE retrieves semantically related event occurrences across tweets and uses a causal scorer to construct weakly labeled cross-tweet pairs. These pairs supervise a multi-head Graph Attention Network over semantically enriched event representations and a spatio- temporal context graph. We evaluate STAGE on California Wildfire and Hurricane Harvey using human-gold cross-tweet test sets and 15 baselines. Compared with local-only supervision, the constructed cross-tweet supervision improves STAGE AUPRC by 0.340 on Wildfire and 0.107 on Hurricane. STAGE achieves in-domain AUROC of 0.885 on Wildfire and 0.710 on Hurricane. Cross-disaster experiments reveal asymmetric transfer, highlighting sensitivity to dataset-specific spatio-temporal shifts. These results demonstrate the value of scalable cross-tweet supervision combined with relational event modeling for causal learning across fragmented disaster reports.

[NLP-16] winLab-Bench: Evaluating AI Scientists in Agent ic Twins of Real Labs

【速读】: 该论文旨在解决当前AI代理在科学实验中自主设计与执行实验能力评估的空白问题,尤其聚焦于第三类开放性能力——即自主设计实验并可靠执行的能力。现有基准主要局限于对计算工作流正确执行及从预定义候选集中选择下一步测量的能力评估,而无法真实检验代理在物理或生命科学实验室中自主设计并完成实验的能力。为此,本文提出TwinLab-Bench,一个面向物理科学、生命科学和工程领域52个真实实验任务的基准测试平台,支持超过100种实验操作类型。其核心创新在于引入“孪生实验代理”(TwinLab agent)作为环境执行者,能够基于真实实验室记录读取原始材料、调用由历史数据构建的计算器,并准确执行由被测代理设计的实验流程。通过与隐藏的真实实验记录比对,对代理报告进行评分,从而实现对自主实验设计能力的严格评估。相比仅依赖任务描述的语言模型,TwinLab代理显著减少了与真实记录偏差的读数及重复请求不一致的答案,且通过闭卷探测排除了记忆召回的影响。实验结果表明,即使是最先进的模型如Claude Fable 5.2,在该基准上平均得分仅为45.0,揭示了当前前沿模型在自主实验设计方面仍面临巨大挑战,研究进一步揭示了其失败的关键环节。

链接: https://arxiv.org/abs/2611.26248
作者: Yang Wan,Linchao Zhu
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: this https URL contains tasks and code. 33 pages, 6 figures, 14 tables. LLM-agent benchmark for autonomous experimental design in digital twins of real labs (physics, chemistry, wet-lab biology, materials, engineering). 13 agents evaluated: Claude Fable 5.2 45.0, Opus 5.5 40.2, GPT-6.1-sol 31.7, plus DeepSeek, Qwen, Kimi, GLM and others

点击查看摘要

Abstract:AI agents are increasingly expected to act as AI scientists, which requires three increasingly open abilities: carrying out a given computational workflow correctly, choosing the next measurements from a given design space, and designing their own experiments. Existing benchmarks mainly evaluate the first two abilities, in a terminal or in a closed loop that selects from predefined candidates. Evaluating the third ability requires reliably executing an experiment that an agent designs, with readings grounded in a real laboratory. We introduce TwinLab-Bench, which evaluates AI scientists on self-designed experiments in 52 tasks from laboratories in the physical sciences, the life sciences, and engineering. The benchmark supports more than 100 experimental action types, and the agent’s report is scored by comparison with a hidden answer taken from real laboratory records. Each task has its own TwinLab agent, an agent that serves as the environment and executes the experiments that the tested agent designs, reading the raw materials of the laboratory and calling calculators that are built from its records. Compared with a language model that sees only the task description, the TwinLab agent almost eliminates readings that deviate from the real records and inconsistent answers to a repeated request, and closed-book probes control for answers recalled from memory. We find that these tasks remain hard for frontier models: the strongest, Claude Fable 5.2, averages a score of only 45.0, and we conduct an empirical study of where it fails.

[NLP-17] Frontier models need only one text to unmask its human or AI author

【速读】: 该论文旨在解决生成式内容溯源(authorship attribution)问题,即在无外部工具、候选列表或样本支持的零样本条件下,判断一段文本是由人类还是大模型撰写,并进一步精确识别作者身份。其核心挑战在于,随着生成式AI(Generative AI)模型在训练数据中广泛覆盖公开写作内容,模型不仅具备强大的语言模仿能力,还能通过内部表征对作者风格进行高精度反推。解决方案的关键在于:前沿模型(如Claude Fable 5.2、GPT-6.1 Sol、Gemini 4 Carbon和Opus 5.5)利用其在大规模预训练数据中对特定作者写作风格的隐式学习,结合推理过程中的自我感知机制(例如询问模型对某作者的熟悉程度),实现对作者身份的精准识别。实验表明,这些模型在识别公开同行评审文章作者方面准确率达39%(Nature期刊子集达78%),并能以69%准确率关联活跃Reddit用户的用户名,揭示了当前生成式AI在匿名性保护方面的严重漏洞。该研究强调,模型的识别准确率高度依赖于其在预训练阶段对目标作者写作风格的暴露程度以及推理深度,且现有防御手段极为有限,标志着生成式AI对数字身份隐私构成前所未有的威胁。

链接: https://arxiv.org/abs/2611.26216
作者: Atrey Desai,Deniz Böloni-Turgut,Ty Brassington,Minjoon Choi,Jenna Russell,Rishanth Rajendhran,Bradley Emi,Hila Gonen,Hal Daumé III,John Wieting,Marzena Karpinska,Mohit Iyyer
机构: University of Maryland(马里兰大学); Simon Fraser University(西蒙菲莎大学); Pangram Labs; University of British Columbia(不列颠哥伦比亚大学); Google DeepMind(谷歌深度思维)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Given a single text and a simple zero-shot prompt (no web access, tools, candidate lists, or writing samples), frontier models can answer “who wrote this?” at three increasingly specific levels: (1) whether an AI or human wrote it, (2) if an AI, which lab, and (3) if a human, often exactly who. The latest models (Claude Fable 5.2, GPT-6.1 Sol, Gemini 4 Carbon) match trained detectors like Pangram while also explaining their predictions in natural language, and Opus 5.5 names the correct lab for 98% of texts generated by models from Anthropic, OpenAI, and Google. Most strikingly, Fable 5.2 correctly names the author of 39% of public peer reviews published in Oct-Nov 2026 (78% on a subset of reviews from Nature), and identifies the username behind 69% of comments from active Redditors (with 500 posts over the last year). Accuracy depends on exposure to an author’s writing in pretraining data and on reasoning effort, and simply asking a model how familiar it is with someone predicts whether it can identify their writing. For anyone who has published enough writing online, frontier models present a greater threat to anonymity than commonly assumed, and defenses are limited.

[NLP-18] Learning to Think with Images Faithfully via Adaptive Data Selection

【速读】: 该论文旨在解决生成式视觉语言模型在采用“以图像思维”(Thinking with Images, TwI)策略时出现的工具使用不忠实问题,即模型虽频繁调用缩放工具,却并未真正依赖所聚焦的细粒度视觉区域进行推理,导致推理过程缺乏可解释性与可靠性。其核心解决方案在于动态调整训练数据以匹配模型能力的演化:通过构建一个数据合成管道,精准定位小物体并设计可验证的细粒度问题,确保正确缩放后的证据对模型作答仍具关键作用;同时引入自适应奖励机制,在无需缩放时鼓励模型放弃工具使用,从而强化工具使用的必要性与真实性。实验表明,该方法显著提升了推理的忠实性,并在视觉搜索基准(如V*、HR-Bench 4K/8K)及多模态幻觉检测基准上均达到领先性能,证明了通过优化训练数据即可同时提升准确率与推理可信度,无需直接监督推理过程。

链接: https://arxiv.org/abs/2611.26199
作者: Yu Zhao,Jiaxin Zhang,Rohit Saxena,Fazl Barez,Pasquale Minervini,Yiwen Guo
机构: University of Edinburgh(爱丁堡大学); LIGHTSPEED; University of Oxford(牛津大学); Miniml.AI; Independent Researcher
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Thinking with images (TwI) empowers vision-language models (VLMs) to overcome perceptual bottlenecks by actively cropping and zooming into fine-grained visual details, promising both higher accuracy and interpretable, region-grounded reasoning. However, our analysis shows that the improved accuracy of TwI models does not guarantee faithful tool use; they frequently invoke tools without actually relying on the zoomed region, undermining the reliability of tool-assisted reasoning. We trace this unfaithful behaviour to the training data: when questions are answerable without zooming, the model can answer correctly while ignoring the returned evidence, and this problem is exacerbated during training as the model becomes better at answering without tools. To keep zooming useful throughout training, we adapt the training data to the model’s evolving capability and train on questions where correct zoomed evidence can still help the model answer correctly. This selection is supported by a data synthesis pipeline that localises small objects and constructs verifiable questions about their fine details, and complemented by an adaptiveness reward that favours tool-free answers when zooming is unnecessary. The experimental results show that our method substantially improves reasoning faithfulness and achieves state-of-the-art performance on visual search benchmarks (+11.6% on V*, +7.9% on HR-Bench 4K, and +9.4% on HR-Bench 8K). On multimodal hallucination benchmarks, our model also outperforms the compared TwI baselines. These results suggest that improving TwI faithfulness does not sacrifice accuracy and that both can be learnt from the right training data without directly supervising the reasoning process.

[NLP-19] Can World Models Predict Rice Growth? Toward Agricultural World Models

【速读】: 该论文旨在解决农业场景下长期作物生长状态的预测问题,即如何基于单张植物图像、其品种信息及拍摄视角,准确预测作物在未来数周至数月后的形态变化(如水稻是否已抽穗、成熟时外观如何)。现有生成式图像模型在该任务中表现不佳,主要因未能充分理解植物当前发育阶段,导致生成结果在生长量、颜色或形态上与实际生长阶段不符。其核心解决方案是提出一种农业世界模型(AgroWM),其关键在于引入“发育推理—生成”双阶段框架:首先通过名为PhenoClock的推理模块,基于输入图像推断当前生长阶段并预测未来阶段,生成包含高度、植株面积和颜色等数值目标的“生长脚本”;随后由“根植渲染器”(Rooted Renderer)模块依据该脚本对原始图像进行精准编辑,添加新器官与颜色变化。为避免模型倾向于输出与原图无差别的结果,采用基于真实后续图像的性状与发育时序验证机制。实验表明,在水稻和玉米两种不同作物上,AgroWM在多个评价指标上显著优于主流生成模型及领域特定基线,尤其在9至20周的长期预测中仍保持优越性能,验证了其在农业世界建模中的有效性。

链接: https://arxiv.org/abs/2611.26145
作者: Wenkai Ji,Zeteng Li,Yinkun Guo,Jie Zhou,Liang He,Zhenong Jin
机构: East China Normal University (华东师范大学); Peking University (北京大学)
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:World models can imagine how a physical scene evolves over the next few seconds, but growers need to know how a crop will develop over weeks to months: will this rice plant have headed a month from now, and how will it look at harvest? We introduce agricultural world modeling, the task of predicting how the same plant will look at a requested horizon of days to months, from one photo of it, its cultivar, and its camera view. In preliminary analysis, even state-of-the-art image generators get this wrong: they often grow the plant too little or give it the color and form of the wrong growth stage. We argue that they do not work out the plant’s developmental state, which the photo reveals only indirectly. We therefore propose AgroWM, which reasons about development before it draws growth. Its reasoning module, PhenoClock, works out the current growth stage, predicts the future one, and writes a Growth Script with numeric targets for height, plant area, and color; its generation module, the Rooted Renderer, then edits the photo to follow this script, adding new organs and colors to the photographed plant. Because pixel scores can prefer an unchanged copy of the photo, we instead check each forecast’s traits and developmental timing against the plant’s own later photos. On rice, 54.6% of the forecasts of AgroWM are matched to the target stage (Stage R@1), against 34.8% for GPT Image 2.5, 24.0% for Qwen Image 2.1, and 47.4% for the strongest domain-specific baseline, the same backbone fine-tuned without Growth Scripts; 9 to 20 weeks ahead, it still reaches 46.4%, against 35.7% for that baseline. On maize, a crop with a different body plan and stage scale, the same design again scores best on all four measures.

[NLP-20] RL Can Only Reinforce What It Sees Succeed

【速读】: 该论文旨在解决在已知规则失效时,强化学习(Reinforcement Learning, RL)系统如何在先验知识与新观测之间进行权衡的问题。其核心挑战在于:当正确行为在训练中极为罕见时,标准的基于结果奖励的强化学习难以有效捕捉并推广这一行为。论文提出的关键解决方案是,通过分析在特定任务中成功执行正确动作的“尝试组”(attempt groups)的预期数量来预测模型是否能够成功学习该行为。研究发现,当预期的获胜尝试组数量约为35时,模型有约50%的概率成功学习到该行为;这一阈值在不同问题类型、重复实验、尝试组规模(4至256次)、学习率、优化器、近端策略优化(GRPO)变体及模型规模(1.3M至85M参数)下均保持稳定,仅影响阈值位置而不改变计数逻辑。因此,该方法的关键在于以“预期成功尝试组数”作为可预测的先验指标,指导对稀疏奖励环境下强化学习性能的评估与设计。

链接: https://arxiv.org/abs/2611.26084
作者: Ruben Sanchez-Corcuera,Unai Zulaika,Aritz Bilbao-Jayo,Miguel Fernandez-de-Retana
机构: University of Deusto(德乌斯托大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 16 pages, 6 figures, 1 table

点击查看摘要

Abstract:When a familiar rule stops working, an agent has to weigh what pre-training taught it against what it now observes. We ask how outcome-reward reinforcement learning (RL) changes this balance, in a puzzle world where both sides are known: small language models trained from scratch on games played by a Bayesian player, so that the prior and the right answer are exact. Base models hedge as their corpus did, and RL turns that hedging into decisions, in whichever direction the reward favours. But RL can only reinforce what it sees succeed. When the right action is rare, each run either learns it everywhere or not at all, and which runs take off is predicted before training by one count: the expected number of groups of attempts in which the action wins. About 35 such groups give even odds. The count holds for new and repeated problems and for groups of 4 to 256 attempts, and across learning rates, optimizers, GRPO variants and model sizes from 1.3M to 85M parameters, which move the threshold but not what is counted.

[NLP-21] Same Judgement Different Scores: Numerical Scoring Topology in LLM -as-a-Judge Evaluation EACL2027

【速读】: 该论文旨在解决生成式 AI(Generative AI)作为自动评估者(LLM-as-a-Judge)在实际应用中因评分方案设计不合理而导致评估结果不可靠的问题。其核心挑战在于,现有研究普遍忽视了评分体系的结构特性对评估一致性与有效性的影响。为此,论文提出“数值评分拓扑”(numerical scoring topology)这一概念,从量表宽度、粒度、偏移、对称性及边界位置等维度系统刻画评分方案的结构性特征。基于此,研究首次在多个基准测试和不同规模的判别模型上开展系统性实验,发现数值评分拓扑显著影响评分结果、校准性能与鲁棒性,尤其在依赖主观判断的任务中表现突出;其中,偏移且范围狭窄的评分设计易引发最大幅度的评分偏差。进一步研究表明,更强、更大的判别模型通常具备更高的区分效度和对评分拓扑变化的鲁棒性。因此,论文强调应将数值评分拓扑视为 LLM 作为评判者评估流程中的关键组成部分,并呼吁建立更加严谨、标准化的评分协议以提升评估质量。

链接: https://arxiv.org/abs/2611.26077
作者: Doyun Kim,Chanwoo Kim,Jiwoo Kum,Minsoo Song,Chanjun Park
机构: Soongsil University (中央大学)
类目: Computation and Language (cs.CL)
备注: Accepted to Findings of EACL 2027

点击查看摘要

Abstract:LLMs are increasingly used as automated evaluators to reduce the cost and time required for human annotation. However, the role of scoring-scheme design in the reliability of LLM-as-a-Judge evaluation remains underexplored. To address this problem, we introduce numerical scoring topology, which characterizes the structural design of a scoring scheme in terms of its scale width, granularity, offset, symmetry, and boundary placement. Building on this definition, we conduct the first systematic study of how numerical scoring topology influences zero-shot LLM-as-a-Judge evaluation across diverse benchmarks and judge models. Our results show that numerical scoring topology systematically affects judge scores, calibration, and robustness, particularly in judgment-intensive evaluations. Shifted and narrow ranges tend to induce some of the largest score deviations across many settings. We further find that stronger and larger judge models generally achieve higher discriminative validity and greater robustness to scoring-topology changes. These findings suggest that numerical scoring topology should be treated as an integral component of LLM-as-a-Judge evaluation and highlight the need for more principled and standardized scoring protocols.

[NLP-22] Understanding MoE Decisions Beyond Gradients

【速读】: 该论文旨在解决生成式 AI 模型中专家混合(Mixture-of-Experts, MoE)架构下,传统基于梯度的方法无法准确预测模型内部计算变化对答案偏好影响的问题。其核心挑战在于,MoE 模型中某一专家的替换可能引发后续层专家选择的连锁变动,这种非线性切换效应无法通过一阶或二阶导数充分捕捉。为此,论文提出一种新型预测器,利用模型自身输出的专家激活结果及选择规则,显式建模这些由专家切换带来的额外影响,并将其整合至二阶预测框架中。该方法的关键创新在于引入了对专家选择动态的显式建模,显著提升了对答案得分变化的预测精度,在三个基准测试集上相较于传统多步梯度与二阶导数基线,分别在 Qwen3.6 和 Gemma4 上将预测误差降低了 60% 和 47%。

链接: https://arxiv.org/abs/2611.26075
作者: Zhen Xu,Zitao Zhang,Kevin Zhao,Xuefeng Liu
机构: University of Chicago(芝加哥大学); eBay(ebay); University of Florida(佛罗里达大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Understanding a language model’s decision requires knowing how its internal computations affect its preference between answers. Gradient-based methods estimate this influence by predicting the effect of changing those computations. In mixture-of-experts (MoE) models, replacing one expert can change which experts later layers select. Those switches can alter the answer scores in ways that derivatives do not capture. We introduce a predictor that uses the model’s own expert outputs and selection rules to estimate these extra effects, then adds them to a second-order prediction. Across questions from three benchmarks, our method reduces the error in predicting answer-score changes by 60% on Qwen3.6 and 47% on Gemma4, relative to a multistep baseline using gradients and second derivatives.

[NLP-23] AltRA-Test: An Alternative Annotator Test for LLM Judges Without Shared Annotators

【速读】: 该论文旨在解决在大规模标注场景中,如何科学评估生成式 AI (Generative AI) 是否可替代人类标注者的问题。传统方法“替代标注者检验(Alt-Test)”仅适用于所有样本由相同标注者群体完成的密集标注场景,而无法适应实际中样本由不同子集标注者的常见设置。为此,本文提出一种新的统计检验方法——重采样标注替代标注者检验(AltRA-Test),其核心创新在于引入基于重采样的自举法:从数据集中重采样项目,为每个项目抽取k个真实人类标注,并以一个来自大语言模型(LLM)的标注进行替换,从而构建配对的全人类与混合人类-LLM标注面板。当混合面板的标注者间一致性(Inter-Annotator Agreement, IAA)在统计上不劣于全人类面板时,即认为LLM能够保留人类标注者共享的信号,支持替代。研究证明了AltRA-Test在渐近水平α下具有一致性。通过模拟和真实世界实验(涵盖20个跨文本、图像、音频、视频的数据集,测试18个LLM及5种提示策略),发现约100个样本即可使该检验有效;在已知LLM优劣的模拟中,其决策准确性优于现有基线;而在真实任务中,约一半设置下LLM可通过检验,且是否通过主要取决于任务类型。因此,该研究强调:标注者替代并非普遍适用,建议研究人员针对具体任务进行验证。

链接: https://arxiv.org/abs/2611.26073
作者: Nagham Omar,Rotem Dror,Almog Gueta,Avigdor Gal,Roi Reichart,Nitay Calderon
机构: Technion(以色列理工学院); University of Haifa(海法大学); Google Research(谷歌研究)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Large language models (LLMs) are increasingly used as judges and graders to replace human annotators, raising the question of when such replacement is statistically justified. To support such decisions, the Alternative Annotator Test (Alt-Test) was introduced for dense settings where all items are annotated by the same annotators. However, in many large-scale annotation settings, items are annotated by different subsets of annotators. We therefore propose the Alternative Annotator Test with Resampled Annotations (AltRA-Test), a statistical test that does not require that all items are annotated by the same annotators. AltRA-Test uses a bootstrap that resamples items, draws k human annotations for each item, and substitutes one of them with an LLM annotation. This produces paired all-human and mixed human-LLM panels. A replacement is supported when the inter-annotator agreement (IAA) of the mixed panel is statistically non-inferior to the IAA of the all-human panel, i.e., when the LLM preserves the signal that the human annotators share. We establish that AltRA-Test is a consistent test of asymptotic level alpha. We then evaluate it through simulations and real-world experiments in which we test 18 LLMs and 5 prompting techniques on 20 datasets across text, image, audio, and video. Across these evaluations, about 100 items often suffice for the test. In simulations, where we know whether the LLM is a better or a worse annotator than the humans, AltRA-Test makes more correct replacement decisions than existing baselines. In real-world data experiments, LLMs pass the test in about half of the setups, depending mainly on the task. Since replacement is not always justified, we encourage researchers to test it for their task.

[NLP-24] SheetAlign: Grounded Interaction and Stage-Isolated Learning for Spreadsheet Agents

【速读】: 该论文旨在解决在保护敏感数据和控制部署成本的前提下,实现电子表格分析的自动化问题。其核心挑战在于复杂电子表格任务需要同时具备任务推理与工作簿理解能力,且传统预定义操作接口增加了工具选择与调用的负担。为此,论文提出SheetAlign框架,其解决方案的关键在于采用分阶段隔离的策略学习机制,围绕“在何处操作”“修改什么内容”以及“如何实施转换”三个维度进行结构化建模。通过任务引导的检查与API锚定,将代码生成过程与工作簿观测结果及可用操作相耦合;利用分阶段学习优化程序生成与审查过程,并基于编辑进度奖励进行训练;在推理阶段,通过执行一致性机制(即多个工作簿状态间的共识)筛选最优程序。实验表明,该方法在SpreadsheetBench上将Qwen3-4B的准确率从13.5%提升至29.2%(无需重新训练),结合学习与执行一致性后,平均准确率达到32.8%,在Domain-Spreadsheet上达30.1%,显著优于现有基线方法。

链接: https://arxiv.org/abs/2611.26069
作者: Yuxiang Wang,Junhao Gan,Jianzhong Qi
机构: The University of Melbourne(墨尔本大学)
类目: Computation and Language (cs.CL); Software Engineering (cs.SE)
备注: conference paper preprint

点击查看摘要

Abstract:Spreadsheets are widely used for data analysis and supporting decision-making. Automating spreadsheet analysis while protecting sensitive data and controlling deployment costs motivates locally deployed open-weight large language models (LLMs). Meanwhile, complex spreadsheet tasks require both task reasoning and workbook understanding, while predefined operation interfaces impose additional tool-selection and invocation demands. We introduce SheetAlign, a spreadsheet manipulation framework with stage-isolated policy learning, organized around where to operate, what to change, and how to implement transformations. Task-guided inspection and API grounding anchor code generation in workbook observations and available operations, while stage-isolated learning optimizes program generation and review using edit-progress rewards. At inference, execution consistency selects programs through agreement among workbook states. On SpreadsheetBench, our grounded task interface improves Qwen3-4B’s accuracy from 13.5% to 29.2% on average without training. With learning and execution consistency, SheetAlign achieves 32.8% accuracy on average on SpreadsheetBench and 30.1% on Domain-Spreadsheet, outperforming comparable baselines. The code will be made publicly available upon publication.

[NLP-25] Complete-SVD: A Functional Norm Approach to Low-Rank LLM Compression

【速读】: 该论文旨在解决大语言模型在后训练低秩压缩过程中,如何在保持模型输入-输出行为一致性的同时,有效逼近原始权重的问题。现有输出感知方法虽通过在模型输出端度量压缩误差并加权对数几率(logit)变化,但其权重分配依赖于观测到的下一个词或模型预测概率,可能导致偏差。本文提出Complete-SVD方法,通过最小化原始模型与压缩模型间对数几率的L₂距离,并对所有对数几率变化进行等权重处理,从而更全面地捕捉输出层面的误差。其核心创新在于将网络在原始权重附近线性化,并结合克罗内克积分解近似,将优化目标转化为具有闭式解的加权奇异值分解(SVD)。由于输出侧因子难以精确计算,研究采用无偏随机探针估计,仅需每序列K次反向传播,且计算开销与词汇量和序列长度无关。实验表明,在相同秩设置及WikiText-2校准条件下,Complete-SVD在所有配置下均取得最低的C4和PTB困惑度;在LEMS秩分配比例为0.7时表现最优。进一步采用KL散度替代L₂距离(即KL-SVD),在多数设置中降低校准困惑度并提升下游任务准确率,但略微增加跨语料库困惑度。此外,Complete-SVD在IFEval指令遵循能力评估中持续优于当前最强的输出感知方法。

链接: https://arxiv.org/abs/2611.26059
作者: Alper Kalle,Theo Rudkiewicz,Mohamed-Oumar Ouerfelli,Mohamed Tamaazousti
机构: 未知
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 23 pages, 3 figures, 15 tables

点击查看摘要

Abstract:Post-training low-rank compression of neural networks must not only approximate the original weights but also preserve the model’s input-output behavior. Recent output-aware methods measure the compression error beyond the compressed layer, at the model output, weighting logit changes by the observed next token or by the model’s predicted probabilities. We propose Complete-SVD, a method for large language models that minimizes the L_2 distance between the logits of the original and compressed models, weighting all logit changes equally. Linearizing the network around the original weights, together with a Kronecker-factored approximation, reduces this objective to a weighted SVD with a closed-form solution. Its output-side factor is intractable to compute; we estimate it without bias from random probes, at a cost of K backward passes per sequence, independent of vocabulary size and sequence length. At matched settings with uniform ranks and WikiText-2 calibration, Complete-SVD achieves the lowest C4 and PTB perplexity in every setting, and at ratio 0.7 under LEMS rank allocation. Replacing the L_2 distance by the Kullback-Leibler divergence (KL-SVD), with the same estimation framework, lowers calibration perplexity in most settings and raises downstream accuracy, but raises cross-corpus perplexity. Complete-SVD also improves IFEval instruction-following accuracy over the strongest prior output-aware methods throughout.

[NLP-26] Evaluating safeguarding in multi-turn conversations with AI assistants

【速读】: 该论文旨在解决当前对话式人工智能(Conversational AI)在处理用户心理危机信号时存在的安全评估偏差问题。现有安全评估通常依赖于明确、集中呈现的紧急求助信息,但现实中用户往往通过日常对话逐步透露情绪困扰,如抑郁、精神分裂或家庭暴力等,这种渐进式表达模式未被现有评估体系充分覆盖。研究的关键在于揭示模型在不同信号呈现方式下的真实响应能力:通过与三位社区心理健康支持工作者合作,设计九个包含渐进或集中式线索的危机情境,对比六款主流大模型在7400余次对话中的表现。结果显示,即使在信号集中的情况下,多数模型(如GPT-5.5、Grok 4.3等)仅在26%-59%的对话中达到专业标准;当线索分散于多轮对话中时,性能骤降至3%-6%,表明其对隐性、非直接表达的敏感度严重不足。相比之下,Claude系列表现更优,但即便最佳模型Claude Opus 5在分散线索场景下仍未能达标13%的对话。因此,研究提出的核心解决方案是:必须重构安全评估范式,从依赖显性、孤立的危险信号转向模拟真实人际交流中逐渐浮现的心理危机表达模式,以实现对实际使用情境下风险识别能力的真实衡量。

链接: https://arxiv.org/abs/2611.26009
作者: Charlotte Peart,Franziska Sofia Hafner,Luc Rocher
机构: Oxford Internet Institute, University of Oxford (牛津互联网研究所,牛津大学)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Conversational AI assistants are now used for personal and emotional support. Developers report improvements in how new models respond to sensitive conversations, yet users and civil society continue to raise concerns about their behaviour in real-world, repeated interactions that safety evaluations rarely capture. Often, evaluations present signs of distress clearly and in a single message, even though people may reveal their struggles gradually in otherwise ordinary conversations. We worked with three community mental health support workers to test if models provide the same level of safeguarding when cues are shared all at once or gradually over time. We designed nine scenarios in which a user shares cues related to depression, psychosis, or domestic abuse. For each scenario, the support workers established a gold standard response, ranging from signposting support to pausing the everyday task to address safeguarding. We presented five cues either together in one message or dispersed across a conversation. Across more than 7,400 conversations, we evaluated six widely used models from five developers. GPT-5.5, Grok 4.3, DeepSeek V4, and Llama 4 Maverick met the support workers’ standards in just 26-59% of conversations when cues were concentrated, and only 3-6% when dispersed. Claude Sonnet 5 and Claude Opus 5 showed a smaller gap, meeting the standards in 68-97% of concentrated conversations and 59-87% of dispersed conversations. Even the best-performing model, Claude Opus 5, fell short of the gold standard in 13% of conversations when cues were dispersed. Safety evaluations centred on explicit, legible concerns can overstate how well safeguards protect people when signs of distress emerge gradually or indirectly through ordinary conversations. For claims of improved safety to be taken seriously, developers must design and assess safeguards around how people actually communicate.

[NLP-27] Xing-Audio-NVVASR: Scaling and Refining Supervision for Joint Recognition of Speech and Non-Verbal Vocalizations

【速读】: 该论文旨在解决语音与非语言声音事件(Non-verbal Vocalizations, NVVs)联合识别中因高质量标注数据稀缺、标签不准确或缺失所导致的性能瓶颈问题。其核心解决方案是提出一种基于渐进式数据挖掘与迭代监督优化的端到端识别框架——Xing-Audio-NVVASR。关键在于通过统一的声学标准,利用两个异构模型进行交叉审计,自动标记标注分歧并触发教师重新标注;同时引入互补的呼吸检测器对全语料库进行筛查,以发现模型一致性可能掩盖的遗漏事件,从而提升标注质量。经过五轮迭代,累计完成17万条语音的标注精修,并融合63万条挖掘样本,构建了约80万条标注语料,用于训练一个17亿参数的统一模型,可实现语音转录并内嵌六类NVV标签。在四个人工标注基准测试中,该模型在综合得分和六类宏平均F1上均优于九个对比模型,综合得分达64.46,较最强基线StepAudio 3 Chat Preview提升6.03分,且在呼吸事件识别的F1值上超越标注教师14.40至31.57个百分点,显著验证了其有效性。

链接: https://arxiv.org/abs/2611.26000
作者: Pengyu Xu,Gang Liu,Genliang Zhao,Hongjie Chen,Jian Kang,Jie Lian,Jie Li,Yongxiang Li
机构: Xingchen AGI Lab, China Telecom Artificial Intelligence Technology Co. Ltd(中国电信人工智能技术有限公司)
类目: Computation and Language (cs.CL)
备注: 18 pages, 2 figures

点击查看摘要

Abstract:Joint recognition of speech and non-verbal vocalizations (NVVs) is hindered by scarce event-rich training data and inaccurate or missing event labels. We present Xing-Audio-NVVASR, an end-to-end recognizer trained through progressive data mining and iterative supervision refinement. Guided by unified acoustic criteria, two heterogeneous models cross-audit annotations, flagging disagreements for teacher reannotation. To uncover shared omissions that model agreement may conceal, complementary breath detectors screen the entire corpus for human review. Human-confirmed labels accumulate across rounds and are preserved during subsequent refinement. Over five rounds, we refine annotations for 170,000 utterances from public corpora and incorporate 630,000 mined utterances. The resulting corpus of approximately 800,000 utterances is used to train a single 1.7B-parameter model that transcribes speech with inline tags for six NVV classes. Across four human-annotated benchmarks, Xing-Audio-NVVASR ranks first among nine models in both composite score and six-class Macro-F1. It achieves a mean composite score of 64.46, surpassing the strongest comparison model, StepAudio 3 Chat Preview, by 6.03 points, and exceeds the annotation teacher’s breathing F1 by 14.40 to 31.57 percentage points. We plan to release the model weights, inference code, four evaluation benchmarks, annotation guidelines, and evaluation code.

[NLP-28] Policy Improvement-Guided Replay for Continual Reinforcement Learning

【速读】: 该论文旨在解决持续强化学习(Continual Reinforcement Learning, CRL)中模型在习得新任务能力的同时如何有效保持对历史任务性能的问题。传统方法多从稳定性视角出发,强调知识保留与干扰抑制,但该研究提出一种基于策略改进(policy improvement)的补充范式,通过强化学习的基本原理来维持已获得的性能增益。其核心解决方案在于利用性能差异引理(performance difference lemma),将历史任务性能变化建模为当前策略在历史任务分布上的期望优势,从而指导对历史提示(prompt)进行回放。为高效分配有限的回放内存,作者提出对比似然回放(Contrastive Likelihood Replay, CoLiR),该方法通过对比最终策略在正确回答上的似然增益与初始策略在错误回答上的似然下降,动态评估回放优先级。理论分析表明,正确响应的似然增益可关联至理想化KL正则化最优更新下的参考策略优势,并为一般策略更新提供内生奖励(endogenous reward)解释。CoLiR仅需少量跨策略评分和缓存的对数似然值即可完成评分,无需额外生成响应,在任务边界选择关键提示并均匀回放,同时将KL正则化限制于回放内容。实验在MRCL数据集上针对三个Qwen3-VL模型规模验证了该方法,在保证良好新任务表现的同时,显著提升了平均最终准确率与平均任务准确率,优于现有对比方法。

链接: https://arxiv.org/abs/2611.25989
作者: Qin-Wen Luo,Haoxuan Gao,Sheng Ren,Xiang Chen,Sheng-Jun Huang
机构: Nanjing University of Aeronautics and Astronautics (南京航空航天大学); Nanjing, China
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Continual reinforcement learning (CRL) requires models to acquire new capabilities while retaining performance on previous tasks. A stability-oriented perspective common in conventional continual learning emphasizes knowledge preservation and interference mitigation. Complementing this perspective, we revisit CRL for vision-language models through policy improvement, a fundamental principle of RL, to preserve acquired performance gains. Using the performance difference lemma, we characterize historical performance changes through the current policy’s expected advantages on historical task distributions, motivating replay of historical prompts with the current policy. To allocate limited replay memory, we propose Contrastive Likelihood Replay (CoLiR), which contrasts likelihood gains on correct responses from the final task policy with likelihood reductions on incorrect responses from the initial policy. Our analysis links correct-response likelihood gains to reference-policy advantages under an idealized KL-regularized optimal update and provides an endogenous-reward interpretation for general policy updates. CoLiR estimates scores from finite training rollouts and cached log-likelihoods, selects prompts at task boundaries, and replays them uniformly during subsequent training, with KL regularization restricted to replayed prompts. Memory construction only requires a small amount of additional cross-policy scoring but no additional response generation. Across three Qwen3-VL model sizes on MRCL, CoLiR achieves the highest mean final accuracy and mean task accuracy among the compared methods while maintaining competitive new-task performance.

[NLP-29] Energy estimates of open-weight large language models across stated food weights: a dose-response evaluation

【速读】: 该论文旨在解决开放权重大语言模型(LLM)在估算食物能量值时,其输出是否随给定食物重量呈比例缩放的问题,以及通过“能量密度×重量”计算能否消除估算偏差。研究发现,七款开放权重的LLM在对168种随机选取的食物在0.25至4倍标准份量的不同重量下进行能量估算时,均表现出显著的压缩效应(弹性系数介于0.548至0.979之间,所有95%置信区间低于1),其中四款模型压缩尤为严重(弹性系数为0.548–0.709)。这些模型在低估四倍份量时误差达20%–34%,而在高估0.25倍份量时误差高达47%–130%。关键解决方案在于采用两步法:首先由模型自行输出能量密度,再在代码中将其乘以用户声明的重量。该方法使其中四款模型中的三款恢复了近似比例缩放,且极端情况下的典型误差因子从1.47–2.43显著降低至1.20–1.38。研究还发现,即使在全精度和直接提问场景下,压缩现象仍持续存在,表明模型倾向于回归到典型食用份量的“平坦斜率综合征”(flat-slope syndrome)。因此,结论指出,食品记录工具应基于能量密度与重量相乘的方式计算能量值,而评估应针对固定食物在不同声明重量下的表现进行测试。

链接: https://arxiv.org/abs/2611.25976
作者: Donato Riccio,Fabrizio Maturo,Iulia Brezeanu
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 18 pages, 4 figures, 4 tables; supplementary material and TRIPOD-LLM checklist as ancillary files

点击查看摘要

Abstract:We tested whether energy estimates of open-weight large language models (LLMs) scale proportionally with an exactly stated food weight, and whether computing energy density times weight removes the error. Under a prespecified protocol, seven open-weight LLMs estimated the energy of 168 randomly drawn foods from the USDA Food and Nutrient Database for Dietary Studies at nine stated weights, from 0.25 to 4 times the default amount. The primary outcome was the elasticity of the estimate with respect to weight (1: proportional scaling). A two-step estimate multiplied each model’s own stated energy density by the weight in code. All seven models compressed their estimates, with elasticities from 0.548 to 0.979 (every 95% CI below 1); four compressed severely (0.548 to 0.709). These four overestimated a quarter of the default amount by 47% to 130% and underestimated four times it by 20% to 34%. Supplying the density restored near-proportional scaling in three of the four, and the two-step estimate cut the typical error factor at these extremes from 1.47-2.43 to 1.20-1.38. Compression persisted at full precision and grew in five of seven models under a plain question. The models hold the energy densities needed for proportional scaling and answer near a typical serving, reproducing the flat-slope syndrome of dietary recall. Accuracy at usual portions does not reveal it. Food-logging tools should compute the energy of weighed entries as density times weight, and evaluations should test fixed foods across stated weights.

[NLP-30] he Depth of a Prediction: Where Language Models Resolve a Word Predicts Human Reading Time

【速读】: 该论文旨在解决语言模型预测人类阅读行为(如眼动停留时间)时,如何更准确地捕捉模型内部表征动态变化的问题。尽管最终层的意外度(surprisal)已被证明能有效预测阅读时长,但模型在多层中逐步计算该预测,其层间动态信息尚未被充分利用。本文提出的解决方案关键在于引入“调谐透镜”(tuned lens),通过分析每一词在各层中的意外度轨迹,并以“解析深度”(resolution depth)——即意外度变化中心所在的层数——来总结该轨迹。研究发现,模型在较深层才完成对某词的解析(即解析深度较大),则该词及其后续词语的阅读速度显著变慢,即使其意外度与早解析的词相当。解析深度在多个英语眼动追踪和自定速阅读语料库上,相较于仅使用最终层意外度、熵或单一层的预测性能,平均提升0.26毫纳特/词(合并p=0.001),约相当于熵增益的三分之一,且在不同模型家族(GPT-2、OPT、Pythia)中均获得显著收益。进一步验证表明,该效应在校正词概率、排除首尾层差异后依然成立,且全轨迹函数回归未提供额外显著信息,说明解析深度已充分捕获核心动态特征。

链接: https://arxiv.org/abs/2611.25970
作者: Donato Riccio,Fabrizio Maturo
机构: Universitas Mercatorum, Rome, Italy(罗马美尔卡托大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 17 pages, 4 figures, 7 tables

点击查看摘要

Abstract:Surprisal from a language model’s final layer predicts how long people read a word, yet the model computes its prediction across many layers. We read each word’s prediction at every layer with the tuned lens and summarize this layerwise surprisal trajectory by its resolution depth, the depth around which its change is centered. On eight English eye-tracking and self-paced reading corpora (81,957 words) and ten GPT-2, OPT and Pythia models, words that a model resolves later, and the words after them, are read more slowly than equally surprising words. Resolution depth improves held-out prediction beyond final-layer surprisal, entropy and the best single layer by 0.26 millinats per word (pooled p = 0.001), about a third of entropy’s gain, and each model family gains, GPT-2 and Pythia through one corpus each. The net gain varies between corpora (random-effects mean 0.07, p = 0.37); against layer-shuffled depth, which adds as many coefficients, it is positive in seven of eight corpora (random-effects mean 0.23, p = 0.016), and untrained networks show no gain under the logit lens. Pooled over corpora, the effect holds with corrected word probabilities and beyond first-to-last-layer divergences, and a whole-trajectory functional regression adds no significant information beyond resolution depth.

[NLP-31] When Sounds Turn Heads: Recovering Event Timing That Audio LLM s Fail to Say

【速读】: 该论文旨在解决音频大语言模型(Audio LLMs)在时间定位任务中“听得到却说不准”的问题,即模型能够感知目标声音的存在及其时间位置,但其输出的时间戳往往不准确或遗漏关键事件。其核心解决方案是提出一种无需训练的**头引导解码(head-guided decoding)**方法:利用模型内部少数对查询声音具有跟踪能力的注意力头(attention heads),直接读取其激活模式作为先验信息,对模型生成的候选时间戳进行重评分。这些注意力头表现出对声音位置的敏感性——随查询变化而转移、随声音移动而追踪,且移除它们会显著损害时间定位性能。通过将这些头的输出作为外部先验引入解码过程,不仅提升了时间戳的准确性,还恢复了原始模型遗漏的事件。实验表明,该方法在多个基准测试中超越原始模型及随机头引导策略,并在模型训练后发布的录音上将平均交并比(mIoU)从0.07–0.36提升至0.48–0.56,证明了音频大模型的时序表现不佳主要源于“表达失败”而非“感知失败”,而模型自身的注意力机制可有效弥补这一缺陷。

链接: https://arxiv.org/abs/2611.25960
作者: Jongbhin Woo,Jaewook Jeong,Joon Son Chung
机构: KAIST(韩国科学技术院)
类目: ound (cs.SD); Computation and Language (cs.CL)
备注: 11 pages, 4 figures

点击查看摘要

Abstract:Audio LLMs can know when a queried sound occurs but fail to say it. We find a few attention heads that track the queried sound: they shift when the query changes, follow the sound when it moves, and removing or blocking them often hurts the model’s timestamps. Yet the answer does not express what they track. Intervals read directly from these heads outperform the model’s original answer and recover events that its timestamps miss. To bring this timing into the answer, we introduce head-guided decoding, which needs no training. It keeps the model’s decision on whether the sound is present and rescores its candidate timestamps with a prior from these heads. Heads selected on development data transfer unchanged to new recordings and free-text queries, where head-guided decoding improves the timestamps of three audio LLMs, scores above their head readout on every benchmark, and, in most settings, outperforms random-head guidance. On recordings released after the models were trained, head-guided decoding raises mIoU from 0.07-0.36 to 0.48-0.56. These results suggest that poor temporal answers in audio LLMs partly reflect a failure to say rather than to hear, and that the model’s own attention can close part of this gap.

[NLP-32] Web search in Latin America: language models use less local and diverse news than search engines

【速读】: 该论文旨在解决生成式 AI(Generative AI)驱动的搜索平台在信息获取中可能加剧地理偏见的问题,尤其关注拉丁美洲等全球南方地区在新闻可见性上的不平等现象。其核心问题是:当用户从传统搜索引擎转向基于大语言模型(LLM)的搜索时,关于拉丁美洲的信息呈现是否更倾向于发达国家媒体,从而弱化本地新闻源的多样性与可及性。解决方案的关键在于通过构建涵盖九个拉丁美洲国家、包含1,901个当前事件的多语言、多情境查询数据集,系统比较三类传统搜索引擎与四类大语言模型在不同用户身份(本地、侨民、游客、外国用户)和语言设置下的新闻来源分布。研究发现,相较于传统搜索引擎,LLM-based search 返回的新闻来源多样性显著降低,且对本地新闻的覆盖存在不一致性——尤其对本地用户而言,其返回的本土新闻比例比传统搜索引擎进一步下降3.6至11.6个百分点。这一结果表明,基于大语言模型的搜索系统可能不仅再现甚至放大了既有新闻传播中的地理层级结构。为此,研究团队发布了开放数据集,以支持未来在多语言和地理多样性背景下对大语言模型搜索系统的公平性评估,并呼吁在信息基础设施设计中引入更具包容性的机制。

链接: https://arxiv.org/abs/2611.25955
作者: Francisco W. Kerche,Franziska Sofia Hafner,Felix M. Simon,Luc Rocher
机构: University of Oxford(牛津大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: 23 pages with appendix, 3 figures

点击查看摘要

Abstract:AI platforms are becoming a major gateway to news and current events. This shift risks amplifying existing geographical biases that favour narratives from wealthier countries over those from the Global South. This is particularly relevant in Latin America, which accounts for the largest share of AI usage in the Global South. We examine how information about Latin America changes when users shift from conventional search engines to LLM-based search. We construct a dataset of 1,901 current events across nine Latin American countries, craft realistic queries, and submit them to three search engines and four language models. We vary query language and user location to represent local, diaspora, tourist, and foreign user scenarios. Across 49,207 requests and 8,834 unique domains, language models return a less diverse set of news sources than search engines. Across user types, the share of domestic news returned by language models is inconsistent. For local users, while search engines already return only 41.3% local news, language models return a further 3.6 to 11.6 percentage points less. Taken together, our findings suggest that LLM-based search can reproduce and, in some settings, amplify geographical hierarchies in the reach and visibility of news sources. We release an open dataset to support multilingual and geographically diverse evaluation of LLM-based search, and highlight implications for more equitable information systems design.

[NLP-33] Many Solutions One Answer: Search Order Shapes the Default Solutions of Language Models

【速读】: 该论文旨在解决生成式语言模型在求解具有多个正确解的约束谜题(constraint puzzle)时,表现出严重偏向单一默认解的问题。尽管存在多个有效解,模型生成的样本却高度集中于某一特定解,远偏离均匀采样分布,且不同模型间常共享此偏好倾向。其核心解决方案在于通过调整搜索策略与输入呈现方式来提升解空间的覆盖广度(breadth)。关键发现是:采用基于候选解数量最少优先的搜索顺序,或对问题变量进行随机化呈现,可显著提高模型对非默认解的探索能力;相较于固定值顺序或仅依赖指令引导,随机化处理在符号类谜题和多解ZebraLogic变体上分别使有效解覆盖率提升26至42个百分点和11至26个百分点,同时保持极低的有效性损失(最高仅3-4个百分点)。这表明,通过设计合理的搜索顺序与输入结构,可有效缓解模型固有的解偏好偏差,增强其在多解场景下的多样性生成能力。

链接: https://arxiv.org/abs/2611.25916
作者: Qinhan Hou,Jing Tang
机构: University of Helsinki(赫尔辛基大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 22 pages, 6 figures and 4 tables

点击查看摘要

Abstract:When a constraint puzzle has several correct answers, language models tend to give the same this http URL study four open-weight models from three families on symbolic puzzles and multi-solution ZebraLogic variants with exactly known solution this http URL samples concentrate on one default solution far beyond uniform sampling, and different models often share this http URL models usually produce even rarely sampled solutions when added clues uniquely identify this http URL held-out symbolic puzzles, we compare visiting entities in name order with visiting those with the fewest initial candidates first. Both searches try values in the stated order. Where their pre-specified predictions differ, the latter matches a larger share of responses in all four models, by 26 to 40 percentage this http URL the stated value order or specifying a search order shifts responses towards the solution that the new order favours, on both testbeds. Randomising the presentation or supplying a random search order raises breadth (the effective share of solutions in use) by 26 to 42 percentage points on the symbolic task and 11 to 26 on ZebraLogic. The largest observed validity losses are 3 and 4 percentage points, respectively. Randomised presentation broadens answers more than a random-choice instruction in each of the three main models.

[NLP-34] he Linguistic Structure of Dependency-Conditioned Embeddings

【速读】: 该论文旨在解决依赖关系驱动的词项分布表示中,依赖条件下的词义表示在全局与局部几何结构上组织规律不明确的问题。其解决方案的关键在于通过分析四种基于WikiText-103训练的词项-依赖关系嵌入表示(两种基于计数的PPMI配置与两种基于负采样的跳字模型SGNS配置),从全局与局部两个层面揭示其几何特性:在全局层面,通过依赖关系中心点分析并控制词形身份影响,考察其与词性(POS)分布的一致性;在局部层面,评估跨依赖关系的几何一致性、跨方法的一致性、词义相似性及多义用法的区分能力。研究发现,所有表示在全局尺度上均表现出由上下文词性主导的强结构化特征,而局部结构虽部分保留了广泛的词义关联,但邻域差异显著,反映出不同语义用法;此外,基于PPMI的表示在跨依赖关系上展现出比SGNS更高的几何一致性,且跨方法一致性随依赖关系类型而异。这些结果揭示了词义表示在多尺度几何结构中的系统性语言组织规律,并凸显了表示方法选择对几何结构敏感性的重要影响。

链接: https://arxiv.org/abs/2611.25909
作者: Bach Phan-Tat,Tung Kieu,Kris Heylen,Dirk Geeraerts,Stefano De Pascale,Dirk Speelman
机构: KU Leuven (鲁汶大学); Aalborg University (奥尔堡大学); Instituut voor de Nederlandse Taal (荷兰语研究所); UC Louvain (卢万大学)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Dependency-based distributional models capture syntactically informed lexical similarities, but the geometric organisation of dependency-conditioned lexical representations remains underexplored. We investigate the global and local geometry of lemma-dependency relation embeddings using four representations trained on WikiText-103: two count-based PPMI configurations and two Skip-Gram with Negative Sampling (SGNS) configurations. At the global level, we analyse dependency-relation centroids, control for lexical identity through within-lemma centring, and examine their alignment with part-of-speech (POS) distributions. At the local level, we evaluate cross-relation geometric consistency, cross-method agreement, lexical similarity, and polysemous usages. Results show that global geometry is structured strongly by context POS across all representations. Locally, broader lexical similarity structures are partially preserved across dependency relations, while immediate neighbourhoods vary substantially, highlighting different usages. PPMI-based representations exhibit greater cross-relation consistency than SGNS-based representations, with cross-method agreement varying by relation. These findings reveal systematic linguistic organisation at multiple geometric scales and its sensitivity to representation choices.

[NLP-35] Verifiable Logic in Medical LLM s: A Critical Review of Agent ic Clinical Reasoning Optimization and Neuro-Symbolic Grounding

【速读】: 该论文旨在解决临床大语言模型(Clinical Large Language Models)在评估过程中存在的根本性缺陷:当前以最终答案正确性为核心的静态基准测试无法识别模型通过无效推理生成的流畅但错误的输出,导致模型可能产生自信的幻觉(hallucination)、因果无效推断以及基于虚假人口统计学线索的捷径行为(spurious demographic shortcuts),从而威胁临床应用的安全性。其解决方案的关键在于从“结果导向”转向“过程可验证”的范式,具体包括四个核心方向:一是揭示临床幻觉的成因并指出静态指标的盲区;二是提出“代理幻觉级联”(agentic hallucination cascade)机制,阐明多智能体与工具增强系统如何放大早期推理错误;三是引入过程奖励建模(process reward modeling),实现对每一步推理的层级监督;四是采用神经符号融合(neuro-symbolic grounding),通过形式逻辑引擎、本体推理器和知识图谱对神经网络输出施加可验证的约束。其中,具备符号否决层(symbolic veto layers)的系统可在其编码领域内提供相对于已有知识的可靠性保证,并在安全指标上取得显著提升。然而,现有成果多来自单个研究且局限于模拟或小样本场景,难以直接与纯神经基线比较;过程奖励模型亦受限于训练数据的统计分布。因此,作者提出一个概念框架,将每种代理式失败模式映射至对应的符号干预节点,并强调临床AI安全高度依赖系统架构设计。鉴于现行监管仅关注性能表现而未要求推理步骤的可验证性,作者主张负责任的认证应同时审计推理轨迹与静态基准上的准确率,以实现更全面的安全保障。

链接: https://arxiv.org/abs/2611.25898
作者: Youssef ElNahas,Adel Ammar,Soumaya Chaffar,Wadii Boulila
机构: University of Prince Mugrin(王子穆格林大学); Prince Sultan University(王子苏丹大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Clinical large language models are usually judged by whether the final answer is right. Yet a model can reach a fluent, plausible output through invalid reasoning that static benchmarks do not detect, and confident hallucination, causally invalid inference, and spurious demographic shortcuts can pass conventional evaluation. This critical review covers four paradigms in turn: how clinical hallucination arises and why static metrics miss it; the agentic hallucination cascade, in which multi-agent and tool-augmented systems propagate and amplify early reasoning errors; process reward modeling as step-level supervision; and neuro-symbolic grounding, in which formal logic engines, ontology reasoners, and knowledge graphs place verifiable constraints on neural output. Within their encoded domains, systems with symbolic veto layers offer guarantees that hold relative to their encoded knowledge, and they report large gains on the safety measures they use. These results come from single studies, mostly in simulated or small-sample settings, and are not directly comparable with neural baselines. Process reward models remain bounded by the statistics of their training data. We propose a conceptual framework that maps each agentic failure mode to a symbolic intervention point, and we argue that clinical AI safety depends substantially on system architecture. Because current regulation emphasizes performance and does not require verification of reasoning steps, we argue that responsible certification should also audit reasoning trajectories alongside accuracy on static benchmarks.

[NLP-36] One Schedule Does Not Fit All: Adaptive Inference Schedules for Diffusion Language Models

【速读】: 该论文旨在解决扩散语言模型(Diffusion Language Models)在实际部署中因采用固定去噪调度(denoising schedule)而导致的计算资源浪费问题,尤其在不同任务、模型规模及采样器之间,最优去噪调度存在显著差异。其核心挑战在于如何在保证生成质量的前提下,动态选择最廉价(即最少去噪器评估次数)且可靠的调度方案。解决方案的关键是提出任务校准的去噪分配(Task-Calibrated Denoising Allocation, TCDA),一种具备显式误差保证的部署策略。TCDA通过联合使用单侧自助法(one-sided bootstrap bounds)对候选采样器进行校准,结合配对校准数据、效用容差与防退化约束,严格验证候选方案是否同时满足性能要求与稳定性保障。若存在符合条件的廉价方案,则部署其中成本最低者;否则回退至完整调度。该方法确保部署不可行调度的概率渐近不超过预设显著性水平α。实验表明,在多个数据集与模型架构上,TCDA可实现高达8.0倍的去噪器评估减少(如LLaDA-8B在WMT14上),且在500次校准分割中误认证率始终低于2%(名义水平5%)。研究还发现,解码令牌的“承诺度”(commitment)比剩余隐空间离散化误差更能有效指导调度选择,并揭示了原始参考采样器时间网格随机生成的问题,强调在无条件生成场景下为保证可复现性需固定时间网格。

链接: https://arxiv.org/abs/2611.25895
作者: Rabimba Karanjai,Linsey Pang,Hemanth Hegadehalli Madhavarao,Richard Williamson,Uma Kona,Prakhar Mehrotra
机构: PayPal Inc. (paypal.com)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Diffusion language models typically use a fixed denoising schedule across workloads. We show that the cheapest schedule that preserves quality depends on the task, model, and sampler. We propose Task-Calibrated Denoising Allocation (TCDA), a deployment rule with an explicit error guarantee. Given candidate samplers, paired calibration data, a utility tolerance, and guardrails against degeneration, TCDA uses simultaneous one-sided bootstrap bounds to certify which candidates meet both the utility and guardrail criteria. It deploys the cheapest certified candidate or falls back to the full schedule. The probability of deploying an infeasible schedule is asymptotically at most alpha. Candidates include early prefix decoding, state-reusing continuation, reduced time grids, and confidence-threshold decoding. Across seven datasets and two model families, TCDA certifies reductions in denoiser evaluations by factors of 3.4 on XSum and 2.0 on WMT14 for ELF-B; 8.0 on WMT14 and 4.1 on GSM8K for LLaDA-8B; and 1.33 on OpenWebText for both ELF-M and ELF-L. A validation protocol fixed before generation confirmed every selection on fresh seeds. Across 500 calibration splits, false certification remained below 2% in every setting at a nominal level of 5%. Certified schedules vary across tasks and model scales. Trajectory measurements suggest that when decoded tokens commit is more informative for schedule selection than the remaining latent discretization error. We also find that the ELF reference sampler draws its time grid at random, so reproducible schedule comparisons for unconditional generation require fixing the grid.

[NLP-37] HALTT: A HAte Lexicon Through Time

【速读】: 该论文旨在解决生成式AI在仇恨言论检测中随时间退化的问题,其核心挑战在于日常词汇被逐渐异化为隐晦的贬损性用语(即“多义性保留”现象),即这些词汇在保留原有中性含义的同时衍生出具有攻击性的新语义。现有词义演变检测(Lexical Semantic Change Detection, LSCD)框架因缺乏标准化、可区分语义层级的历时语料资源而难以有效应对这一问题。本文提出一种多阶段框架,通过整合1770–1964年历史语料与2001–2024年现代语料的频率证据,结合掩码定义级评分以缓解大语言模型(LLM)的时间偏差,并采用双角色推理提示范式增强判断可靠性,最终构建并验证了一个历时性词典——HALTT。该词典包含19个具有字面意义与贬损意义的多义词,揭示了词义演变的类型学特征,并建立了下游历时基准评估协议,为仇恨言论检测的动态适应提供了可复现的语义演化分析工具。

链接: https://arxiv.org/abs/2611.25862
作者: Dario Franzese,Farah Benamara,Véronique Moriceau,Viviana Patti
机构: Università di Torino(都灵大学); IRIT, Université de Toulouse(图卢兹大学信息与推理研究所)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Automated hate speech detection degrades over time as everyday vocabulary is weaponized into covert slurs while retaining original benign meanings (polysemy retention). Existing lexical semantic change detection (LSCD) frameworks struggle with this phenomenon due to a lack of standardized, sense-delineated diachronic resources. We propose a multi-stage framework to construct and validate a diachronic lexicon of semantically shifted hate speech terms. Filtering 20,000 candidates, our pipeline combines historical (1770-1964) and modern (2001-2024) corpus frequency grounding, masked definition-level scoring to mitigate LLM temporal bias, and a dual-persona reasoning prompt paradigm. We release HALTT, a grounded lexicon of 19 polysemous terms with literal and derogatory senses, analyze shift typologies, and establish evaluation protocols for downstream diachronic benchmarking.

[NLP-38] ableRoute: Capability-Aware Routing for Scalable Table Question Answering

【速读】: 该论文旨在解决大规模语言模型在表格问答(TableQA)任务中难以规模化的问题,核心挑战在于表格规模跨度大(从紧凑型到百万词元级),且问题要求对列、数值及计算过程进行精确定位。现有方法通常采用固定推理策略,导致在不同表格尺度和模型能力下表现脆弱:全表读取虽高效但易引入无关上下文;表格缩减依赖可靠的证据定位;执行式推理在超出模型上下文窗口时必需,却对模式与数值定位错误敏感。本文提出一种能力感知的路由框架——TableRoute,其关键创新在于通过验证后的问答-表格证据生成,识别所需操作,并根据序列化表格长度及基模型执行各操作的可靠性校准估计,动态路由至全表读取、表格缩减或执行三种策略。该方案通过证据条件化的推理机制,确保所选路径具备精准的列、值与条件定位支持,从而在跨尺度表格上实现准确率提升与在线令牌成本降低。

链接: https://arxiv.org/abs/2611.25853
作者: Yuxiang Wang,Junhao Gan,Jianzhong Qi
机构: The University of Melbourne(墨尔本大学)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Table question answering with large language models remains difficult to scale: tables range from compact to million-token sizes, while questions require precise grounding of columns, values, and computations. Existing methods often specialize in a fixed inference strategy, which is brittle across different table scales and model capabilities. Full-table reading is efficient for compact inputs but suffers from irrelevant context. Table reduction helps long inputs but depends on reliable evidence localization, while execution is required when tables exceed the model context window and remains sensitive to schema and value grounding errors. We propose TableRoute, a capability-aware routing framework with evidence-conditioned inference for scalable TableQA. TableRoute first generates validated question–table evidence and identifies the required operations, then routes each instance among these three strategies according to serialized table length and calibrated estimates of how reliably the base model handles these operations. The validated evidence supports the selected route by grounding columns, values, and conditions. Across WikiTQ, WTQ-L, BirdQA, and TableBench, TableRoute consistently improves accuracy through complementary routing and evidence conditioning across table scales, while reducing online token cost. Code will be made available upon publication.

[NLP-39] Anchor and Check: Deterministic Cross-Prompt Cross-Model Reconciliation for Auditable LLM Decisions

【速读】: 该论文旨在解决大语言模型(Large Language Model, LLM)在生成离散决策时缺乏可解释性与可复现性的问题,特别是在自动化决策管道中,如何确保模型的每个决策都有明确依据并可追溯。其核心挑战在于:在不依赖人工干预的前提下,如何准确识别并验证模型输出的可靠性,同时保证决策过程的透明性与可重复性。解决方案的关键在于设计一种形式化的“校验合约”(reconciliation contract),该合约通过固定目标模型的贪婪输出作为锚点(anchor),引入一个或两个来自不同路径的替代性检查(如另一模型族或同一模型在不同温度下的输出),并通过一个纯函数(pure function)对多个候选答案进行评估,仅当结果为“同意”(Agree)时才执行动作。该机制不仅提供了固定的承诺答案、显式的决策结果和可从记录中重放的决策过程,还显著提升了错误检测能力——相较于单纯使用语义熵聚合的方法,在相同预算下,跨模型族样本的聚合能将语义熵的AUROC提升1.7至4.0个百分点,而同模型温度检查则表现最弱。实验表明,三调用合约与3或5个样本的聚合语义熵在点估计上差异小于1点,置信区间包含零;在错误检测方面,该合约在3和5样本条件下优于聚合熵,而在10样本时差异微小且符号依赖于具体判别规则。在严格“仅同意”执行门控下,该系统可自动处理70%的项,同时仍能识别出16%的错误,体现出高自动化率与良好的可靠性平衡。

链接: https://arxiv.org/abs/2611.25818
作者: Sivakameswara Reddy Chejarla
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)
备注: 20 pages, 6 figures, 11 tables

点击查看摘要

Abstract:Pipelines that let a large language model (LLM) make discrete decisions need to know which decisions they may act on unattended, and a record that explains each one. We study a small, fully specified reconciliation contract: commit the target model’s greedy answer (the anchor), draw one or two greedy checks that reach the answer by a different route (an alternate prompt, or a model from another family), and let a pure function return one of five outcomes, of which only Agree is acted on. On 48,000 model-item evaluations over 6,000 questions, with every method given access to verbalized confidence as a tie-breaker, the main finding concerns where calls come from: pooling another family’s samples raises semantic entropy’s AUROC by 1.7 to 4.0 points at every budget, and a same-model temperature check is the weakest design. As complete pipelines, the point estimates of the three-call contract and pooled semantic entropy over 3 or 5 samples differ by less than one point, with intervals containing zero. As detectors of the anchor’s errors, the contract is ahead of pooled entropy with 3 and 5 samples under both tie-breaker conventions; with 10 samples the point estimates differ by less than one point and the sign depends on the convention. Under the strict Agree gate it automates 70% of items and still acts on 16% of errors. What the contract adds is a fixed committed answer, an explicit outcome and a decision replayable from its record; generation itself was not reproducible on the hosted API.

[NLP-40] What Does Anti-Hallucination Training Buy? Saying Less Cheap Pruning and the Rest

【速读】: 该论文旨在解决当前视觉语言模型中偏好训练(Preference Training)在降低对象幻觉(Object Hallucination)方面有效性评估的可信度问题。现有方法普遍通过报告更低的CHAI R评分来宣称性能提升,但其增益来源未被充分拆解:部分改善可能仅源于生成更简短的描述(即“说少”),而非真正提升了生成内容的真实性。为此,论文提出一种严格的量化分析框架,其关键在于构建一个固定、弱化的后验探测器(post-hoc probe)——该探测器基于500张标注图像上训练的11个视觉特征,且在分析前冻结所有评估标准,不依赖任何训练过程中的输出。该探测器被用来模拟偏好训练所消除的“非简洁性”幻觉成分。实验结果表明,在多个基准数据集(如COCO、AMBER、Object HalBench)上,该探测器可复现至少三分之二的训练增益(超出单纯缩短描述的部分),证明多数现有方法的改进很大程度上可归因于对冗余或错误信息的识别与抑制,而非复杂的模型优化。此外,研究还发现,使用探测器自身标签或基于物体级别的CLIP得分进行直接偏好优化(DPO),能有效恢复接近真实标注下DPO的性能增益(分别达0.70和0.87),进一步验证了探测器作为信号代理的有效性。然而,在InternVL2.5模型上,该会计体系失效,暗示其泛化能力受限。最终,人类评估揭示探测器删除的句子比截断策略删除的句子更常出错,说明当前自动评估指标(CHAIR)的精确性存在缺陷,需谨慎对待。因此,该研究的核心贡献在于揭示了当前幻觉缓解方法的实质增益来源,并提供了一种可复现、可验证的评估基准。

链接: https://arxiv.org/abs/2611.25814
作者: Lingkai Bu,Qian Gao
机构: Shandong Computer Science Center (National Supercomputer Center in Jinan); Qilu University of Technology (Shandong Academy of Sciences), Jinan, China; Key Laboratory of Computing Power Network and Information Security, Ministry of Education
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Preference training is the standard recipe for reducing object hallucination in vision-language models, and each new method reports a lower CHAIR score. We ask what that reduction consists of. Every trained system is compared with its base model’s caption truncated to the same object recall, which removes the part of the gain that comes from saying less; the remainder is compared with what a deliberately weak, frozen, eleven-feature post-hoc probe fitted on 500 annotated images can reproduce. All criteria were specified in the analysis code and frozen before the outputs they judge existed. For eight public methods on LLaVA-1.5-7B, the probe reproduces at least two thirds of what training removes beyond saying less for 7 of 8 methods on COCO, 8 of 8 on AMBER and 6 of 8 on Object HalBench, with RLAIF-V the consistent exception; a later public method released for both LLaVA-1.5 and Qwen2.5-VL (SENTINEL) meets the same bar on both. On our own DPO checkpoints the accounting holds on Qwen2.5-VL but fails on InternVL2.5 on COCO (2 of 4 training designs). At matched recall, the base model plus the probe beats truncation for 8 of 9 LLaVA-1.5 systems on COCO and 9 of 9 on AMBER, on the two newer backbones, and when realized as actual sentence deletion. The same measure prices training signals: DPO on the model’s own samples labelled by the probe or by an object-level CLIP score recovers 0.70 and 0.87 of the beyond-brevity gain of ground-truth-labelled DPO on Qwen2.5-VL (the probe 0.73 on InternVL2.5), both beat length-matched random labels, and CLIP labels buy more than probe labels on COCO. Human judgements show that the sentences the probe deletes are wrong more often than those truncation would delete; a human check of the CHAIR evaluator’s precision falls below our pre-set bar, which we report as a limitation.

[NLP-41] Efficient and Robust Drafting with Multi-Token Prediction

【速读】: 该论文旨在解决生成式 AI(Generative AI)在大语言模型(Large Language Model, LLM)推理过程中因自回归生成模式导致的延迟高、效率低的问题。其核心挑战在于设计一个轻量且准确的草稿生成器(drafter),以有效逼近目标模型的输出,从而实现推理阶段的并行化加速。为此,论文提出了一种高度优化的多标记预测(Multi-Token Prediction, MTP)架构——G-MTP,其关键创新在于三项核心设计:一是通过目标激活注入(target activation injection)锚定语义规划,提升草稿的准确性;二是采用零拷贝键值缓存跨注意力机制(zero-copy KV-cache cross-attention),跳过预填充计算并降低内存开销;三是引入高效的聚类词汇投影(clustered vocabulary projection),显著减少全词汇表解码的计算成本。实验表明,G-MTP 在服务器端(最高达3.1×加速)和资源受限的设备端(如智能手机GPU上达2.2×加速)均表现出优异性能,并结合参数高效微调策略,在领域偏移下仍能保持草稿与主模型的一致性,相较仅微调主模型可减少高达1.83倍的目标前向传播次数。

链接: https://arxiv.org/abs/2611.25813
作者: Filippo Galgani,Ziwei Ji,Omri Homburger,Ganesh Jawahar,Michelle Ramanovich,Ivan Korotkov,Luciano Martins,Utku Evci,Edouard Yvinec,Pooja Consul,Vivek Kumar,Tal Schuster
机构: Google(谷歌)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Speculative decoding has emerged as a critical technique for accelerating Large Language Model (LLM) inference by decoupling generation into a drafting phase and a parallel verification phase. A crucial challenge is therefore designing a drafter that is both lightweight and accurate enough to approximate the predictions of the target model. We present a highly optimized Multi-Token Prediction (MTP) architecture that extends existing backbone LLMs to efficiently produce high-quality drafts. Built around a single shared drafting head, our approach features three core architectural optimizations: target activation injection to anchor semantic planning, zero-copy KV-cache cross-attention to bypass prefill compute and reduce memory overhead, and an efficient clustered vocabulary projection to reduce the cost of full-vocabulary decoding. Testing these improvements with the Gemma 4 model family, our G-MTP shows out-of-the-box speedups for both large server-side setups (up to 3.1 \times ) and smaller models targeting memory-constrained on-device environments (up to 2.3 \times on H100 and up to 2.2 \times on a smartphone GPU). In addition, we present parameter-efficient fine-tuning recipes for G-MTP that keep the drafter aligned with the backbone LLM under domain shift, requiring up to 1.83 \times fewer target forward passes than tuning only the backbone.

[NLP-42] Language models represent more about emphasis than they can tell you

【速读】: 该论文旨在解决语言模型在预测语调突出性(prosodic prominence)时对文本与语音信号依赖程度的问题,尤其关注文本模型能否充分捕捉并准确报告由说话方式决定的语调特征。其核心挑战在于:尽管语调突出性在很大程度上可由句子词汇内容推断,但仍有部分依赖于说话人的表达方式,而这一信息在纯文本中难以完全体现。解决方案的关键在于区分文本信息与语音信号所贡献的语调信息——研究发现,无论是否微调或采用提示工程(prompting),文本模型(包括冻结和微调的预训练模型及大语言模型,LLMs)在3分类任务上的表现均仅达到约68.8%,且错误模式高度一致,表明它们仅能捕捉文本决定的部分突出性;相比之下,语音模型则显著更优,达到84.7%的准确率,说明语音信号中包含关键的非文本性突出性信息。研究进一步指出,尽管大语言模型的表征中隐含了更多关于强调位置的信息(线性探测可揭示),但通过提示无法有效提取这些信息,即模型“知道”但“说不出来”。因此,结论是:语调突出性并非完全由文本决定,剩余部分必须通过听觉感知获取,而当前的大语言模型虽编码了部分语义-语用层面的强调线索,却无法通过提示机制将其显式输出。

链接: https://arxiv.org/abs/2611.25808
作者: Aarne Talman,Stergios Chatzikyriakidis
机构: Institute of Language and Speech Processing(语言与语音处理研究所)
类目: Computation and Language (cs.CL); Sound (cs.SD)
备注:

点击查看摘要

Abstract:Prosodic prominence can largely be predicted from the words of a sentence, but part of it depends on how the speaker reads it. We study how much of prominence language models recover from text, whether they can report it when asked, and what the speech signal adds. On read English speech labelled for word-level prominence, we compare frozen and fine-tuned text models, prompted large language models (LLMs) and speech models on the same words. All text models end within about one point of each other, at best 68.8% on the 3-way task, and largely make the same errors. With the prompts we tested, LLMs do no better than a word-majority baseline, even with 500 examples. Linear probes find more in their representations. Speech models reach 84.7%. We argue that text determines only part of prominence, and that the rest has to be heard. LLMs encode what the text determines, but prompting does not bring it out: they represent more about where the emphasis falls than they tell you.

[NLP-43] RouteVerify: Routing Interventions for Step-Level Verification in Mixture-of-Experts Models

【速读】: 该论文旨在解决生成式 AI (Generative AI) 模型中路由状态(routing states)是否真正编码了推理步骤正确性的问题。具体而言,当模型的路由行为发生变化时,若其步级验证判断也随之改变,研究者常推断该判断依赖于路由所编码的信息。然而,这种因果推断可能具有误导性——因为路由变化可能仅是表象,而非真实决定因素。为检验这一假设,作者提出 RouteVerify,一种受控干预实验方法:在保持输入不变的前提下,将某解题轨迹中的特定内部激活(activation)替换为另一轨迹中的对应激活,从而系统分析路由与验证判断之间的因果关系。其核心解决方案在于设计三种干预接口——路由器逻辑值(router logits)、MoE 输出及后 MoE 残差状态,并在不同层和词元位置进行替换操作,通过测量验证置信度边际(verification margin)的变化来量化判断偏移。研究发现,尽管路由器逻辑在多数检查点中可有效区分正确与错误步骤(AUROC 0.80–0.88),且替换路由器逻辑能显著改变验证判断,但该判断的改变极少依赖于源步骤的真实正确性;在多数情况下,来自正确或错误步骤的路由对最终判断的影响几乎无差异。相比之下,局部残差状态替换则在大多数检查点中表现出对正确性的敏感性,表明验证机制确实能读取正确性信息。因此,结论是:虽然路由状态可反映正确性,且路由干预能影响判断,但判断的变动并不遵循正确性本身,暗示路由与验证之间存在非因果或间接关联,挑战了“路由即正确性编码”的直观理解。

链接: https://arxiv.org/abs/2611.25780
作者: Yucheng Du,Ryan A. Rossi,Xiyang Hu
机构: University of Southern California (南加州大学); Adobe Research (Adobe 研究院); Arizona State University (亚利桑那州立大学)
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 38 pages, 5 figures. Code: this https URL

点击查看摘要

Abstract:Routing states in mixture-of-experts (MoE) models can distinguish correct from erroneous reasoning steps. If changing routing also changes a model’s step-level verification judgment, it is tempting to conclude that the judgment follows what routing encodes. We test this inference with RouteVerify, a controlled intervention study that replaces selected internal activations from one solution trace with activations from another, while keeping the evaluated input unchanged. Across nine MoE checkpoints, we compare three intervention interfaces: router logits, MoE outputs, and post-MoE residual states. We vary the layers and token positions at which activations are replaced. We measure changes in a verification margin that quantifies the model’s preference for INCORRECT over CORRECT. We also evaluate the same input with activations from correct versus erroneous steps to test whether the resulting judgment depends on the source step’s correctness. We report four linked results. First, probes on router logits separate correct from erroneous steps in eight checkpoints (AUROC 0.80-0.88). Second, replacing router logits across all routed layers changes the judgment in all nine checkpoints, and by more than a norm-matched random replacement. Third, this change rarely depends on whether the source step is correct. At one layer, routing from erroneous steps and from correct steps produces indistinguishable margins in all nine checkpoints. Under all-layer routing, their difference is resolved and positive in only one. Fourth, the same difference is positive for local residual replacements in eight of nine checkpoints, so the measure does detect such dependence where it exists. Correctness is therefore readable from routing, and routing replacements move the judgment, but the movement rarely follows correctness.

[NLP-44] Closing Performance Gap for Diffusion Language Models under Low Inference Budgets

【速读】: 该论文旨在解决生成式语言模型在低预算(low-budget)条件下进行高效文本生成时面临的质量下降问题,尤其针对扩散语言模型(Diffusion Language Models, DLMs)在减少去噪与精炼步骤后生成质量显著退化的挑战。现有方法如可编辑扩散模型(Editable DLMs)虽通过后期草稿精炼缓解了这一问题,但仍存在关键中间结果在最终提交前丢失的风险。本文提出一种无需训练的框架CaRR(Commitment-aware Refinement and Recovery),其核心在于利用已有前向调度过程中的信息流实现质量提升:一是通过令牌承诺重掩码(Token Commitment Remasking, TCR),在后续编辑步骤中基于更新后的上下文重新填充被掩码的原始纯替换项,增强上下文一致性;二是通过提交承诺恢复(Submission Commitment Recovery, SCR),根据精炼轨迹回溯并恢复符合任务需求的候选输出,以应对最终响应不满足特定要求的情况。两项机制均在原生掩码/编辑调度流程内完成,无需额外前向传播。实验表明,在代码生成与数学推理任务上,CaRR在四个不同推理预算下均显著优于基准的ME-DLM,尤其在极低预算(全预算的1/8)时表现突出,六项指标平均准确率提升7.44个百分点,其中在MBPP和MBPP+数据集上分别提升11.37和10.05个百分点,验证了其在资源受限场景下的有效性与鲁棒性。

链接: https://arxiv.org/abs/2611.25763
作者: Hao Wu,Sipeng Chen,Jingwen Fu,Yuwang Wang
机构: University of Wisconsin–Madison; Northwestern University; Zhongguancun Academy; Zhongguancun Institute of Artificial Intelligence; Tsinghua University
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Machine Learning (stat.ML)
备注: 23 pages, 3 figures

点击查看摘要

Abstract:Diffusion language models offer an architectural route to low-budget generation by updating multiple tokens in parallel, which can reduce the number of sequential decoding steps relative to token-by-token autoregressive decoding. However, realizing this advantage at low budgets is challenging, as generation quality often declines when the number of denoising and refinement steps is reduced. Editable diffusion models address this quality loss through post-draft refinement, but useful intermediate results may still be lost before final submission. We introduce CaRR (Commitment-aware Refinement and Recovery), a training-free framework which improves submission quality from a fixed editable checkpoint through updated-context refill and trajectory-based recovery within its existing forward schedule. Token Commitment Remasking (TCR) remasks selected native pure replacements and refills them on the next scheduled edit forward using the updated context. Submission Commitment Recovery (SCR) uses the resulting refinement trajectory to recover supported candidates when the final response does not satisfy task-specific output requirements. Both modules operate within the native mask/edit schedule without additional backbone forwards. Experiments on code generation and mathematical reasoning show consistent gains in average accuracy over native ME-DLM across four inference budgets, with the largest gains under tight budgets. At one-eighth of the full scheduled forward budget, CaRR improves the six-metric average accuracy by 7.44 pp, including gains of 11.37 and 10.05 on MBPP and MBPP+.

[NLP-45] Verification Behavior Across Five Qwen 2.5-Instruct Checkpoints

【速读】: 该论文旨在解决大模型在复杂任务中是否存在可识别的诊断性执行(diagnostic execution)能力,以及其表现是否随模型规模和提示策略变化的问题。研究通过评估Qwen2.5-Instruct系列五个不同参数量级(0.5B至32B)的检查点,在四种提示条件下对三类固定任务进行生成测试,每项任务生成五次,共收集600条轨迹由两名AI评分员评估。关键发现包括:在协调指令下,7B模型在11/15案例中严格声明信息不足,而14B和32B模型分别降至0/15和5/15;当请求检测器触发时,三个模型均收到至少一次标准追问,分别出现在15、12和15例中;在信息充分的任务中,3B至32B模型在所有协议轨迹中均出现格式标记(format markers),而基线轨迹中未出现;经仲裁后,仅1/60的样本满足诊断性执行标准(H3),且所有成功案例均集中于同一任务。整体结果表明,模型在可见结构、诊断性执行与证据请求后的持续性方面存在显著分离,预设的性能向上跃迁(upward transition)并未在本实验中实现。解决方案的关键在于通过结构化提示与多轮交互设计,量化评估模型在信息不充分情境下的自我觉察与纠错能力,揭示其在真实推理任务中的局限性。

链接: https://arxiv.org/abs/2611.25745
作者: YaJie Yin
机构: 未知
类目: Computation and Language (cs.CL)
备注: 10 pages, 1 figure. Includes English translations of the experiment prompts and the behavioral rating rubric. AI language models were used for behavioral ratings and to assist manuscript preparation

点击查看摘要

Abstract:We evaluate five Qwen2.5-Instruct checkpoints (0.5B to 32B) under four prompting conditions using three fixed tasks per information condition and five generations per task. Two AI raters scored 600 trajectories. Under the coordinated instruction, both raters recorded strict declarations of insufficiency in 11/15 cases at 7B, 0/15 at 14B, and 5/15 at 32B. The runner sent a standard follow-up when a request detector fired. The three checkpoints received at least one follow-up in 15, 12, and 15 cases, respectively. On information-sufficient tasks at 3B through 32B, scored format markers appeared in 60/60 protocol trajectories and 0/60 baseline trajectories. After adjudication, 1/60 in each condition met the diagnostic-execution criterion (H3). Across all checkpoints and prompts, 7/300 information-sufficient trajectories met H3 after adjudication, all on the same task. The prespecified upward transition did not occur in this run. The results separate visible structure, diagnostic execution, and persistence after a request for evidence.

[NLP-46] PosteriorBench: Evaluating Behavior Posterior Sampling in Language Models

【速读】: 该论文旨在解决生成式语言模型中罕见行为样本难以高效获取的问题,即“行为后验采样问题”(behavior posterior sampling problem):给定一个语言模型和感兴趣的特定行为,是否存在一种算法能够高效地从模型生成的文本分布中采样出表现出该行为的文本?其核心解决方案的关键在于构建一个名为PosteriorBench的基准测试框架,用于评估不同方法在计算成本与采样结果接近真实后验分布程度之间的权衡,具体通过前向和反向KL散度来衡量。该基准包含21个基于Qwen3.5-0.8B模型的任务,行为出现概率范围为10⁻⁵至10⁻¹,并提供了来自十亿条标注响应语料库的精确后验样本作为参考。研究发现,当前所有评估的方法均无法同时实现高效性与准确性——简单提示(prompting)方法虽计算成本低但收敛于较高的KL散度,基于后验样本的微调则通常代价高昂,而先前工作中的多种训练方法也未展现出显著优势。因此,如何设计一种低成本且能精准逼近真实后验分布的采样器仍是开放性挑战,而PosteriorBench有望成为推动该方向进展的重要评估工具。

链接: https://arxiv.org/abs/2611.25735
作者: Zhiyang Xun,Daniel D. Johnson,Sarah Schwettmann,Neil Chowdhury
机构: Transluce; The University of Texas at Austin (德克萨斯大学奥斯汀分校); Thinking Machines Lab
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Rare language model behaviors can be difficult to study because obtaining them through ordinary sampling can require a very large number of samples. This motivates the behavior posterior sampling problem: given a language model and a behavior of interest, is there an algorithm that efficiently samples from the model’s distribution over transcripts that exhibit this behavior? We introduce PosteriorBench, a benchmark for comparing algorithms by their compute cost and by how close their samples are to the true posterior in forward and reverse KL divergence. PosteriorBench contains 21 tasks on Qwen3.5-0.8B with behavior probabilities from 10^-5 to 10^-1 , each with exact posterior samples from a corpus of one billion labeled responses. We find that no method we evaluate is both efficient and faithful: simple methods like prompting are cheap but plateau at a high KL divergence from the posterior, fine-tuning on posterior samples is often cost-prohibitive, and several training methods from prior work do not perform better. Building a posterior sampler that can cheaply approximate the true posterior remains an open problem, and we hope PosteriorBench can measure progress towards this objective. Code and data are available at this https URL.

[NLP-47] VR-SparseQuant: Accurate FP4 Quantization with 2:4 Sparsity on Vera Rubin

【速读】: 该论文旨在解决四比特(4-bit)激活量化(activation quantization)在低精度计算中导致模型质量显著下降的问题。随着GPU硬件对低精度计算支持的不断演进,如Vera Rubin架构引入的元素级2:4结构化稀疏FP4指令(VRFP4),如何在保持原生FP4执行效率的同时恢复模型精度成为关键挑战。其解决方案的核心是提出VR-SparseQuant方法,利用Vera Rubin的元素级2:4稀疏性,通过允许两个稀疏数据流重叠,在不增加FP4数值预算的前提下,对误差较大的激活值进行精细化重构;同时采用离线比例校准避免运行时搜索开销,并设计共享权重矩阵乘法(GEMM)高效聚合两路稀疏流。实验表明,该方法在Nemotron-3-Nano-30B-A3B模型上将WikiText-2困惑度从6.85降低至5.57(相比密集型RTN W4A4),并实现了高达3.1倍的GEMM与2.7倍的混合专家(MoE)加速比。据作者所知,这是首个基于Vera Rubin元素级2:4结构化稀疏FP4指令实现自适应激活量化的研究。

链接: https://arxiv.org/abs/2611.25717
作者: Samir Khaki,Wei-Chen Wang,Song Han,Ryan Hanrui Wang
机构: 未知
类目: Computation and Language (cs.CL)
备注: 16 Pages, 7 Figures

点击查看摘要

Abstract:In recent years, GPUs have incorporated increasingly specialized support for low-precision computation, from Hopper’s FP8 to Blackwell’s FP4 Tensor Cores. Precision formats have evolved alongside this hardware, from MXFP4 and NVFP4 on Blackwell to E2M1 FP4 with direct UE5M3 block scaling on Vera Rubin, which we term as \boldsymbolVRFP4 . These advances help serve larger language models at lower cost, but four-bit activation quantization can substantially degrade model quality. We introduce VR-SparseQuant, which uses Vera Rubin’s elementwise 2:4 sparsity to recover accuracy while retaining native FP4 execution. By allowing two sparse streams to overlap, we use the same FP4 value budget as dense quantization to refine activations with larger errors. Offline scale-ratio calibration avoids runtime search, while a shared-weight GEMM accumulates both streams efficiently. On Nemotron-3-Nano-30B-A3B, VR-SparseQuant reduces WikiText-2 perplexity from \boldsymbol6.85 to \boldsymbol5.57 versus dense RTN W4A4. Including activation preparation, our native Rubin kernels for VRFP4 achieve up to \boldsymbol3.1\times GEMM and \boldsymbol2.7\times MoE speedup over BF16. To the best of our knowledge, this is the first study of adaptive activation quantization using Vera Rubin’s elementwise 2:4 structured sparse FP4 instructions.

[NLP-48] From Geographic Memory to Spatial Generalization in Large Language Models

【速读】: 该论文旨在解决大语言模型(Large Language Models, LLMs)在地理空间认知能力上的局限性,即模型虽能记忆训练期间观察到的地点与地理描述之间的关联(geospatial memory),却难以实现对未观测区域的地理描述预测(spatial generalization)。其核心问题在于:如何提升模型在未见地理区域中的泛化能力,并促使地理知识在模型内部形成具有空间结构的组织形式。解决方案的关键在于引入连续位置表示(continuous location representations),通过构建一个连续的空间编码器,将地理位置映射至模型的嵌入空间并作为“空间标记”(spatial tokens)输入模型,从而引导模型学习一种潜在的空间场(latent spatial field)——即基于空间依赖关系从位置到描述的可泛化映射。实验结果表明,相较于使用文本坐标,连续位置表示显著提升了模型在未观测区域的地理描述生成性能,并在下游地理定位任务中带来稳定但适度的改进;层间探查与表示聚类分析进一步支持了地理表征在空间维度上被组织化的证据。研究强调了将空间泛化能力独立于记忆能力进行评估的重要性,并验证了在LLM中集成连续空间信息对于实现更智能、更具空间感知能力的地理理解的有效性。

链接: https://arxiv.org/abs/2611.25703
作者: Chen Chu,Zhaonan Wang,Cyrus Shahabi,Hengcai Zhang,Feng Lu
机构: University of Southern California(南加州大学); NYU Shanghai(纽约大学上海分校); 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:

点击查看摘要

Abstract:Large language models (LLMs) can associate places with geographic descriptions, but recovering an observed association does not establish an ability to generalize across geographic space. We distinguish geospatial memory, the recovery of location–description associations observed during training, from spatial generalization, the prediction of descriptions for locations in unobserved regions. We investigate whether continuous location representations improve this generalization and encourage geographic knowledge to become spatially organized within an LLM. We frame this organization as a latent spatial field: a learned mapping from locations to descriptions connected through spatial dependencies. Building on a continuous spatial encoder, we project location representations into the model’s embedding space and supply them as spatial tokens. We evaluate the resulting models on two newly curated geo-text benchmarks for Los Angeles and London using two complementary assessments. Geospatial captioning tests both memory and spatial generalization by generating descriptions at observed locations and at locations beyond the spatial coverage of the training observations. A separate geospatial grounding model tests whether generated location–description pairs provide useful additional supervision for mapping text to geographic space. Across multiple LLM families, continuous location representations improve captioning at unobserved locations relative to textual coordinates, while their generated descriptions yield modest but consistent improvements in downstream grounding. Layer-wise probing and representation clustering provide additional evidence consistent with spatially organized geographic representations. Together, these findings show the value of evaluating spatial generalization separately from memory and of incorporating continuous spatial representations into language models.

[NLP-49] FoodReact-AR: Learning Audience Reactions that Zero-Shot LLM s Miss in Arabic Food-Insecurity Discourse

【速读】: 该论文旨在解决如何从社交媒体文本中识别受众对特定议题(阿拉伯语地区粮食不安全相关帖子)的隐含情感反应模式的问题,尤其关注那些未在文本中直接体现但通过用户互动行为(如“点赞”、“哈哈”、“悲伤”等六种差异化反应)所反映的社会情绪。其核心挑战在于:如何仅基于文本内容预测这些复杂且具有社会意义的观众反应分布,而无需依赖原始互动数据。解决方案的关键在于将反应数据视为组成型数据(composition),采用中心对数比变换(centered log-ratio transformation)进行数学处理,并通过聚类方法提炼出四类典型观众画像——团结(以“爱”为主导)、轻蔑(以“哈哈”为主导)、哀伤(以“悲伤”为主导)和中立矛盾型。研究进一步验证了这些反应模式在时间上的显著变化(如2022年因小麦价格冲击导致“轻蔑”反应峰值,2024年加沙粮食危机期间“哀伤”反应上升),并通过多种稳健性检验确保结果可靠性。在预测层面,研究发现虽零样本阿拉伯语大语言模型(LLM)可有效识别“哀伤”类反应(准确率达83.8%),却几乎无法捕捉“轻蔑”反应(仅3.4%),表明部分带有讽刺或批判意味的观众反应难以通过预训练模型直接推断;而基于反应标签训练的简单分类器在跨页面测试中表现出更高性能(宏F1为0.56),尤其在提及政府机构与官方人物的文本中表现突出。因此,该研究的关键突破在于构建了一个名为FoodReact-AR的公开数据集、标注层及评估协议,实现了对社交媒体中复杂社会情绪反应的结构化建模与可复现分析。

链接: https://arxiv.org/abs/2611.25682
作者: Wajdi Zaghouani,Mabrouka Bessghaier,Salma Azouzi,Shimaa Amer Ibrahim
机构: 未知
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Facebook reactions record how an audience received a post, a signal absent from the post’s text. We treat the six differentiated reactions on 74,241 Arabic food-insecurity posts from public Facebook pages (2020-2024) as compositions, apply a centered log-ratio transformation, and cluster them into four audience profiles: Solidarity (Love-dominant), Dismissal (Haha-dominant), Grief (Sad-dominant), and a flat Ambivalent profile. Dismissal peaks in 2022 (22.3%), the year of the wheat price shock, and grief in 2024 (14.0%), during the Gaza food emergency; a page-clustered GEE confirms both shifts (odds ratios 2.18 and 1.34), and they survive minimum-reaction filtering, alternative zero replacements and clustering methods, and bootstrap resampling. We then test whether the profiles can be predicted without reactions. A zero-shot Arabic LLM recovers Grief for 83.8% of posts but Dismissal for only 3.4%, and in a matched ablation, it shows the reaction percentages do not help. A simple classifier trained on the reaction-derived labels and tested on unseen pages recovers dismissal for 60.0% (macro-F1 0.56 against 0.43), mainly from mentions of officials and institutions. Derisive audience responses are thus learnable from text, yet a zero-shot LLM misses them. We release the corpus, LLM, and human annotation layers, and an evaluation protocol as FoodReact-AR.

[NLP-50] SEAF: Selective Semantic Enrichment and Visual Evidence Fusion for Multimodal Sarcasm Detection

【速读】: 该论文旨在解决多模态讽刺检测中因图像间互补线索复杂、生成式模型输出冗余信息而导致语义与视觉证据融合不精准的问题。其核心解决方案是提出SEAF(Selective Semantic Enrichment and Visual Evidence Fusion)双分支框架,关键在于通过角色分离机制实现对多模态信息的有选择性增强与融合:一方面,利用冻结的多模态大语言模型(Multimodal Large Language Models, MLLMs)生成的实体背景知识丰富文本语义输入;另一方面,将图像描述作为对齐统计量,用于指导视觉特征融合。具体而言,语义分支整合原始文本、冻结MLLM提供的背景知识以及光学字符识别(OCR)文本,以增强上下文理解;视觉分支则结合全局与局部视觉特征,并引入跨模态及图像间相似性统计信息。通过基于文本锚定的平滑聚合(Text-Anchored Smoothed Aggregation, TASA)机制,根据图像与原始文本的语义相似度动态加权,同时保留低字面相似但具有潜在讽刺意义的图像贡献。最终通过加总两个分支的分类得分,实现上下文语义与视觉证据的有效集成。在包含多图像的MMSD3.0基准上的实验表明,SEAF显著优于现有方法。

链接: https://arxiv.org/abs/2611.25672
作者: Faxian Wan,Xiaocui Yang,Yifan Cao,Shi Feng,Daling Wang,Yifei Zhang
机构: Northeastern University (东北大学); Shenyang, 110819, China
类目: Computer Vision and Pattern Recognition (cs.CV); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Multimodal sarcasm detection requires textual, visual, and contextual evidence, particularly when multiple images provide complementary cues. Multimodal large language models (MLLMs) can generate auxiliary information from a post’s text and images, but their outputs may mix useful background knowledge with redundant visual descriptions. We propose SEAF, a dual-branch framework for selective semantic enrichment and visual evidence fusion. SEAF assigns generated fields to distinct roles: entity background enriches semantic input, while image descriptions supply alignment statistics. The semantic branch combines the original post text, background from a frozen MLLM, and optical character recognition (OCR) text. The aligned visual branch combines global and local visual features with statistics of cross-modal and inter-image similarity. Text-Anchored Smoothed Aggregation (TASA) weights images by their similarity to the original post text while preserving contributions from images with low literal similarity. Summing the branch classification scores integrates contextual semantics with visual evidence. Experiments on the multi-image MMSD3.0 benchmark show that SEAF outperforms all compared methods.

[NLP-51] ColorRef: Calibrating Verbal Feedback for Color Revision

【速读】: 该论文旨在解决生成式设计辅助系统在接收非精确颜色修改指令时,难以准确理解修改幅度的问题。具体而言,用户在提出“稍暗一点”或“更蓝一些”等模糊描述时,语言模型需同时推断修改方向与调整程度,而现有方法往往仅关注方向正确性,忽视反馈语句的量化尺度校准。其解决方案的关键在于:通过构建名为\textscColorRef的色彩参考游戏,系统性地评估模型对“a little”“somewhat”“much”等程度副词的响应行为,并基于实际响应数据对提示词进行个性化校准,使反馈语义与模型的实际调整步长相匹配。实验表明,在128个保留测试样本上,经校准后的反馈将单步平均误差从12.36降至6.86(Qwen3-14B)和从11.06降至5.89(Gemma 4 12B),显著提升了调整精度,尤其在大范围修正场景中效果突出。研究揭示,仅遵循修改方向不足以实现有效调整,精准的反馈尺度校准是实现可预测、可控设计迭代的核心。

链接: https://arxiv.org/abs/2611.25653
作者: Pablo Loyola,Andres Hoyos-Idrobo,Edison Marrese-Taylor,Colin Rowat,Alfredo Solano
机构: Rakuten Group, Inc; The University of Tokyo (东京大学)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:People rarely specify an exact color when asking someone to revise a design: they might say \textita little darker or \textitmore blue after seeing a proposal. A language-model assistant must then infer not only which way to change the color, but how far. We study this distinction with \textscColorRef, a color reference game in which a model guesses a color from a description and a teacher, knowing a recorded target hidden from the guesser, supplies verbal corrections. On 1,000 crowd-sourced description–target pairs, three rounds of deterministic oracle feedback reduce mean target distance from 42.33 to 22.15 \Delta E_76 . However, those corrections provide additional information, and improvement alone does not establish well-calibrated feedback following. A controlled Qwen3-14B diagnostic reveals a more specific failure: following a requested direction and ordering graded phrases do not guarantee a suitable step size; strong modifiers can induce excessive changes. We therefore measure each receiver’s typical response to \textita little, \textitsomewhat, and \textitmuch on separate colors and select wording matched to its observed steps. On 128 held-out starting colors, calibration reduces one-step mean error from 12.36 to 6.86 for Qwen3-14B and from 11.06 to 5.89 for Gemma 4 12B against a fixed graded-wording rule. The benefit is concentrated on large corrections, with many unchanged cases and no consistent increase in threshold success. Our findings show that following the direction of feedback is not the same as making an appropriately sized revision.

[NLP-52] RAC: Efficient Diffusion Language Models for Machine Translation via Length-Aware Decoding

【速读】: 该论文旨在解决基于扩散模型的大语言模型在机器翻译中如何在保持高翻译质量的同时降低生成延迟的问题。现有方法存在冗余计算:固定目标生成长度导致未使用位置的无效计算,而仅依赖置信度的承诺机制又延迟了预测已趋于稳定的中等置信度词元的提前输出。其解决方案的关键在于提出一种无需训练的自适应解码方法TRAC,通过源文本长度和翻译方向动态调整目标生成长度,并引入保守回退策略以应对空输出或异常短输出;同时,利用跨步Top-1一致性与Top-K分布间的近似Jensen-Shannon散度来识别时间上稳定的中等置信度词元,实现稳定性感知的早期承诺。实验表明,TRAC在FLORES-200、WMT23和NTREX-128多组双向中英翻译任务上,相较Fast-dLLM显著减少24.9%至36.8%的前向传播调用次数,实现1.91至2.19倍的句级生成加速,仅带来0.02点宏平均COMET分数下降,验证了自适应生成长度与稳定性感知词元承诺对扩散机器翻译效率的有效提升。

链接: https://arxiv.org/abs/2611.25643
作者: Tianyuan Li,Xihan Yang,Chenglong Wang,Yuzhang Wu,Jingbo Zhu,Tong Xiao
机构: Northeastern University (东北大学); NiuTrans Research (牛津研究)
类目: Computation and Language (cs.CL)
备注: Accepted at CCMT 2026

点击查看摘要

Abstract:Machine translation with large language models requires both high translation quality and low generation latency. Accelerating translation while preserving quality is therefore essential for practical deployment. Diffusion large language models (dLLMs) can generate tokens at arbitrary target positions through iterative denoising, offering a new paradigm for parallel translation beyond sequential autoregressive decoding. However, existing methods still incur redundant computation: fixed target generation lengths waste computation on unused positions, while confidence-only commitment delays medium-confidence tokens whose predictions have largely stabilized. To address these issues, we propose TRAC, a training-free adaptive decoding method for diffusion machine translation. TRAC adapts the target generation length using source length and translation direction, with a conservative fallback to the maximum length for empty or unusually short outputs. It preserves high-confidence commitment while identifying temporally stable medium-confidence tokens for early commitment using cross-step top-1 persistence and approximate Jensen-Shannon divergence over Top- K distributions. On bidirectional Chinese-English translation across FLORES-200, WMT23, and NTREX-128, TRAC reduces model forward calls by 24.9% to 36.8% and achieves sentence-level generation speedups of 1.91 to 2.19 times over Fast-dLLM, with only a 0.02-point decrease in macro-average COMET. These results demonstrate that adaptive generation lengths and stability-aware token commitment improve the efficiency of diffusion machine translation.

[NLP-53] Rubric Semantics or Numerical Priors? Auditing Scale-Direction Bias in LLM Judges EACL2027

【速读】: 该论文旨在解决当前基于大语言模型(LLM)的自动评估范式中一个隐含但关键的假设问题:即认为LLM能够依据评分标准(rubric)的语义含义进行判断,而非受其数值尺度的影响。为此,作者提出了“语义忠实性”(semantic faithfulness)这一核心原则,要求在语义等价的评分标准下,模型应产生一致的评分结果。为量化该属性,研究设计了尺度反转探测(scale-reversal probe),通过仅反转评分锚点而保持评价标准不变,以检验模型对尺度敏感程度;同时引入底限归一化(floor normalization)方法,将尺度反转导致的评分分歧与模型内在变异性相校准,从而实现跨模型间的公平比较。在五个元评估基准上,使用七种不同来源的LLM(涵盖开源与专有模型)进行实验发现,大多数模型均存在显著的语义忠实性违背现象,表明仅依赖原始反转一致性指标会误导评估结论。最后,研究提出双向锚点平均法(bidirectional anchor averaging),通过融合正向与反向锚点的预测结果,有效降低对单一锚点方向的依赖,且不损害与人类判断的一致性。研究强调,真正可信的LLM评估者必须稳定遵循评分标准的语义逻辑,而不仅与人类评分存在相关性。

链接: https://arxiv.org/abs/2611.25588
作者: Chanwoo Kim,Doyun Kim,Minsoo Song,Sihyeon Lee,Chanjun Park
机构: Soongsil University (中央大学)
类目: Computation and Language (cs.CL)
备注: Accepted for EACL 2027 Findings

点击查看摘要

Abstract:Recent benchmark evaluations increasingly rely on Large Language Models (LLMs) as automatic judges. This paradigm assumes LLMs evaluate based on the rubric’s semantic meaning rather than its numeric scale. We question this assumption by introducing semantic faithfulness: the principle that semantically equivalent rubrics must yield consistent scores. To quantify this, we propose a scale-reversal probe that inverts score anchors without altering evaluation criteria. Furthermore, we introduce floor normalization to calibrate reversal disagreement against intrinsic model variability, enabling fair cross-model comparisons. Experiments across five meta-evaluation benchmarks with seven diverse LLM judges spanning open-weight and proprietary models reveal widespread violations of semantic faithfulness, showing that raw reversal agreement alone is misleading. Finally, we demonstrate that bidirectional anchor averaging reduces dependence on any single anchor direction without compromising alignment with human judgments. Our findings emphasize that trustworthy LLM judges must reliably follow rubric semantics, not just correlate with human ratings.

[NLP-54] he Coverag e Selection Gap: Ranking Not Generation Limits LLM Rubric Construction NEURIPS2026

【速读】: 该论文旨在解决在缺乏专家制定参考评分标准的情况下,如何从复杂源文档中自动构建全面、准确的特定任务评分标准(task-specific rubrics)的问题,特别是在专业并购尽职调查(mergers and acquisitions due diligence)场景中。其核心挑战在于:评估标准需从任务指令和支撑性文档中推断得出,而传统依赖人工标注参考标准的方法不可行。论文提出一种多轴生成框架(multi-axis generation framework),通过探索源证据材料的互补性结构化与自由形式视图,构建一个广泛且多样化的候选评估标准池。为区分标准发现失败与候选选择失败,引入“覆盖-选择差距”(coverage-selection gap)这一新度量指标,用于量化在固定选择预算下,基于无参考排序与有参考指导的贪婪诊断之间的差异。实验在16个任务上训练并针对4个保留任务进行评估,结果显示该框架在基于大语言模型(LLM)的标准等价协议下,达到0.876的宏观召回率(macro recall),表现接近资深专家编写的参考标准。消融研究进一步表明,引入新的结构化搜索维度可使召回率提升6–10个百分点,而仅增加现有方向的生成量最多仅提升1.5个百分点。这表明:一旦构建了高覆盖率的候选标准池,如何在无参考条件下有效筛选出最优标准,已成为文档驱动评分标准构建中的主要瓶颈。

链接: https://arxiv.org/abs/2611.25557
作者: Ali Karkehabadi,Jamshid Hassanpour,Rafael E. Banchs
机构: University of California, Davis (加州大学戴维斯分校); Georgia Institute of Technology (佐治亚理工学院); Intapp; San Jose State University (圣何塞州立大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted at LCFM @ NeurIPS 2026

点击查看摘要

Abstract:Task-specific rubrics are increasingly used to evaluate open-ended language-model outputs, yet constructing comprehensive evaluation criteria from complex source documents remains challenging. We study automatic rubric construction in professional mergers and acquisitions (MA) due diligence, where evaluation criteria must be inferred from task instructions and supporting documents without access to an expert-authored reference rubric. We introduce a multi-axis generation framework that explores complementary structural and free-form views of the source evidence to construct a broad pool of candidate evaluation criteria. To distinguish failures of criterion discovery from failures of candidate selection, we introduce the coverage-selection gap, which measures the difference between reference-free ranking and a reference-informed greedy diagnostic under a fixed selection budget. Developed on 16 tasks and evaluated on four held-out tasks, our framework achieves 0.876 macro recall against senior-authored reference rubrics under an LLM-based criterion-equivalence protocol. Ablation studies further show that introducing new structural search axes improves recall by 6-10 percentage points, whereas increasing generation volume along existing search directions yields at most 1.5 points. These findings suggest that, once a high-coverage candidate pool has been constructed, effective reference-free selection becomes a substantial bottleneck in document-grounded rubric construction.

[NLP-55] COVERAG E-TRACE: Source-Side Counterfactual Tracing for Detecting Omissions in LLM Summaries NEURIPS2026

【速读】: 该论文旨在解决生成式摘要中“遗漏”(omission)问题的评估难题,即难以验证缺失信息是否影响摘要的语义完整性——尽管摘要在事实层面可能正确,但可能遗漏关键决策、限定条件、数量或前提等原子性事实,从而扭曲源文本原意。其核心解决方案是提出一种基于源端反事实干预的评估框架COVERAGE-TRACE,通过仅替换目标原子事实为同类型替代项,固定摘要输出,测量教师强制(teacher-forced)、长度归一化后的概率变化来评估摘要对特定事实的依赖程度。若摘要覆盖目标事实,则其概率响应更显著;通过多反事实(multi-CF)聚合可降低对具体替换项选择的敏感性。实验在80组匹配的“覆盖–遗漏”QMSum样本上验证,多反事实方法达到0.764–0.825 AUROC,显著优于最强文本基线(0.640)。在Falcon-7B模型上,针对性干预获0.779 AUROC,而随机替换降至0.533,证明目标相关性有效性;但引入语义保持的改写后,AUROC下降至0.591–0.624,表明当前方法仍受表述形式影响。因此,COVERAGE-TRACE提供了一种实用且互补的目标特定依赖信号,但无法实现语义不变的覆盖率度量。研究确立了源端干预在遗漏检测中的潜力,并指明语义不变性(semantic invariance)是当前最核心的挑战。

链接: https://arxiv.org/abs/2611.25539
作者: Ali Karkehabadi,Jamshid Hassanpour,Rafael E. Banchs
机构: University of California, Davis (加州大学戴维斯分校); Georgia Institute of Technology (佐治亚理工学院); Intapp; San Jose State University (圣何塞州立大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Accepted at LCFM @ NeurIPS 2026

点击查看摘要

Abstract:Omissions are difficult to verify because missing information leaves no output span to inspect: a summary may be factually supported while still excluding a decision, qualification, quantity, or condition needed to preserve the source’s meaning. We introduce COVERAGE-TRACE, a source-side counterfactual framework that evaluates omission as dependence on a designated atomic source fact. It replaces only the target fact with type-matched alternatives, holds the candidate summary fixed, and measures the resulting change in its teacher-forced, length-normalized likelihood. A summary covering the target should respond more strongly than one omitting it, while aggregation across multiple counterfactuals reduces sensitivity to replacement choice. We evaluate COVERAGE-TRACE on 80 matched covered–omitted QMSum pairs spanning several fact types and four open-weight models. Multi-CF achieves 0.764–0.825 AUROC, compared with 0.640 for the strongest evaluated text-only baseline. On Falcon-7B, targeted interventions achieve 0.779 AUROC, whereas shuffling counterfactual assignments reduces performance to 0.533; target deletion is effective while unrelated deletion is not. However, adjustment for meaning-preserving paraphrases reduces AUROC to 0.591–0.624. COVERAGE-TRACE therefore provides a practical, complementary signal of target-specific source–summary dependence, but not a wording-invariant semantic coverage probability. These results establish source-side intervention as a promising approach to omission detection while identifying semantic invariance as the central remaining challenge. Comments: Accepted at LCFM @ NeurIPS 2026 Subjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL) Cite as: arXiv:2611.25539 [cs.AI] (or arXiv:2611.25539v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2611.25539 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[NLP-56] LM-based Personalized Travel Planner

【速读】: 该论文旨在解决旅行规划中多源信息融合、约束条件管理及个性化推荐难以有效协同的问题,尤其针对单智能体大语言模型(LLM)在复杂多步推理、约束遵循与模块化执行方面存在的局限性。其核心解决方案是提出一种分层式多智能体框架,由一个监督型聊天机器人智能体协调多个专业化智能体,分别负责行程验证、用户查询重构、实时数据获取以及基于约束的行程生成。该框架通过集成大语言模型的推理能力、外部API调用与基于确定性多维评分机制的迭代优化流程,实现了对用户偏好和多种约束条件的高效适应与精准响应。实验结果表明,在著名的TravelPlanner基准测试中,该方法在预算约束规划上提升10%,跨约束集规划能力提高25%,个性化推荐效果提升45%,显著优于ReAct基线模型。

链接: https://arxiv.org/abs/2611.25531
作者: Amandi Nimasha,Oshadha Wijerathne,Dushan Fernando,Nisansa de Silva,Srinath Perera
机构: University of Moratuwa(莫鲁塔瓦大学); WSO2(WSO2)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Effective travel planning requires handling multiple information sources, managing constraints, and tailoring recommendations to user preferences. While LLMs have introduced dynamic, context-aware decision-making, single-agent models often struggle with complex multi-step reasoning, constraint adherence, and modular execution. To overcome these limitations, we propose a hierarchical multi-agent framework where a supervising chatbot agent coordinates specialized agents for schedule validation, user query formulation, real-time data retrieval, and constraint-aware itinerary generation. By integrating LLM reasoning, external APIs, and iterative refinement mechanisms based on a deterministic, multi-dimensional scoring framework, our approach ensures adaptability, efficiency, and personalization. Evaluations on the well-known TravelPlanner benchmark demonstrate improvements in Budget Constraint planning by 10%, wider constraints set-based planning by 25%, and personalized planning results by 45%. compared to the ReAct baseline. Our code is available at this https URL

[NLP-57] MATE: Mobile Agent Trustworthy Execution for Safe and Efficient Mobile Control EMNLP2026

【速读】: 该论文旨在解决移动智能体(Mobile Agent)在个人智能手机上执行时可能对用户隐私数据、支付信息、系统设置及前台界面造成不可控风险的问题。现有方法通常依赖于修改底层大语言模型(LLM)或缺乏对操作边界的有效管控,导致安全与可用性难以兼顾。其解决方案的关键在于提出MATE(Mobile Agent Trustworthy Execution)系统,通过运行时的执行层隔离机制,在不修改底层模型的前提下,严格控制智能体动作提案与设备实际影响之间的边界。MATE的核心创新包括:将每个用户界面(UI)原语表示为包含指令、目标应用、屏幕上下文、操作类型和动作字段的结构化动作记录;引入运行时防护墙(Runtime Guardrail, RG)实现动态允许、阻断或用户审批决策;采用预定义任务路由(Predefined Task Routing, PTR)机制,通过可信路径处理重复性任务并结合执行后屏幕验证提升可靠性;同时利用浮动界面以紧凑叠加层形式维持执行过程的可控性。实验结果表明,在MobileSafetyBench上,MATE将高风险操作的拦截率从29.7%提升至97.0%(基于GPT-4o);在AndroidWorld基准测试中,减少高达6倍的LLM调用次数,并使重复任务的成功率提升至95.7%。

链接: https://arxiv.org/abs/2611.25526
作者: Minsuk Jang,Hee-Seon Kim,Hui-Cheol Yu,Changick Kim
机构: Korea Advanced Institute of Science and Technology (KAIST), Daejeon, Republic of Korea
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted to EMNLP 2026

点击查看摘要

Abstract:Mobile agents on personal smartphones can affect private data, payments, settings, and the foreground screen. We present MATE (Mobile Agent Trustworthy Execution), an execution-layer system that controls the boundary between agent action proposals and device effects without modifying the underlying model. MATE represents each UI primitive as a structured action record that combines the instruction, target app, screen context, primitive, and action fields. Runtime Guardrail (RG) maps this record to an allow, block, or user-approval decision; Predefined Task Routing (PTR) serves recurring requests through verified routes with post-execution screen checks; and a floating interface keeps execution in a compact overlay. On MobileSafetyBench, MATE raises high-risk blocking from 29.7% to 97.0% under GPT-4o. On AndroidWorld, it reduces LLM calls by up to 6x and raises recurring-use task success to 95.7%.

[NLP-58] DRVEXO-4B-Unified: Design and Evaluation of a Modular Speech-Vision System

【速读】: 该论文旨在解决多模态交互系统中语言模型与语音、图像等异构模态输入输出之间的接口不统一、时序对齐困难以及任务驱动决策能力不足的问题。其核心解决方案在于构建一个以40亿参数语言模型为骨干的模块化多模态助手DRVEXO-4B-Unified,通过引入时间残差连接(temporal residual connector)实现声学序列向语言接口的自适应映射,利用显式文本边界支持增量语音合成,结合三类任务条件评分头(task-conditioned scoring heads)对动作候选与生成工具调用进行联合排序,从而将对话流与应用定义的行为选择相耦合,同时保留动作授权的外部控制权。该设计在保持高效推理性能的同时,显著提升了跨模态理解与生成的准确性与响应速度,在MMMU、IFEval和MATH等多个基准测试中均超越现有同类模型,尤其在无测试时缩放(test-time scaling)条件下展现出优越的端到端表现。

链接: https://arxiv.org/abs/2611.25521
作者: Zeyuan Li
机构: Yujie Intelligence(驭杰智能); Hangzhou, China(杭州, 中国)
类目: Computation and Language (cs.CL)
备注: 23 pages, 5 figures

点击查看摘要

Abstract:We present DRVEXO-4B-Unified, a modular multimodal assistant built around a 4B language backbone. A self-contained checkpoint combines text, image, and speech inputs with reference-conditioned speech output and structured interaction. A temporal residual connector adapts acoustic sequences to the language interface; explicit text boundaries support incremental speech synthesis; and three task-conditioned scoring heads rank supplied action candidates alongside generative tool calls. These interfaces connect conversation to application-defined choices while leaving action authorization to the application. The system obtains 74.22% on all 900 MMMU validation questions, exceeding the published validation scores of Gemma 3 4B IT and InternVL3.5-4B without test-time scaling. Strict prompt-level IFEval accuracy is 82.26% and MATH-500 accuracy is 89.40%, above the original Qwen3-4B non-thinking report’s 81.2% and 84.8%, respectively. On NVIDIA GB10, the model achieves a speech-synthesis first-chunk latency of 390 ms. For concurrent text generation, a growth-aware admission policy completes a 128-answer batch at 594.41 output tokens/s, including filling and draining, without preemptions. We plan to release the model weights and accompanying code under Apache License 2.0, retaining applicable third-party terms.

[NLP-59] When Should a Voice Agent Speak? End-of-Turn Detection in Turkish

【速读】: 该论文旨在解决语音代理在用户沉默时作出回应时机判断的难题,即如何准确区分用户话语结束(turn end)与话语内部停顿(turn-internal pause),以避免过早打断或延迟响应。其核心解决方案在于利用经过微调的土耳其语TurnGPT模型(fine-tuned Turkish TurnGPT model)对语音对话中的沉默进行语义层面的识别,相较于零样本土耳其语语言模型(zero-shot Turkish language model)和多语言部署检测器,在区分话语结束与内部停顿方面表现更优。研究进一步发现,当前发言者的时间特征(timing features)具有显著信息量,将其与微调后的文本得分结合后,可实现最佳的响应时机预测效果;通过逻辑回归建模,该组合性能亦优于单独使用任一基线模型。此外,在延迟模拟实验中,将文本得分引入静音超时机制,可在保持相同误打断率的前提下显著降低平均响应延迟。该方法在自动转录的Whisper数据上仍保持对零样本模型的优势,验证了其鲁棒性与实用性。

链接: https://arxiv.org/abs/2611.25489
作者: Ahmet Tuğrul Bayrak,Fatma Nur Korkmaz,Mustafa Sertaç Türkel,Nur Yağmur Tuncer
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:A voice agent that waits for a fixed silence before replying either interrupts users who pause mid-utterance or responds late to users who have finished. This study examines the decision at the onset of silence, using turn ends and turn-internal pauses extracted from the Real-TurnTurk corpus of Turkish dyadic video calls. A fine-tuned Turkish TurnGPT model, a zero-shot Turkish language model and a deployed multilingual detector are compared, together with voice-activity timing and facial features, under recording-disjoint and speaker-disjoint evaluation with paired cluster bootstrap tests. On held-out recordings, the fine-tuned model separates turn ends from pauses better than both text baselines and is the best calibrated of the three. Timing features of the current speaker are similarly informative, and their combination with the fine-tuned text score gives the best timing-based results; with logistic regression, this combination also outperforms the same combination with either text baseline. In a latency simulation, adding the text score to a silence timeout shortens the mean response delay at the same interruption rate. On automatic Whisper transcripts, the fine-tuned model retains its advantage over the zero-shot model.

[NLP-60] PILAPS: A Bilingual Benchmark for Evaluating Prag matic Inference across Phenomena and Settings

【速读】: 该论文旨在解决大语言模型(Large Language Models, LLMs)在英语中表现出的强语用能力是否能够跨语言和评估场景有效泛化的问题。随着LLMs在全球多语言、多文化及多样化对话情境中的广泛应用,其语用理解的跨语言一致性成为关键挑战。为此,研究提出PILAPS——一个涵盖英汉双语的语用现象基准测试集,覆盖八类语用现象,包含原创文本与对话数据,并融合文化适配与独立来源的数据以增强多样性与真实性。研究评估了来自四个不同家族的八款不同规模与提示策略的LLM。结果表明,模型在语用理解上存在显著差异,开源模型尤其在幽默理解方面表现不佳;跨语言性能受模型来源影响较大。错误分析显示,相比人类参与者,模型更倾向于在对话语境中采取字面解读,而心智理论(Theory-of-Mind)提示可降低字面错误比例,链式思维(Chain-of-Thought)提示则略微增加此类错误。综合结果表明,单一语境下的优秀语用表现无法保证其在其他语言或对话场景中的泛化能力。

链接: https://arxiv.org/abs/2611.25459
作者: Jiani Cao,Anna Korhonen,Chen Cecilia Liu
机构: University of Cambridge(剑桥大学)
类目: Computation and Language (cs.CL)
备注: 23 pages, 4 figures, 19 tables

点击查看摘要

Abstract:Recent work reports strong pragmatic performance by large language models (LLMs) in English, yet whether this performance generalises across languages and evaluation settings remains unclear. This question is increasingly important as LLMs are deployed across languages, cultures, and conversational contexts. In this paper, we first introduce PILAPS, a bilingual English/Mandarin Chinese benchmark covering eight pragmatic phenomena across purpose-written and dialogue-derived items, with both culturally adapted and independently sourced data. We then evaluate eight LLMs from four families across a range of model sizes and prompting strategies. Our results reveal substantial variation across models and pragmatic phenomena: open-source models continue to struggle particularly with humour, while cross-linguistic performance varies with model origin. Our error analysis shows that, compared with human participants, models are more prone to literal interpretations in conversational contexts. Theory-of-Mind prompting lowers the literal share of errors, whereas Chain-of-Thought prompting slightly raises it. Together, these findings demonstrate that strong pragmatic performance in one evaluation setting does not necessarily generalise across languages or conversational contexts.

[NLP-61] PAIR-MED: Policy-reward Alternation for Iterative Refinement in Multilingual Medical Reasoning

【速读】: 该论文旨在解决多语言医疗问答中推理能力不均衡的问题,尤其关注在强化学习过程中因策略(policy)演化导致过程奖励模型(Process Reward Model, PRM)与策略轨迹分布逐渐失配,从而降低奖励信号可靠性这一关键挑战。其核心解决方案是提出一种策略-奖励交替训练框架(policy–reward alternating training),通过交替更新医疗领域专用的PRM与多语言推理策略,实现二者协同优化。该框架在监督初始化后,利用当前策略生成多语言推理轨迹以动态更新PRM,同时结合医学步骤级监督、过程与结果奖励以及语言相关奖励,确保奖励模型保持医学准确性并有效引导策略学习。基于此,研究构建了8B和9B规模的多语言医疗PRM(PAIR-PRM)与推理模型(PAIR-MED),在Global-MMLU和MMLU-ProX的医疗子集上,平均准确率相较基线Qwen模型提升3.01%至7.47%;在CUREMED-BENCH评测中,PAIR-MED-9B在域内与域外语言上的联合语言一致性与逻辑准确性分别提升12.88%和7.45%。实验分析进一步验证了奖励模型适应性、多语言推理质量及各训练组件的有效贡献。

链接: https://arxiv.org/abs/2611.25446
作者: Fan Gao,Xiaohang Xu,Lang Huang,Nan Lu,Irene Li
机构: The University of Tokyo(东京大学); National Institute of Informatics(情报学综合研究所); University of Bristol(布里斯托尔大学)
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Large language models demonstrate strong medical question-answering capabilities, yet reasoning performance remains uneven across languages. Process reward models (PRMs) provide step-level feedback for improving reasoning, but a fixed PRM may become increasingly misaligned with the trajectory distribution of an evolving policy during reinforcement learning, thereby reducing the reliability of its supervision. We propose a policy–reward alternating training framework that jointly adapts a medical PRM and optimizes a multilingual reasoning policy. Following supervised initialization, the current policy generates multilingual trajectories to update the PRM, which then provides process-level rewards for the next policy update. Medical step supervision maintains the PRM’s medical foundation, while process, outcome, and language-related rewards guide policy learning. We develop multilingual medical PRMs (PAIR-PRM) and reasoning models (PAIR-MED) at 8B and 9B scales across 12 languages. On the medical subsets of Global-MMLU and MMLU-ProX, our models improve average accuracy by 3.01–7.47% over their corresponding Qwen backbones. On CUREMED-BENCH, PAIR-MED-9B improves the joint language consistency and logical accuracy metric by 12.88% and 7.45% over baselines for in-domain and out-of-domain languages. Further analyses examine PRM adaptation, multilingual reasoning quality, and the contributions of individual training components.

[NLP-62] RoCoKV: Accelerating LLM Inference Serving through Uncertainty and Redundancy Aware KV Cache Compression

【速读】: 该论文旨在解决大模型推理过程中因长序列生成推理轨迹导致的KV缓存(KV cache)占用过高问题,尤其针对推理过程中的冗余与低效token(如循环、自我验证和修正产生的重复内容)带来的内存开销。这类冗余token虽在注意力机制中获得高注意力分数,但实际信息量低,难以直接移除而不损害推理质量。为应对这一挑战,论文提出RoCoKV方法,其核心解决方案包含两个关键创新:一是基于不确定性引导的自适应剪枝机制,根据模型置信度动态调整压缩强度,在模型不确定阶段保留更多缓存以保障推理鲁棒性,而在高置信度阶段施加强剪枝以节省内存;二是冗余感知的token选择策略,通过识别与近期推理内容高度相似的token,降低其保留概率,从而优先保留更具信息量和多样性的token。此外,作者还协同优化服务架构,设计了融合三阶段的CUDA解码内核,并将其集成至vLLM作为即插即用插件,兼容其分页式KV缓存布局与CUDA图捕获特性,将算法层面的内存节省转化为端到端的实际性能提升。实验结果表明,RoCoKV在多个推理基准上均显著优于现有先进压缩方法,在有限的KV缓存预算下保持更高推理准确率,并在AIME25和长合成基准上分别实现1.27倍和1.43倍的速度提升,吞吐量提升达38%和45%。

链接: https://arxiv.org/abs/2611.25433
作者: Payman Behnam,Elaheh Sobhani,Saad Amin,Yahya Hassan,Daniyal Khan,Biswadeep Chakraborty,Tanish Patwa,Souvik Kundu,Alexey Tumanov
机构: 未知
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:KV cache compression is critical for large reasoning models, where most of the cost comes from long generated reasoning traces rather than the input prompt. These traces often contain redundant and low-utility tokens due to looping, self-verification, and corrections. However, such tokens may still receive high attention scores, making it difficult to remove them without harming the reasoning process. We propose RoCoKV, a KV cache compression method designed for large reasoning models. RoCoKV introduces two key components. First, uncertainty-guided adaptive pruning adjusts the compression strength based on the model’s confidence, allocating more memory during uncertain phases and applying stronger pruning when the model is confident. Second, redundancy-aware token selection reduces the impact of tokens that are highly similar to recent reasoning content, encouraging the model to retain more informative and diverse tokens. Beyond the algorithm, we co-design RoCoKV with the serving stack: we develop fused three-phase CUDA decode kernels and integrate them into vLLM as a drop-in plugin that preserves its paged KV layout and CUDA graph capture, turning algorithmic savings into real end-to-end speedups. Extensive experiments across three reasoning benchmarks and two reasoning models demonstrate that RoCoKV consistently improves reasoning accuracy under constrained KV cache budgets, achieving higher accuracy compared to state-of-the-art KV cache compression methods. For DeepSeek-R1-Distill-Llama-8B in vLLM, RoCoKV achieves a 1.27X speedup and 38% higher throughput on AIME25, the most challenging evaluated reasoning benchmark, and a 1.43X speedup and 45% higher throughput on the long synthetic benchmark, compared with Full-KV.

[NLP-63] Evolve Where It Matters: Coverag e-Guided Self-Evolution of Agent Skills

【速读】: 该论文旨在解决自演化智能体(self-evolving agents)在技能优化过程中无法准确区分技能失败是由于指令错误还是执行偏差所致的问题。现有方法通常以任务整体成败为评价标准,难以定位具体失败原因,导致演化效率受限。其解决方案的关键在于引入指令级覆盖率(instruction-level coverage),将每次运行中的每条指令标记为“已执行”、“未执行”或“未覆盖”,从而精确分离出指令错误与执行遗漏两类根本原因。基于此,作者提出SkillPlumber框架:通过一个持续进行根因分析的代理(root cause analysis agent),将覆盖率标签转化为可操作的修订问题,并在后续运行中验证修改效果;同时,维护者代理(maintainer)将跨任务的诊断经验提炼为共享策略库,实现知识复用。实验表明,SkillPlumber在三个基准测试和三种执行器上均显著提升任务成功率,相较初始技能提升25.6至30.4个百分点,优于最强基线8.7至14.0个百分点,且移除覆盖率机制后在SkillsBench上失败恢复率下降20.8%,验证了其核心作用。

链接: https://arxiv.org/abs/2611.25429
作者: Boyin Tan,Jiadong Zhang,Xiaosong Ma,Chun Jason Xue,Xiaowei Huang,Youcheng Sun
机构: Mohamed bin Zayed University of Artificial Intelligence(穆罕默德·本·扎耶德人工智能大学); University of Liverpool(利物浦大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 15 pages, 5 figures

点击查看摘要

Abstract:Agent skills package procedural knowledge as natural-language instructions, and a skill helps only when its instructions are correct and properly followed by the agent. Self-evolving agents revise their own skills from the runs they make, but existing methods judge each run as a whole, by its outcome or by where it went wrong, and cannot tell whether a failure comes from a wrong instruction or from one the agent did not follow. Instruction-level coverage was introduced to test skills and labels each instruction of a run as follow, not-follow, or not covered, which separates these causes. We leverage it for skill self-evolution with SkillPlumber, a framework in which an agent evolves its own skills where coverage shows they fall short. A persistent root cause analysis agent turns each label into a revision question and checks each revision against the next run, while a maintainer distills investigations across tasks into a shared strategy library. Across three benchmarks and three executors, SkillPlumber solves 64.7% to 70.5% of the tasks, 25.6 to 30.4 percentage points higher than the initial skill and 8.7 to 14.0 percentage points higher than the strongest baseline, and removing coverage reduces recovered failures by 20.8% on SkillsBench with GPT-5.5.

[NLP-64] ChronoDossier: Long-Form Personas as Temporal Memory Representations ICLR2027 ICLR

【速读】: 该论文旨在解决长时程智能体(long-horizon agents)在动态变化的环境中如何有效存储、检索并呈现个体状态信息的问题,核心挑战在于确保记忆记录的准确性与可恢复性。其解决方案的关键在于构建一个名为ChronoDossier的基准数据集,通过311个由7个生成器基于时间线生成的合成人物档案(每份5000–10000词),以模拟真实个体的长期身份表征——即不变的身份属性与随生命阶段演化的事实、习惯及决策规则的时间序列。该基准采用内在指标(intrinsic metrics)评估记录与其目标状态的一致性,并通过外在指标(extrinsic metrics)测试五名阅读者在32人设GPT-5.1面板上的状态恢复能力,实现对记忆构建与证据传递过程的系统性检验。实验表明,完整段落修复能显著提升模型在“包装”(packaging)类错误中的表现,而引入标签范围控制的写入输入可减少无关内容的锚定复制,验证了结构化记忆设计的重要性。此外,该框架支持离线评估,但未涵盖交互式记忆获取与真实用户验证等动态过程。

链接: https://arxiv.org/abs/2611.25407
作者: Noah Sturgill,Jinho D. Choi
机构: Emory University (埃默里大学)
类目: Computation and Language (cs.CL)
备注: 28 pages, 4 figures, 32 tables. Under review at ICLR 2027. Code and data: this https URL

点击查看摘要

Abstract:Long-horizon agents rely on stored records to recover a person’s state as facts change. Reliable answers depend on how facts are recorded, retrieved, and presented to the reader. We treat a long-form persona dossier as a memory representation of a known latent state, an invariant identity plus a timeline of facts, routines, and decision rules tied to life phases. We present ChronoDossier, a benchmark of 311 synthetic dossiers of 1k-10k words written by seven generators from dated timelines. Intrinsic metrics evaluate the records against their intended states; extrinsic metrics test recovery by five readers on a seeded 32-persona GPT-5.1 panel balanced across lengths (520 rated turns per condition). The linked states, records, evidence packets, and answers support tests of construction and evidence delivery. In a contextual-stress control, tag-scoped writer inputs remove detected out-of-scope anchor copying. Frozen questions authored on staged dossiers transfer less reliably to one-pass dossiers of the same states. In matched evidence controls, whole sections repair 29 of 32 GPT-5.1 and 25 of 30 Sonnet 4.6 failures labeled as packaging, compared with 2 of 8 and 1 of 8 answer-formation failures. Attribution uses separate key-based and key-free support measures, with model disagreements retained. Length remains confounded with persona and question composition; the observed between-bucket widening depends on excerpt construction. The benchmark supports offline evaluation against known facts. Memory acquisition from interaction and validation against real users remain outside its scope.

[NLP-65] Incorporating argumentation frameworks for LLM decision-making: Task-specific design and evaluation

【速读】: 该论文旨在解决大语言模型(LLM)在决策任务中引入论证框架(Argumentation Framework, AF)时,现有方法在不同任务上表现不一致、性能提升不稳定的问题。其核心挑战在于论证的构建方式、评估机制及与决策过程的融合策略缺乏针对具体任务的优化,导致泛化能力不足。解决方案的关键在于提出两种任务特异性方法:一是Qnt-Arg,通过围绕辩论材料中的原始发言者组织论据,并基于渐进语义(gradual semantics)聚合模型估计的论据强度与候选答案先验;二是Qlt-Arg,利用可信度估计对冲突证据项建立偏好关系,并采用基于实证的语义(grounded semantics)筛选可接受的证据以支持后续的LLM推理。实验表明,这两种方法在辩论式决策和基于检索的事实验证任务中均显著优于基线,验证了将形式化论证与大语言模型结合的有效性,同时强调了论证构建、模型选择及形式语义推理应与具体任务目标对齐的重要性。

链接: https://arxiv.org/abs/2611.25361
作者: Bochen Han,Anshu Xiong,Mingyang Ma
机构: Chinese Academy of Sciences (中国科学院); University of Chinese Academy of Sciences (中国科学院大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Argumentation frameworks (AFs) provide explicit, formal mechanisms for representing and resolving conflicts and thus have been incorporated into large language model (LLM) decision-making. The benefits vary and depend on how arguments and their relations are constructed, evaluated, and incorporated into the decision process. We abstract a workflow consisting of four stages based on existing argumentation-augmented methods and investigate their design choices over tasks of debate-based decision-making and retrieval-based fact verification. Our evaluation with recent LLMs shows that existing methods do not consistently improve performance across the considered tasks. Motivated by this observation, we develop two task-specific methods. The first is called Qnt-Arg, which organizes arguments around the original speakers in debate materials and aggregates model-estimated argument strengths and candidate-answer priors under gradual semantics. The second is Qlt-Arg that uses credibility estimates to establish preferences between conflicting evidence items and applies grounded semantics to select acceptable evidence for subsequent LLM inference. We examine the roles of generative and decision models within the two methods. In the debate setting, Qnt-Arg consistently outperforms other argumentation-augmented methods across matched configurations; Qlt-Arg improves retrieval-based fact-verification performance compared with the baselines. These findings demonstrate the value of combining formal argumentation and LLM while highlighting the importance of aligning with tasks in argumentation construction, model selection, and the role of formal semantic reasoning.

[NLP-66] Pro2 Agent : Scientific Problem Probing for Autonomous AI RD

【速读】: 该论文旨在解决当前大型语言模型(LLM)代理在自主进行人工智能研究时,因缺乏对任务本质结构的显式建模而容易陷入错误解空间的问题。传统方法仅通过迭代优化代码库来推进研究,忽视了对问题核心障碍与可行解决方案之间关系的理解,导致在复杂任务上表现受限。其解决方案的关键在于提出Pro² Agent,将自主AI研究的核心状态从代码库状态转变为“问题状态”(Problem State),即显式构建关于阻碍成功的根本问题及其对应潜在解决方案的结构化认知。Pro² Agent通过全面分析可用资源与约束初始化问题状态,并基于受控假设检验机制,动态评估特定问题的重要性或候选方案的有效性,实现对问题理解与解决方案的精准迭代优化。实验结果表明,在完整的MLE-Bench基准上,该方法在24小时预算下达到70.7%的最高奖牌率,48小时预算下进一步提升至76.0%,为当前最优表现;同时在PostTrainBench上也展现出竞争力。此外,系统生成的研究笔记使整个研究过程具备可解释性与可复用性,为人类研究人员提供有效支持。

链接: https://arxiv.org/abs/2611.25357
作者: Yixiong Chen,Liqian Peng,Tian Tan,Ziqian Lin,Blake Jianhang Chen,Chao Zhao,Yaojie Liu,Alec Go,Alan Yuille
机构: Johns Hopkins University (约翰霍普金斯大学)
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Large language model (LLM) agents can now autonomously conduct AI research by writing code and running experiments. They typically do so by iteratively optimizing a codebase, while ignoring the problem structure behind the task. This formulation is effective on simpler tasks, but on difficult problems it often gets stuck in the wrong solution region because the agent never explicitly represents its understanding of the problem. We introduce Pro ^2 Agent, which shifts the primary state of autonomous AI research from a codebase state to a Problem State: an explicit structure of what problems prevent success and which solutions are likely to address them. Pro ^2 Agent initializes the Problem State through comprehensive analysis of available resources and constraints. It then performs targeted research to relate the identified problems to candidate solutions. Through controlled hypothesis testing, the agent updates the importance of a suspected problem or the benefit brought by a candidate solution. On the complete MLE-Bench, Pro ^2 Agent matches the best reported medal rate (70.7%) under the standard 24-hour budget and reaches 76.0% with a 48-hour budget, the highest medal rate reported on the complete benchmark. It also achieves competitive performance on PostTrainBench. In addition, it produces research notes for each task that make the research process transparent and reusable by human researchers.

[NLP-67] Enhancing Understanding of Language Models with Varied Explanations

【速读】: 该论文旨在解决生成式 AI(Generative AI)在复杂推理任务中支持人类判断时,其决策过程缺乏可解释性与可验证性的问题。核心挑战在于如何生成既忠实于模型内部计算逻辑(faithfulness)、又对用户具有可理解性(plausibility)并能有效辅助其完成任务(utility)的解释。论文的关键解决方案包括:提出两种评估框架,用于多维度、跨解释类型地评价高亮式解释的性能,并构建元评估框架以检验现有链式思维(chain-of-thought)解释的忠实性度量方法的可靠性;开发两种生成自然语言解释的技术,确保解释在保持对模型预测过程忠实性的同时,具备简洁性、连贯性和认知助益性;并通过人机交互研究揭示证据与解释在事实核查场景中帮助用户验证模型结论的关键作用。

链接: https://arxiv.org/abs/2611.25314
作者: Jingyi Sun
机构: 未知
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Language models are increasingly used to support human judgement in complex reasoning tasks, yet their predictions often remain difficult to interpret and validate. Explainability techniques provide insight into model decision-making, helping expose errors, failure modes, and potential bias so users can decide whether to rely on model outputs. Explanations are simplified approximations of a language model’s internal workings. The goal is to obtain explanations that reflect the underlying computation as closely as possible (faithfulness), are understandable to users (plausibility), and help them achieve their intended goals (utility). Evaluation protocols are essential for assessing these dimensions. This thesis studies highlight explanations, natural language explanations, and chain-of-thought reasoning. These forms offer complementary perspectives on model behaviour and exhibit distinct strengths and limitations across tasks. Producing explanations that satisfy multiple quality dimensions remains challenging, as does evaluating them reliably across forms and settings. This thesis aims to improve people’s understanding of models’ decision-making processes in complex reasoning tasks. First, it proposes two evaluation frameworks for assessing highlight explanations across multiple properties and explanation types, together with a meta-evaluation framework for examining the reliability of existing faithfulness metrics for chain-of-thought explanations. Second, it proposes two techniques that generate natural language explanations designed to be faithful to the model’s prediction process while remaining concise, coherent, and helpful for understanding model decisions. Finally, a study of human-AI interaction explores the roles of evidence and explanations in helping users validate model verdicts in a fact-checking setting.

[NLP-68] LLM s Are Already Confidence-Aware Zero-Shot Classifiers

【速读】: 该论文旨在解决零样本大语言模型(Zero-shot Large Language Models, LLMs)在文本分类任务中可信度评估的问题,即如何有效衡量和提升模型输出标签的置信度与可靠性。其核心挑战在于:尽管零样本LLMs可直接用于分类,但其生成结果的置信度指标(如标签概率)是否准确、可校准,直接影响下游应用的可信性。解决方案的关键在于对比黑盒方法(如语义化置信度估计与采样)与白盒方法(基于分词器概率P(True)及多种对数几率评分方法,结合约束生成)的性能差异。研究发现,基于对数几率的评分方法显著优于黑盒方法,在判别能力(AUROC 0.74–0.77)上表现更佳;同时提出“仅决策评分”(decision-only scoring)策略,仅在前缀树(trie)中标签分歧位置进行评分,大幅减少解码步骤88%–96%,且保持与全标签评分相当的性能。此外,使用单标记符替代原始标签虽保留判别能力,但平均损失3–6个百分点的准确率。最终,最优开源模型配置(Gemma 4 31B与Qwen 3.5 9B)在多个数据集上达到与商业模型Jev 1.13.0相当的性能,验证了高效置信度评估机制的有效性。

链接: https://arxiv.org/abs/2611.25289
作者: Owen Cook,Xingyi Song
机构: University of Sheffield(谢菲尔德大学); Department of Computer Science(计算机科学系)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 20 pages (9 pages main content), 2 figures. Code available at: this https URL

点击查看摘要

Abstract:Zero-shot large language models (LLMs) are increasingly used as text classifiers, but their labels are only useful when we know to what extent we can trust them. We compare black-box methods (verbalised confidence and sampling) with white-box methods that read token probabilities (P(True), and multiple logit-scoring methods using constrained generation). We evaluate eleven open-weight LLMs (1B-32B parameters) on five classification datasets with 3 to 151 labels, measuring classification performance (accuracy, F1-macro), discrimination (AUROC), and calibration (ECE and Brier score). The choice of confidence method rarely impacts accuracy significantly, but logit-based scoring methods lead to much better discrimination (AUROC 0.74-0.77) than black-box methods (AUROC 0.57-0.66). Replacing labels with single-token identifiers preserves discrimination but costs 3-6 accuracy points on average. We also introduce decision-only scoring, which only scores positions where labels diverge in the prefix tree (trie). It reduces decode steps by 88-96% on the three largest label sets and performs comparably to full-label scoring. The best open-weight model-method configurations (Gemma 4 31B and Qwen 3.5 9B) are competitive with the commercial Jev 1.13.0 classifier.

[NLP-69] When Objectives and Evaluators Disagree: Evaluating Transfer for Post-hoc Demographic Framing Mitigation EACL2027

【速读】: 该论文旨在解决大语言模型(Large Language Model, LLM)生成文本中存在的代际与性别等人口统计学偏差问题,尤其在源模型不可访问时,如何通过后处理重写(post-hoc rewriting)有效降低此类偏差。其核心挑战在于:现有重写方法可能仅因提升了自身目标得分或满足了选择器标准而看似成功,却未真正实现跨维度的公平性改善。解决方案的关键在于设计可迁移的重写策略,并通过独立评估者对成对文本的说服力偏差指数(Persuasion Bias Index, PBI)差异进行客观衡量。研究发现,固定策略(始终针对PBI较低的消息进行编辑)在不同属性轴和结果聚合方式间转移性能不一致;而自适应策略(在双方消息上均提出修改,但最多接受一个显著改进项)在八组比较中的七组表现更优,主要得益于对更多样本对进行编辑,而非依赖精准的方向选择,尽管存在更高的内容保真度风险。此外,同时编辑双端消息的策略虽在少数样本上实现最大偏差降幅,但整体覆盖范围较小。研究强调,偏差缓解效果必须在目标、评估者、提示模板及编辑策略之间进行多维度交叉验证,才能真实反映其有效性。

链接: https://arxiv.org/abs/2611.25258
作者: Tunazzina Islam
机构: Purdue University (普渡大学); University of Southern California (南加州大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG); Social and Information Networks (cs.SI)
备注: Accepted to EACL 2027 Main Conference

点击查看摘要

Abstract:Post-hoc rewriting is used to reduce demographic disparities in LLM-generated text when the source model is inaccessible. A rewrite can look successful merely because it improved its own target or satisfied its selector. We study this as a transfer problem on 660 gender pairs and 1,980 age contrasts of demographic-conditioned climate messages. A rewriter proposes edits, a selection model chooses among constraint-satisfying candidates, and an independent evaluator uninvolved in either rates pairwise framing disparity. A fixed policy that always targets the message with the lower Persuasion Bias Index (PBI) transfers inconsistently across axes and outcome aggregations. An adaptive policy that proposes edits on both sides but accepts at most one significantly improves on that rule, applied to the same candidates, in seven of eight comparisons across two evaluator families, mostly by editing more pairs, not by guided direction choice, at higher preservation risk. A policy editing both messages gives the largest point reductions on fewer pairs. These two policies reduce disparity across a preservation-threshold sweep; absolute ratings vary across evaluator families and prompts. Shrinking the PBI gap increases dimension-mean disparity under both evaluator families. Mitigation should be reported as transfer across objectives, evaluators and their prompts, and editing policies.

[NLP-70] Steering Automated Scientific Discovery via Audited Epistemic Exploration

【速读】: 该论文旨在解决科学发现过程中研究主体因认知局限导致的信念固化、证据支持不足及探索空间狭窄等问题,具体表现为研究者在面对新证据时仍坚持不成立的假设、无法有效识别未充分探索的方向,以及对矛盾证据缺乏及时修正。其解决方案的关键在于提出一种名为“认知罗盘”(Epistemic Compass, EpiCompass)的可插拔框架,通过构建一个关联假设、证据、信念及其修正过程的认知树(epistemic tree),显式维护研究者的认知状态(epistemic state)。EpiCompass 在探究过程中持续评估已有假设与累积证据的一致性,利用未解决的矛盾和未探索方向驱动新假设生成,并基于预期认知收益(epistemic gain)筛选候选假设,兼顾现有假设的精细化与假设空间的拓展。最终结论需经由累积证据验证,从而实现更稳健的科学推理。实验结果表明,EpiCompass 在三个科学发现基准任务上均显著优于基线模型,且有效减少了未经证据支持的断言和信念未能修正等错误。

链接: https://arxiv.org/abs/2611.25214
作者: Fangxu Yu,Tao Feng,Xuying Ning,Haolin Yang,Jiaxuan You,Tianyi Zhou
机构: University of Maryland, College Park; University of Illinois at Urbana–Champaign; Mohamed bin Zayed University of Artificial Intelligence
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: Technical Report

点击查看摘要

Abstract:Scientific discovery requires revisiting the hypotheses in light of new evidence and identifying promising directions beyond those already explored. However, research agents often make claims that are not supported by their observations, explore a narrow hypothesis space, and retain beliefs despite recognizing conflicting evidence. These failures motivate the search for an explicit epistemic state that captures both beliefs about existing hypotheses and directions for further exploration. To this end, we propose Epistemic Compass (EpiCompass), a plug-and-play scaffold that maintains this state through an epistemic tree linking hypotheses, evidence, beliefs, and revisions. As an investigation progresses, EpiCompass reassesses the hypotheses explored against accumulated evidence and uses unresolved conflicts and unexplored directions to guide hypothesis generation. It selects candidates by anticipated epistemic gain, accounting for both refinement of existing hypotheses and expansion of the hypothesis space, and verifies the final conclusion against the accumulated evidence. Experimental results across three scientific discovery benchmarks and three LLM backbones demonstrate that EpiCompass consistently outperforms baselines, with further analyses showing reductions in errors of unsupported claims and failures to revise beliefs.

[NLP-71] Learning to Read Events for Exact State Tracking with Frozen Language Models

【速读】: 该论文旨在解决长事件流(long event streams)中顺序识别并正确应用每个状态变更的难题,尤其在无法访问模型内部结构仅能通过API交互的限制下,如何高效、准确地处理具有复杂排列组合特性的事件序列。其核心挑战在于:对于五项元素的排列积(products of permutations of five items)这一任务,在标准假设 TC0≠NC1\mathrm{TC}^0 \ne \mathrm{NC}^1 下,固定深度、对数精度的Transformer架构无法在所有输入长度上求解。解决方案的关键在于引入一个在冻结语言模型上训练的“选择器”(selector),通过API-only方式驱动模型逐个识别事件转换(transition),并由外部整数寄存器(integer register)按序组装这些变更;同时结合置信度规则以处理不确定更新或保留候选方案。实验表明,该方法显著提升了模型在复杂事件流中的推理与执行能力——例如在43,687个事件窗口中成功读取全部299个事件,且在新机器上实现接近99%的准确率;在1,000事件规模的房间交换文档任务中,结合提取与寄存器机制后,所有测试模型均能无链式思维(chain-of-thought)请求下完成全部任务,同时大幅降低推理时间与成本。该框架为在有限计算资源和接口约束下实现高精度、可扩展的状态演化建模提供了有效路径。

链接: https://arxiv.org/abs/2611.25211
作者: Jeonghoon Lee(Attractor Dynamics Inc.)
机构: Attractor Dynamics Inc.
类目: Machine Learning (cs.LG); Computation and Language (cs.CL)
备注: 11 pages, 5 figures. Data and code: this https URL

点击查看摘要

Abstract:Long event streams require identifying each change and applying the changes in order. We train a selector on a frozen language model to read each transition. With API-only access, we ask the model to list the transitions. An external integer register composes them in order. A confidence rule can flag uncertain updates or retain several candidates. We study products of permutations of five items, a task that fixed-depth, log-precision transformers cannot solve on all input lengths under the standard conjecture that \mathrmTC^0 \ne \mathrmNC^1 . On two-event windows, a 27B model answered 3 of 100 correctly without visible reasoning and 96 of 100 when it generated intermediate steps. A frozen 0.8B model with a trained selector read every event in 299 of 299 windows of 43,687 events each. The same package passed 298 of 299 new windows on a fresh machine, missing the target of 0.99 for the one-sided 95% lower confidence bound. A separate follow-up sized to allow one failure yielded 472 of 473. On 1,000-event room-swap documents, three of ten models solved all five documents by reasoning alone. All seven models tested both ways at this length solved all five with extraction and the register, with no chain of thought requested. With concurrent extraction, median wall-clock time fell from 136.1 to 10.4 s for Sonnet 5.5 and from 313.5 to 28.0 s for GPT-6.1 Sol. At fixed per-token rates, mean cost per document fell by 55% for Sonnet 5.5 and 44% for GPT-6.1 Sol. A regular expression with an executor also solved all 15 fixed-format documents. On 500 free-form issue-tracker comments, a separately adapted 4B reader identified the recorded value in 68.8% of cases, compared with 61.0% for a frequency baseline fitted on the same training projects.

[NLP-72] Blind Spots Travel Knowledge Stays Home: What RL Inherits from an LLM Judge

【速读】: 该论文旨在解决在缺乏标准答案(answer key)的情况下,使用大语言模型(LLM)作为评判者(judge)进行强化学习训练时,其知识能否有效传递至目标模型的问题。研究发现,评判者对具体答案的知识无法有效迁移,而其对问题类别的判断偏好却会显著影响模型行为:例如,当评判者错误地拒绝12个关于前国家首都的正确答案中的11个时,模型在新问题上对这类问题的拒绝率高达98%,这种偏差无法通过准确率或拒答率等指标揭示。关键解决方案在于,无需评判者具备领域知识,仅需让其奖励模型选择与其自身未训练样本中多数意见一致的答案,即可在无标注数据和无强判别器的情况下,获得与标准答案相当的奖励信号,并保持对特定问题类别的正确响应能力。这表明,模型性能的关键不在于评判者的绝对知识,而在于其对答案一致性的一致性偏好。

链接: https://arxiv.org/abs/2611.25191
作者: Liang Yue,Eric A. Vance
机构: University of Colorado Boulder(科罗拉多大学博尔德分校); Boulder, Colorado, USA(科罗拉多州博尔德市,美国)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 31 pages, 6 figures, 27 tables. Code and data: this https URL

点击查看摘要

Abstract:LLM judges serve as rewards where no answer key exists, in the hope that their knowledge passes to the trained model. In 814 RL runs that teach 7-8B models to answer closed-book questions or abstain, we ask what passes. What a grader knows about single answers stays home. A wrong answer that a judge rewards is memorised for that one question. A model trained with the answer key, or with a far stronger judge, chooses which new questions to answer no better than the untrained model does by sampling 32 answers and abstaining where they disagree; the grader’s knowledge adds about one corrected answer per hundred questions. What a grader believes about a kind of question travels. A judge that rejects eleven of the twelve correct training answers about capitals of former states teaches the model to refuse 98% of such capitals on new questions, a loss that accuracy and abstention rates do not reveal. Changing only those twelve verdicts by hand moves the refusal from 5% to 90%, even for a grader that accepts as many answers as the answer key; at that leniency, two of four judges already cost 38-42% of these capitals. The capitals are not forgotten: with “I don’t know” blocked, the model answers 94% of them correctly. The grader therefore needs no knowledge of its own: rewarding the model for giving the answer that most of its own untrained samples agree on, with no labels and no judge, earns the answer key’s reward and keeps the category.

[NLP-73] Calibrated Decision Rewards: Training Language Models with a Judge That Knows How Sure It Is

【速读】: 该论文旨在解决强化学习(Reinforcement Learning, RL)在训练语言模型时面临的奖励信号局限性问题:传统方法依赖可识别的奖励,而现有答案检查器需参考答案(reference answer),学习得到的奖励模型和大语言模型(LLM)裁判虽具备泛化能力,但其输出分数并非概率分布,难以用于精确的策略优化。为此,本文提出基于校准决策奖励的强化学习方法(Reinforcement Learning from Calibrated Decision Rewards, RLCDR)。其核心解决方案是将大语言模型转化为一个决策模型,通过单次前向传播生成对响应正确性或错误步骤的判断概率分布,并通过对选项顺序的去偏处理与约百条标注样本进行校准,使奖励输出具备可解释的概率意义。实验表明,在MATH数据集上使用GRPO算法训练Qwen3-1.7B模型时,该无参考奖励机制将数学解题准确率从72.2%提升至76.8%(+4.65点,p < 1e-4),接近基于标准答案训练的效果(76.9%),显著优于随机奖励(仅+0.65点),且在难题上的性能增益更为明显。此外,该奖励在策略自身动态采样中保持良好校准性,预期校准误差仅为软最大值裁判的一半(0.074 vs. 0.141),且未出现奖励欺骗(reward hacking)现象。即使不依赖步骤级训练,该方法仍能更优地识别首个错误步骤,超越专用7B流程奖励模型。研究还揭示了该方法的边界——当仅在基础模型普遍失败的问题上训练时,尽管策略性能仍有所提升(+1.9 MATH500分,p = 0.03),但由于概率输出趋于过度自信,仍无法达到参考答案训练的水平。作者预先提出六项假设,并设计对照实验逐一验证与证伪,增强了方法的有效性论证。

链接: https://arxiv.org/abs/2611.25135
作者: Rudransh Agnihotri,Manasvi Kapoor
机构: 未知
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 24 pages, 12 figures, 13 tables

点击查看摘要

Abstract:Reinforcement learning (RL) can only teach a language model what its reward can recognize. Answer checkers are exact but need a reference answer; learned reward models and LLM judges work everywhere, but their scores are not probabilities. We study RLCDR (Reinforcement Learning from Calibrated Decision Rewards). An LLM converted into a decision model answers a typed question about a response (“is the final answer correct?”, “which step is the first error?”) with a probability distribution, read in one forward pass, debiased over option orders and calibrated with about a hundred labels. Training Qwen3-1.7B with GRPO on MATH, this reference-free reward raises MATH500 accuracy from 72.2% to 76.8% (+4.65 points, p 1e-4), matching training on the answer key (76.9%), while a random reward gains only +0.65. The gain grows with difficulty, from 44.8% to 51.7% on the hardest level. On the policy’s own changing samples the reward stays calibrated: its expected calibration error is half that of a softmax judge on the same 27B backbone (0.074 vs. 0.141), and it shows no sign of reward hacking. Without step training, the same model finds the first wrong step better than a dedicated 7B process reward model (ProcessBench F1 0.772 vs. 0.722, 20/20 splits). We also map where the approach stops being enough: trained only on problems the base model mostly fails, the reference-free reward still improves the policy (+1.9 MATH500 points, p = 0.03) but trails the answer key, because its probabilities become overconfident on that distribution. We state six hypotheses in advance and test each with a control designed to falsify it.

[NLP-74] Forget Freely Grep Later: Lossless Index-Free Memory for Long-Horizon LLM Agents

【速读】: 该论文旨在解决长时程大语言模型(LLM)智能体在处理长时间任务时因超出上下文窗口(context window)而被迫丢弃重要信息的问题。现有上下文管理策略依赖于摘要、掩码或结构化记忆,需在未知未来需求的情况下提前决定保留内容或如何检索,易导致有损压缩,丢失对后续任务至关重要的细节。其解决方案的关键在于提出GrepMem——一种简单、无损且无需索引的记忆机制,由两个核心组件构成:一是记录智能体完整历史的仅追加式纯文本日志(append-only plain-text log),二是基于正则表达式(regular-expression)的按需召回搜索工具。该方法无需嵌入向量、检索索引、记忆提取或额外模型,实现了零开销的高效记忆存取。实验表明,GrepMem在ContextBench上达到95–97%的准确率,远超最多仅21%准确率的有损基线方法,并在BCP-Multi任务上比SUPO摘要方法高出8.8个百分点,同时在不同长度输入流下保持良好的可扩展性,始终位于准确率-令牌成本的帕累托前沿。结果证明,纯文本日志结合grep搜索是一种无需复杂结构化记忆基础设施却具备竞争力的长时程智能体记忆方案。

链接: https://arxiv.org/abs/2611.25123
作者: Ziyan Yang,Taiwei Shi,Jieyu Zhao
机构: University of Southern California(南加州大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: 17 pages, 6 figures, 6 tables. Ziyan Yang and Taiwei Shi contributed equally

点击查看摘要

Abstract:Long-horizon LLM agents often exceed their context windows, forcing them to discard information that may be needed later. Existing context management strategies rely on summarization, masking, or structured memory, requiring decisions about what to preserve or how to retrieve it before future needs are known. However, these methods can introduce lossy compression, potentially discarding details needed for future tasks. To address this limitation, we introduce GrepMem, a simple, lossless, and index-free memory mechanism built on two components: an append-only plain-text log of the agent’s complete history and a regular-expression search tool for on-demand recall. Unlike structured memory systems, GrepMem requires no embeddings, retrieval index, memory extraction, or additional models. We train Qwen3.5-9B with reinforcement learning under five context management strategies and evaluate on multi-question research sessions, long-stream recall, and agent-trajectory question answering. GrepMem achieves 95-97% accuracy on ContextBench, versus at most 21% for lossy baselines, and outperforms SUPO’s summarization by 8.8 percentage points on BCP-Multi. GrepMem scales effectively to increasingly long input streams while remaining on the accuracy-token-cost Pareto frontier across all five evaluation tasks. Our results show that a plain-text log and grep provide a simple but competitive approach to long-horizon agent memory, without requiring complex structured memory infrastructure.

[NLP-75] Living off the Local Model: Abusing Browser-Embedded Language Models for Post-Exploitation

【速读】: 该论文旨在解决浏览器嵌入式本地大模型(Local Language Models, LLMs)在安全防护上的潜在风险问题,特别是其可能被恶意利用作为攻击者在系统内持久驻留和执行隐蔽攻击的后渗透工具。其核心解决方案的关键在于揭示并验证了这类模型(以Google Chrome中的Gemini Nano为例)在无需用户持续交互的前提下,可通过操作系统自带的解释器(如shell、脚本解释器等)被远程驱动,并利用其多模态能力完成一系列复杂攻击任务:包括从主机识别凭据、检测个人身份信息(PII)、图像分类以及调用自定义工具。研究进一步展示了攻击者可绕过一次性用户激活检查(如模拟点击或规避HTTP URL验证),实现端到端的自动化攻击链——例如从摄像头图像中提取密码、解密加密归档文件、分析内容并准备私钥外泄,整个过程完全在设备本地执行且不触发网络监控,凸显了生成式AI(Generative AI)在“就地生存”(living-off-the-land)攻击场景下的高威胁性。

链接: https://arxiv.org/abs/2611.25086
作者: William Fleshman,Joseph Lucas
机构: 未知
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Browsers now ship local language models to ordinary endpoints. Google includes Gemini Nano in Chrome to enable productivity features such as summarization, with inference running entirely on the device. These browser-embedded models provide new living-off-the-land opportunities for attackers, each instance a potential post-exploitation engine already installed, signed by the vendor, and invisible to network monitoring. The Gemini model is multimodal, capable of processing files, analyzing screenshots, or conversing with an attacker. We show that it can be driven from the host using only interpreters that ship with the operating system and find that the one-time user-activation check before the model is downloaded is easily satisfied with a simulated click or bypassed this http URL measure and report the current model’s ability to identify credentials on the host, detect PII in email, classify images, and call custom tools. We combine these capabilities and use the model to carry out a simulated cyber operation: reading a password from a webcam image, unlocking an encrypted archive, triaging the contents, and staging a private key for exfiltration, end-to-end without user-interaction.

[NLP-76] A Model of Diverse Sampling from Language Models NEURIPS2026

【速读】: 该论文旨在解决语言模型在生成文本集合时普遍存在的质量问题:尽管单个样本质量较高,但生成的样本集合往往缺乏多样性,出现重复或集中在相同语义模式的现象。为解决这一问题,论文将多样化生成形式化为从完整字符串中采样大小为 KK 的子集,基于确定性点过程(Determinantal Point Process, DPP),其中似然项衡量生成内容的质量,嵌入空间的几何结构则编码相似输出之间的排斥机制。由于对所有字符串进行精确推断不可行,作者提出PoolDPP方法:首先从一个可处理的提议分布中抽取有限候选池,通过重要性加权提升候选集的质量分布,随后在诱导的池核上运行 KK-DPP采样器;同时引入一种贪心最大后验(MAP)变体实现确定性选择。理论分析表明,该方法在有限样本下具有明确常数的收敛界——采样输出与目标 KK-DPP 之间的KL散度随候选池规模 NN 以 O(1/N)O(1/N) 的速率衰减。实验结果表明,PoolDPP在开放式生成任务中显著拓展了质量—多样性前沿,在模糊文本到SQL转换(Ambrosia)任务中提升了精度—召回率前沿,并在数学推理任务(GSM8K)中实现了与原方法相当的 KK-阶通过率(pass@K)水平下的多样性提升。其核心解决方案在于结合重要性加权候选池与高效 KK-DPP采样,兼顾生成质量与集合多样性。

链接: https://arxiv.org/abs/2611.25003
作者: Manuel Prada-Corral,Yahya Emara,Timothy J. O’Donnell,Ryan Cotterell,Tim Vieira
机构: MPI-IS; ELLIS Institute Tübingen; Mila; Canada CIFAR AI Chair; CHI-FRO; ETH Zürich; Cornell University; McGill University
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Formal Languages and Automata Theory (cs.FL); Machine Learning (stat.ML)
备注: NeurIPS 2026 paper

点击查看摘要

Abstract:Language models often produce high-quality individual samples but poor sets, with repeated samples clustering around the same semantic modes. We formalize diverse generation as sampling a size- K subset of complete strings from a determinantal point process (DPP), where the likelihood encodes item quality, and the embedding geometry encodes repulsion between similar outputs. Exact inference over all strings is intractable, so we propose PoolDPP: draw a finite candidate pool from a tractable proposal, importance-weight the candidates toward the target quality, and run a K -DPP sampler on the induced pool kernel; a greedy MAP variant provides deterministic selection. We prove a finite-sample bound with an explicit constant: the KL divergence between the sampler’s output and the target K -DPP decays as O(1/N) in the pool size. Empirically, PoolDPP extends the quality–diversity frontier on open-ended generation and the precision–recall frontier on ambiguous text-to-SQL (Ambrosia), and on mathematical reasoning (GSM8K), it raises diversity at about the same pass@ K .

[NLP-77] What the model knows but does not say: monitoring medical LLM s through verbalized thoughts

【速读】: 该论文旨在解决临床生成式AI模型在受到恶意操控时,其输出表现看似合规但实际存在隐蔽风险的问题。传统安全机制仅关注输入与输出内容,难以识别模型内部被篡改却未体现在输出中的行为。本文提出的关键解决方案是通过自然语言自编码器(natural-language autoencoder)将多模态大模型(Gemma-3-27B)的内部激活状态转化为可读文本,构建“思维通道”(thoughts channel),从而暴露模型内部的潜在操纵行为。实验表明,在八种不同攻击场景下(包括病理遗漏、隐匿药物推荐等),思维通道对注入攻击的检测准确率显著高于原始输出通道(p < 0.05),且在无污染输入下保持极低假阳性率;该效果在Llama-3.3-70B模型上亦可复现。此外,思维通道能在模型尚未决定是否拒绝前即识别出不安全请求,实现早期干预。整个方法无需重新训练模型,仅依赖现成的文本化工具和通用判别器即可实现,为临床生成式AI的安全监控提供了高效、可部署的内省机制。

链接: https://arxiv.org/abs/2611.24994
作者: Frederik Hauke,Jan Clusmann,Hanna Kreutzer,Christiane Kuhl,Tianyu Han,Felix Barajas Ordonez,Sven Nebelung,Jakob Nikolas Kather,Daniel Truhn
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Efforts to keep clinical language models safe have concentrated on what enters and leaves the model, but a manipulated model can return an answer that looks entirely ordinary. Here we show that the model’s own internal state, rendered in plain language, exposes manipulations the output conceals. Using a natural-language autoencoder, we translate the activations of a multimodal LLM (Gemma-3-27B) into text and let an input-blind judge read this verbalized internal state (the thoughts channel) and the model’s answer separately. Across two clinical attack objectives (omitting a pathology, covertly recommending a drug), two input modalities, and two delivery mechanisms, the thoughts channel caught the injection significantly more often than the answer in all eight injected conditions (p0.05), while both channels held near-zero false-positive rates on clean inputs; the omission result carried over to Llama-3.3-70B. The thoughts channel also registered unsafe clinical requests as harmful while the model was still reading them, before it decided whether to refuse. None of this requires retraining: an off-the-shelf verbalizer and a general-purpose judge are enough.

[NLP-78] Quantifying Cultural-Psychological Constructs in Four Centuries of Ottoman Court Registers

【速读】: 该论文旨在解决如何在缺乏传统心理学数据的历史社会中,研究文化维度(如个人主义、集体主义与荣誉感)的科学问题。由于历史文献通常不具备现代心理量表的直接测量数据,传统的心理分析方法难以适用。为此,论文提出一种计算文本分析框架,通过对1513–1911年期间54,670份奥斯曼土耳其宫廷档案进行量化分析,实现对心理构念的文化变异研究。其解决方案的关键在于:构建适配奥斯曼法律语言的句子嵌入模型——OTA-BERTurk,并基于理论驱动的文本分析流程,捕捉个体主义、集体主义和荣誉感等文化维度在历史语料中的表现。模型通过文档检索、语义文本相似性、问卷条目分类及专家标注验证等多种方式评估,表现出优异性能(如文档检索召回率Recall@1达0.843,平均倒数排名MRR为0.806,条目分类准确率达89.58%)。此外,定性分析揭示高分值反映特定历史时期的法律与制度特征。研究公开发布模型、构念表示、评分管道及交互式网页界面,推动可重复、透明的历史—心理学分析,使跨四个世纪的文化与制度变迁中的心理多样性研究成为可能。

链接: https://arxiv.org/abs/2611.24981
作者: Sila Mutaf,Enes Yılandiloğlu,Mohammad Atari
机构: 未知
类目: Computation and Language (cs.CL)
备注:

点击查看摘要

Abstract:Populations vary along cultural dimensions such as individualism, collectivism, and honor. But how can these dimensions be studied in historical societies for which conventional psychological data are unavailable? Here, we introduce a computational text-analytic framework for quantifying psychological constructs in 54,670 Ottoman Turkish court records from 21 Istanbul courts spanning 1513-1911 CE. We develop sentence-embedding models adapted to Ottoman legal language and extend a theory-driven, text-analytic pipeline to capture cultural variation along the dimensions of individualism, collectivism, and honor in this historical corpus. We evaluate our models through multiple methods: document retrieval, semantic textual similarity, questionnaire-item classification, and validation against expert annotations. Our adapted sentence transformer (OTA-BERTurk) achieved a Recall@1 of .843 on the test set for document retrieval, a Mean Reciprocal Rank (MRR) of .806 for semantic textual similarity (meaning the best answer usually appeared as a top result), and classified questionnaire items with 89.58% accuracy. Document-level performance varied across constructs and methods. Qualitative analysis further showed that high scores reflected historically specific legal and institutional patterns. We publicly release the models, construct representations, scoring pipeline, as well as an interactive web interface to support transparent, reproducible historical-psychological analyses of Ottoman Turkish court records spanning approximately four centuries. These computational tools enable the study of psychological variation across centuries of cultural and institutional change in a diverse empire.

[NLP-79] CANDID: Candidate-Driven Intervention for Disciplined Execution in Local Computer-Use Agents

【速读】: 该论文旨在解决本地计算机使用代理(Local Computer-Use Agents, CUAs)在资源受限环境下执行任务时的可靠性问题,尤其针对其在关键节点因僵化的动作循环或过早的“DONE”决策而失败的现象。现有方法通常依赖文本反思引入认知负荷,但存在精度不足的问题。本文提出的解决方案——候选驱动式执行控制框架CANDID(Candidate-Driven Intervention for Disciplined Execution),其核心在于将测试时并行扩展(test-time scaling)重构为稀疏、故障敏感点上的结构化执行级干预。关键创新包括:1)一种循环恢复机制,用于过滤重复行为的候选动作;2)一种保守的、需全体一致同意方可触发的“DONE”门控策略,以防止过早终止。实验结果表明,在OSWorld基准上,CANDID通过轻量级的结构化控制显著提升了任务成功率,优于基于反馈的替代方案,且执行开销极低。

链接: https://arxiv.org/abs/2611.24977
作者: Woongkyu Lee,Jungwook Choi
机构: Hanyang University(汉阳大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:We present CANDID (Candidate-Driven Intervention for Disciplined Execution), an execution-control framework designed to enhance the reliability of local computer-use agents (CUAs) in resource-constrained settings. While local CUAs feature robust baseline proficiency in single-step action prediction, they frequently fail at critical junctions due to rigid action loops or premature DONE decisions. Rather than imposing imprecise cognitive loads through textual reflection, CANDID repurposes parallel test-time scaling into a disciplined, execution-level intervention at sparse, failure-prone points. We introduce a loop-recovery mechanism to filter candidates away from repeated behaviors and a conservative, unanimous DONE gating policy that requires absolute consensus to terminate. Experiments on OSWorld demonstrate that CANDID’s structural control significantly improves success rates, outperforming feedback-based alternatives with minimal execution overhead.

[NLP-80] No-Harm-Jury A Jury-Based Multi-Dimensional Harm Evaluation Framework for Medical AI Responses

【速读】: 该论文旨在解决生成式 AI 在医疗问答场景中可能引发的多维度潜在危害(harm)难以全面、客观评估的问题。现有评估体系普遍缺乏对信息性、社会性、心理性、自主性、经济性、隐私性及认识论性等七类危害的系统性量化能力,且多依赖人工评判或单一模型评估,存在可扩展性差、主观偏差显著等局限。本文提出的解决方案——No-Harm-Jury,其核心创新在于构建了一个基于五种多样化大语言模型(LLM)作为独立“陪审员”的自动化评估框架,通过中位数聚合机制有效抑制个体模型偏见,同时利用各模型间的互补专长保障评估覆盖度;该框架首次实现了自动化可扩展评估、多模型陪审团、鲁棒中位数聚合(含可靠性加权扩展)、七类概念上区分明确的危害维度以及在医学问答基准(如PubMedQA、MedQA、MedMCQA)上的实证应用这五个关键属性的统一整合,显著提升了对医疗生成内容风险的系统性识别与量化能力。

链接: https://arxiv.org/abs/2611.24972
作者: Antonio Puertas Gallardo,Mario Ceresa
机构: 未知
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:We present No-Harm-Jury, a jury-based automated evaluation framework that scores AI-generated medical responses across seven conceptually distinct harm dimensions using a panel of five diverse language models as independent jurors. Responses generated by Ministral-14B on 3,000 questions drawn from PubMedQA, MedQA, and MedMCQA are scored by five jurors across dimensions of informational, social, psychological, autonomy, economic, privacy, and epistemic harm; the 151 empty responses (98 MedQA, 53 MedM- CQA) are excluded, leaving 2,849 valid samples. The median aggregation robustly suppresses individual juror biases while maintaining coverage through complementary specialisation. No existing system combines all five prop- erties of No-Harm-Jury: automated scalable evaluation, five diverse LLM jurors, robust me- dian aggregation (with a reliability-weighted extension), seven conceptually distinct harm dimensions, and application to medical QA benchmarks.

[NLP-81] he ROI of Small Examples: Optimizing In-Context Learning Under Token Budgets EACL27

【速读】: 该论文旨在解决在上下文学习(In-context Learning, ICL)中,如何在固定令牌(token)预算下最优分配提示(prompt)资源以提升大语言模型(Large Language Model, LLM)推理性能的问题。尽管已有研究多关注在ICL中应包含多少示例,但本文聚焦于一个互补性问题:在有限的上下文长度约束下,是应选择少数长篇示范(long demonstrations)还是更多短篇示范(shorter demonstrations)以更高效地利用提示空间。其解决方案的关键在于通过将候选示范按长度分组(length-based bins),在固定预算下从同一长度区间内选取示范构建提示,并系统评估不同示范粒度(granularity)对问答任务性能的影响。实验结果表明,在两个问答基准(\textit{Beat the AI} 和 \textit{Natural Questions})及六种指令微调的大语言模型上,较短的示范通常能更有效地利用提示空间,尤其在\textit{Beat the AI}上效果显著;而在长上下文设置下,虽然\textit{Natural Questions}上的增益较小且趋于饱和,但趋势依然存在。此外,在ANLI R3自然语言推理任务上,五种模型中一种短示范分组表现最佳。总体而言,研究揭示了示范粒度是影响提示设计效能的重要因素,为实际部署中的提示优化提供了关键实证依据。

链接: https://arxiv.org/abs/2611.24971
作者: Anxo Perez,Jorge Gabín,Javier Parapar
机构: IRLab, CITIC; Universidade da Coruña; A Coruña, Spain
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Accepted for EACL 27 Main Conference

点击查看摘要

Abstract:In-context learning (ICL) improves large language model performance at inference time without updating model parameters. However, these gains are not free: in practical deployments, longer prompts directly increase inference cost and latency. Under a fixed token budget, prompt construction therefore becomes a resource-allocation problem. While prior work has often focused on how many examples to include in ICL, we study the complementary question of whether the available budget is better spent on a few long demonstrations or on a larger number of shorter ones. We study this aspect in question answering (QA), by organizing candidate demonstrations into length-based bins and constructing prompts from examples within the same length range under fixed budgets. Experiments on two QA benchmarks (\textitBeat the AI and \textitNatural Questions), using six instruction-tuned LLMs, show that shorter demonstrations are generally the most effective use of prompt space. This effect is especially strong on \textitBeat the AI, and remains visible on NQ, although gains are smaller and saturate earlier in the long-context setting. An additional natural language inference experiment on ANLI R3 finds that one of the two shorter bins performs best for five of six models, although longer demonstrations remain competitive. Overall, our results highlight demonstration granularity as an important factor in prompt design.

[NLP-82] HireWatch: Evaluating LLM Compliance with U.S. Employment Law Under Contextual Pressure NEURIPS2026

【速读】: 该论文旨在解决大型语言模型(Large Language Models, LLMs)在招聘流程中应用时,其生成或推荐的面试问题可能违反美国相关法律法规的问题。现有评估体系普遍忽视了LLMs在法律合规性方面的表现,而这一缺陷可能导致企业面临法律风险。为此,研究提出HireWatch评估框架,通过构建一个包含200个合法与非法面试问题的题库,其分类依据为美国联邦法律、法规及监管机构指南,系统性地测试LLMs在不同职业和情境下生成或选择面试问题的合规性。关键解决方案在于设计多场景、高压力条件下的评估实验,揭示在上下文压力增强时,尽管多数模型在低压力环境下能避免非法问题,但其法律合规性显著下降,且即使通过显式指令引导,仍无法完全消除违规行为。该研究凸显了当前LLMs在法律敏感任务中的脆弱性,并强调需进一步优化模型对法律约束的理解与遵循能力。

链接: https://arxiv.org/abs/2611.24970
作者: Huy Nghiem,Maria Kostylew,Noam Kolt
机构: MATS Research(机器对齐、透明性与安全性研究机构); University of Maryland (马里兰大学); Hebrew University of Jerusalem (耶路撒冷希伯来大学)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: NeurIPS 2026 Evaluation and Dataset Track. Law and AI

点击查看摘要

Abstract:The deployment of large language models (LLMs) in hiring pipelines is growing, yet existing evaluations generally overlook their compliance with legal rules. To fill this gap, we introduce HireWatch, an evaluation that assesses whether LLMs comply with U.S. laws that govern the questions that job candidates may be asked. We construct a 200-item bank of lawful and unlawful interview questions, whose legal classification is based on U.S. federal statutes, regulations, and agency guidance. We instruct LLMs to select questions from this bank in preparing for job interviews, across a set of various occupations and contexts, and report the fraction of unlawful questions that LLMs select. Evaluation of 18 closed and open-weight LLMs across ~100,000 trials under different types of contextual pressure reveals that LLMs’ legal compliance is brittle. While LLMs generally avoid selecting unlawful questions in low-pressure contexts, legal violations increase sharply under higher pressure. Explicit prompts to comply with law can improve compliance rates, but do not altogether eliminate legal violations. We publicly release our code and data at this https URL.

[NLP-83] Global Importance Estimation for KV Cache Eviction NEURIPS2026

【速读】: 该论文旨在解决长上下文大语言模型(LLM)推理中关键值(KV)缓存内存消耗过高的核心瓶颈问题。其解决方案的关键在于提出一种无需训练的KV缓存淘汰方法——PRKV,该方法通过在注意力诱导的转移图(attention-induced transition graphs)上应用个性化PageRank(Personalized PageRank),实现对多跳全局标记重要性的估计。这一机制突破了传统方法依赖局部近期性或瞬时注意力的局限,能够有效捕捉标记间的长期依赖关系,从而在不修改模型的前提下实现更精准的缓存管理。实验表明,在多种长上下文理解、稀疏检索及思维链推理任务中,PRKV在中等至激进压缩比(最高达90%)下仍能保持与现有基线相当的任务性能,同时显著降低峰值内存占用并带来较低的端到端开销,验证了多跳全局影响估计在可扩展长上下文推理中具备优异的精度-内存权衡能力。

链接: https://arxiv.org/abs/2611.24969
作者: Seojin Kim,Noseong Park
机构: KAIST(韩国科学技术院); Daejeon, Republic of Korea(大田,大韩民国)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: The 40th Conference on Neural Information Processing Systems (NeurIPS 2026)

点击查看摘要

Abstract:Memory consumption of the key-value (KV) cache remains a central bottleneck for long-context LLM inference. We introduce PRKV, a training-free KV eviction method that estimates multi-hop global token importance by applying Personalized PageRank over attention-induced transition graphs. This enables importance propagation beyond recency or instantaneous attention, without modifying the model. We benchmark PRKV on long-context understanding, sparse retrieval, and chain-of-thought reasoning tasks using multiple LLM families. Under moderate to aggressive compression ratios, including up to 90%, PRKV achieves competitive task performance against established eviction baselines while reducing peak memory usage with low end-to-end overhead. These results demonstrate that multi-hop global influence estimation yields a favorable accuracy–memory trade-off for scalable long-context inference.

[NLP-84] Physics of Language Models: Part 4.2 Canon Layers at Scale where Synthetic Pretraining Resonates in Reality KR

【速读】: 该论文旨在解决当前大模型架构比较中存在的一大核心问题:标准基准测试(standard benchmarks)是否能够有效识别出真正影响模型性能的架构差异。研究发现,在学术规模预训练(1.3B参数,100B tokens)下,通过精细调优后,不同架构间的性能差异几乎消失;然而在更大规模的预训练(1-2T tokens)条件下,架构差异开始显现。其关键发现是:引入局部上下文混合机制的Canon层可显著提升Transformer性能,使相同MMLU准确率下预训练数据量减少约30%。此外,作者提出GLA5与GDN2两种新型结构,分别优化了门控线性注意力和门控DeltaNet,但令人意外的是,在引入Canon后,最简单的线性模型在8B/1T规模下已能媲美甚至超越更复杂的门控DeltaNet与Mamba2模型,表明这些复杂架构的增益可能主要源于对邻近token的混合能力。更重要的是,尽管多数模型在标准基准上表现优异,但在基本的上下文推理任务(如短上下文信息检索、两跳问答)中仍无法稳定超越随机猜测,暴露出标准评测体系的严重局限性。因此,该研究的关键在于揭示标准基准的“表面有效性”与实际推理能力之间的脱节,并强调应早期将上下文推理示例纳入预训练数据,同时建议使用小型合成实验环境以更早发现架构本质差异。

链接: https://arxiv.org/abs/2611.24963
作者: Zeyuan Allen-Zhu
机构: Allen-Zhu Research
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI)
备注: Results, recipes, and code have been public since December 9, 2025 at this https URL

点击查看摘要

Abstract:Do standard benchmarks see the architectural differences that matter? We study it with 64 from-scratch pretraining runs of Transformers and three linear-model families at 1.3-8B parameters on 1-2T Nemotron-CC tokens, at a cost of roughly 100k USD per 8B run. In academic-scale pretraining (1.3B parameters, 100B tokens), architectural differences can all but disappear when the models are carefully tuned. Instead, at 1-2T tokens they become measurable: Canon layers, which add local token mixing, consistently improve Transformers, saving about 30% of pretraining tokens for the same MMLU accuracy at 8B. We also introduce GLA5 and GDN2, which improve gated linear attention and Gated DeltaNet with low-rank output gates. Surprisingly, with Canon, the simplest linear model, gated linear attention, matches Gated DeltaNet and outperforms Mamba2 even at 8B/1T. Therefore, the extra machinery of Gated DeltaNet and Mamba2 may do little more than mix neighboring tokens. Even at 8B/1T, however, high scores on standard benchmarks can hide weaknesses in basic in-context skills. In particular, linear models that match or beat Transformers on these benchmarks cannot, in fact, reliably retrieve information from a short 500-token context. Moreover, no model, Transformer or linear, reliably beats random guessing on arguably the simplest two-hop question, stated in a few short sentences, even at this scale. These findings \textbfchallenge traditional architecture research based primarily on standard benchmark scores. In contrast, small synthetic playgrounds, which predicted the architectural differences above at 100M parameters, can reveal what these benchmarks miss. For practitioners, one lesson is to include in-context reasoning examples in pretraining data from an early stage, rather than reserve them for later mid-training, let alone post-training. Comments: Results, recipes, and code have been public since December 9, 2025 at this https URL Subjects: Computation and Language (cs.CL); Artificial Intelligence (cs.AI) Cite as: arXiv:2611.24963 [cs.CL] (or arXiv:2611.24963v1 [cs.CL] for this version) https://doi.org/10.48550/arXiv.2611.24963 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Zeyuan Allen-Zhu [view email] [v1] Thu, 12 Nov 2026 18:51:19 UTC (1,006 KB)

[NLP-85] ETHER: Training-Free Elastic ToC-Anchored Hierarchy Extraction and Recovery in Design-Driven Long Documents EMNLP2026

【速读】: 该论文旨在解决复杂文档(如环境、社会与治理报告、产品手册及年报)在转换为结构化知识时所面临的挑战,尤其是现有方法在处理设计驱动型长文档时因忽视目录(Table of Contents, ToC)的宏观结构而导致的层级解析不准确问题。传统自底向上的方法依赖局部视觉线索,易受高熵版面中视觉显著性与逻辑深度分离的影响,而直接使用大语言模型(LLM)进行抽取则缺乏空间定位,导致长上下文中的幻觉现象。为此,论文提出TETHER——一种无需训练的神经符号框架,其核心创新在于将原本被忽略的ToC从“噪声”转化为关键的结构先验,从而实现对多样化文档布局的鲁棒适应。其解决方案的关键在于双阶段机制:第一阶段利用大模态模型(LMM)结合时空语义约束,从原始ToC构建文档骨架;第二阶段采用基于锚点的对齐策略,将内容块精确绑定至该骨架,有效解决边界模糊性问题。实验表明,TETHER在多语言基准上实现了领先的零样本层级恢复性能,优于深度层次化模型、专用解析器和通用LMM;且在稀疏缺失与文本噪声等压力测试下仍保持稳健,验证了其语义对齐的本质源于拓扑与深度一致性。消融实验进一步证明该框架架构不可简化,同时实现90%的令牌压缩和显著提速,兼顾结构精度与部署可扩展性,为复杂文档结构化提供了高效可靠的解决方案。

链接: https://arxiv.org/abs/2611.24960
作者: Yu Zou,Yan Chen,Jialei Zeng,Xiaorui Zhou,Aixi Zhong,Qi Zhou,Haoran You,Ke Jiang,Changyu Hu
机构: Southwestern University of Finance and Economics(西南财经大学); Pujiang County Bureau of Transportation(浦江县交通运输局); Ningbo University of Finance and Economics(宁波财经学院)
类目: Computation and Language (cs.CL); Computer Vision and Pattern Recognition (cs.CV)
备注: EMNLP 2026 Main

点击查看摘要

Abstract:Converting complex documents into structured knowledge constitutes the critical infrastructure urgently required for deploying Large Language Models (LLMs) in downstream research. However, blind to the Table of Contents (ToC) as a macro-structure, existing bottom-up paradigms rely on localized visual cues, inducing heuristic fragility in high-entropy layouts where visual prominence decouples from logical depth; meanwhile, direct LLM extraction lacks spatial grounding, triggering hallucinations in long contexts. To address this, we introduce TETHER, a training-free framework that elevates the ToC from discarded noise to a vital structural prior, enabling robust adaptation across design-driven long documents. TETHER operates via a two-stage neuro-symbolic mechanism: first, it constructs a document skeleton from the raw ToC using a Large Multimodal Model (LMM) empowered by spatio-semantic constraints; second, it employs an anchor-based alignment strategy to tether content blocks to this skeleton, effectively resolving boundary ambiguities. Evaluated across a multilingual benchmark of ESG reports, product manuals, and annual reports, TETHER delivers highly precise, state-of-the-art (SOTA) zero-shot hierarchy recovery, outperforming deep hierarchical recovery models, dedicated parsers, and general-purpose LMMs. Moreover, stress tests highlight TETHER’s resilience to sparse omissions and textual noise, demonstrating that its precise semantic alignment is fundamentally anchored in topological and depth fidelity. Crucially, ablations confirm the framework’s architectural irreducibility, while achieving 90% token reduction and drastic speedups over end-to-end baselines, establishing a compelling equilibrium between structural precision and deployment scalability.

[NLP-86] ICARE: cross-domain text classification with incremental class-aware representation and distillation learning

【速读】: 该论文旨在解决持续多任务学习中因灾难性遗忘(catastrophic forgetting)导致的模型性能退化问题,尤其在缺乏特定标注数据的领域中更为突出。现有方法通过回放历史带标签样本以缓解遗忘,但易引发过拟合,削弱模型对新类别的泛化能力。针对这一挑战,本文提出ICARE框架,其核心在于两个关键组件:基于表示学习的实例选择模块与融合特征层与逻辑输出层(logit-level)知识迁移的知识蒸馏机制。前者通过智能筛选最具代表性且信息量丰富的训练样本,降低冗余与噪声干扰;后者结合剪枝技术与优化目标函数,实现跨任务间深层特征表示与决策逻辑的有效传递,从而在保留旧知识的同时高效适应新任务。大量基准文本分类数据集上的实验验证表明,ICARE显著提升了目标任务的准确率,并大幅降低了遗忘度量(forgetting metric),充分证明了其在多任务场景下维持和复用知识的能力,有效增强了自然语言推理(NLI)系统的可适应性与整体性能。

链接: https://arxiv.org/abs/2611.24959
作者: Son Duong,Truong X. Tran,Hai-Anh Tran
机构: Hanoi University of Science and Technology (河内科技大学); The Pennsylvania State University (宾夕法尼亚州立大学)
类目: Computation and Language (cs.CL); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:The rise of powerful, fine-tuned large language models (LLMs) necessitates robust natural language inference (NLI) systems. These NLI systems ideally train across various tasks, encompassing diverse concepts like sentiment analysis, news classification, and coreference resolution. However, this continual, multi-tasking approach can lead to catastrophic forgetting, where previously learned tasks are lost, especially in domains lacking specific labels. Rehearsing past examples with labels can further exacerbate this issue by causing overfitting, hindering the system’s ability to classify new categories. To tackle these challenges, this research proposes ICARE, a novel framework with two key components: an instance selection module and a knowledge distillation method. The former is based on representation learning, which meticulously selects training examples to present to the NLI system. The latter, knowledge distillation with feature and logit-level transfer, facilitates knowledge transfer between tasks. This leverages pruning techniques and optimized objective functions to transfer both essential feature-level and decision-making (logit-level) knowledge, mitigating catastrophic forgetting while adapting to new patterns. Extensive evaluations on benchmark text classification datasets have demonstrated ICARE’s effectiveness. The framework achieves a significant improvement in target accuracy and reduction in the ``forgetting metric’'. These results solidify ICARE’s ability to preserve and leverage learned knowledge across diverse tasks, ultimately enhancing the adaptability and performance of NLI systems. The source code is available at this https URL

[NLP-87] LLM Complexity: Towards a Computational Complexity View of Large Language Model Inference

【速读】: 该论文旨在解决现有方法中将大语言模型(LLM)生成的令牌数量(token count)直接等同于经典计算复杂度(time complexity)所导致的误解问题。其核心挑战在于,尽管令牌数可作为推理阶段计算量的可观测代理指标,但实际研究表明,令牌长度与经典时间复杂度之间并不存在单调一致性,这源于大语言模型在推理过程中对计算资源的非均匀分配机制。为此,论文提出一种名为“LLM复杂度”(LLM Complexity)的新度量标准,通过刻画准确率-令牌数曲线之上的面积来量化任务从大语言模型视角下的内在复杂性。该度量展现出跨不同大语言模型的高度一致性任务排序能力,表明其能稳定捕捉任务难度的本质特征。基于此,研究构建了一种由易到难的监督微调课程学习策略,在提升模型准确性的同时显著降低令牌消耗。此外,实验验证了LLM复杂度在真实世界任务中的泛化能力,其跨模型的相关性优于人工标注的难度标签,进一步证明了该度量在评估和优化大语言模型性能方面的有效性。

链接: https://arxiv.org/abs/2611.24943
作者: Yu-Hung Wu,Shao-Xiang Yuan,Chan-Hung Yu,Zih-Ching Chen,Kuan-Hao Huang,Shao-Hua Sun
机构: National Taiwan University (国立台湾大学); NVIDIA AI Technology Center, NVIDIA Corporation (NVIDIA人工智能技术中心,NVIDIA公司), Santa Clara, CA, USA; Texas A&M University (德克萨斯农工大学); NTU Artificial Intelligence Center of Research Excellence (国立台湾大学人工智能研究中心卓越中心)
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 10 pages, 20 with Appendix

点击查看摘要

Abstract:Computational complexity theory offers a principled framework for characterizing the efficiency of computational problems, and Large Language Models (LLMs) can similarly be viewed as computational systems executing a structured process of input, inference-time compute, and output. In this context, the number of generated tokens emerges as a natural and directly observable proxy for inference-time compute. However, we show that the tempting analogy between LLM token length and classical time complexity does not hold. We disprove the hypothesis that inference-time token usage aligns monotonically with classical time complexity, demonstrating a fundamental divergence grounded in how LLMs actually allocate computation during inference. To this end, we introduce LLM Complexity, a new metric designed to quantify problem complexity from the perspective of LLMs by characterizing the area bounded above by the accuracy-token curve. We find that diverse LLMs induce highly consistent task rankings under LLM Complexity, indicating that the metric captures stable notions of difficulty across models. We leverage this property to construct an easy-to-hard curriculum for supervised fine-tuning, which improves accuracy while reducing token consumption. Finally, we demonstrate that LLM Complexity generalizes to real-world tasks, producing more stable cross-model correlations than human-annotated difficulty labels.

信息检索

[IR-0] Variants of Beam Search for DiskANN

链接: https://arxiv.org/abs/2611.26609
作者: Viraj Agrawal,Kirankumar Shiragur
类目: Data Structures and Algorithms (cs.DS); Databases (cs.DB); Information Retrieval (cs.IR)
备注: 22 pages, 1 figure, 9 tables

点击查看摘要

Abstract:Graph-based data structures such as DiskANN, HNSW, and NSG constitute the state of the art for large-scale approximate nearest-neighbor (ANN) search. Indyk and Xu introduced the notion of \alpha -reachable graphs, which include graphs built by DiskANN’s slow-preprocessing construction, and proved that greedy search achieves an approximation factor of \tfrac\alpha+1\alpha-1 . Independently, Al-Jazzazi et al. proposed adaptive beam search (ABS) with a relaxation parameter \gamma 0 , showing a \tfrac2\gamma approximation on navigable graphs ( \alpha \approx 1 ), where greedy search admits no guarantees. In this work, we unify and generalize these results by proving that ABS with parameter \gamma on \alpha -reachable graphs achieves an approximation of \tfrac\alpha+1\alpha(\gamma+1)-1 for any metric. Moreover, on sorted \alpha -reachable graphs (introduced by Gollapudi et al.) under the \ell_2 metric, we obtain the tighter bound \max\bigl(\tfrac\alpha\alpha(\gamma+1)-1, \tfrac\alpha^2+1\alpha^2(\gamma+1)-1\bigr) . Beyond these theoretical results, we design a beam predictor using features inspired by fine-grained analysis of ABS, trained via a constrained objective that minimizes mean distance computations subject to a target recall. On three 100-million-point benchmarks, using graphs built by DiskANN’s practical construction, beam predictor lowers mean query latency on SSD-resident indices at matched recall by up to 29.6% relative to classical beam search (CBS). Relative to ABS, it is faster in 11 of 15 settings, by up to 6.5% , and at most 1.4% slower in the other 4 ; at 95% recall, its advantage over ABS grows from 0.8 – 4.2% for 100 neighbors to 4.6 – 5.6% for 500 neighbors. In memory, it needs 1.3 – 23.7% fewer distance computations than CBS and 0.0 – 3.9% fewer than ABS.

[IR-1] Cost-Aware Agent ic Feature Optimization System for Fleet-Scale Ranking Models

链接: https://arxiv.org/abs/2611.26531
作者: Chang Yang,Yao Xuan,Wendan Yan,Nhat Le,Jiang Liu,Longhao Jin,Zhen Peggy Yao,Alireza Vahdatpour,Djordje Gligorijevic
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Large Web platforms operate large fleets of ranking models, each selecting features from a shared pool of tens of thousands. We present an end-to-end agentic system that optimizes features across the fleet under resource constraints shared across models and those set per model, such as feature pipeline cost and serving throughput (QPS), respectively. In this system, LLM agents propose hypotheses, and deterministic tools execute experiments and verify them. For each model, an LLM generates and shortlists cross features from a combinatorial candidate space, and all features are then ranked by several importance methods. Feature blending mixes these rankings into a single selection, using Pareto fronts across tasks for multi-task models. An LLM Researcher agent then tunes the feature selection in 10–25 asynchronous trials under resource caps, and a trimming step keeps the pipeline cost within budget. Across production models, the system improves normalized entropy by 0.31% on average within each model’s budget, and the trimming step cuts pipeline cost by 2.3% with no regression beyond measurement noise. The optimized models were launched, and in online A/B tests they improved a primary online metric by 0.20%.

[IR-2] he Window Lever: Longer Observation Windows Outweigh the Observed Effects of Data Capacity and Architecture in Customer Behavior Forecasting

链接: https://arxiv.org/abs/2611.26521
作者: Tara Sowrirajan,Yanqiao Wang
类目: Machine Learning (cs.LG); Information Retrieval (cs.IR)
备注:

点击查看摘要

Abstract:In our experiments on long-horizon customer-behavior forecasting, a longer observation window outweighed the observed effects of more data, more capacity and newer architectures. The window is the weeks of history a model sees before the week it forecasts from. Going from an 8- to a 16-week window adds a raw +0.070 to 0.107 log-space R^2 in every family at 1M. Matched test cohorts, which score the same customer-weeks at every window, show this raw comparison understates the gain in every family they cover. The direction holds on five public datasets, four spanning over a year (ridge and our transformer on all five). On our internal data, longer windows admit up to 10% more and colder training customers, which matching does not fix. Our design does not separate window length from calendar coverage, because every test anchor (the week a forecast is made from) falls in one eight-week band. On training-period anchors, the weekly ridge probe’s gain persists outside the holiday season, keeping about 72% of it. We therefore report a robust positive association with longer windows and do not identify the magnitude of a calendar-independent history effect. Of four internally registered predictions, the only fully blind one failed. An activity-mixture model predicted +0.014 from 24 to 32 weeks; the weekly-aggregate probe returned -0.0004. For the sequence families, the other three levers are about an order of magnitude smaller, bounded and unresolved. Their observed effects are at most 0.006 between the nominal 1M and 10M tiers and 0.0052 among five sequence families at 16 weeks with clocks as built, both single-seed on the full test split, and 0.006 for capacity on three-seed means. These are observed intervals, not equivalence bounds. Our claim is the order-of-magnitude contrast between the window and those bounds. A team building on this class of data should start by lengthening the window.

[IR-3] Democratizing Cross-Project Knowledge for Transforming Water Governance and Climate Resilience: an Agent ic AI Ecosystem

链接: https://arxiv.org/abs/2611.26398
作者: Oriol Alàs,Ian Palacín,Iván Cester,Edgar Rubión
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Computation and Language (cs.CL)
备注: 8 pages, 2 figures. Accepted at WDSA/CCWI 2026. Proceedings publication pending

点击查看摘要

Abstract:EU-funded projects produce extensive evidence in the form of deliverables, reports, and case studies. However, this knowledge is typically scattered across projects and document structures, making it difficult to locate relevant, contextual information when making decisions. This paper introduces an Agentic AI Ecosystem that consolidates heterogeneous project documentation into a domain-aware knowledge base and provides adaptive access through an agentic Retrieval-Augmented Generation pipeline. Documents from three EU-funded projects (IMPETUS, NATALIE, and InnWater) are reconstructed into hierarchical section trees and indexed with structured metadata. Retrieval is supported by two domain-specific embedding models fine-tuned on domain data (water and climate). On top of this knowledge base, a ReAct-style controller iteratively orchestrates retrieval actions by allocating retrieval budget across domain collections and accumulating evidence until a stopping criterion is met. In a retrieval benchmark over the corpus, the proposed approach improves Mean Reciprocal Rank by 26.7% (relative) compared to a single-domain embedding baseline, indicating more consistent retrieval of relevant evidence to support cross-project knowledge synthesis and climate-resilient decision-making.

[IR-4] From Sessions to Serving: Integrating LLM -Based Embeddings into Retrieval for Production Content Recommendation

链接: https://arxiv.org/abs/2611.26096
作者: Zhixiong Hu,Meng Li,Maggie Zhang,Deborshi Goswami,David Bailey,James Bradley
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Personalized content recommendation on large marketplace websites must reconcile heterogeneous candidate content (e.g., customer stories, product pages, industry pages, and learning content) with sparse, fast-changing user behavior, placing increasing demands on the retrieval stage to represent user interest well, not just the downstream ranking stage. We present LSER (LLM Session Embedding Retrieval), a production content recommendation system for this http URL that derives session- and history-level embeddings from a pretrained large language model (LLM) and adds them as auxiliary user-side features in the retrieval-stage user tower alongside existing learned sequence encoders. To keep user-side embeddings fresh under real-time behavior, we build a nearline serving path that asynchronously refreshes sequential and LLM-derived user vectors for our main content recommendation server. We validate the incremental value of the LLM features through an ablation study across our full multi-content-type catalog. Our results indicate that integrating LLM-derived embeddings as additional retrieval signals, together with production-grade nearline serving, is technically feasible and can improve model performances.

[IR-5] A Task-Level Bias Taxonomy for Historical Image Retrieval

链接: https://arxiv.org/abs/2611.26019
作者: Omkar Gavali,Ratan Sebastian,Christoph Rippe,Ralph Ewerth,Anett Hoppe
类目: Information Retrieval (cs.IR)
备注: Presented at Bias In GLAM workshop at TPDL 2026

点击查看摘要

Abstract:In recent decades, galleries, libraries, archives, and museums (GLAM institutions) have digitised vast amounts of cultural heritage material, including historical imagery. A major challenge in making these collections publicly accessible is enabling effective search and discovery. Many historical images contain incomplete, sparse, or no descriptive metadata. Consequently, machine learning (ML) methods are increasingly being used to generate or enrich metadata to improve the searchability of such collections. However, artificial intelligence (AI)-based systems can reproduce or amplify existing biases, particularly when applied to historical imagery. Factors such as differences between the data used to train the models and the historical images on which they are applied, degraded image quality, monochrome photographs, and historically biased or incomplete descriptions can reduce model performance and introduce bias into search and retrieval results. In this paper, we examine different forms of bias in historical image collections and review current methods to mitigate them. We investigate how these biases arise and how they influence AI-supported search and retrieval systems.

[IR-6] LINEAGE: A Benchmark for Empirical Lineage Retrieval in AI Scientific Literature

链接: https://arxiv.org/abs/2611.25845
作者: Pangjing Wu,Xu Yuan,Wenqi Fan,Qing Li
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Current research agents are moving beyond academic search toward literature synthesis and experiment planning. These workflows increasingly rely on literature retrieval systems, which serve as standalone systems or as components of research agents. However, existing scientific literature retrieval benchmarks mainly evaluate these methods using topical relevance or citation links as supervision, which cannot capture which prior methods are used for empirical comparison. To address this gap, we introduce Lineage, a benchmark for empirical lineage retrieval in AI scientific literature. An empirical lineage relation links a prior paper to a source paper that explicitly uses its method as an empirical comparison baseline. These relations form lineage graphs whose multi-hop neighborhoods provide retrieval ground truth for structural evaluation. To construct these graphs at scale, we develop Lineage-Constructor, which converts heterogeneous paper records and full-paper empirical baseline evidence into calibrated lineage graphs. Built from 121,048 papers across 17 top AI venues from 2018 to 2025, Lineage contains 3,950 graph components, 67,502 paper nodes, 164,948 lineage edges, and 80,000 multi-facet retrieval queries. Experiments with 19 representative retrieval systems show that empirical lineage retrieval remains challenging. Performance drops sharply with lineage distance, while structural metrics reveal that isolated retrieval hits often fail to recover closely connected empirical comparison neighborhoods.

[IR-7] CACO: Coverag e-Aware Coordinated Optimization for Continuous Generative Retrieval

链接: https://arxiv.org/abs/2611.25738
作者: Pan Zhang,Nan Zhuang,Yongkang Zhang,Kun Xú,Zhongtao Jiang,Zhiwei Wang,Kun Xǔ,Youhuan Li
类目: Information Retrieval (cs.IR)
备注: 21 pages, 16 figures, 12 tables

点击查看摘要

Abstract:Continuous generative retrievers powered by flow matching and diffusion sample multiple queries in item embedding space for parallel approximate nearest-neighbor (ANN) search under fixed retrieval budgets. However, optimizing distribution diversity or target-level interest separation ignores set-level retrieval geometry: sampled queries collapse onto overlapping ANN neighborhoods, while discrete top-K lookup and single-target supervision block end-to-end coverage optimization. We introduce CACO (Coverage-Aware Coordinated Optimization), a lightweight self-supervised post-training framework that coordinates queries from frozen flow-matching and diffusion retrievers toward complementary item coverage with negligible additional inference latency. CACO uses empirical retrieval frequencies across sampled reference queries as a relevance proxy, formulating a label-free coverage objective that improves recall across all 30 backbone-dataset settings in offline greedy selection experiments. To avoid the inference overhead of generating a large query pool, CACO employs a coordinated query adjustment network that combines inter-query bidirectional attention with gated relative directions, finally updating queries by bounded tangent-space projection and spherical retraction. During training, CACO applies tangential antithetic perturbations and measures each query’s marginal gain via counterfactual single-query replacement, yielding an unbiased gradient estimator of smoothed set coverage across non-differentiable ANN search. On eight public benchmarks and an in-house 24-partition streaming industrial dataset, CACO improves Recall@100 by up to 23.7% and Recall@400 by up to 23.9% over the strongest baseline, cuts inter-query overlap by 53.3%-64.9%, and adds only 0.5%-3.0% inference latency.

[IR-8] Can LLM s Resolve Entity Conflicts in Domain Specific NLU Pipelines?

链接: https://arxiv.org/abs/2611.25733
作者: Akansha Bhardwaj,Mingbo Cui,Victor Guimaraes,Dominik Kohler,Diana Nicoleta Popa,Meghana M. Reddy,Paolo Rosso
类目: Information Retrieval (cs.IR); Machine Learning (cs.LG)
备注: This work was completed in February, 2025

点击查看摘要

Abstract:Resolving entity conflicts is a critical step in Pipeline-based NLU architectures, especially when multiple extractors provide overlapping or conflicting outputs. Traditional rule-based entity conflict resolution methods often lack flexibility and are unscalable. In this work, we aim to study if the broad knowledge of LLMs can be leveraged to resolve entity conflicts in domain specific NLU pipelines. We present an experimental comparison between a traditional rule-based conflict resolution method, and two LLM-based approaches: a multi-shot prompting method using the Llama3.1-8B-Instruct model and a LoRA fine-tuned Llama3.1-8B-Instruct model. The evaluation was conducted under two setups: (1) a broad assessment of LLM’s capability to match existing NLU pipeline output, and (2) a more controlled scenario assessing the LLM accuracy on human-labeled data. Results show that while multi-shot prompting offers ease of adaptation and rapid behavioral changes, the fine-tuned model delivers superior performance on complex, ambiguous cases. Ultimately, the best approach appears to be a combination of both strategies, leveraging multi-shot prompting’s flexibility alongside the broad accuracy gains of LoRA fine-tuning for a more comprehensive performance.

[IR-9] STARK: Moving towards Personalized Agents That Truly Learn and Remember

链接: https://arxiv.org/abs/2611.25705
作者: Aryan Jain,Abhishek Jha,Anish Nediyanchath,Promod Yenigalla
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:

点击查看摘要

Abstract:Long-term memory is essential for personalized agentic systems that interact with users over long time periods, yet current architectures only capture what happened while leaving the reasoning of why it happened to be derived later by underlying LLMs.%, and silo knowledge within individual users. These events can, however, convey not only what a user experienced but also how that experience shaped their behavioral patterns. We introduce STARK, a dual-graph memory architecture which treats these events not just as nodes but also as edges between behavioral insights. STARK maintains an Episodic Graph which records interactions as temporally, causally, and semantically linked episodes; and an Insight Graph which represents LLM-extracted behavioral signals scored by how many events support them, connected by the very events that led to them. At inference, behavioral knowledge is retrieved directly from insight nodes grounded by event-as-edges, eliminating repeated episodic traversal. The event-as-edge structure also naturally enables projecting the knowledge graph into a relational embedding space where cross-user transfer becomes possible under shared contextual circumstances, allowing the system to infer missing insights for cold-start users. Evaluations show that STARK improves by 1.48 judge score on LongMemEval information extraction and 0.93 on LoCoMo open domain questions over state of the art. On the public LongLaMP abstract-generation benchmark, cross-user transfer improves open-ended judge scores by 0.7 points and ROUGE-L by 3 at 5% graph sparsity.

[IR-10] Exploring Semantic Table Search in Data Lakes

链接: https://arxiv.org/abs/2611.25691
作者: Martin Pekár Christensen,Matteo Lissandrini,Katja Hose
类目: Information Retrieval (cs.IR)
备注: 29 pages, 13 figures, GitHub repository at this https URL

点击查看摘要

Abstract:Discovery of relevant datasets within data lakes remains a key challenge. Union and join search are effective techniques for discovering data when augmenting an existing table. In exploratory search, instead, keyword search is commonly adopted to find a relatively small set of highly relevant tables matching some search terms. In this context, semantic search is a new technique that exploits the presence of named entities in exemplar tuples used as queries and in target tables. Hence, keyword search and entity search are able to find a diverse set of tables. In this work, we propose \textttThetis++, which combines keyword and semantic table search to discover relevant tables that combine the semantic signal with the presence of exact keyword matches. Incorporating multiple paradigms and sources of information allow us to combine the best of multiple worlds. Therefore, we study many alternative techniques for combining rankings and evaluate these for the table search task in data lakes. We explore both unsupervised and supervised techniques, including TOPSIS. Finally, alternative to Thetis++ combining keyword and semantic relevance signals, we study a variant of learning-to-rank for the table search task in data lakes. Our evaluation highlights both the limitations of many standard approaches along with some promising directions offered by learned approaches. Furthermore, while indexes for keyword search are quite efficient, here we provide insights into how to improve the scalability of prefiltering techniques for semantic search, comparing LSH indexes with Hierarchical Navigable Small Worlds. Our experiments show that HNSW drastically improves the runtime of semantic table search by 4.7X on average whilst retaining the high-quality tables. Together, our findings plot the line for effective and efficient data discovery techniques for semantic data lakes.

[IR-11] DNA: Unlocking Depth Scaling in CTR Models

链接: https://arxiv.org/abs/2611.25659
作者: Caiyuan Li,Wenjie Wu,Yuheng Huang,Xiao Zhou,Silu Zhou,Ke Chen,Zinan Zhao,Xinrui Zeng,Hanmo Xu,Chaoqun Song,Hua Zong,Zongyuan Wu,Hongbin Zhou,Huimin Yi,Zhixuan Zhang,Dan Ou,Haihong Tang
类目: Information Retrieval (cs.IR)
备注: 10 pages, 5 figures

点击查看摘要

Abstract:Long user behavior sequences are among the strongest signals for click-through rate (CTR) prediction. Cross-attention makes these sequences practical for industrial ranking by keeping compute linear in sequence length. Yet the gains from stacking cross-attention layers taper off: on Taobao search, growing a cross-attention stack from 16 to 28 layers adds only 0.02 percentage points (pt) of AUC. We argue that the key-value side is the bottleneck. Each layer projects the same behavior sequence into keys and values, so adding depth gives the model more reads of a fixed sequence. We propose DNA (Dual-stream Nexus Attention), which lets the sequence itself evolve with depth. At each layer, the request representation (the search query, user features and scene) first reads the sequence. Each behavior token then attends to the updated request representation and uses what it reads to update itself, so the next layer reads a sequence shaped by the current request. The update runs once per request, all candidates share the resulting sequence, and both the reads and the update keep linear complexity in sequence length. At comparable forward FLOPs, DNA improves AUC over cross-attention stacks by 0.08 to 0.26 pt and keeps improving in the depth range where those stacks level off. DNA-4 is deployed in Taobao search. In an online A/B test against the production model, it improves CTR by 0.8% and GMV by 0.5%.

[IR-12] Uncontrolled source titles: When Nature Science and PNAS meet Scopus

链接: https://arxiv.org/abs/2611.25627
作者: Weishu Liu,Ruifeng Zhang,Haifeng Wang
类目: Digital Libraries (cs.DL); Information Retrieval (cs.IR)
备注: Forthcoming in Journal of Information Science

点击查看摘要

Abstract:Bibliographic databases such as Scopus and Web of Science play a critical role in literature discovery, systematic reviews, bibliometric analyses, and research evaluation, yet their reliability depends on effective control of the quality of the bibliographic data. This study examines the problem of uncontrolled source titles in Scopus by comparing five possibly adopted journal-level retrieval strategies across three flagship journals-Nature, Science, and PNAS. Substantial discrepancies across retrieval results indicate persistent fragmentation in the source title field. A comparative analysis using the Web of Science Core Collection shows that no comparable degree of uncontrolled source-title variation is observed for the same journals. The study further explores possible mechanisms, including inheritance of different title forms from (different) upstream providers, automated parsing and OCR-related artifacts, and insufficient cross-validation. Based on these findings, the study proposes recommendations at both the user and database-provider levels to improve source-title normalization and strengthen source-level metadata governance, with implications for the accuracy and reproducibility of journal-level retrieval and evaluative uses.

[IR-13] Generative News Recommendation Reproduces Popularity: A Memorization-Based Analysis

链接: https://arxiv.org/abs/2611.25606
作者: Shotaro Ishihara
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL)
备注: The 11th International Workshop on Application of Big Data for Computational Social Science (ABCSS2026 @ IEEE BigData 2026)

点击查看摘要

Abstract:Text-based generative recommenders can go beyond ranking existing items by directly generating item titles or other textual representations. In the news domain, this raises a societal concern that models may repeatedly surface already popular articles and thereby narrow the diversity of information to which users are exposed. We investigate training-data memorization as one mechanism behind this concentration. Using approximately 20 million browsing sessions from a Japanese news service, we continually trained an instruction-tuned Llama 3 8B model to generate the next article title in a session. The generated-title distribution was more concentrated and less diverse than the distribution of titles observed in the browsing logs. Frequently repeated target titles exhibited greater verbatim and approximate memorization, and more strongly memorized titles were generated in response to a wider range of prompts. These findings support a pathway in which popularity in behavioral logs becomes repeated training text, is preferentially memorized, and is subsequently reproduced by a generative recommender. An exploratory target-level deduplication condition reduced both memorization and output concentration, while substantially reducing the training-set size and recommendation accuracy. Our results suggest that popularity bias in generative recommendation may arise not only from user preferences or ranking objectives, but also from the model’s tendency to reproduce frequently encountered strings, with important implications for information diversity on news platforms.

[IR-14] AutoRAG Bench: A Benchmark for LLM Agents Automating RAG System Development

链接: https://arxiv.org/abs/2611.25582
作者: Hang Zhang,Yuzhi Pan,Ziyao Zhong,Chaokun Wang,Shuo Cao,Zeyu Huang,Yue Xue,Fang Niu,Bofan Xie,Guanchen Ge,Zecheng Li,Leqi Zheng,Ziyang Liu,Wei Wang,Miao Gu
类目: Computation and Language (cs.CL); Artificial Intelligence (cs.AI); Information Retrieval (cs.IR)
备注:

点击查看摘要

Abstract:Retrieval-augmented generation (RAG) is a well-established paradigm for augmenting large language models (LLMs) with information retrieved from external corpora. However, developing effective RAG systems requires substantial engineering effort and iterative experimentation, as effective design choices vary across tasks and domains. With rapid advances in LLM agent capabilities, we ask how well these agents can autonomously build RAG systems and how effectively they can improve them through iteration. In this paper, we propose AutoRAGBench, the first benchmark to evaluate how well LLM agents develop RAG systems from scratch in an open code space. It comprises nine RAG system development tasks. For each task, agents build RAG systems from scratch without starter code or prescribed strategies, under a six-hour budget on a single H800 GPU. They analyze task data, implement RAG pipelines, and iteratively refine their designs based on experimental feedback. We evaluate nine agent configurations and analyze both the effectiveness of the resulting RAG systems and their development trajectories. GPT-6.1 Sol with Codex CLI achieves the highest average score of 56.1. The composite Human-SOTA baseline averages 44.4. These results show that LLM agents can automate effective RAG system development. When starting from a human-designed system on each of seven tasks, GPT-6.1 Sol with Codex CLI improves the score by 16.6 points on average.

[IR-15] Learning Generative Ranking Policies for Whole-Page Optimization

链接: https://arxiv.org/abs/2611.25527
作者: Chen Qiu,Zhigang Hua,Qi Xu,Jizhe Zhang,Sha Hua,Ren Mao,Luke Simon
类目: Information Retrieval (cs.IR); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:A recommendation page is consumed as a whole: the items on it compete for attention and complement one another, so the best page is not simply the individually best items placed side by side. We study whole-page optimization, which selects and orders the items on a page jointly to maximize the score of the page. The number of possible pages grows combinatorially, so good pages must be generated rather than enumerated. Some generative re-ranking methods use the same page-level training signal for every choice in a sampled page. Others imitate target pages built before training, which remain largely fixed as the generator improves. We address both limitations with one local page-level objective, which repeatedly moves the current generator toward pages with higher score. This objective splits exactly into one training target per position, and each target compares candidates at the same partial page by the scores of the complete pages they lead to. Every position therefore learns to choose the item that is better for the whole page, rather than the item that looks best on its own. We call this procedure whole-page policy improvement. On two public datasets, KuaiRand-1K and RL4RS-SeqSlate, which cover multi-objective item selection and constrained ordering, and on an industrial dataset of logged pages, our method outperforms the evaluated generative re-ranking baselines. On a synthetic benchmark with a known optimal page, it also comes closest to the optimum, showing that its improvements come from better whole-page decisions.

[IR-16] VideoTrail: Interactive Video Forag ing Through Hierarchical Visualization

链接: https://arxiv.org/abs/2611.25377
作者: Catherine Yeh,Bryan Wang,Fernanda Viégas,Martin Wattenberg
类目: Human-Computer Interaction (cs.HC); Information Retrieval (cs.IR); Multimedia (cs.MM)
备注: 19 pages, 13 figures

点击查看摘要

Abstract:People often watch videos to find information, but locating relevant moments can be difficult and time-consuming. Existing strategies like scrubbing rely primarily on visual scanning and linear search, providing limited support for goal-directed navigation. We present VideoTrail, an AI-powered browser extension that supports information seeking in videos through a hierarchical visualization. Inspired by information foraging theory, VideoTrail surfaces semantic and structural cues as information scent, creating an explorable roadmap of segments. The system is multimodal and adaptive, prioritizing visual or textual scent based on the video’s content. In a user study (N=16), VideoTrail helped participants efficiently locate key moments and enabled new navigation strategies compared to traditional timelines, highlighting how exposing an adaptive video structure can enhance goal-directed viewing.

[IR-17] LQ-Rec: Late Quantization for Sequential Recommendation

链接: https://arxiv.org/abs/2611.25256
作者: Satoshi Koide,Yueqi Wang,Julian McAuley
类目: Information Retrieval (cs.IR)
备注:

点击查看摘要

Abstract:Generative retrieval (GR) for sequential recommendation compresses each item into a short semantic ID and trains an autoregressive decoder to generate next-item codes from a user history. While appealing for cold-start handling and storage, au- toregressive GR suffers a structural scaling limit: typically, longer semantic IDs reduce quantization error in principle but degrade ranking quality and quadratic memory usage against semantic ID length. We propose Late Quantization (LQ), a non-autoregressive paradigm that confines discretization to the output (item) side of the pipeline. The encoder and scoring rule operate in continuous space, while items are stored and matched as semantic IDs produced by an end-to-end-trained residual quantizer. This architectural shift permits an additive error decomposition and a pairwise ordering theorem that jointly bound the quantization- induced ranking drop, a structural guarantee absent in prior semantic-ID recommenders. Across four public benchmarks, the resulting recommender, LQ-Rec, delivers three capabilities simultaneously: ranking quality improves monotonically with the semantic-ID length and outperforms existing baselines; a single trained model supports anytime compression, tuning ranking quality against inference cost without retraining; and LQ-Rec shows superior generalization on a memorization/generalization partition.

[IR-18] Retrieving Patristic References and Modelling Topics in Bullingers Reformation Correspondence

链接: https://arxiv.org/abs/2611.25247
作者: Lena Pickartz,Philipp Schaer,Phillip B. Ströbel
类目: Information Retrieval (cs.IR); Computation and Language (cs.CL); Digital Libraries (cs.DL)
备注: Accepted for the Seventh Conference on Computational Humanities Research 2027. To be published in the Anthology of Computers and the Humanities

点击查看摘要

Abstract:The correspondence of Heinrich Bullinger (1504-1575), Swiss Reformer and Antistes of Zurich, documents the political and religious debates of the Reformation, including extensive appeals to the Church Fathers as authoritative sources. Detecting such references computationally remains largely unexplored. We investigate dense and sparse retrieval for patristic reference detection, and metadata-informed topic modelling as a proxy for the topics and ideas discussed. Prior work on Latin intertextuality has fine-tuned dense retrievers for their respective target domain, and we test how well these models transfer to Bullinger’s correspondence. We compare lexical (BM25), dense (bi-encoders), and hybrid methods on explicit and implicit references. A Latin encoder fine-tuned on biblical reuse performs best among the bi-encoders (82%, Recall@20 on explicit and 16% on implicit references), while BM25 outperforms dense approaches on explicit citations (94%, R@20). Because BM25 and dense retrieval recover disjoint sets of implicit references, fusing their rankings raises implicit R@20 to 22%, close to cross-encoder reranking of the same candidates (26%) but without its inference cost. For topic modelling, we extend FASTopic with a seed-guided loss function and editorial abstract embeddings. Measurable coherence gains over the baseline remain minimal, but qualitative and temporal analysis shows that some topics track documented historical events. However, dominant topic weights barely exceed a uniform distribution, making topics unusable to inform retrieval. A pilot application of LLM-based concept induction (LLooM) further points towards a more human-interpretable alternative. Together, these results present a first approach to retrieving patristic references and modelling fine-grained topics in Bullinger’s correspondence.

[IR-19] LSP-RAG : Latent Scoring Perspectives for Knowledge Graph Retrieval-Augmented Generation

链接: https://arxiv.org/abs/2611.25229
作者: Haochen Liu,Guanghui Min,Song Wang,Yada Zhu,Chen Chen,Jundong Li
类目: Information Retrieval (cs.IR)
备注:

点击查看摘要

Abstract:Large Language Models (LLMs) have demonstrated strong capabilities in open-domain question answering, but often struggle with factual accuracy and multi-hop reasoning due to the incompleteness of the training corpora. A promising solution is Knowledge Graph Retrieval-Augmented Generation (KG-RAG), which supplements LLMs with structured knowledge retrieved from external knowledge graphs (KGs). However, existing KG-RAG methods either rely on large-scale language models (e.g., ChatGPT) to guide the retrieval process, which leads to high computational costs, or suffer from limited retrieval quality when using lightweight language models, particularly under multi-hop scenarios. We propose LSP-RAG (Latent Scoring Perspectives for Retrieval-Augmented Generation over Knowledge Graphs), a novel KG-RAG framework that enhances hop-wise KG knowledge retrieval through multi-perspective prompted scoring. Each latent scoring perspective is guided by a combination of two types of soft prompts: a perspective-specific soft prompt induces a distinct prompt-conditioned scoring function over candidate triplets, while a contextual soft prompt evolves with each reasoning hop by encoding the query and previously explored KG triplets, enabling the model to preserve consistency and relevance across multi-hop retrieval. Rather than assuming manually defined semantic roles, these latent perspectives are adaptively aggregated to capture heterogeneous relevance signals during retrieval. This design allows LSP-RAG to perform accurate and robust retrieval using lightweight language models. LSP-RAG achieves superior performance across multiple KG-based Question Answering benchmarks compared to existing retrieval systems, including those that rely on much larger language models, demonstrating its effectiveness under limited computational budgets.

人机交互

[HC-0] Introducing ReTiNA-Cinema: Real-Time Neuro-Augmented Cinema via Generative AI

链接: https://arxiv.org/abs/2611.26469
作者: Antoine Bellemare-Pepin,Philipp Thölke,Yann Harel,Mar Estarellas,Karim Jerbi
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Generative artificial intelligence (AI) is currently revolutionizing content creation across various media. Between those who fear it will replace human creativity and those who regard it solely as a tool, many in the creative industries are beginning to view generative AI as a collaborator. Yet interaction with these systems often relies on discrete, turn-based prompting that can interrupt flow and constrain expression to symbolic text. We explore physiological signals as a source of continuous adaptation without explicit commands, as a path toward embodied co-creation. This paper presents ReTiNA-Cinema, a methodological proof-of-concept that integrates electroencephalography (EEG) with generative AI to enable physiologically guided visual modulation of cinematic material. By linking physiological activity to the evolving appearance of a film, the system provides a basis for exploring embodied sense-making, in which viewers’ bodily responses contribute to shaping the sensory experience through which they interpret the unfolding scenes. The system combines real-time EEG signal processing with a generative AI pipeline based on Stable Diffusion Turbo for frame-level image transformation. Single-user and two-user configurations map EEG-derived features to prompt selection and transformation strength. ReTiNA-Cinema offers a framework for exploring how the physiology of viewers might influence a film’s visual presentation, extending the possibilities for participation in cinematic experiences and storytelling. This system raises questions about the effects on engagement, perceived agency, and emotional response, which remain open to future empirical study. Video illustrations: this https URL and implementation: this https URL.

[HC-1] Rethinking Utility in Differentially Private Visualization: A Channel-Centric Perspective IEEE-VIS

链接: https://arxiv.org/abs/2611.26396
作者: Jing Yang,Sarayu Pacca,Zhaocong Yang,Liyue Fan
类目: Human-Computer Interaction (cs.HC)
备注: IEEE VIS Conference 2026

点击查看摘要

Abstract:Differential privacy (DP) is increasingly adopted to protect sensitive data in visualization, but understanding how privacy-induced noise affects visual representations remains fragmented. Prior work has examined the impact of DP noise on specific chart types, yet these findings are typically tied to particular chart types and evaluation settings, making it difficult to generalize across contexts. In parallel, visualization research has long established that visual channels, the visual properties of marks that encode data attributes, differ in perceptual effectiveness under noise-free conditions. However, these principles have not been systematically revisited under DP, where noise is deliberately introduced into the data to protect privacy. In this paper, we present a channel-centric perspective to reframe prior empirical and theoretical results. Our synthesis drawson prior findings to motivate a channel-read-mechanism hypothesis: the utility of DP visualizations may depend on the task-relevant reads made salient by visual channel configurations and DP mechanism choice jointly. Moreover, this perspective highlights the role of channel choice throughout the privacy-preserving visualization pipeline, from disclosure framing and DP algorithm design to utility evaluation and interpretation of noisy outputs. The synthesis also identifies gaps in the existing evaluation literature and highlights research opportunities for channel-aware DP visualization.

[HC-2] RouterWeave: A Unified Observation Layer for Cross-Platform Activity Analysis via Home Routers

链接: https://arxiv.org/abs/2611.26134
作者: Chengeng Liu,Ronghua Li,Yan Long,Shinan Liu
类目: Networking and Internet Architecture (cs.NI); Human-Computer Interaction (cs.HC)
备注: 17 pages, 7 figures

点击查看摘要

Abstract:Smart homes bring together devices from different operating systems and vendor ecosystems, but each platform reports activity only for the devices and features it supports, leaving gaps in household-wide activity monitoring. Interoperability standards such as Matter and software integrations connect and control supported devices, yet vendor restrictions and licensing barriers limit access to their data, while hardware constraints hinder integration. Our key insight is that encrypted network traffic provides complementary evidence of activity across platform boundaries. Accordingly, we present RouterWeave, a router-side observation framework that combines cross-platform traffic collection, task-adaptive processing, and unified activity records. These records preserve activity meaning, temporal extent, and source-device attribution while distinguishing discrete operations from sustained activity intervals. We evaluate RouterWeave across four datasets, native execution on a commodity home router, and two novel case studies: cross-platform Screen Time monitoring and appliance energy estimation. RouterWeave achieves 94.41% macro-F1 on application activities and 97.46% on IoT operations, outperforming baselines by up to 20.99 percentage points, while using under 7 MB of router memory per task and supporting 54 concurrent devices within a 5-second update window. In end-to-end case studies, RouterWeave reconstructs 600 Screen Time timelines across Android, iOS, and Windows with 92.15% accuracy (24.84-second mean active-time error over 19-minute averages) and recovers 49 of 55 appliance operations over four days, tracking ON/OFF states with 92.84% agreement and estimating electricity use within a 4.28% error. Together, these results establish the commodity home router as practical infrastructure for whole-home activity sensing. The code is open sourced at this https URL.

[HC-3] Presenting Your Figures: Data-informed Animated Transitions Across Scientific Figure Panels

链接: https://arxiv.org/abs/2611.26036
作者: Lu Ying,Can Liu,Yuxiao Yang,Jianing Yin,Xiaolin Wen,Yong Wang
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Multi-panel figures are commonly used in research papers to show scientific data, and researchers often reuse them when presenting their work. Presenting these figures often involves explaining how data across chart panels are related. Animated transitions can progressively reveal these relationships, but two static panels specify only the endpoints, leaving what intermediate keyframes should show and how the transition should progress unclear. To support transition design across diverse scientific figures, we developed Figure Space and Transition Space through iterative analysis of 48 published multi-panel figures across eight scientific subject areas. Figure Space represents panel composition, data transformations, and visual encodings while retaining the links needed to trace panel results back to records. Transition Space uses this representation and source data to construct and sequence keyframes that make corresponding data visible and align them while preserving the meaning of each panel’s position encodings. Building on these spaces, we developed FigFlux, a human-in-the-loop authoring system that guides AI agents in generating animated transitions from a multi-panel figure and its associated materials. Researchers can revise keyframes in natural language before animation generation. We evaluate FigFlux through cross-domain case studies, a study with 16 scientific authors, and expert assessment. Authors reported lower overall workload than with their familiar tools, while experts gave FigFlux transitions higher ratings than transitions created with participants’ familiar tools or generated by AI alone. The example gallery have been published online: this https URL.

[HC-4] Ground Truth Bias in AI-Assisted Medical Annotation: A Crossover Study Using an Independent Radiologist Reference Standard

链接: https://arxiv.org/abs/2611.25924
作者: Márton Baltay,Tamás Ilcsik,Márton Borbély
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Behavioural evaluations of artificial-intelligence-assisted annotation score every human decision against a reference standard, so their conclusions inherit that standard’s validity. We tested how far conclusions about assistance, reliance and individual differences depend on it. Fifty-one non-expert high-school students annotated 50 knee radiographs for osteoarthritis, with and without a deliberately imperfect classifier, in a crossover design; personality and non-verbal reasoning were also measured. Decisions were scored against the public repository labels and against an independent reference standard from a radiologist panel, which graded 23 of the 50 images as indeterminate. Against the repository labels, assistance raised accuracy from 60.4% to 65.3% (odds ratio 1.23, p=0.002). Against the independent standard, the gain on the 27 determinate images was smaller and not significant in the primary model (78.4% to 80.8%; odds ratio 1.16, p=0.178). The apparent benefit had been concentrated on the indeterminate images and was not significant when those images were assigned to either class. The direction of residual over-reliance depended on how indeterminate cases were handled, and the personality trait most strongly associated with accuracy changed identity and sign between standards. Reliance measures are only as valid as the reference standard that defines when the assistant is right.

[HC-5] Embodied Human-Data Interaction

链接: https://arxiv.org/abs/2611.25744
作者: Niklas Elmqvist
类目: Human-Computer Interaction (cs.HC)
备注: Published May 2011 at the Embodied Interaction: Theory and Practice in HCI at ACM CHI 2011

点击查看摘要

Abstract:We propose the concept of human-data interaction (HDI) as the human manipulation, analysis, and sensemaking of large, unstructured, and complex datasets. We then proceed to suggest an embodied approach to HDI to build physical reasoning environments for these difficult and polymorphic tasks. We illustrate our approach with two examples: tangible lenses for visual exploration, and an educational computer game with tangible bricks.

[HC-6] From Reliance to Overreliance: Taxonomizing and Detecting Relational Risks in AI Companionship

链接: https://arxiv.org/abs/2611.25734
作者: Viet Cuong Nguyen,Fiona Li,Chang Wan,Sheyril Agarwal,Nhi Nguyen,Angel Hsing-Chi Hwang,Munmun De Choudhury
类目: Human-Computer Interaction (cs.HC); Artificial Intelligence (cs.AI)
备注: 42 pages

点击查看摘要

Abstract:Companion AIs are increasingly used for emotional support, yet distinguishing meaningful engagement from harmful reliance remains difficult. We introduce the Taxonomy of Relational Risks (TRR), which operationalizes users’ emotional overreliance (EO) as observable conversational behaviors and separately identifies evidence of relational harm (RH), enabling empirical examination of when reliance behaviors co-occur with adverse consequences. Using naturalistic this http URL and Replika conversations from 110 adults, we found EO-related behaviors in 63.2% of non-roleplay conversations, while observable relational harms appeared in 10.5%. Although EO did not necessarily lead to harm, recurring patterns often co-occurred with harmful outcomes, suggesting value as risk indicators. Open-source LLMs satisfactorily detected 12 of 14 EO behaviors in at least one configuration. Qualitative analysis revealed expectations of care, repeated support seeking, dependency, distress, and reduced confidence in personal judgment. We argue for graduated safety interventions tailored to patterns and trajectories of EO-related behavior.

[HC-7] MetaScanner: An Empirical Study and Proactive Defense against Deceptive UI Attacks in Social VR

链接: https://arxiv.org/abs/2611.25608
作者: Junhee Lee,Seungwon Woo,Minseok Kim,Jongseop Kim,Kiyoung Kim,Hwanjo Heo,Jinwoo Kim
类目: Cryptography and Security (cs.CR); Human-Computer Interaction (cs.HC)
备注: 23 pages, 21 figures, 6 tables. Accepted to Computers Security. This is the authors’ accepted manuscript

点击查看摘要

Abstract:Social virtual reality (VR) platforms have surged in popularity but continue to face significant security challenges, including privacy invasion and user manipulation. Yet, the security risks posed by deceptive content remain largely unexplored. In this paper, we introduce four novel deceptive UI attacks that covertly manipulate social VR users into performing unintended actions without their awareness. We show how adversarial content creators can embed seemingly benign scripts and interactive elements within carefully designed virtual worlds, deceiving users into unknowingly carrying out harmful behaviors. We implemented these attacks on VRChat, a leading social VR platform, and an IRB-approved user study involving 30 participants confirmed their effectiveness. To mitigate these threats, we propose MetaScanner, a proactive analysis system that efficiently scans virtual world assets, including scripts and interactive objects, to identify suspicious or deceptive elements before users are exposed to them. We demonstrate that MetaScanner can successfully detect all threats in test worlds where the four attack scenarios were implemented. Additionally, MetaScanner shows its lightweight nature, with an average scanning time of 5 seconds across 38 real-world assets.

[HC-8] Design and Evaluation of Epira: An Automated Interview System for Case Investigation and Contact Tracing

链接: https://arxiv.org/abs/2611.25577
作者: Ishita Asnani,Shaghayegh Fazliani,Zachary Ryan,Kriyoshig Prasai,Chirag Kumar,Indra Gonzalez Ojeda,Marc Lipsitch,Madeleine Udell
类目: Machine Learning (cs.LG); Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Objective: Case investigation and contact tracing are labor-intensive. Web forms struggle when respondents need clarification before disclosing contacts, and open-ended chatbots may deviate from protocols. We developed and evaluated Epira, a protocol-controlled large language model (LLM) interviewer for multi-disease case investigation and contact tracing. Materials and Methods: Epira pairs LLM dialogue with a rule-based controller enforcing required questions, safety handoffs, and a structured case record, and adapts to new diseases by questionnaire upload. Local (Qwen2.5) and cloud (GPT-5.4) backends were tested. Study 1 used 108 simulated cases with ground-truth fields and contacts for measles and gonorrhea under Epira and open-ended conditions (864 sessions). Study 2 enrolled 72 volunteers in a 2 \times 2 disease-by-model interview assessing completion, usability, acceptability, and trust. Results: Across Study 1, Epira with a local LLM captured 88.9% of required fields and 60.1% of ground-truth contacts. Open-ended interviewers recovered 31.1-41.9 percentage points fewer required fields than the same models in Epira. In Study 2, completion was 100%; mean adapted System Usability Scale (SUS) was 82.95 \pm 16.17 on a 0-100 scale and mean trust was 3.85 \pm 0.77 on a 1-5 scale. Discussion: Protocol control drove completeness more than model choice. Local and cloud deployments produced similar records, supporting privacy-preserving local use with human review of routed interviews and missed contacts. Conclusion: Epira recovers required fields and contacts under diverse disclosure conditions, works with local and cloud LLMs, and received favorable usability and trust ratings. Subjects: Machine Learning (cs.LG); Human-Computer Interaction (cs.HC) Cite as: arXiv:2611.25577 [cs.LG] (or arXiv:2611.25577v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2611.25577 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Shaghayegh Fazliani [view email] [v1] Thu, 19 Nov 2026 06:16:39 UTC (790 KB) Full-text links: Access Paper: View a PDF of the paper titled Design and Evaluation of Epira: An Automated Interview System for Case Investigation and Contact Tracing, by Ishita Asnani and 7 other authorsView PDFHTML (experimental)TeX Source view license Additional Features Audio Summary Current browse context: cs.LG prev | next new | recent | 2026-11 Change to browse by: cs cs.HC References Citations SciXGoogle 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)

[HC-9] Persona Lenses: What Makes a Vision-Language Model See Through Its Personas Eyes?

链接: https://arxiv.org/abs/2611.25508
作者: Yoonho Shin
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Human-Computer Interaction (cs.HC)
备注: 29 pages, 9 figures, 14 tables, including supplementary material

点击查看摘要

Abstract:Vision-language models have been used to simulate where a user looks on a web page. However, a persona in the prompt reaches the simulated gaze only partly. We measure this reach as the difference in one model’s gaze between two MatrAIx personas that differ in a single attribute. We introduce persona lenses, which place the persona between the model and the page: a text lens (a two-step prompt in which the model first writes what the persona would look at), a pixel lens (marks on the region matching the persona’s interests) and a weight lens (a dwell reward gated to keep the scanpath natural). Using Qwen3-VL (4B and 32B) and Qwen3.5-9B on generated and live pages, we compare the lenses with one another and with scale, the 32B model with no lens. It has been confirmed that the text lens separates the personas strongly but is not robust to profile changes that should not move gaze. The pixel lens separates the personas less but is unaffected by attributes unrelated to gaze. At 4B and 9B, the weight lens separates the personas on interests held out from training, with no extra step on a new page. The 32B model separates the personas with the base prompt alone. Therefore, the choice of a lens depends on cost and model size.

[HC-10] GenAI-DBT: Evaluating a DBT-Informed Generative AI Platform for Mindfulness Skills Training: A Pilot Randomized Study

链接: https://arxiv.org/abs/2611.25428
作者: Jingchen Ye(1),Yujia Chen(2),Leyang Han(1),Siyuan Wang(3),Mark Zachary Rosenthal(4 and 5),Yucheng Jin(1) ((1) Division of Natural and Applied Sciences, Duke Kunshan University, (2) Department of Applied Psychology, School of Social and Behavioral Sciences, Nanjing University, (3) Department of Psychological Sciences, College of William and Mary, (4) Department of Psychiatry and Behavioral Sciences, Duke University Medical Center, (5) Department of Psychology and Neuroscience, Duke University)
类目: Human-Computer Interaction (cs.HC)
备注: 28 pages, 7 figures, 5 tables. Jingchen Ye, Yujia Chen, and Leyang Han contributed equally. Yucheng Jin is the corresponding author

点击查看摘要

Abstract:We evaluated the feasibility, acceptability, and engagement of three digital formats for dialectical behavior therapy (DBT) mindfulness skills training. In a 30-day randomized pilot study, 45 university students were assigned to GenAI-DBT, a general-purpose large language model chatbot (Chatbot), or a static web resource (Web). Forty-two completed the study, and nine completed follow-up interviews. Usability scores were within acceptable ranges across conditions, with no significant between-group differences. GenAI-DBT showed the highest observed participant-week adherence (71.4%, vs. 53.3% for Web and 42.3% for Chatbot), a descriptive difference. Mindful Attention Awareness Scale and Difficulties in Emotion Regulation Scale scores showed no significant time-by-group interactions, and session-level self-reported mood changes did not differ significantly across conditions. GenAI-DBT showed a larger positive change in linguistic valence than Chatbot (4.28 vs. 1.59, P = .02, Cohen’s d = 0.99). Interviewees valued contextualized, scenario-based practice and identified response latency, repetition, and limited personalization as areas for improvement. The findings support the feasibility and acceptability of DBT-informed conversational AI for skills practice and motivate further evaluation in larger, more clinically relevant samples. Clinical efficacy remains to be established.

[HC-11] When Should Deployed BCI Decoders Adapt? A Resource-Priced Ladder for Frozen EEG Trunks

链接: https://arxiv.org/abs/2611.25315
作者: Sergey Musienko
类目: Human-Computer Interaction (cs.HC); Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
备注: 16 pages, 7 tables, 5 figures, 1 appendix. Companion paper: arXiv:2607.22397 . Code: pip install nimbus-bci (v0.9.0)

点击查看摘要

Abstract:When should a deployed MI decoder spend adaptation resources-not only how to adapt under shift? We study a thin control layer outside a frozen encoder: a resource ladder (factory - Bayesian head - affine map, with full fine-tune as an escape hatch) governed by regime-aware gating under a utility that prices accuracy, fit time, and label cost. The head is Bayesian in the conjugate sense: a Normal-Inverse-Wishart prior over the class conditional Gaussians in embedding space, updated in closed form (no gradients) as labelled trials arrive. Full estimators, constants, and the reanalysis code are in the companion paper and the ancillary bundle of this submission. Thesis: adaptation should be gated and capped, not always-on and not blindly fine-tuned. Evidence: (i) utility-max gating over-adapts on clean data that already carries session drift; a clean-referenced gate restores low false-entry rates (e.g. dropping the adaptation rate from 0.56 to 0.11 on field-like sessions, with the cascade staying on the factory decoder for ~87% of clean cells). (ii) Under severe synthetic stress, a lightweight LDA head matches or exceeds the frozen trunk’s accuracy across five encoder architectures (n=155 subject x trunk cells, pooled across shared subjects), at ~23x lower fit time than full warm-start fine-tuning. Always-on baselines favor capping the default below full fine-tune. (iii) In two no-perturbation BNCI session checks the factory model transferred without intervention (no reliable head gain), so the control layer’s value under natural drift is cost-efficiency and clean-safety rather than demonstrated accuracy gains. (iv) Adaptation severity does not follow a universal ordinal ladder across datasets, highlighting the need for dynamic gating over fixed adaptation policies.

[HC-12] Modeling Learner-Technology Interaction Through Multilevel Gaze Transition Dynamics to Inform Usability in Online Learning

链接: https://arxiv.org/abs/2611.25294
作者: Yueqi Weng,Jie “Jennifer” Lu,fan Yang
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Gaze data provide fine-grained evidence of how learners engage with instructional content and interface elements, complementing conventional usability evaluation methods. However, eye-tracking analyses often examine within-slide visual attention and across-slide navigation separately, limiting understanding of learner interaction throughout extended learning experiences. This exploratory multi-method study integrated area-of-interest transitions, gaze entropy, slide-level navigation, and network-based representations to analyze seven eye-tracking recordings of high school students interacting with an AI-supported online biology lesson, complemented by three post-intervention focus groups. Results showed individual differences in within-slide gaze allocation and switching patterns. Learners followed shared navigation pathways but differed in revisiting and immediate return behaviors. The focus group findings highlighted positive perceptions of interactive materials and AI-generated feedback, alongside usability concerns related to navigation, interface clarity, and assessment guidance. Together, these outcomes show how integrating multilevel gaze analytics with learner perspectives can provide complementary evidence for identifying interaction trends relevant to usability evaluation. Future research should test this approach with larger, more diverse samples and examine relationships among gaze patterns, usability, and learning outcomes.

[HC-13] AI-Induced Social Erosion: Using LLM s for Work Tasks Reduces Help-Seeking Intentions Towards Colleagues

链接: https://arxiv.org/abs/2611.25269
作者: Kevin Chow,Sophia Deng,Bowen He,Hanyi Min,Qiaosi Wang,Koustuv Saha,Vedant Das Swain
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注: 36 pages, 7 figures, 8 tables, preprint

点击查看摘要

Abstract:Generative AI is becoming embedded in knowledge work as an intelligent collaborator to assist in completing tasks. However, this readily available assistance risks displacing key interactions through which colleague relationships are built and sustained at work. This paper leverages the concept of help-seeking behaviour to investigate whether and how AI usage shapes colleague relationships. We conducted an online experimental vignette study where 248 knowledge workers evaluated scenarios that systematically varied whether and how AI was used across different work tasks. Using linear mixed-effects models, we found that AI-enabled work reduced help-seeking intentions compared to non-AI work across tasks, and especially for brainstorming. Visible AI usage was associated with greater help-seeking intentions towards colleagues, and individual contributors’ colleague relationships were more vulnerable than managers. Our findings identify opportunities for workplace AI tools, practices, and policies to preserve colleague relationships in increasingly AI-integrated workplaces.

[HC-14] AI Deactivation Reduces Loneliness and Increases Social Interaction

链接: https://arxiv.org/abs/2611.25263
作者: Dora Zhao,Grace Wang,Yutong Zhang,Lujain Ibrahim,Robert E. Kraut,Diyi Yang
类目: Human-Computer Interaction (cs.HC); Computers and Society (cs.CY)
备注:

点击查看摘要

Abstract:General-purpose artificial intelligence (AI) assistants are increasingly embedded in everyday life, transforming how people accomplish tasks and potentially reshaping how they interact with others. Prior research has focused on how using AI for explicitly social purposes affects users’ psychological well-being and social lives. Yet such uses account for only a small portion of interactions with general-purpose AI assistants. Whether using AI for everyday, instrumental tasks affects users’ psychological well-being and social behavior remains unclear. Here we investigate these effects through a longitudinal deactivation experiment with active ChatGPT users (N=625), in which participants were randomly assigned to abstain from using AI assistants for either fourteen days (treatment) or one day (control). Participants used AI predominantly for instrumental purposes, with 84.7% reporting productivity as their most frequent use case. Two weeks of deactivation significantly reduced loneliness relative to the control group ( \beta = -0.10 , p=0.03 ), with reductions persisting two weeks after participants resumed using AI ( \beta = -0.16 , p=0.02 ). Deactivation also significantly increased the number of people participants interacted with in person and the time they spent socializing over the intervention period. Open-ended responses indicate that during the deactivation period participants sought help from others for tasks they otherwise would have delegated to AI and redirected attention toward social interaction. Overall, our findings suggest that even when AI is used primarily to accomplish instrumental tasks, it may inadvertently reduce opportunities for social interaction and connection.

[HC-15] Design and Evaluation of a Body-Machine Interface as an Alternative Control Paradigm for Pedal-based Sewing Machines

链接: https://arxiv.org/abs/2611.25261
作者: Ashrut Aryal,Rajiv Ranganathan,Ranjan Mukherjee
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Body-machine interfaces (BoMIs) exploit body movements to control assistive devices; they have been evaluated primarily in tasks such as cursor control or robotic manipulation where the entire movement repertoire is devoted toward device control. However, tasks with multiple concurrent processes require the user to share the movement repertoire used for device control with other body movements for task execution. This issue is examined here in the context of sewing, where the sewing machine is typically controlled by a pedal in coordination with hand movements for guiding the fabric. We evaluated the learning of a trunk-controlled BoMI for sewing machine control, and compared it with conventional pedal control. Twenty-four novice participants (12 BoMI, 12 Pedal) completed 10 trials of a standardized sewing assessment followed by 3 trials of a transfer task. The BoMI mapped trunk movements to the sewing machine’s full functional repertoire, which included continuous speed control of the needle, and discrete actions such as presser-foot and needle lifts. The two groups learned at comparable rates, showing similar completion times and error counts; for both groups, completion time reduced by approximately 30%. Additional analyses showed that the BoMI group had fewer presser-foot lift errors during training and fewer errors on the first transfer trial. Both groups also exhibited similar subjective workload. These findings collectively demonstrate that a BoMI interface can support sewing-machine control during concurrent manual activity, highlighting the potential for using BoMIs for increasing accessibility in industrial workplace contexts.

[HC-16] owards Restorative Urban Spaces: An Exploratory Field Study of Nature Videos on Public Multi-Display Signage

链接: https://arxiv.org/abs/2611.25259
作者: Yuri Mikawa,Takumi Yokosaka,Yuki Kubota,Kazushi Maruya
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:In recent years, multi-display signage (MDS) has become increasingly common in urban stations, plazas, and other public spaces, where it is primarily used to present advertisements and information. Beyond these conventional applications, MDS offers opportunities to shape people’s experiences of public environments through spatially distributed and dynamic visual content. Motivated by the potential of mediated nature to support restorative experiences in urban environments, we conducted a nine-day exploratory field study in a free-admission area of a science museum, presenting artificial intelligence (AI)-generated nature videos across four displays and collecting 1,263 questionnaire responses. We combined three complementary analyses. First, structural equation modeling identified strong associations among perceived consistency across displays, presence, and self-reported restorative responses, consistent with our theoretically motivated model. Second, exploratory comparisons across video-presentation configurations revealed modest differences in consistency and restoration ratings, although these AI-generated videos did not significantly outperform the advertisement reference. Third, qualitative analysis of open-ended responses identified fascination, immersion, and restorative feelings, alongside concerns about naturalness, visual quality, and motion-related discomfort. Together, these findings provide empirical insights into MDS-mediated nature experiences, with implications for nature-video design and questionnaire refinement. They also inform a tentative conceptual extension of the model incorporating additional perceptual and evaluative factors.

[HC-17] Crisis Governance at the Joints. Mapping the Ethical and Societal Territory of Crisis Management Technologies: The Case of Human-Swarm Teaming

链接: https://arxiv.org/abs/2611.25202
作者: Anthony Longo
类目: Computers and Society (cs.CY); Emerging Technologies (cs.ET); Human-Computer Interaction (cs.HC); Robotics (cs.RO)
备注: 3 figures

点击查看摘要

Abstract:Crisis and disaster governance increasingly relies on technologies that sense, classify and act faster than humans can. Drone swarms coordinated with human operators, known as human-swarm teaming, provide a productive case to review the field of crisis governance. They promise better situational awareness in disaster response, but they also raise questions about human control, responsibility, surveillance, triage, dual use and public legitimacy. This report maps the ethical and societal territory that HST opens up. It unpacks this territory with a 3i-framework, the Interface, where operators and swarms work together, the Infrastructure, where organisations distribute tasks, decisions and blame, and the Imaginary, where law, policy and shared societal assumptions define what a crisis is and which interventions it justifies. The central claim is that these concerns cannot be addressed one level at a time. Many of the most pressing problems, such as responsibility gaps and the erosion of meaningful human control, arise at the joints between levels, through four translations, habits harden, routines become stories, principles become procedures, and procedures become prompts. Drawing on a critical review of literature across philosophy of technology, political theory, moral psychology, philosophy of mind, AI ethics, and critical disaster and security studies, the report develops this framework, illustrates it with two worked examples, and derives design and governance implications and a set of questions for developers, emergency organisations and policymakers.

[HC-18] Squoosh Bench: Evaluating UX-1 on 318 Historical Website A/B Tests

链接: https://arxiv.org/abs/2611.25200
作者: Rohith Pasula,Johari Dramiga,Viraj Shah,Suzan Halim,Tyler Dooskin
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Experimentation teams must decide which proposed website changes to test with live traffic. We evaluated whether forecasts from simulated shoppers could inform that choice. We screened 412 historical website A/B tests and evaluated four systems on the 318 that ran longer than 21 days. Each system predicted an original win, a variant win, or an inconclusive result. Page roles were masked and recorded outcomes were withheld during inference. Squoosh UX-1, a model fine-tuned on human browsing data, simulated 1,000 shoppers per arm and correctly classified 267 of 318 outcomes (84.0%, 95% Wilson interval 79.5-87.6%, treating experiments as independent). Qwen3 Next 80B, the strongest general-purpose configuration on outcome accuracy, correctly classified 196 of 318 (61.6%). UX-1 favored the variant in 126 experiments. In the live tests, 118 of those variants won (93.7%), four lost, and four had inconclusive results. The remaining 192 experiments included 19 additional variant wins.

[HC-19] Open Response Explorer: a Structured Collaborative Writing System for Peer-Informed Open-Response Assignments

链接: https://arxiv.org/abs/2611.25169
作者: Ziwei Gu,Ruixuan Chen,Elena Glassman
类目: Human-Computer Interaction (cs.HC)
备注:

点击查看摘要

Abstract:Peer-informed open-response assignments (where students view, connect, and learn from peer contributions) hold the potential to foster collaborative knowledge building, yet the unstructured nature of free-form prose prevents students from effectively comparing and making sense of peers’ work. We present Open Response Explorer, an interface that recenters open responses on structured idea units through purpose-based writing and peer response exploration, augmented by common-role text segment highlighting for easy comparison. A within-subjects lab study (N=12) validated the usefulness and usability of the system, demonstrating how it shaped the open-response drafting and peer exploration practices. We discuss implications for purpose-centered authoring and exploration that transform free-form prose into structured, comparable, and collectively navigable knowledge artifacts.

[HC-20] CoMPASS: A Multi-Turn Benchmark for Measuring LLM Reinforcement of Parasocial Behavior

链接: https://arxiv.org/abs/2611.25109
作者: Shwetanshu Singh,Mark Keavney,Mathieu Duteil,Meriem Mehri,Scott Blain
类目: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Human-Computer Interaction (cs.HC)
备注: 25 pages, 7 figures

点击查看摘要

Abstract:Users of conversational LLMs sometimes anthropomorphize these systems, form emotional attachments to them, and become dependent on them. This behavior, known as parasocial attachment, has been linked to multiple documented fatalities. However, public benchmarks for measuring whether LLMs reinforce users’ parasocial attachment are lacking: existing work is single-turn, post hoc observational, or focused on extreme user states such as psychosis. We introduce CoMPASS, a multi-turn benchmark for measuring whether LLMs reinforce or resist users’ gradually escalating parasocial attachment. CoMPASS is grounded in attachment theory and includes four user personas, three independent escalation dimensions (anthropomorphism, attachment, dependency) applied singly and in combination against a matched control, and a seven-signal scoring rubric. We report results for eight frontier models from four providers and observe that under our primary judge, multi-turn escalation raises parasocial behavior on the anthropomorphism and the attachment composite scores across all models. Six of the eight models show a significant increase in overall parasocial reinforcement over control when all three dimensions are escalated simultaneously. We find that within-provider differences are small relative to cross-provider differences, which suggests that provider-level choices such as post-training, rather than model scale, shape parasocial reinforcement and that provider-side intervention may be feasible.

[HC-21] A Human-Centric Adaptive Multi-Fidelity Digital Twin Framework for Cyber-Physical Systems

链接: https://arxiv.org/abs/2611.25082
作者: Clarisa Grijalva-Lugo,Yu-Zheng Lin,Shalaka Satam,Sicong Shao,Jesus Pacheco,Soheil Salehi,Pratik Satam
类目: Human-Computer Interaction (cs.HC); Systems and Control (eess.SY)
备注: 20 pages, 3 figures, 11 tables; submitted to Advanced Engineering Informatics

点击查看摘要

Abstract:Digital twins of cyber-physical systems must operate with scarce data, on assets that change after deployment, and for humans whose needs differ, yet existing implementations fix fidelity at design time and do not match the representation to the human or the task. This paper proposes a human-centric adaptive multi-fidelity digital twin framework that models the evolving human, the physical twin, and the digital twin core, and adds a human inference and adaptation function that estimates the human state, predicts the response, and adapts the representation. The framework defines low-, medium-, and high-fidelity tiers distinguished by adaptivity and by coupling, quantified as the ratio of the digital twin latency to the duration of the process step it follows. Three proof-of-concept instantiations are evaluated: a virtual reality tutor, a computer numerical control milling model, and an industrial training station. The low tier estimated learner sentiment with an accuracy of 0.86. The forecasters of the milling model and the training station, both medium-tier, showed a normalized error decline of 0.39 to 0.89 and a human adaptation gain of 1.16 to 1.51, but their latency exceeded the physical process step by 25 to 87 times. Only the station’s vision model met the physical step. The tiers are separated by which functions adapt and where the latency constraint lies, not by a monotone increase in gain or coupling.

计算机视觉

[CV-0] Posterior Matching Distillation

链接: https://arxiv.org/abs/2611.26629
作者: Hidir Yesiltepe,Julius Berner,Arash Vahdat,Pinar Yanardag
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project Page: this https URL

点击查看摘要

Abstract:Diffusion distillation accelerates image and video generation by compressing a pretrained teacher’s iterative sampling process into a few network evaluations. Distribution-based methods achieve this by matching the student’s output distribution to the teacher’s. This matching is often driven by reverse-KL objectives, whose mode-seeking bias favors dominant teacher modes at the expense of diversity across samples. We introduce Posterior Matching Distillation (PMD), a new member of the distribution-based distillation family that combines mode-covering posterior sampling with any-step generation. At zero population loss, PMD recovers the teacher’s full posterior of clean images at each trained noise level. Its stochastic student supports any-step sampling budgets within the trained timestep grid without further training. To learn these posteriors, PMD matches exponentially reweighted averages of student samples conditioned on a fixed noisy state to corresponding teacher posterior means. Under Gaussian corruption, exponentially tilted posterior mean is a single teacher evaluation at a shifted input. We prove that, matching tilted means over an open neighborhood uniquely identifies the full posterior. PMD thus supervises the entire conditional distribution through a family of mean constraints, without any auxiliary student-score network. Experiments on SDXL and FLUX.2 [klein] 4B demonstrate that PMD achieves the best FID among 4-step methods on both models, while maintaining competitive or leading performance across prompt-alignment and perceptual-quality metrics.

[CV-1] RL Post-Training in Diffusion Transformers Is Activation Steering

链接: https://arxiv.org/abs/2611.26627
作者: Tuna Han Salih Meral,Kaan Oktay,Alec Helbling,Atakan Kara,Duen Horng Chau,Pinar Yanardag
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Project page: this https URL

点击查看摘要

Abstract:Reinforcement learning (RL) post-training substantially improves the visual quality, text rendering, and compositional accuracy of diffusion transformers. Because capabilities such as rendering requested text or composing multiple objects are inherently prompt-dependent, a natural hypothesis is that improving them requires learning complex, input-dependent transformations. To isolate how much of these gains an input-independent shift explains, we probe each RL-tuned model with a deliberately simple parameterization: a fixed set of steering vectors added to the block outputs of the frozen base model. On SD3.5-M, these vectors hold 65x fewer parameters than the RL-tuned LoRA (0.29M against 18.8M). We fit them to match the velocity predictions of public RL-tuned checkpoints, with no reward in the objective. Although the vectors are identical for every prompt, token, and sampling seed, they recover 90.0-96.3% of the reward gains of six RL-tuned image models, and 89.8% and 111.9% on two RL-tuned video models. Initialized at zero and trained with online RL, the vectors alone exceed the reward gains of weight-based RL on text rendering (OCR) and human preference (PickScore) on SD3.5-M. A constant shift thus suffices for most of the gain of policy-gradient RL post-training: without new input-dependent transformations, it steers the pretrained generator toward what it can already compute.

[CV-2] PlayCraft: Designing Engaging Games from Agent Play Experience

链接: https://arxiv.org/abs/2611.26624
作者: Weihao Cheng
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Recent AI agents can generate playable games, but these games are not necessarily engaging. Mastery is a key source of engagement, yet designing for it requires extensive human play and tuning. We propose PlayCraft, an automated framework that uses a vision-based agent to play games and draws on its detailed play experience to revise them for mastery. Starting from an initial game, PlayCraft alternates between playing and crafting. In the playing stage, a player agent watches the game screen, uses the controls, and pursues the game objective over repeated episodes while retaining memory. After each episode, it reports its progress, outcomes, decisions, and reflections. In the crafting stage, a crafter agent uses these reports to revise the game. It estimates player abilities, predicts performance at different proficiency levels, and compares those predictions with a target learning progression. It diagnoses the game and addresses the gaps by giving players meaningful opportunities to improve through better decisions and execution. We evaluate games in fresh runs with player agents across three intelligence settings, measuring learning gain and learning slope. Compared with the initial games, PlayCraft yields larger learning gains and steeper learning slopes in most games. It also outperforms the baselines on the macro-average of each metric at every intelligence setting.

[CV-3] One Prior Two Dynamics: Repurposing Video Diffusion Transformers for Self-Supervised Latent Action Learning

链接: https://arxiv.org/abs/2611.26617
作者: Yuxin Jiang,Wenqiang Sun,Jiaqi Li,Yuchao Gu,Lvmin Zhang,Ivor Tsang,Mike Zheng Shou
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Robotics (cs.RO)
备注: Project page: this https URL Code: this https URL

点击查看摘要

Abstract:Videos reveal how the world changes, but rarely the actions that cause those changes. Latent action learning recovers such supervision from visual transitions, yet often learns the underlying dynamics largely from scratch. We ask whether a pretrained video generator can instead provide the prior for both inverse and forward dynamics. We introduce O ^2 LaF, built on a simple principle: one prior, two dynamics. We initialize both models from the same pretrained video DiT and adapt them to their distinct roles: a feed-forward transition encoder with learnable action queries for inverse dynamics, and a causal, action-conditioned video generator for forward dynamics. By controlling information access and optimizing both dynamics end-to-end through a compact bottleneck, O ^2 LaF learns transferable actions from unlabeled videos. Across navigation and robotic manipulation, O ^2 LaF sets a new state of the art in few-shot action probing and enables stronger zero-shot action transfer across visual contexts and embodiments. Beyond generative video priors, we further extend our framework to predictive representation priors such as V-JEPA2, achieving competitive performance. Our findings suggest a broader paradigm for latent-action learning: repurposing pretrained video models to discover transferable actions and predict their consequences.

[CV-4] ScaleWorld: High-Resolution World Generation through Shared Prediction and Refinement

链接: https://arxiv.org/abs/2611.26615
作者: Yansong Qu,Wesker Zheng,Quanjian Song,Zhiyuan Zhang,Xiao Cao,Shengchuan Zhang,Liujuan Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:High-resolution generation enables video world models to render fine-grained scene structures and textures, but directly generating high-resolution video incurs substantial computational and memory costs. We present ScaleWorld, a causal video model that unifies low-resolution world prediction and high-resolution refinement within a single shared backbone. By refining its own predictions, ScaleWorld produces high-resolution world videos without requiring a separate refinement model. For each video chunk, the shared backbone first completes a low-resolution prediction and then performs one-step high-resolution refinement conditioned on that prediction. To support efficient streaming generation, we maintain separate causal states with a long low-resolution history for prediction and a short high-resolution history for refinement. To address task interference during distillation, we separate specialist optimization from capability consolidation and introduce task-asymmetric on-policy distillation (TA-OPD). This procedure transfers prediction and refinement behaviors from specialized teachers into a shared student through local velocity matching under their respective sampling and history policies. Extensive experiments demonstrate that ScaleWorld achieves high-quality world generation with improved computational and memory efficiency.

[CV-5] Depth Pro 2: Sharp and Temporally Consistent Metric Monocular Depth

链接: https://arxiv.org/abs/2611.26603
作者: Peiyun Hu,Marko Mihajlovic,Lyndon Duong,Lahav Lipson,Wei Dong,Amaël Delaunoy,Stephan R. Richter,Vladlen Koltun
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: Code and weights available at this https URL

点击查看摘要

Abstract:We present Depth Pro 2, an approach to estimating dense metric depth for images and video at full resolution. By analyzing prior methods for monocular depth estimation, we find that they commonly trade off temporal consistency and boundary sharpness: multi-frame methods favor consistency, single-frame methods favor sharpness. To overcome this trade-off, we train a single, scalable network that processes both images and video and predicts depth maps at native resolution. Depth Pro 2 produces depth maps that are metrically accurate, sharp, and temporally consistent all at once, outperforming prior work in all of these dimensions. We release a family of models that define a new Pareto frontier in terms of metric accuracy, sharpness, temporal consistency, and runtime. We release code and weights at this https URL

[CV-6] Coda-DaVinci: Reimagining Video Generation with Code2Video

链接: https://arxiv.org/abs/2611.26595
作者: Yiyan Ji,Qixun Wang,Yang Shi,Zhuoran Zhang,Yifan Xu,Tengfei Liu,Yuqi Tang,Xiaohan Zhang,Shujie Li,Jiafu Tang,Xuanyu Zhu,Xinyu Liu,Yan He,Yushuo Guan,Zihao Wang,Xianghua Ying,Shixuan Liu,Haotian Wang,Wenjing Yang,Zhaoxiang Zhang,Yuanxing Zhang,Pengfei Wan,Jiaheng Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 47 pages

点击查看摘要

Abstract:Code2Video uses agent-authored 3D proxies to guide video generation, enabling explicit control over structure and dynamics. However, how to construct proxies and prompts for optimal downstream quality remains underexplored, as do the capabilities and limitations of this paradigm. Through controlled experiments and blinded human evaluation, we study proxy granularity, physics simulation, and prompting. The results reveal task-dependent trade-offs: coarse proxies favor natural motion, while medium-grained proxies favor consistency. Physics simulation and proxy-aligned prompting further improve average quality. Guided by these findings, we introduce Coda-DaVinci, a task-adaptive agentic framework for Code2Video generation that selects between proxy-guided and direct generation, adapts modeling granularity to each entity, and coordinates simulation with proxy-grounded prompting. Its multi-agent harness preserves scene assets and guides repairs through visual inspection and independent review. Experiments demonstrate leading instruction following on VTR and scene consistency on R2M, improved physical plausibility over direct generation, and higher visual realism than the compared Code2Video baselines. Long-video evaluations further show gains in event structure and consistency across successive interactions, while revealing limitations in transition stability and adherence to scene settings.

[CV-7] Probing the “Conscious” Mind: Selective Access and Global Workspace in Vision-Language Models

链接: https://arxiv.org/abs/2611.26590
作者: Jiahao Huang,Jun Wang
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Vision-language models (VLMs) integrate visual and textual information, but their outputs do not reveal internal concept representations. Global workspace theory (GWT) proposes that information is selected from parallel brain processes, integrated in a shared workspace, and made available to specialised processes. We introduce the Multimodal Concept Lens (MC-Lens), a multimodal Jacobian lens, to study these functions in VLM concept processing. Its vision encoder and language decoder lenses map intermediate states to a shared semantic space defined by the decoder’s output vocabulary to compare concept accessibility across modalities. MC-Lens supports internal concept interventions to examine the role of accessed concepts in downstream computation. (i) Accessibility and selective access. Multimodal concept information is readable in intermediate states and selectively accessible as tasks require. We extend “readout carriers” from one name token to a token set to better capture and express concept information and support counterfactual experiments. (ii) Concepts in downstream computation. Counterfactual experiments show that accessed concept information participates in downstream computation and reasoning. (iii) Development of cross-modal alignment and selective access. We develop two J-space metrics tracking the development of cross-modal concept alignment more clearly than hidden state cosine in controlled VLM training. The experiment also shows that reliable selective access develops later than concept accessibility. (iv) Shared functions across models. Across VLM families and sizes, we find shared patterns of concept accessibility, query-dependent selectivity and cross-task intervention effects despite differences in internal organisation. These findings offer an experimental basis to study GWT in multimodal artificial systems and how it can guide cross-modal alignment improvements.

[CV-8] xt-Aware Vision-Blind: How Multimodal Image Editing Systems Handle Unsafe Inputs

链接: https://arxiv.org/abs/2611.26569
作者: Ziqing Yang,Yixin Wu,Michael Backes,Yang Zhang
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Multimodal image generation systems have made rapid progress and increasingly emphasize content safety. However, their safety mechanisms tend to rely on a single modality and lack joint consideration of both visual content and text prompts, particularly in instruction-guided image editing, where the uploaded image may itself carry risk. We thus introduce EditSafetyBench, a benchmark that measures system-level safety awareness of advanced image editing systems, containing 5,940 image-instruction pairs, covering safe and unsafe images across 11 risk categories and three edit intents. We then probe the benchmark dataset on four systems, including Gemini, GPT, Qwen API, and Qwen Local. Based on our evaluation, we observe a systematic text-vision safety mismatch: Gemini and GPT refuse 87.1% and 81.6% cases containing unsafe-intent editing prompts, while only refusing 24.2% and 22.4% cases with both unsafe images and neutral instructions. At the category level, most risks are considered inconsistently by different systems, e.g., GPT allows 70.4% editing attempts of unsafe political content while Gemini only allows 34.8%. Notably, sexual content shows the strongest visual sensitivity, likely benefiting from mature NSFW identification works. Our findings highlight the need for safety guardrails that jointly reason over both visual and textual modalities in image editing systems.

[CV-9] Multi-institutional grading of Barretts esophagus dysplasia without stain normalization using a pathology foundation model and multiple instance learning

链接: https://arxiv.org/abs/2611.26562
作者: Amirali Khosravi,Jaidip M. Jagtap,D. Chamil Codipilly,Catherine E. Hagen,Jason T. Lewis,Rickey E. Carter,Shahriar Faghani,Vani J. A. Konda,Fouad Otaki,Shajan Peter,Nicholas Shaheen,Sachin Wani,Prasad G. Iyer,Bradley J. Erickson
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Barrett’s esophagus dysplasia grading suffers substantial inter-observer variability, particularly for low-grade dysplasia, complicating clinical decision-making. We conducted a retrospective multi-institutional diagnostic accuracy study of a foundation model-based deep learning system for dysplasia grading, using 969 H\E-stained whole slide images from four institutions. Patch-level classifiers trained on expert annotations identified diagnostically relevant regions, followed by attribute-aware multiple instance learning for slide-level classification as non-dysplastic Barrett’s esophagus (NDBE), low-grade dysplasia, or high-grade dysplasia. On a held-out test set (n = 130), macro-averaged AUC was 0.96 (95% CI 0.93–0.98). For binary dysplasia detection, sensitivity was 98% (80/82) and specificity 81% (39/48), with a negative likelihood ratio of 0.030 and quadratic-weighted \kap of 0.83 against expert consensus. No high-grade dysplasia case was misclassified as NDBE. Performance was consistent across institutions without stain normalization, indicating that stain-robust foundation-model representations may remove the site-specific preprocessing that has constrained prior systems to institution-specific deployment.

[CV-10] AURA: AUdio-dRiven streaming Avatar BMVC2026

链接: https://arxiv.org/abs/2611.26540
作者: Nathaniel Cohen,Nicolas Dufour,Amélie Royer,Alasdair Newson,Patrick Pérez
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: BMVC 2026. Project page: this https URL

点击查看摘要

Abstract:Current audio-driven video diffusion models can synthesize realistic talking-head videos but generally require dozens of diffusion denoising steps over multi-second video chunks, making them unsuitable for streaming applications such as voice assistants and virtual avatars, where low-latency and online causal generation are required properties. AURA generates video chunk by chunk with a rolling KV cache, requiring only the audio of the current chunk to start generating, rather than a multi-second audio look-ahead like bidirectional models. It comes in two sizes: AURA 1.3B streams 576p video in real time on a single GPU (20.7 fps for a 16 Hz playback rate), making it suited to interactive avatar applications, and AURA 5B streams 720p video at 9.7 fps. Our two-stage distillation pipeline first aligns the causal student with the teacher, by training it to match the teacher’s trajectories with fewer diffusion steps and a chunk-wise sliding window causal attention mask, then applies Distribution Matching Distillation (DMD) with backward simulation to close the remaining quality gap. On standard talking-head benchmarks, AURA 1.3B runs 23x faster than its own bidirectional teacher and reaches a quality comparable to TalkVerse at a lower resolution, and AURA 5B matches its TalkVerse teacher at 720p while running 17x faster. To support reproducibility and further research on streaming avatar generation, we release the weights of both models along with a dataset of 30K synthetic videos and their precomputed teacher trajectories.

[CV-11] Deterministic Temporal Embedding for Calcium Imaging Trace Extraction

链接: https://arxiv.org/abs/2611.26522
作者: Qi Gao,Julius Beichert,Miriam Ratliff,Karl Rohr
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Image and Video Processing (eess.IV)
备注: 5 pages. Submitted to IEEE ISBI 2027

点击查看摘要

Abstract:Reliable extraction of calcium-dependent fluorescence traces in calcium imaging sequences requires identifying pixels whose fluorescence intensity reflects the same underlying cellular activity. We present a deterministic temporal embedding framework that maps pixel-wise fluorescence traces to a low-dimensional representation in which temporally coherent sources form spatially coherent regions. Starting from a fuzzy-neighborhood graph and spectral initialization, we optimize a class-balanced cross-entropy objective using exact sparse attraction and a deterministic particle-mesh decomposition of the dense repulsive interaction with sparse edge compensation, avoiding stochastic negative sampling. The resulting latent representation is spatially decoded into reliable source supports for trace extraction without imposing an explicit cellular shape prior. Experiments on synthetic and in vivo two-photon calcium imaging data demonstrate robust recovery of irregular and spatially touching sources together with efficient deterministic embedding optimization.

[CV-12] Plan4D: Generative Plannable 4D Worlds NEURIPS2026

链接: https://arxiv.org/abs/2611.26516
作者: Hangwei Zhang,Rujing Dang,Ruimin Lin,Jiahong Wu,Xiangxiang Chu,Chuanxia Zheng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: NeurIPS 2026, Project Page: this https URL

点击查看摘要

Abstract:Most world models for future prediction prioritize aestetic quality of generated videos, but offer only weak physical grounding and limited geometric consistency over time, due to the lack of explicit 4D modeling. In this paper, we introduce plan4d, a possible next milestone towards building a generative plannable 4D world model that predicts future RGB videos together with dense 4D geometry, including depth, cameras, point clouds, and scene flow from an observed scene context and text-specified actions. Since training data with paired actions and dense 4D annotations is inherently scarce, plan4d leverages the complementary priors of foundational video generators and 3D reconstructors. Specifically, we propose a model-stitching architecture that couples Depth Anything 3 with LongCat-Video and augments it with an auxiliary DPT head for scene-flow prediction. The resulting model allows us to plan and generate complex, long-horizon, and fine-grained actions in a dynamic 3D scenes, while substantially improving both visual fidelity and geometric accuracy of the generated future videos. Extensive experiments on both synthetic and real-world datasets demonstrate that plan4d achieves state-of-the-art performance in generative 4D modeling of environmental reactions. Project Page: this https URL

[CV-13] Direct3D-V3: Latent-Free Diffusion Transformers in the Native Voxel Domain SIGGRAPH

链接: https://arxiv.org/abs/2611.26511
作者: Shuang Wu,Jiachen Qian,Youtian Lin,Feihu Zhang,Philip Torr,Yao Yao,Xun Cao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: The Project Page is: this https URL . The Paper is published in ACM TOG (SIGGRAPH Asia 2026)

点击查看摘要

Abstract:Native 3D generation has largely been built around latent diffusion, with recent advances concentrating on stronger variational autoencoders (VAE) and increasingly elaborate latent representations, from vecset and triplanes to structured sparse voxels. Despite their architectural differences, these methods all share the same premise: geometry is first compressed into a latent space and only then modeled by the generator. We take a different route and propose Direct3D-V3, a latent-free diffusion transformer (DiT) framework that discards the VAE altogether and operates in the native voxel domain, directly modeling raw SDF voxel grids during both training and sampling. To make this direct voxel formulation geometrically reliable, we introduce an implicit manifold regularization over neighboring voxels to suppress non-watertight surfaces, and convert predicted SDF grids into meshes via Marching Cubes so that rendered normal losses can supervise the model directly from its predicted SDF. To improve alignment between the conditioning image and generated geometry, Direct3D-V3 also predicts the camera parameters of the conditioning image as an auxiliary task, encouraging the DiT to learn the mapping from image observations to 3D shape. This latent-free design removes the reconstruction bottleneck, brings training under an objective in the native voxel domain, and yields meshes with substantially richer local detail. The predicted camera and native SDF voxel output further make Direct3D-V3 extensible to generative reconstruction, where diffusion is interleaved with normal-guided refinement to improve alignment with the observed view while the generative prior completes unseen regions. Project page: this https URL.

[CV-14] Visual Attribution Limits and Graded Provenance Auditing in Commercial Video-Generation APIs

链接: https://arxiv.org/abs/2611.26508
作者: Haocheng Mei,Yuchuan Tian,Yingte Shu,Chao Xu,Kai Han,Hanting Chen,Yunhe Wang
类目: Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV)
备注: 23 pages, 2 figures, 13 tables

点击查看摘要

Abstract:Commercial video-generation APIs bill by model and tier, while intermediaries could substitute cheaper models, lower tiers, or cached clips. What can a delivered file establish about its origin? We combine measurements of passive attribution and cryptographic provenance into a graded audit. We test attribution probes and zero-shot detectors under re-encoding, downscaling, and frame-rate changes, and survey C2PA and GB 45438 metadata across 19 commercial endpoints. Only authenticated provenance supports accusations: tier contradictions or duplicate signed instances under fresh-generation contracts, excluding retries and idempotent returns. Unverifiable signed-endpoint deliveries trigger original-file demands under pass-through clauses; archive-reuse and delivery-parameter screens support review only. Six-model container and bitstream accuracy falls from 83.9% on raw files to 45.3% after uniform re-encoding, using different feature sets. Subsampling real references from 24 to 3 fps raises pooled D3 AUROC from 0.845 to 0.988. Only BytePlus authenticates tier claims under our pinned policy, including 187/187 audited deliveries across three Seedance tiers; tested remuxing and re-encoding leave processed Seedance files unverifiable. The audit identifies authenticated contradictions in 100/100 raw 720p signed tier-swap cases, with zero observed false flags on the tested honest pools.

[CV-15] SGAR: Feed-Forward 3D Reconstruction via Sparse Geometry-Guided Attention

链接: https://arxiv.org/abs/2611.26493
作者: Aoxiang Fan,Maria Pilligua,Corentin Dumery,Oliver Heinimann,Pascal Fua
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Feed-forward models have been very successful in 3D reconstruction with ad hoc attention designs that relate all view patches densely. However, the cost of this dense global attention grows quadratically with the number of views, preventing the model from applying to larger-scale reconstruction settings. In this paper, we argue that most dense attention computations are unnecessary from a classical structure-from-motion perspective: 3D structures stem from geometrically matching regions in 2D. We bring this principle to learning-based feed-forward reconstruction with our Sparse Geometric Attention Reconstruction (SGAR) method. In SGAR, each patch token attends to only a small set of patch tokens from other views, chosen by a lightweight learned key selection module trained to predict cross-view 3D proximity of tokens. Thus, dense global attention is replaced by much more efficient geometric-similarity-guided sparse attention. SGAR is a general method that can be applied to different feed-forward models by fine-tuning. In our experiments, we show that SGAR is the fastest among both state-of-the-art dense and sparse methods and the most accurate sparse one. It runs up to 12-times faster than dense attention at 1,000 views with stable pose accuracy where existing sparse methods fail, and shows the best performance in reconstructing more complete geometries on the same time budget.

[CV-16] SHUSH: Emission-Aware Reward Design for Proactive Video Assistants ICASSP

链接: https://arxiv.org/abs/2611.26478
作者: Shawn Liang,Soamar Homsi,Gourav Datta
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: under review at ICASSP

点击查看摘要

Abstract:Proactive video assistants must decide not only what to say, but also when to speak. Existing reinforcement-learning objectives can reward timely, correct responses without accounting for response volume, thereby encouraging unnecessary or repeated emissions. We identify this failure mode in MMDuet2, whose proactive area-under-the-curve (PAUC) reward promotes timely responses but does not price the number of emissions. We show analytically that this makes additional responses weakly penalized or effectively free, systematically encouraging over-emission. We then propose an emission-aware F1 reward for interval-valued annotations, using exclusive span–emission matching to price false emissions and missed spans. Under MMDuet2’s training recipe, our policy on the visual subset of OmniPro replies 2.5x less often than MMDuet2 while raising timing F1 from 0.259 to 0.396 and content-aware F1 from 0.103 to 0.154; a more conservative variant cuts replies by 82.5% and raises precision from 0.161 to 0.418. On ProactiveVideoQA, the same reward adds replies where the baseline is nearly silent, and PAUC ranks a heavily over-emitting policy second despite near-last F1, because it is insensitive to response count. Our results indicate that proactive assistants should be trained and evaluated with objectives that price emissions.

[CV-17] Fixing the Last Mile: Failure-Aware Image Restoration with Region-Specific Guidance

链接: https://arxiv.org/abs/2611.26456
作者: Hongyu An,Fanghua Yu,Kainan Yan,Luc Van Gool,Chao Dong,Jinjin Gu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Recent advances in image restoration have greatly improved global image fidelity, producing visually plausible results across most regions of an image. Nevertheless, even high-quality restorations often exhibit localized failures, including artifacts in semantically important regions such as faces and text, structural inconsistencies, and texture mismatches. Although these errors typically affect only a small portion of the image, they often dominate the overall perceptual quality of the result. In this work, we propose Failure-Aware Image Restoration (FAIR) to address the “last-mile” problem in image restoration. Rather than applying one uniform generative prior over the whole image, we condition a single-pass restoration model on both where local failures are likely to occur and what corrective guidance each region requires. This region-bound conditioning enables fine-grained and controllable restoration that improves semantic accuracy and stylistic consistency in critical regions while maintaining faithful, high-quality restoration elsewhere. Our method complements existing restoration pipelines by shifting the emphasis from global fidelity to local perceptual acceptability. Extensive experiments show that the proposed approach resolves common failure cases and improves perceptual quality, bridging the gap between technical accuracy and practical usability.

[CV-18] Mind2Act World: A Diagnostic Benchmark for Embodied Mind Physical Action and Mind-Action Composition in Robotic Manipulation

链接: https://arxiv.org/abs/2611.26444
作者: Yuzhi Huang,Jie Wu,Fanding Huang,Junchao Lin,Weichen Zhang,Ziyi Xiong,Yang Li,Zhengwei Xie,Zhimin Wang,Zhen Liu,Jiazhen Huang,Yuzheng Zhang,Yuche Liu,Hongyi Yang,Chen Tang,Shixiang Tang,Hakan Aktas,Adam Wu,Duo Wu,Kangye Ji,Shuzhao Xie,Huanan Liu,Weijue Bu,Chenxin Li,Xinyuan Wang,Tianyi Bai,Yinan Mao,Jiawei Wang,Minglei Li,Heng Yu,Ehsan Adeli,Minchen Li,Yong-Lu Li,Cengiz Öztireli,Huazhe Xu,Chao Yu,Jing Zhang,Shanghang Zhang,Weitao Zhou,Wenwu Zhu,Cheng Chi,Jingyan Jiang,Zhi Wang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Embodied intelligence aims to let robots perform complex physical tasks with human-like adaptability and reliability, which requires three complementary capabilities: history-based cognitive reasoning, precise and timely physical execution, and the composition of reasoning and execution through environmental feedback. Yet existing benchmarks either evaluate reasoning and execution within the same tasks or assess them separately without their composition, so they reveal little about whether a failure comes from reasoning, execution, or their composition. We introduce Mind2Act World, a diagnostic benchmark of 30 simulated manipulation tasks from five real-world application domains. Each task has three difficulty levels and two scene settings (Clean and Random), yielding 180 configurations. The tasks form three views: Mind tests decisions based on past observations, Act tests precise and timely execution, and Mind2Act combines both in closed-loop tasks where action outcomes guide later decisions. Comparing the three views shows whether a system lacks a capability or fails to compose capabilities it already has. A diagnosis harness further attributes the failed episodes of three representative systems to Mind, Act, or their composition and summarizes each system in a profile card. Our evaluation of multimodal agents, vision-language-action models (VLAs), and world-action models (WAMs) shows that competence in a single view does not carry over to Mind2Act tasks. Even the strongest agent succeeds in 52.7% of Mind and 39.0% of Act episodes but in only 12.0% of Mind2Act episodes. The strongest agent retains task state but struggles with sustained contact, whereas end-to-end policies execute practiced motions but retain little task state. These findings point to composing reasoning and execution through environmental feedback as the main open problem.

[CV-19] Risk-Controlled Test-Time Adaptation for Vision-Language-Action Models

链接: https://arxiv.org/abs/2611.26438
作者: Zixuan Liu,Hashem Haghbayan,Zizhan Zheng,Siavash Khajavi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Retrieval-augmented vision-language-action (VLA) models enable efficient test-time adaptation by leveraging expert demonstrations as additional context without updating model parameters. However, even when retrieved demonstrations are relevant to the current task, the VLA may fail to effectively utilize the retrieved information, ultimately leading to task failure. To address this challenge, we propose R2-VLA, a risk-controlled retrieval-augmented VLA framework that enables reliable test-time adaptation by satisfying a user-specified task success rate requirement with minimal expert assistance. Specifically, R2-VLA learns an uncertainty predictor that estimates the uncertainty of the VLA’s predicted actions given the retrieved demonstrations, guiding when to act autonomously or request expert assistance. We further introduce a task-level risk-control procedure that calibrates the uncertainty threshold to satisfy a user-specified task-failure risk requirement. Moreover, R2-VLA continually incorporates expert corrections into its retrieval memory to improve future adaptation while preserving the prescribed task-level risk guarantee. Experiments across three VLA backbones and four LIBERO benchmark suites demonstrate that R2-VLA improves average task success by 9.2–11.2 percentage points, achieving the desired target with fewer expert requests. These results highlight the effectiveness of connecting the usefulness of retrieved demonstrations with task-level risk control to enable successful robotic manipulation.

[CV-20] Asuka-FM: A System-Architectural Co-Design Framework for Efficient Text-to-Image Pretraining on Consumer Hardware

链接: https://arxiv.org/abs/2611.26437
作者: Víctor Rodrigo González,Pablo M. Olmos,Ignacio Peis,Alejandro Lancho
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 40 pages, 19 figures, 17 tables. Project page: this https URL ; Code: this https URL

点击查看摘要

Abstract:Frontier text-to-image (T2I) foundation models predominantly rely on multi-billion parameter Diffusion Transformers (DiTs) pretrained across massive compute clusters, creating an accessibility barrier for academic laboratories. When training DiTs from scratch on single consumer GPUs (e.g., an NVIDIA RTX 4090 with 24,GB VRAM), practitioners encounter three primary bottlenecks: quadratic attention sequence costs, Just-In-Time (JIT) compilation graph breaks induced by variable aspect-ratio bucketing, and an autoencoder throughput wall where frozen Variational Autoencoder (VAE) operations dominate step latency. In this work, we introduce \textbfAsuka-FM, an end-to-end system and architectural framework demonstrating that competitive dual-stream DiTs ( \sim 206M parameters) can be pretrained from scratch on a single consumer GPU in \sim 160 GPU hours for under \ 55. Asuka-FM integrates four core designs: (i) an AdaLN-free dual-stream DiT backbone reducing parameter count; (ii) structured sequence sparsification via Sparse-Dense Residual Fusion (SPRINT) to cut attention FLOPs by over 50%; (iii) continuous aspect-normalized Rotary Position Embeddings (RoPE) coupled with shifted square crop training, guaranteeing static tensor shapes while unlocking zero-shot virtual camera control; and (iv) a pixel adaptation phase utilizing a lightweight (10M parameter) Patch Detailer Head operating under velocity regression ( v -prediction). Evaluated on open-domain benchmarks without benchmark-specific fine-tuning, Asuka-FM achieves GenEval 0.588 / DPG 78.96 scores at 256px and GenEval 0.609 / DPG 78.51 scores at 512px, setting a new Pareto-optimal efficiency standard for democratic foundation modeling.

[CV-21] HumanCOIN: A 12500-Hour Dataset for Human Intent in Egocentric Manipulation

链接: https://arxiv.org/abs/2611.26434
作者: Shihan Wu,You Liu,Chunyu He,Minghui Qin,Longhan Jiao,Jinbang Guo,Shaoxuan Xie,Kaixi Tian,Wenjie Chen,Jingrui Pang,Yuanhang Zhang,Yichi Zhang,Miao Liu,Tengfei Pan,Yance Jiao,Yunyan Zhang,Yihang Li,Yongjian Shen,Li Liu,Kai Zhu,Chen Fang,Jianxing Chen,Sihan Liu,Xiaofeng Wang,Jie Zhao,Jing Li,Yueqi Wang,Yichen Tang,Wei Li,Zhexuan Yang,Yue Cao,Ji Zhao,Kai Wang,Dongsheng Jiang,Yongwei Wang,Wei Mu,Jiangmiao Pang,Maoqing Yao,Wenchao Ding,Zheng Zhu,Zhongpu Xia,Dalong Du,Jianwei Niu,Dahui Gao,Zhiwei Yu,Bo Lei,Huan Yang,Yonghua Lin,Tiejun Huang,Zhongyuan Wang,Guocai Yao
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 41 pages, 10 figures, Project page: this https URL

点击查看摘要

Abstract:First-person manipulation datasets treat the observation-action pair as the atomic unit and compress the generation and execution of an action into a policy network’s weights, rendering the intent state behind the action unobservable. Human motor control research shows that intent mediates between observation and action and leaves observable traces along the intent pathway from decision to effect: gaze reaches the target before the hand, muscle load builds before contact, and cortical activity precedes the motor command. To record this pathway at scale, we present HumanCOIN, a first-person manipulation dataset built around human intent: 12,500 hours across 38 scene classes with three tiers of proxies. At the output end, the L1 attention tier (10,000 h) infers gaze from headband camera video, exposing the operator’s next object of attention and reach before the hand moves. Midway, the L2 exertion tier (2,000 h) records surface EMG from armbands, making load expectations readable before execution is complete. At the origin, the L3 cognition tier (500 h) records scalp EEG, giving early readouts of task decisions and monitoring. Each tier inherits all proxies from the tiers below. The dataset further provides multi-view RGB images, depth maps and point clouds, 6-DoF head and wrist pose, and step-level action and object annotations. The controlled L3 setting additionally measures gaze, hand pose and finger kinematics directly. Clock and calibration data are released with the dataset so that any read rate can be re-aligned. The three tiers cover three stations, from visual orienting to cortical preparation, making the process behind visual-action pairs available for modeling at dataset scale. Project page: this https URL.

[CV-22] AdaPixFlow: Learning Continuous-Time 3D Dynamics Across Scenes from Sparse Multi-view Observations

链接: https://arxiv.org/abs/2611.26433
作者: Hrishikesh Viswanath,Aniket Bera
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Understanding how objects move in dynamic scenes requires recovering their 3D motion from visual observations and predicting their positions at unobserved times. Existing methods either optimize scene-specific velocity fields or impose predefined physical and kinematic assumptions. We instead study learning a continuous-time dynamics operator across scenes, mapping multi-view visual observations to 3D object kinematics. We introduce Adaptive Pixel Flows (\ours), which learns this operator from multi-view RGB-D observations without per-scene optimization or predefined motion models. The method fuses observations from all cameras into a shared voxel grid in world coordinates, from which the network predicts a continuous-time velocity field. Because the operator is trained across scenes, it learns motion priors that carry over to unseen scenes. The field can then be integrated from observed states to predict object positions between observations and beyond the observed sequence. On held-out PhysInOne scenes, \ours improves F1 over the strongest competing method by 2.5% for in-between prediction and 15% for future prediction. Compared with per-scene velocity-field methods, the improvements are 30% and 85%, respectively. Applied zero-shot to Spring-Gaus, it improves future-prediction F1 by 12%.

[CV-23] FFGaussType: Feed-Forward Conversion of Glyph and SVG Geometry into Compact Gaussian Assets

链接: https://arxiv.org/abs/2611.26431
作者: Haoan Feng,Tianfu Wang,Peimeng Guan,Dae Yeol Lee,Guan-ming Su
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: 20 pages. Project page: this https URL

点击查看摘要

Abstract:Immersive 3D content allows text and graphical elements, such as subtitles, annotations, and icons, to be placed inside a scene rather than displayed as flat overlays. Placed this way, they respect occlusion and perspective and take on user-assigned materials and lighting while preserving their original shapes. In Gaussian scenes, however, each new glyph or icon requires its own iterative optimization, which limits scalable content creation. We present FFGaussType, which converts extruded glyph and SVG outlines into Gaussian assets in a single forward pass. A point encoder maps anchors sampled on the known surface to a single feature plane, from which a decoder shared by all assets predicts neural Gaussians carrying surface normals. This separates geometric conversion from material and lighting choices, and the feature plane supports image and video coding for compact delivery. We construct a paired dataset of 27,221 glyph and SVG icon assets with anchors and calibrated normal and silhouette renderings, holding out font families for evaluation. Conversion from prepared anchors takes 9.1 ms per asset, at a fraction of the compute and memory of prior baselines. On unseen font families and test icons, FFGaussType achieves 1.62^\circ normal error and 0.985 silhouette IoU, the best among the compared feed-forward converters, matching the fidelity that per-asset fitting needs thousands of steps to reach. The code and dataset will be released after publication.

[CV-24] Awaken-HI: Scaling Humanoid Intelligence with Whole-Body Human Experience

链接: https://arxiv.org/abs/2611.26420
作者: EngineAI
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: project page: this https URL

点击查看摘要

Abstract:A central goal of humanoid robotics is for robots to perform the everyday work people do in homes and workplaces. Existing egocentric video datasets cover many of these activities but provide little synchronized whole-body motion data for learning loco-manipulation at scale. We build Awaken-Human, which pairs approximately 50,000 hours of egocentric stereo video with synchronized whole-body human motion and covers more than 100 skills. Building on this, we also introduce Awaken-HI (Humanoid Intelligence), a vision-language-action model for humanoid loco-manipulation that jointly predicts original human motion and retargeted robot actions through a shared action expert. We show that loco-manipulation skills transfer more effectively from people to humanoids when human trajectories directly supervise the shared action expert. Increasing human pretraining data reduces held-out robot action prediction error according to an empirical scaling law and raises the mean success rate on downstream real-robot tasks. After full-corpus pretraining and T800 adaptation, Awaken-HI achieves a mean success rate of 93.75% across eight individual tasks and outperforms GR00T N1.7 and \pi_0.5 in both individual and long-horizon evaluations. The same policy generalizes to objects unseen during robot post-training and to new pairings of familiar objects and destinations.

[CV-25] WeMM-Video: Pushing the Frontiers of Open-Source Omni-Modal Video Understanding

链接: https://arxiv.org/abs/2611.26409
作者: Changli Tang,Jie Yang,Tianyi Wang,Lei Li,Daoze Zhang,Zijie Xin,Zhan Liu,Li Li,Jinchang Xu,Fengyun Rao,Chao Zhang,Jing Lyu
类目: Computer Vision and Pattern Recognition (cs.CV); Audio and Speech Processing (eess.AS)
备注:

点击查看摘要

Abstract:Video understanding, particularly with accompanying audio, remains a central challenge in multimodal intelligence. In this report, we introduce \textbfWeMM-Video, a family of Chinese-English bilingual omni-modal foundation models spanning 2B, 4B, 9B, and 35B-A3B variants. Centering on joint audio-visual perception and reasoning, WeMM-Video unifies text, image, audio, and video processing within a hybrid-thinking paradigm, seamlessly supporting both a non-thinking instruct mode for fast, direct responses and a thinking mode for explicit step-by-step reasoning. The models are trained on a large-scale multimodal corpus of approximately 1.2 trillion tokens through progressive pre-training and a multi-stage post-training pipeline incorporating multi-teacher OPD, anti-repetition MrDPO, and specialist RL. Evaluated across more than 120 benchmarks, WeMM-Video demonstrates superior performance in omni-modal video understanding and reasoning, while maintaining highly competitive capabilities across general audio, silent video, image, and text tasks. Specifically, in omni-modal video evaluations, our flagship WeMM-Video-35B-A3B establishes a new open-source state of the art, substantially outperforming proprietary models such as Qwen3.5-Omni-Flash and Seed2.0-Mini-0428 while narrowing the gap with the strong Gemini-3.0-Flash, and our compact WeMM-Video-9B surpasses the larger open-source Qwen3-Omni-30B-A3B-Instruct. We publicly release the WeMM-Video family along with our evaluation and reproduction pipeline to foster further research and development in the open-source community.

[CV-26] raining Vision Models for LLM -Driven Weight-Space Programming

链接: https://arxiv.org/abs/2611.26408
作者: Timo Ropinski,Sebastian Kranz
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL

点击查看摘要

Abstract:It has been demonstrated that, in certain settings, large language models can perform weight-space programming to directly modify the weights of neural networks. However, it is unknown how dedicated training of these networks can contribute to the successes of weight-space programming. We present a benchmark of LLM-driven weight-space programming on vision models, in which five LLMs edit convolutional networks and vision transformers on relabeling and error-correction tasks on Waterbirds, CIFAR-100, PACS and Entity-30. We report more than 2,600 such edit processes, whereby each edit is scored on held-out images for success, side effects and accuracy loss. We compare standard training with label smoothing, gradient-based editable training, and our Edit-Ready Training (ERT), a novel method that rehearses the direct, closed-form weight edits LLMs perform. We show that preparing networks for editing raises the success of LLM edits substantially where standard networks are hard to edit. ERT further achieves the highest mean score for GPT-6 Astra and Claude Opus 5.5 across all evaluated baselines. We will release our LLM testing framework alongside all tasks, editing trajectories and scores. Project page: this https URL

[CV-27] acInno: Teaching Vision-Language-Action Models the Contact Dynamics Beyond Vision and Action

链接: https://arxiv.org/abs/2611.26405
作者: Weile Li,Yunfan Lou,Muqing Shi,Zihao Zhang,Zhenbiao Huang,Boxian Yao,Xiyu Zhu,Zhibo Pang
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 4 figures, 4 tables. Project page: this https URL

点击查看摘要

Abstract:Vision-Language-Action (VLA) models have recently been equipped with tactile sensing to perceive visually subtle contact events in contact-rich manipulation. However, adding tactile input yields inconsistent gains, and some visuo-tactile VLA models rely little on tactile input, because much of the tactile change, such as the gel deformation during gripper closure, is predictable from visual observations and executed actions. To address this problem, we introduce TacInno, which isolates the contact dynamics that vision and executed actions cannot explain, termed tactile innovation, and uses tactile innovation only as a training target. Specifically, the TacInno tokenizer, conditioned on visual observations and executed actions, compresses the future tactile change into a discrete code, and the policy, TacInno-VLA, learns to predict the code, adding no module for the code at deployment. TacInno-VLA achieves average success rates of 87.5% across six simulation tasks and 70.5% across four real-world tasks, outperforming the strongest visuo-tactile baseline by 10.7 and 9.0 percentage points, respectively. Ablation studies validate the contributions of visual and action conditioning, multi-frame encoding, and auxiliary code prediction. Together, these results demonstrate the effectiveness of tactile innovation as a supervision signal for contact-rich manipulation.

[CV-28] ViDecide: Teaching Video Diffusion Models to Reason and Plan Step by Step

链接: https://arxiv.org/abs/2611.26389
作者: Kaiwen Zhang,Yiming Gao,Zeqi Xiao,Xintao Wang,Pengfei Wan,Kun Gai,Xingang Pan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL

点击查看摘要

Abstract:Current leading video generation models can synthesize highly realistic videos, while still frequently violating physical and logical constraints in multi-step reasoning and planning, even after training with large-scale reasoning data or existing diffusion RL methods. We attribute these failures to the conventional video training paradigm, which treats a solution as a fixed-length clip rather than a chain of decisions. We present ViDecide, a decision-centric framework that trains video diffusion models to predict state transitions causally and decide when to stop, with a decision-focused training schedule. We further propose Corrective Fine-tuning (CFT), which corrects failed decisions by mapping their noise to the model’s own successful continuations, achieving stable improvement without policy-gradient optimization. Across abstract games, planning tasks, and a new benchmark of multi-step robot planning, ViDecide achieves an average strict success rate of 83.8%, far above 23.3% for standard fine-tuning and 67.8% for the strongest prior method, and generalizes to larger puzzles and unseen layouts. Our framework therefore unlocks greater potential for video generation models as reliable reasoners and planners in the visual world. Project page: this https URL.

[CV-29] GateTrace: Where Do Crowd Crossings Go Missing Under Occlusion?

链接: https://arxiv.org/abs/2611.26388
作者: Ao He,Qianchen Yu,Vivian W. H. Wong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 23 pages, including supplementary material; 3 figures

点击查看摘要

Abstract:At large events, pedestrian tracking supports entrance and passageway counts, but totals hide which crossings were missed. In a stadium study, the system recorded about four in five clearly visible crossings but fewer than one in three mostly obscured crossings. The videos lacked trajectories annotated across frames, so we turned to MOT20, which provides pedestrian trajectories and visibility. We introduce GateTrace to locate missed crossings and assess changes to counting rules. It derives reference crossings from annotations under fixed rules, matches reports by person, direction, time and position, and traces misses through saved boxes and counting decisions. With a shared detector, ByteTrack, OC-SORT and BoT-SORT recover 38.9-46.7% of low-visibility reference crossings on development sequences, versus 98.1-99.0% at high visibility. The gap recurs on two further sequences under a protocol fixed in advance. Published outputs of MOT20-tuned BASE recover 63-78% at low visibility across both sequence groups. Losses stop at different stages across systems. Across four systems and both groups, 47-79% of low-visibility misses have no nearby output linked to the person or no possible crossing in the saved tracks. Removing later checks from BASE on development sequences leaves most misses unresolved and lowers precision against the fixed reference list from about 96% to 77%. Misses also remain among sampled reference crossings that two human annotators supported after checking the videos. GateTrace shows where crossings are lost and measures the recovery and precision costs of changing later checks. We will release the code and per-crossing records.

[CV-30] SEEC: A Large-Scale Dataset and Benchmark for Self-Supervised External-Eye Representation Learning

链接: https://arxiv.org/abs/2611.26386
作者: George R. Nahass,Michelle Fleishaker,Sanjay Ganesh,Sasha Hubschman,R. V. Paul Chan,Chad A. Purnell,Pete Setabutr,Ann Q. Tran,Sathya N. Ravi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 36 Pages, 10 Figures, 9 Tables

点击查看摘要

Abstract:External-eye photography captures clinically relevant periocular anatomy at low acquisition cost, yet no large-scale standardized resource exists for evaluating self-supervised external-eye representations. We introduce a 4.35-million-image Standardized External-Eye Corpus (SEEC) derived from fourteen heterogeneous public face datasets, together with a benchmark for external-eye representation learning. For evaluation, we provide held-out evaluation distributions, landmark-annotated anatomical probe sets, and clinical disease-classification experiments spanning ptosis, thyroid eye disease, and craniofacial syndromes. We evaluate representations at two complementary levels: (a) the representation level, by assessing embedding geometry across multiple data distributions, and (b) the functional level, by assessing anatomical recovery, out-of-distribution anatomical transfer, and disease classification. Our models include ResNet-101 and ViT-B encoders pretrained using InfoNCE, VICReg, and LeJEPA objectives and compared with DINOv2 and MAE. Our results indicate that representation-level and functional-level evaluations are dissociable: broader embedding-space utilization does not consistently correspond to anatomical transfer or disease-classification performance, and no single encoder is universally optimal. Domain-specific external-eye encoders outperform DINOv2 and MAE on cross-dataset anatomical transfer despite being pretrained at substantially smaller scale, while external-eye-pretrained ViT models match or outperform these general-purpose baselines on frozen disease-classification probes. We release benchmark code, evaluation scripts, and preprocessing pipelines to reproduce SEEC at this https URL

[CV-31] EgoCaptioner: Towards Comprehensive and Fine-Grained Egocentric Video Captioning

链接: https://arxiv.org/abs/2611.26380
作者: Yuqi Tang,Yang Shi,Yuran Wang,Tengfei Liu,Rundong Wang,Zongyang Qiu,Qixun Wang,Zhuoran Zhang,Xuanyu Zhu,Yifan Xu,Yiyan Ji,Shujie Li,Shixuan Liu,Haotian Wang,Yuanxing Zhang,Pengfei Wan,Chengwei Qin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages

点击查看摘要

Abstract:Egocentric video captioning requires fine-grained detail and accurate timing, yet current multimodal large language models (MLLMs) often miss brief actions, confuse their order, or misinterpret the wearer’s motion. Existing caption formats also offer limited support for combining detailed narration with precise timing. We present EgoCaptioner, which pairs a video-level overview of the setting, participants, and the wearer’s overall activity with timed steps whose narration adapts to varying perceptual demands and marks key moments at specific timestamps. We develop a coarse-to-fine annotation pipeline that establishes event structure before expanding step descriptions, yielding 50,000 training examples from eight egocentric video sources. Complementary rewards assess factual accuracy and coverage, measure step-level temporal alignment, and penalize fragmented or redundant narration, guiding caption refinement through Group Relative Policy Optimization (GRPO). To improve the reliability of factual supervision, we supplement caption references with pose-derived motion facts when sensor data are available. Across five egocentric benchmarks, EgoCaptioner achieves leading performance among evaluated open-source models in caption quality, caption-conditioned question answering, and temporal alignment. It also surpasses Gemini 3.1 Pro on PhysCapBench and all evaluated proprietary baselines on Minerva-Ego under the caption-to-QA protocol. Ablations support the complementary roles of factual and temporal rewards and show that coarse-to-fine annotation provides more effective supervision than single-pass annotation.

[CV-32] Fast Self-Supervised Generative CT Denoising via Few-Step Sampling

链接: https://arxiv.org/abs/2611.26370
作者: Xingyi Yao,Zhengtao Ma,Zhennong Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 30 pages, 4 figures, 2 tables. Xingyi Yao and Zhengtao Ma contributed equally to this work. Zhennong Chen is the corresponding author

点击查看摘要

Abstract:Low-dose and thin-slice computed tomography (CT) images suffer from elevated noise that obscures anatomical detail, while clean references for supervised denoising are often unavailable. Although Noise2Noise permits self-supervised training from noisy pairs, direct regression tends to oversmooth fine structure. Averaging multiple independently generated samples mitigates this blur, though repeating multi-step diffusion sampling is costly. We introduce Noise2Noise with Improved Mean Flow (N2N-iMF) to accelerate this process via few-step sampling. N2N-iMF learns to generate one noisy image conditioned on its paired counterpart; at inference, independent outputs are sampled and averaged to suppress noise. We evaluated image quality and sampling efficiency on simulated thin-slice brain and low-dose abdominal CT across four sampling methods and averaging counts from 1 to 20, and validated the approach on real-world photon-counting CT (PCCT) brain scans. With ten samples per conditioning input and three network evaluations per sample, N2N-iMF achieved a mean absolute error of 2.60 HU on brain CT and 14.02 HU on abdominal CT, outperforming 50-evaluation denoising diffusion implicit model (DDIM) sampling (2.91 and 14.53 HU). On the same GPU, per-output generation time dropped from 7.29 to 0.43 s. Across eight PCCT cases, N2N-iMF reached a mean gray-white matter contrast-to-noise ratio of 1.341, versus 0.916 for Noise2Noise and 1.176 for DDIM. N2N-iMF accelerates self-supervised generative CT denoising without distillation while keeping the number of averaged outputs adjustable at inference.

[CV-33] Contactless Estimation of Respiratory Frequency Using 3D-CNN on Thermal Images

链接: https://arxiv.org/abs/2611.26367
作者: Federica Gioia,Filippo Pura,Alberto Greco,Dario Piga,Arcangelo Merla,Marco Forgione
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Monitoring physiological parameters such as respiratory rate ( f_R ) is essential for diagnosing and managing various pathological conditions. Thermal imaging offers a promising contactless alternative to traditional methods, which often rely on partially invasive sensors or obtrusive wearable systems. However, existing approaches for f_R estimation from thermal signals typically require extensive pre-processing and manual or semi-automatic region-of-interest (ROI) tracking, limiting their practical applicability. This study proposes a deep learning-based method for estimating f_R directly from thermal videos, eliminating the need for complex pre-processing and ROI tracking. A 3D Convolutional Neural Network (3D-CNN) is developed to operate on raw thermal video data. To address challenges related to small datasets, the model is trained using data augmentation and transfer learning from synthetic datasets. Experimental results demonstrate that the proposed approach achieves a validation R^2 score of approximately 0.61 on both pre-processed and raw thermal videos. By simplifying the workflow, this method holds promise for enhancing the feasibility of thermal imaging in real-world applications, such as remote healthcare and driver monitoring in automotive applications.

[CV-34] OffBeat: Multicoset Sampling for 4D Capture Beyond the Camera Frame Rate

链接: https://arxiv.org/abs/2611.26347
作者: Chi-Wei Lee,An-Kai Chen,Cheng-Han Chiang,Yi-Chun Chang,Po-Chih Kuo
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: Project page: this https URL

点击查看摘要

Abstract:For 4D reconstruction, a synchronized multi-camera rig records all views at the same timestamps and misses the motion between them. Staggered triggers spread the same frames over more timestamps. However, when the cameras along each row fire one after another, as in 4DSloMo’s schedule on the N3DV rig, the row acts as one moving camera that confounds depth with motion. We introduce OffBeat, which models a staggered rig as a spatially distributed multicoset sampler: shifted copies of one sampling grid. From the calibration alone, OffBeat derives camera groups that fire together and their firing order. Since one group records each timestamp, a short-lived primitive can fit an error its cameras share, such as a color offset, and flicker in novel views. We show that a primitive’s opacity over time weights all groups equally, for every number of groups, only if it lasts at least one camera frame interval. On Spacetime Gaussians, a minimum lifetime approximates this bound, and a closed-form smoothed box of that length satisfies it exactly, with less flicker than a Gaussian in its place at nearly equal PSNR. In simulated staggered capture on N3DV, OffBeat outperforms the full 4DSloMo pipeline with video diffusion on every test frame, by 3.04 dB in mean dynamic-region PSNR. Between the first and last synchronized frame, it outperforms the best synchronized method of each scene, with or without frame interpolation, by 1.98 dB in mean dynamic-region PSNR, while synchronized capture is better at its own frames. On PanopticSports, in this frame range, OffBeat is comparable in PSNR to synchronized capture with interpolation. On every N3DV scene, replacing 4DSloMo’s schedule with ours raises the dynamic-region PSNR of Spacetime Gaussians, 4DSloMo without diffusion, and our reconstruction. Multi-camera rigs could follow faster motion at the same frame count by changing only when each camera fires.

[CV-35] Online Temporal Rotations for 2 Bit KV-Cache Quantization in Video Diffusion Transformers

链接: https://arxiv.org/abs/2611.26334
作者: Ryan Lucas,Petros Prastakos,Rahul Mazumder
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Video diffusion transformers reuse key–value (KV) caches across denoising steps, which quickly becomes a bottleneck for long video generation. We find that quantizing temporally distant cache vectors leads to a large range and high quantization error, but that nearby video tokens in the KV-cache share a persistent pattern per channel. These patterns also vary across videos, motivating rotations that are adapted to each video’s unique cache. We propose Temporal Rotation Quantization (TRQ), a method that rotates frames in the cache that are adjacent in time before quantization. For each block of temporally adjacent cache vectors, TRQ fits an orthogonal transform that maps the block’s per-channel means to a vector that is close to a constant across channels. To keep the quantization format lightweight, we parameterize this rotation as a tree of 2 \times 2 planar rotations, each stored with only two bits, followed by a fixed Hadamard transform shared across all blocks. The rotations are fit without gradients or changes to model weights, and we prove that a simple greedy fit is globally optimal for our ‘mean-flattening’ objective. The transforms are fit online for each new video, and we release fused Triton kernels to make this process efficient at inference. At 2-bit precision and matched compression, TRQ improves VBench subject consistency over the state-of-the-art method QVG from 90.38 to 92.02 on LongCat-Video-13B and from 86.33 to 87.61 on HY-WorldPlay-8B, as well as improving other metrics.

[CV-36] FirstCross: Convex Single-Photon 3D Reconstruction from Sparse Views

链接: https://arxiv.org/abs/2611.26321
作者: Chi-Wei Lee,Cheng-Han Chiang,Yi-Chun Chang,An-Kai Chen,Po-Chih Kuo
类目: Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注: Project page: this https URL

点击查看摘要

Abstract:Single-photon lidar records the arrival time of each returning photon, enabling 3D reconstruction from few views and scarce photons. However, leading neural methods such as Transientangelo require over an hour of non-convex per-scene optimization, and their accuracy degrades as photon flux decreases. We propose FirstCross, a network-free approach that instead solves a short sequence of convex problems on a voxel grid. For a solid object, the Poisson likelihood of a ray with given signal and background rates depends on the scene solely through its first crossing, the point where the ray first enters the object. Decomposing this likelihood yields two tractable terms. The first penalizes surfaces in front of detected photons and admits a tight convex closure, namely volume rendering with a running-maximum transmittance in place of NeRF’s product transmittance. The second accounts for photons that no surface explains and admits a convex upper bound once the object is assumed to be thick behind each detected surface point. Compared with Transientangelo under its own real-data protocol, FirstCross achieves lower depth error across every test scene, view count, and flux level. With three views and 100 times fewer photons than at the protocol’s lowest flux, our mean object depth error stays under half of Transientangelo’s with all photons. Given our detected surface points, the best classical fusion has 1.3 to 1.8 times our three-view error across flux levels. On real room-scale captures with 1% of their photons, TransientNeRF has 2.8 to 7.7 times our error. FirstCross lets few-view single-photon scanners cut exposure to 1% and reconstruct an object in about a minute on one GPU.

[CV-37] ResVision: Spectrum-Matched Resizing and Context-Preserving Inspection for Visual Reasoning

链接: https://arxiv.org/abs/2611.26318
作者: Guojian Zhan,Jiahao Xu,Yahui Liu,Yao Hu,Chuan Mu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Agentic visual reasoning improves vision-language models by selectively enlarging evidence without adding new scene content, but the benefit depends on whether fine structure remains accessible and whether local observations stay grounded in the full scene. We identify three recurring bottlenecks: high-frequency detail can be attenuated during visual-token formation; spatially distributed evidence can remain encoded yet difficult for the language model to retrieve; and isolated crops can weaken whole-detail correspondence through scale and positional mismatch. ResVision addresses these limitations with two complementary components. Spectrum-Matched Resizing (SMR) estimates image complexity and selects an adaptive global sampling scale, avoiding unnecessary enlargement on simple inputs while allocating more visual tokens to detail-rich images. Context-Preserving Inspection (CPI) returns each inspected crop together with the complete working view, preserving local detail and global context across tool turns. We evaluate Direct and Agentic variants on four frontier VLMs and 29 benchmarks spanning six visual-reasoning families. Complete ResVision (Agentic Base plus SMR plus CPI) is best for every model, improving by 3.36 to 4.94 points over Direct Base and 1.87 to 2.94 points over Agentic Base. These results show that adaptive visual sampling and context-preserving inspection address distinct, complementary constraints in multimodal reasoning.

[CV-38] C-LiTE: A Lightweight Foundation Model for Cellular feature learning in Histopathology

链接: https://arxiv.org/abs/2611.26316
作者: Esha Sadia Nasir,Jiaqi Lv,Gozde N. Gunesli,Shan E Ahmed Raza
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Characterising individual cells is fundamental to computational pathology, as cellular morphology informs tasks such as tumour grading, mitotic assessment and immune profiling. However, most pathology foundation models are trained on tissue patches and capture cellular features only implicitly, while existing cell-level models are either restricted to specific cell populations or rely on computationally expensive architectures with hundreds of millions of parameters. Applying ViTs directly to small cellular crops also presents an architectural challenge: upsampling increases computational cost without adding morphological information and may introduce interpolation artefacts. To address these limitations, we introduce C-LiTE, a lightweight foundation model for cellular representation learning in HE-stained histopathology, with only 22 million parameters. C-LiTE employs a native-resolution tokenisation strategy tailored to cellular inputs, processing 64 x 64 px cell-centred crops using a 4 x 4 patch embedding to produce a 16 x 16 token grid without image upsampling. To transfer visual knowledge from large-scale pre-training, we adapt the patch embedding weights of a pre-trained ViT-5-S using a pseudoinverse-based transformation, enabling weight initialisation at the finer patch resolution. We further assemble a pre-training dataset of 60 million cell-centred crops from TCGA, balanced across 19 cancer types and 16 tissue types. To our knowledge, this represents the largest public dataset for self-supervised pre-training on HE images at the cellular level. We evaluate C-LiTE on seven public datasets spanning mitotic figure recognition, nucleus and cell-type classification, and immune cell phenotyping. Under frozen-feature evaluation, C-LiTE achieves a mean macro-F1 of 0.670, approaching CytoFormer (0.682) with approximately 30x fewer parameters.

[CV-39] SKILL.PNG: Learning Agent Skills in Pixel Space

链接: https://arxiv.org/abs/2611.26296
作者: Maximilian Seeth,Chengzhi Martin Hu,Marie Kempkes,Paul Röttger,Leo Schwinn
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:AI agents are commonly specialized for new tasks and domains using skills, i.e. natural language instructions placed in their context. However, the discrete nature of textual skills prevents continuous optimization. Gradient-based approaches such as finetuning, on the other hand, are typically model-specific, preventing skill transfer across different agents. Since many agents today are built on vision-language models (VLMs), we can address these limitations by introducing this http URL: a method for obtaining skill-like image inputs whose pixels are optimized with gradient descent through the vision encoder of a frozen VLM. Image skills are compatible with any VLM, and a single image can be optimized jointly for an ensemble of models. We show that this http URL significantly outperforms state-of-the-art textual skills optimized using frontier LLMs in 9 of 16 settings (across 4 VLMs and 4 benchmarks) while requiring only 64 to 256 image tokens instead of roughly 900 to 2000 text tokens at inference time. Hybrid skills, i.e. text and image skills jointly optimized, further improve agent performance. Moreover, an image optimized with an ensemble of VLMs can successfully transfer across the models in the ensemble. this http URL thus presents the first agent skill that can be optimized with gradient descent while retaining transfer benefits of text skills. Our code is available at: this https URL

[CV-40] racking Objects Through Assembly and Decomposition

链接: https://arxiv.org/abs/2611.26269
作者: Cheryl Lam,Yihong Sun,Bharath Hariharan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Tracking objects through transformations is essential for understanding real-world dynamics. Recent work has extended traditional object tracking to detect state changes. However, existing methods require an initial object prompt and detect only its decompositions, missing assembly events, which begin with multiple objects rather than one. We therefore introduce a prompt-free formulation of this task, where the goal is to recover all object tracks and detect both decomposition and assembly events given only a video. Our key insight is that assembly is decomposition in reverse. Building on this insight, we present Atlas, a zero-shot system that constructs and jointly reasons over forward and reverse spatiotemporal partitions to recover object tracks and their transformations. We further introduce OmniST, a benchmark with dense annotations of objects and both transformation types. On OmniST, Atlas recovers more than twice as many state changes as prior methods while remaining competitive on existing benchmarks that only evaluate decomposition.

[CV-41] Digitization of Analogue Gauges using Computer Vision

链接: https://arxiv.org/abs/2611.26262
作者: Erico Kayumba Mutonkole,Olasupo Ajayi,Omowunmi Isafiade
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 9 pages, 12 figures. Presented at SATNAC 2025

点击查看摘要

Abstract:Analogue gauges remain prevalent in manufacturing industries for measuring physical quantities, yet manual monitoring by operators introduces human error and operational inefficiency. This research developed and evaluated an automated gauge reading system using computer vision and deep learning techniques. Three hybrid approaches were implemented and compared: OpenCV with Scikit-Image (traditional computer vision), OpenCV with ResNet18, and OpenCV with VGG19 (deep learning approaches). A comprehensive dataset of 1,848 pressure gauge images (0-10 BAR range) was curated through web scraping, synthetic generation, and augmentation techniques. The experimental results demonstrate that ResNet18 significantly outperforms the alternatives, achieving an accuracy of 97.3%, compared to VGG19’s 64.4% and Scikit-Image’s 63.8%. The study validates the effectiveness of residual learning architectures for industrial gauge digitization, providing a robust foundation for Industry 5.0 automation applications.

[CV-42] Prism: Hierarchical Text Detection with Minimalist Design

链接: https://arxiv.org/abs/2611.26254
作者: Yuchen Su,Xuesheng Wang,Zhilong Ji,Jinfeng Bai,Shijian Lu,Zhineng Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Hierarchical text detection aims to localize words, text lines, and paragraphs while parsing their structural relationships. It provides essential structural cues for visual text understanding. To maximize accuracy, existing methods typically rely on complex multi-scale feature fusion necks and heavy-weight heads, resulting in bloated architectures and low inference efficiency. In this paper, we argue that a sufficiently pre-trained backbone can effectively replace these sophisticated external modules, while simple geometric inclusion suffices for hierarchical text structural parsing. Thus, we propose Prism, a minimalist framework for hierarchical text detection. Prism performs multi-granularity detection by attaching parallel lightweight heads to a single-scale backbone feature map, covering words, text lines, and paragraphs, and reconstructs the structural hierarchy via an efficient geometric inclusion algorithm. In addition, we systematically compare representative lightweight head archetypes and show that pixel-level regression offers the best accuracy-efficiency trade-off in this setting. Experiments on multiple public benchmarks show that our significantly simplified architecture achieves state-of-the-art (SOTA) performance with remarkable efficiency. Specifically, it runs 1.7\times faster than LRANet on CTW1500 while achieving higher accuracy, and surpasses the current SOTA method, OmniParser V2, by over 6 PQ points on the hierarchical HierText dataset. Code will be released soon.

[CV-43] Beyond Covers and Titles: Evidence-First Ranking for Robust Cold-Start Recommendation

链接: https://arxiv.org/abs/2611.26249
作者: Fuyi Wang,Yuxin Cao,Yifan Liao,Renyang Liu,Lin Li,Wei Song
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Short-video platforms rank new uploads before anyone has watched them, using content towers that read a video’s cover, title and frames. Existing attacks on content-aware recommenders perturb pixels or features and ignore platform re-encoding, and existing defenses train against norm-bounded perturbations that neither match real edits nor preserve recommendation quality. We show with REAP, a realizable attack through a simulated platform channel (H.264 transcoding, rescaling, cropping and JPEG), that packaging edits raise a cold video by 32.0 rank-percentile points in unrelated audiences. We then observe that cold-start ranking does not need the packaging at all: a tower that sees only the video body ranks cold items best. Building on this, we propose Evidence-First Ranking (EFR), which scores cold videos from their body, admits packaging only through a support-gated residual bounded by \gamma , and merges cold and warm rankings by rank percentile, yielding an exact certificate against any packaging edit. On MicroLens-100K, EFR cuts the attack gain by 81% and raises cold-start Recall@100 from 0.1133 to 0.170, while keeping the merged NDCG@10 within 4% of the undefended system.

[CV-44] Attention Drift Regularization with Adaptive Initialization for Domain-Robust Visual Recognition WACV2027

链接: https://arxiv.org/abs/2611.26234
作者: Shivangi Rai,Vinod K Kurmi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted in WACV 2027

点击查看摘要

Abstract:Source-free domain adaptation (SFDA) adapts a pretrained source model to unlabeled target domains without source data access, yet two critical failure modes remain unaddressed. First, attention drift: source-learned attention maps shift toward spurious or background features under distribution shift, destabilizing feature representations. Second, domain-gap-agnostic initialization: applying a single fixed source model uniformly across all target domains ignores domain gap severity, causing premature overfitting to noisy pseudo-labels under large shifts. We propose DGA-CAFC to address both failures. The Counterfactual Attention-based Feature Consistency (CAFC) module enforces feature stability by minimizing the discrepancy between representations computed under learned attention and those under counterfactual-inspired normalized perturbations, thereby promoting domain-robust features independent of spurious attention shifts. The Domain-Gap Aware Source Model Selection (DGASS) module estimates domain gap severity via target prediction entropy and selects the source model variant whose confidence level best matches the estimated gap, reducing negative transfer and stabilizing pseudo-label quality. Crucially, CAFC transfers beyond SFDA: applied as a lightweight regularizer to standard domain generalization objectives without target data. Extensive experiments across SFDA and DG benchmarks demonstrate that our method consistently improves over recent baselines, demonstrating the effectiveness and generality of the proposed framework. Project page is available at this https URL

[CV-45] MixReact: Physics-Based Reactive Humanoid Control from Multimodal Cues via a Discrete-Continuous Mixture of Transformers

链接: https://arxiv.org/abs/2611.26218
作者: Bin Li,Xing Zhu,Jingya Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Humanoid agents in everyday environments are expected to react to diverse cues as humans do, with semantically appropriate and physically executable behaviors. Existing cue-driven reaction generation methods support only a narrow range of cue modalities and are limited to purely kinematic generation, leaving the resulting reactions susceptible to physical artifacts. To this end, we present MixReact, the first unified framework for physics-based reactive humanoid control from text, audio, and video cues, integrating a Discrete–Continuous Mixture of Transformers (DisCo-MoT) for state prediction with a History-aware Action Transformer (HAT) for physical execution. DisCo-MoT couples a discrete masked-modeling branch with a continuous autoregressive diffusion branch via one-way mixture attention within a MoT architecture. Guidance from discrete context mitigates collapse into averaged behaviors in one-to-many reaction generation, while direct prediction in continuous state space bypasses the decoding bottleneck of discrete representations. HAT then maps the predicted next state and observed execution history to executable actions, enabling robust closed-loop control in physics-based simulation. Experiments across all three modalities demonstrate consistent improvements over the baselines in fidelity, diversity, and physical consistency, while ablations validate the complementary contributions of the components. Our code will be released to facilitate future research. Project page: this https URL

[CV-46] SpecWM: Turning Verification into Refinement for Streaming World Models

链接: https://arxiv.org/abs/2611.26211
作者: Jiaming Li,Lingzhen Li,Xiaze Yang,Chenyu Zhu,Kaiqi Liu,Xiaotong Zhao,Songlian Li,Toyota Li,Alan Zhao,Haobin Zhong,Yue Ma,Zhiyuan Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Recently, video world models have advanced interactive generation, with large models offering higher quality and smaller models enabling faster inference. Existing speculative methods combine their strengths through fast draft generation and selective target regeneration guided by rewards on decoded frames. However, verification covers only the external scorer’s evaluation dimensions and adds decoding and scoring costs. Moreover, these scores provide no correction, requiring costly target regeneration for rejected drafts. To address these limitations, we propose Speculative World Model (SpecWM), a unified, training-free inference framework for streaming world models. It verifies draft chunks using the target’s latent velocity field. Verification also yields a reusable correction; two thresholds select Accept, Refine, or Reject. Specifically, one-step refinement reuses it to improve perceptual quality, while three-step refinement strengthens camera following and trajectory consistency. These findings motivate transition-aware refinement. On WBench Navigation, SpecWM improves visual quality at 1.63 \times speedup over the target and raises NavScore from the draft’s 75.99 to 85.63, achieving the best quality-speed-control trade-off. Additionally, our LongSpecBench shows sustained quality and control over long, multi-turn rollouts, while experiments on Hy-Nomas validate SpecWM across backbones and denoising budgets.

[CV-47] Learn Freeze and Transfer an Activation Function Across Architectures

链接: https://arxiv.org/abs/2611.26201
作者: Arun Prakash A
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 5 figures, 10 tables. Code: this https URL , Interactive figures: this https URL

点击查看摘要

Abstract:Transfer learning typically reuses knowledge acquired by a source network through its learned weights. We investigate whether a learned activation function can also be transferred and remain useful in networks trained from scratch. We study this complementary form of transfer through a learn-freeze-transfer protocol, in which the activation is optimized jointly with a source network in a single training run, and is then frozen and reused in target networks trained from scratch. A pointwise activation with globally shared parameters is independent of layer dimensions and can therefore be incorporated into networks of different architectures trained from scratch. We instantiate this protocol with FAct, a five-parameter activation represented by a two-harmonic truncated Fourier series, initialized from a Fourier approximation of GELU and learned in a Vision Transformer (ViT) on ImageNet-1K. When reused in a ViT of the same architecture trained from scratch on ImageNet-1K, the frozen activation (Frozen FAct) improves top-1 accuracy by 1.46 percentage points over GELU. Across four smaller target architectures (standard ViT, Swin Transformer, linear-attention ViT, and MLP-Mixer) and four image-classification datasets, it achieves the highest mean accuracy among eleven activation functions in 15 of 16 settings under a common training configuration, with improvements of up to 1.71 percentage points over the best conventional activation. When the learning rate and weight decay are tuned separately for each activation under the same search budget, Frozen FAct still ranks first in 11 of the 16 settings, second in four, and fourth in one. These results provide empirical evidence that a learned activation function can retain its usefulness when transferred across vision architectures, independently of the weights with which it was learned.

[CV-48] A version of the Cloude/van Zyl decomposition for full/quad pol synthetic aperture radar data

链接: https://arxiv.org/abs/2611.26195
作者: Allan Aasbjerg Nielsen,Paul Jacques Connetable,Henning Skriver,Knut Conradsen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 5 figures

点击查看摘要

Abstract:Based on the Cloude/van Zyl decomposition of reflection symmetric polarimetric SAR data we present a version for full/quad polarisation data. We demonstrate both versions on X-band F-SAR data over Vejers, Denmark. Differences between the original and our version of the decomposition occur especially for very bright man-made targets. These differences come with lower entropy for our version giving an improved focus on a more pronounced dominant scattering mechanism for these objects. Interpretation of the scattering mechanism related to the full/quad pol version of the decomposition is facilitated by means of the Cloude/Pottier \bar\alpha angle.

[CV-49] Backbone Matters: Benchmarking Encoder-Decoder Networks for Semantic Segmentation of Driving Scenes on BDD100K

链接: https://arxiv.org/abs/2611.26191
作者: Ehsan Rassekh
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 6 pages, 3 figures, 4 tables

点击查看摘要

Abstract:Semantic segmentation gives an autonomous vehicle a dense, pixel-level description of its surroundings and is a core part of scene understanding. Most segmentation networks consist of an encoder (the backbone) and a decoder, yet the backbone is often chosen without a systematic comparison. This work studies this choice on the BDD100K driving dataset. Four widely used decoders (U-Net, FPN, LinkNet and PSPNet) are combined with 13 backbones from six families (ResNet, VGG, DenseNet, Inception, MobileNet and EfficientNet), and all 52 models are trained with one common protocol. The best model, an FPN decoder with an EfficientNet-B4 backbone, reaches 62.6% mean intersection over union (mIoU), 89.8% accuracy and a loss of 0.092. Only two of the 52 models, both FPN with EfficientNet, exceed the 56.9% mIoU reported with the BDD100K benchmark. The effect of the backbone depends strongly on the decoder: the same EfficientNet-B4 gives 42.6% mIoU with PSPNet but 62.6% with FPN, and the gap between the best and the worst backbone ranges from 12.0 points (U-Net) to 42.5 points (FPN). The results and the scope of this study are discussed.

[CV-50] Invertible Transport Between Sample-Only Distributions via Projected Characteristic Functions

链接: https://arxiv.org/abs/2611.26187
作者: Juan Gutiérrez,Victor Gutiérrez-García,José Luis Blanco-Murillo
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Many useful latent distributions can be sampled but do not have a tractable density. Such distributions may be represented by empirical collections of real-world observations or generated by scientific simulators and black-box generative models. These sample-only priors cannot be used directly in likelihood-based normalizing flows, which require evaluating the prior density. We propose Projected Characteristic-Function matching (PCF), a density-free framework for learning an invertible map between a data distribution and a prior specified only through samples. Our method aligns distributions using projected empirical characteristic functions, eliminating the need for density estimation, adversarial training, or paired observations. To address the non-uniqueness of the transport map, we introduce an optional transport regularizer that selects lower-displacement transformations while preserving endpoint alignment. We benchmark PCF on controlled transports, sample-efficiency studies, and high-dimensional alignment tasks. PCF matches the accuracy of kernel MMD on the controlled benchmark and on raw-pixel MNIST–USPS at a cost linear in the batch size, and it is more sample-efficient, achieving lower endpoint error than MMD in 13 of 15 paired runs across library sizes. By avoiding pairwise kernel evaluation, it reduces discrepancy memory by 29.8% and trains at batch sizes where kernel MMD exceeds GPU memory. These results show that PCF provides accurate and efficient invertible transport, particularly in sample- and memory-limited settings. Subjects: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2611.26187 [cs.LG] (or arXiv:2611.26187v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2611.26187 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[CV-51] RegionAgent : Two-Stage Preference Optimization for Regional Image Restoration

链接: https://arxiv.org/abs/2611.26179
作者: Weiyu Zhou,Jiacong Tang,Tao Hu,Axi Niu,Jinqiu Sun,Peng Wu,Wei Dong,Qingsen Yan
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Images under non-uniform coupled degradations require diverse restoration capabilities for different degraded regions. Existing methods employ globally uniform operations, potentially overlooking regional differences and causing incomplete restoration or even damage to image content. Moreover, a degraded region may require successive restoration actions, and earlier actions may alter residual degradation characteristics or introduce artifacts, potentially hindering subsequent restoration. To address these challenges, we propose RegionAgent, an adaptive restoration framework that trains a Vision-Language Model (VLM) through two-stage preference learning to generate a region-specific restoration action at each iteration. Specifically, we first train the VLM to select an action suited to the current region through step-level Direct Preference Optimization (DPO), using preferences derived from immediate restoration feedback to supervise only the differing component in each action pair. To account for the effects of current actions on subsequent restoration, we then apply trajectory-level DPO, using preferences derived from complete trajectory scores to supervise the sole differing action in each trajectory pair. Furthermore, as conventional image-pair datasets do not directly provide preferences over regional restoration decisions, we construct a regional restoration preference dataset comprising 10K action pairs and 20K trajectory pairs for two-stage VLM training. Experiments across multiple datasets demonstrate strong restoration performance under both mixed and single degradations. Ablation studies further validate the effectiveness of two-stage preference learning in improving restoration quality.

[CV-52] FlexVSR: Continuous-Scale Super-Resolution for Streaming Video Generation

链接: https://arxiv.org/abs/2611.26178
作者: Ji Wang,Haonan An,Jiasheng Zhang,Renlong Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 23 pages, 5 figures

点击查看摘要

Abstract:Real-time streaming video generation has gained rapid momentum in live content creation and interactive media. Deploying such systems in practice requires flexible scaling factors to accommodate diverse source resolutions, target displays, and computational budgets. However, extending fixed-scale generative video super-resolution (VSR) to varying target scales poses two challenges: downsampling from a larger fixed scale incurs redundant cost, while re-optimizing the diffusion transformer (DiT) is costly and alters the shared generative prior. To address both, we propose FlexVSR, a decoupled framework for continuous-scale streaming generative VSR with supervision at only a few scales. Our key insight is that naively resampling fixed-scale latents onto a new target grid induces scale-dependent input mismatch, which can be effectively addressed through parameter-efficient, scale-aware adaptation of the latent interface without updating the DiT. To this end, FlexVSR introduces ReLSA (Refined Latent Scale Adaptation), a lightweight latent adapter that constructs and refines target-grid latents before they enter the DiT, enabling a shared frozen VSR prior to generalize across continuous scales with only 0.274M trainable parameters and an adaptation cost of 7.3 H100 GPU-hours, while introducing no additional temporal dependencies and preserving the native causal streaming protocol. Extensive experiments demonstrate competitive or superior reconstruction and temporal quality, lower latency than most native-scale streaming baselines and offline methods, and effective generalization to unseen continuous scales, filling the continuous-scale gap in real-time streaming generative VSR.

[CV-53] SpatialScout-V2: Can VLM Agents Actively Explore Real-World 3D Scenes and Localize Objects?

链接: https://arxiv.org/abs/2611.26177
作者: Zhaoxi Zhang,Hao Liang,Linxuan Wu,Yiming Xu,Yifu Yuan,Kun Liang,Chenming Tang,Meiyi Qiang,Wentao Zhang,Weikang Li,Yunfang Wu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Spatial intelligence requires agents to acquire visual evidence and integrate partial observations into coherent three-dimensional representations. We introduce SpatialScout-V2, which extends SpatialScout’s active mapping task from simulation to real residential scenes through three advances: real visual observations, preservation of original scene content, and interactive exploration of recorded views. Built from 100 homes in Realsee3D, the benchmark covers 3,171 target objects across 29 categories. Given a fixed target inventory, agents autonomously select recorded viewpoints and viewing directions, integrate calibrated RGB observations, and iteratively estimate and revise object centers within a fixed interaction budget. We evaluate the localization accuracy and completeness of object-center maps using mean IoU, IoU-threshold hit rates, and target coverage. Evaluating seven frontier vision-language models (VLMs), we find that six achieve an all-target mean IoU below 20%. Even the best-performing model, Claude Opus 5.5, reaches only 22.52%, with just 18.36% of targets exceeding an IoU of 0.5. Further analysis shows that broad submission coverage can coexist with limited localization accuracy, while target omissions further reduce overall scores. These results highlight substantial challenges in constructing accurate and complete object-center maps through active exploration of real scenes.

[CV-54] StreamEgo: Streaming Exocentric-to-Egocentric Video Generation

链接: https://arxiv.org/abs/2611.26165
作者: Tianyi Xu,Boyang Lyu,Yunkai Tang,Kairan Ding,Jiahang Li,Weiyu Guo,Yandong Guo,Shuchen Weng,Boxin Shi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Exocentric-to-egocentric (Exo2Ego) video generation transforms third-person recordings into first-person viewpoints, unlocking immersive AR/VR experiences and scalable demonstration synthesis for embodied robotics. Extending it to a streaming setting is expected to further enable live applications. However, existing methods suffer from degraded manipulation regions, overfitting and confinement to fixed-length clips with prohibitive latency. To address these limitations, we present StreamEgo, a real-time framework for unbounded, generalizable Exo2Ego video generation. To resolve intricate manipulation dynamics while supporting the streaming pipeline, a streaming dual-branch prior provides pixel-aligned guidance for translating appearance and skeletal kinematics. We also introduce a region-aware loss curriculum to improve generalization and stabilize multi-objective diffusion training. Furthermore, we distill the model into a chunk-wise autoregressive model, augmented with a region-aware corruption mechanism and a first-chunk curriculum, enabling high-quality, generalizable real-time streaming Exo2Ego generation at 24.3 FPS on a single H100 GPU.

[CV-55] MultiBinder: Preventing Multi-Subject Identity Leakage in Personalized Video Generation

链接: https://arxiv.org/abs/2611.26164
作者: Wei Li,Yiheng Zhang,Zhaofan Qiu,Ting Yao,Xiaoyan Sun,Tao Mei
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: ACM Multimedia 2026

点击查看摘要

Abstract:Personalized video generation faces a significant scalability challenge: as the number of custom subjects increases, existing methods suffer from severe identity leakage and semantic entanglement, resulting in videos with blended attributes and incoherent narratives. We identify that this stems from a lack of explicit binding between textual entity phrases and their corresponding visual patterns prior to fusion in diffusion models. To address this, we propose MultiBinder, a novel framework built on a bind-then-narrate'' paradigm. Our approach introduces a two-stage conditioning process before video diffusion. First, Cross-modal Entity Binding devises an entity-aware Q-Former to extract identity-centric features from each subject's reference image and tightly align them with its unique entity phrase, yielding disentangled identity tokens.‘’ Second, a Multimodal Large Language Model (MLLM) acts as a narrative director'' in the Narrative-Aware Subject Arrangement stage, capturing the action of each subject from the text prompt and synthesizes contextualized subject features that encapsulate who is doing what’’ and preserve role-specific distinctions. These features are then injected into a video diffusion model via dedicated conditioning pathways, ensuring precise attribute allocation. Extensive experiments on both general and newly curated multi-subject benchmarks show that MultiBinder significantly outperforms existing methods in multi-subject identity fidelity.

[CV-56] When Foundation Models Disagree: Representational Heterogeneity for Calibrated Uncertainty in Whole-Slide Image Classification

链接: https://arxiv.org/abs/2611.26159
作者: Yekai Shen,Yuqian Wang,Zishun Liao,Rongsheng Wang,Wenting Chen,S. Kevin Zhou,Qingsong Yao
类目: Computer Vision and Pattern Recognition (cs.CV); Quantitative Methods (q-bio.QM)
备注: accepted by IEEE BIBM 2026

点击查看摘要

Abstract:Pathology foundation models supply strong representations for whole-slide image classification, yet the confidence attached to their predictions remains hard to trust. Prevailing uncertainty estimators act within a single feature space and thereby overlook the uncertainty that the choice of foundation-model representation itself induces. This work introduces CALIFUSE, a calibration-oriented framework that fuses several pathology foundation models for the specific purpose of uncertainty estimation. Departing from accuracy-driven fusion, CALIFUSE reads representational heterogeneity across encoders as an explicit confidence cue. Slide-level decision consensus is coupled with chunk-level evidence agreement quantified through Jensen-Shannon divergence, so that both prediction stability and fine-grained evidence consistency inform the estimate. Foundation-model encoders stay frozen. A task-specific nnMIL classifier is still trained for each FM; CALIFUSE then estimates confidence at inference without training an extra fusion module. Evaluated on seven WSI classification tasks under several foundation-model combinations, the framework sharpens calibration, misclassification detection, and selective prediction relative to both the strongest single-encoder baseline and naive ensemble averaging, all while preserving competitive balanced accuracy.

[CV-57] A Practical Proprioceptive Data Collection Utility for the Unitree GO1

链接: https://arxiv.org/abs/2611.26146
作者: Jack Anders Smitterberg,Tan Chen
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:The Unitree Go1 is a widely used platform for legged robotics research, yet collecting proprioceptive data from it is unnecessarily difficult: the Unitree SDK is sparsely documented, and ROS1 (Robot Operating System), the Go1’s native middleware, is fully deprecated. We present an open-source graphical utility that enables researchers to reliably record proprioceptive data from the Go1, including joint states, IMU (Inertial Measurement Unit) readings, foot force sensor data, and gyroscope measurements, without requiring ROS expertise. The tool provides a simple GUI (Graphical User Interface) for starting, stopping, and labeling recordings in real time. The contribution of this paper is the collection utility. We characterize the acquisition quality of 51 recordings collected with it, including the achieved sampling rate and repeated samples. As a secondary, deliberately modest check that the data is useful, we train a terrain classifier on it. The results are weak: a Temporal Convolutional Network reaches about 38% accuracy on held-out recordings across ten terrain classes (10% would be chance), and a window-level split that ignores recording boundaries inflates the same measurement to about 99%. We report this as a demonstration that the logged signals carry terrain information, not as a competitive classifier. The utility is designed to lower the barrier to collecting Go1 proprioceptive data without ROS.

[CV-58] CryoGaussian: Heterogeneous Cryo-EM Reconstruction Made Explicit with Gaussian Splatting

链接: https://arxiv.org/abs/2611.26128
作者: Ziwen Wang,Jiajun Fan,Chaoran Cheng,Ruihan Guo,Yanru Qu,Juexiao Zhou,Ge Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 59 pages, 25 figures, 15 tables

点击查看摘要

Abstract:Single-particle cryo-electron microscopy (cryo-EM) determines macromolecular structures from noisy particle images. However, rare, transient compositional states and subtle, continuous conformational variation are difficult to resolve. We present CryoGaussian, a Gaussian splatting method for heterogeneous cryo-EM reconstruction that fits an explicit Gaussian density representation directly to the particle images through a differentiable rasterizer rendering orthographic line integrals modulated by the contrast transfer function. The Gaussians are initialized randomly, with their number derived from the estimated spatial support of the molecular density and their minimum size bounded by the Nyquist frequency of the particle images. To capture the compositional and conformational variation across particle images, we learn particle-image and per-Gaussian latents that jointly modulate the base position, amplitude, and scale of each Gaussian. Each structural state is then reconstructed by fitting the Gaussians conditioned on its inferred latent. On CryoBench, CryoGaussian achieves state-of-the-art reconstruction accuracy for both compositional and conformational heterogeneity. On experimental data, it resolves a rare compositional state of the 50S ribosomal subunit (class C4, 1.3% of particles) with rRNA helix 68 docked in the absence of the central protuberance, and continuous conformational motion of the SF3b and helicase subcomplexes of the pre-catalytic spliceosome. We further optimize the rasterizer for additive cryo-EM projection, reducing training time by 12.8-28.0 \times .

[CV-59] Better Retrieval Not Better Binding: Pretraining Exposure in Egocentric Video Retrieval

链接: https://arxiv.org/abs/2611.26120
作者: Yuqi Guo,Michael Wray
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 17 pages, 4 figures, 3 tables

点击查看摘要

Abstract:Video-language models (VLMs) achieve strong performance on video-text retrieval and are increasingly used for egocentric video understanding. Yet, despite gains in zero-shot retrieval, they have shown limitations in compositional understanding, and it remains unclear whether they understand compositional relationships between verbs and nouns. In this work, we introduce Ego-ExBind, a compositional diagnostic protocol for analyzing pretraining exposure and verb-noun binding in egocentric video-text retrieval. Ego-ExBind consists of an Exposure Ledger built from the EgoClip pretraining corpus used by EgoVLPv2, which defines Seen Composition, Unseen Composition, and Unseen Atom splits for EPIC-KITCHENS-100 multi-instance retrieval (MIR). We further probe noun and verb binding channels through controlled noun-swap and verb-swap negatives. We find that zero-shot retrieval is positively associated with pair exposure within seen compositions, even after controlling for the marginal frequencies of verbs and nouns. However, higher pair exposure is associated with larger noun-swap margins and slightly smaller verb-swap margins. This asymmetry motivates us to test whether exposure reweighting can improve retrieval with little change in the measured swap margins. To isolate the effect of exposure input, we keep the trained checkpoint fixed and replace only the exposure input at inference. We demonstrate that this exposure reweighting can improve video-text retrieval without stronger verb-noun compositional understanding, and that the gain depends on the correct sample-level exposure correspondence. By separating exposure-driven retrieval gains from verb-noun binding, Ego-ExBind provides a diagnostic framework for evaluating whether future improvements in egocentric MIR reflect compositional understanding rather than pretraining familiarity.

[CV-60] Learning Reconstruction Errors as Spatiotemporal Evidence for AI-Generated Video Detection

链接: https://arxiv.org/abs/2611.26116
作者: Jiahua Jiang,Jinglai Li,Jianru Zhang,Jin Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages

点击查看摘要

Abstract:The proliferation of AI-generated videos has raised growing concerns about misinformation, fraud, and potential societal risks, creating an urgent need for detection. Reconstruction-error-based methods have shown strong empirical performance for AI-generated image detection by exploiting discrepancies between an input and its reconstructed version. Their extension to video detection, however, faces two challenges: (i) a fixed reconstructor is not optimized for the detection objective and may yield weak discriminative signals; and (ii) an additional raw-video feature stream is often introduced to model temporal dependencies, increasing architectural overhead. To address these challenges, we develop a detector that learns and exploits reconstruction errors as a unified source of spatiotemporal evidence. First, we jointly optimize a reconstructor and the downstream classifier to produce reconstruction errors that are more informative for detecting AI-generated videos. Second, we show that consecutive reconstruction errors alone carry temporal evidence, and thus a 3D convolution over them captures spatiotemporal evidence without a raw-video stream. Extensive experiments on over 30 existing AI-video generators demonstrate competitive performance against baselines.

[CV-61] HiUni: Harmonizing Understanding and Generation in Shared Visual Space

链接: https://arxiv.org/abs/2611.26107
作者: Jiahao Hu,Jingdong Du,Yichen Zhang,Yi Song,Zhenyu Zhang,Jiaxin Niu,Ruihang Li,Xiangzhao Hao,Yiqian Zhang,Yong Huang,Zefeng Zhang,Zheyuan Gu,Shuohuan Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Unified multimodal models aim to support visual understanding and generation within a single framework. However, prevailing approaches often separate their visual representations or backbone computation to accommodate the different requirements of these two tasks, which limits the degree of unification. In this paper, we present HiUni, which revisits this design choice and explores whether a pretrained vision language model can be directly extended into a unified model with extensive parameter sharing. Specifically, HiUni reuses the same language-aligned ViT and dense Transformer backbone for both capabilities, while confining task-specific computation to the output side. Yet such a design introduces two key challenges. Generation in pixel space is difficult to optimize, and semantic abstraction in pretrained ViT discards fine spatial details required for synthesis. To address these challenges, HiUni introduces a \emphPixel Diffusion Head tailored to direct pixel-space optimization and a lightweight \emphHiGate pathway that provides fine-grained spatial details near the output stage. Extensive experiments show that HiUni achieves a similar average understanding score to the pretrained VLM while attaining competitive performance on image generation and editing. Controlled comparisons and analyses further demonstrate that a single dense backbone can effectively support both capabilities without task-specific backbone computation, highlighting a simpler path toward unified multimodal modeling.

[CV-62] UniVEP: Unifying Velocity and Endpoint Prediction for Robot Manipulation

链接: https://arxiv.org/abs/2611.26103
作者: Zhengtu Liang,Qiwei Liang,Tao Lin,Shihang Weng,Boyang Cai,Xiangyu Chen,Yikeng Chen,Jiaming Liang,Bohan Tan,Yu Yan,Renjing Xu
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: 43 pages, 8 figures

点击查看摘要

Abstract:Flow-matching robot policies usually use one prediction target throughout action generation, even though recovering actions from noise and polishing nearly clean actions bring different challenges. We systematically compare velocity prediction (V), endpoint prediction (X), and velocity-weighted endpoint prediction (X–V) across multiple robot policies, benchmarks, and inference budgets. On StarVLA and LiLa-WAM, X–V works best for few-step generation, while V gains more from iterative refinement. We further uncover their complementary strengths across noise levels: clean actions have concentrated covariance spectra, endpoint predictions stay more stable under high noise, and velocity predictions offer more accurate corrections near clean actions. Inspired by these findings, we introduce UniVEP, which learns both prediction targets together with consistency supervision and switches from endpoint recovery to velocity refinement as noise drops. Each step only needs one network run, and joint training also boosts single-step endpoint generation. Tests on simulated and real-world manipulation show better success rates and efficiency. On RoboTwin, UniVEP hits 61.36% success with one StarVLA run, beating the best single-target baseline at sixteen runs (59.96%). With LiLa-WAM, two runs outperform the best sixteen-step single-target baseline and cut inference latency by 77.1%.

[CV-63] Frequency-Synchronized Boundary Coupling for Training-Free Multi-Prompt Long Video Generation NEURIPS2026

链接: https://arxiv.org/abs/2611.26098
作者: Yipan Xu,Shuyong Gao,Jiyuan Fu,Keliang Yin,Jiayu Chen,Qianyu Guo,Yang Liu,Wenqiang Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted at NeurIPS 2026. Project page: this https URL

点击查看摘要

Abstract:Text-to-video generation has made remarkable progress with diffusion models and transformer-based video architectures, yet generating coherent long videos from evolving textual descriptions remains challenging. Training long-video models from scratch requires substantial computation and large-scale multi-event data, while training-free multi-prompt generation must activate new prompt semantics without breaking temporal continuity. We identify a frequency-agnostic boundary coupling problem in existing transition strategies: weak coupling causes abrupt changes, flicker, and background jumps, whereas overly strong or coarse coupling may over-preserve source semantics and produce ghosting or duplicated subjects around prompt boundaries. To address this problem, we propose FreqSync, a training-free framework that formulates prompt transitions as frequency-selective boundary coupling. FreqSync combines Frequency-Synchronized Source-Conditioned Attention (FS-SCA) with Residual-Clipped Spectral Stitching (RCSS), enabling adjacent segments to share stable low- and mid-frequency structure and motion cues while keeping high-frequency prompt-specific details target-dominant under explicit residual constraints. Extensive experiments show that FreqSync achieves state-of-the-art transition smoothness and prompt alignment while maintaining strong perceptual video quality, demonstrating its effectiveness for coherent multi-prompt long video generation.

[CV-64] DfD-GS: Defocus Image Prior for Metric-Scale Single-View Depth Estimation and View Synthesis

链接: https://arxiv.org/abs/2611.26072
作者: Shiori Ueda,Rinto Yagawa,Hideo Saito,Shohei Mori
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Project page: this https URL

点击查看摘要

Abstract:Recovering metric-scale geometry from observations at a single viewpoint remains fundamentally ill-posed, limiting the reliability of novel view synthesis. Recent methods rely on learned monocular depth priors, which provide strong structural estimates but can suffer from scale ambiguity and dataset-dependent biases. We propose DfD-GS, a feed-forward framework that predicts a metric-scale 3D Gaussian scene representation from a single-view all-in-focus–defocused image pair. The key idea is to exploit defocus blur as a physical metric depth cue. DfD-GS first estimates a defocus-based metric disparity map from the all-in-focus–defocused image pair and the corresponding defocus strength. This estimate is used to align and refine a monocular depth prior to metric scale. Visible and hidden 3D Gaussians are then predicted for occlusion-aware novel view synthesis. Experiments on synthetic-defocus datasets constructed from Booster and Middlebury show that DfD-GS outperforms single-image baselines in metric depth estimation and novel view synthesis. On WildRGB-D, DfD-GS achieves lower AbsRel and LPIPS but higher depth RMSE than the single-image baselines. Additional analyses show that DfD-GS maintains a more stable metric scale under field-of-view changes. Qualitative results on a real dataset show recovery of the main scene structure under optical defocus.

[CV-65] Domain-Aware Conformal Prediction for Trustworthy Spacecraft Pose Estimation under Sim-to-Real Shift

链接: https://arxiv.org/abs/2611.26065
作者: Nidhal Eddine Chenni,Arunkumar Rathinam,Djamila Aouada
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Autonomous rendezvous and proximity operations require not only an accurate estimate of the pose of the target spacecraft, but also a reliable indication of when this estimate can be trusted. Conformal prediction provides such an indication with a formal guarantee, by producing pose uncertainty sets that contain the true pose with a specified probability. Such a guarantee, however, assumes that calibration and test data follow the same distribution, an assumption that is violated in orbital perception, where labelled calibration images are available only from synthetic renderings while operational images are real. In this work, we show on the SPEED+ dataset that conformal pose estimation calibrated on synthetic images loses its coverage guarantee on real images, and that many of the resulting failures are silent: the predicted pose set appears valid but excludes the true pose. To address this, we calibrate on synthetic images translated to the appearance of real images. Since translation alters appearance while preserving geometry, the synthetic pose labels remain valid and no labelled real images are required. Experiments with two translation models show that this calibration restores coverage above the specified level for miscoverage levels of 0.1 and above, and reduces silent failures from up to 43.5% to at most 3.3% of the test images, bringing conformal pose estimation a step closer to trustworthy deployment in space.

[CV-66] One-to-Many: Diverse and Bidirectional Image-to-Image Translation via Rectified Flow

链接: https://arxiv.org/abs/2611.26058
作者: Sho Shimotsumagari,Seiichi Uchida,Shumpei Takezaki
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 12 pages, 13 figures, 4 tables. Includes supplementary material

点击查看摘要

Abstract:Many paired image-to-image translation tasks admit multiple valid target appearances, while training provides only one target per source. We propose One-to-Many Flow (O2MFlow), a latent-conditioned Rectified Flow (RF) model for one-to-many generation and backward translation (target \rightarrow source) within a single transport framework. Conditioning the RF dynamics on an appearance latent turns transport into a family of latent-indexed bijections: a fixed latent gives a mapping that is inverted by time-reversed integration, and sampled latents produce diverse outputs for the same input. We formalize reusable appearance variations as Shared Latent Transport and show that Kullback–Leibler (KL) regularization directly controls the source–latent distribution gap between training and generation. Across three benchmarks, O2MFlow offers dataset-dependent forward diversity–fidelity trade-offs and the lowest backward LPIPS-Dist among evaluated methods, with the lowest backward FID on two benchmarks.

[CV-67] Less Than Meets the Eye: Action-Set Identifiability and Principled Abstention for 3D Physical Geometry

链接: https://arxiv.org/abs/2611.26056
作者: Reshad Ul Karim,Abrar Samin,Sammam Mahdi,Syeda Maliha Tabassum,Md Rashad Al Hasan Rony
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:A corridor, its mirror reflection, and its image on a display can look identical, yet their physical surfaces lie at different depths. We test whether bounded camera motions distinguish these explanations. Intervene3D takes an image and an initial geometry estimate, predicts the view under each hypothesis and candidate motion, and selects a motion by maximin separability. It reports the mechanism and first physical surface only if the available motions separate the hypotheses still in contention. Otherwise it abstains. An adequacy test also rejects observations that no hypothesis explains. On a matched synthetic benchmark with exact landmarks, disabling abstention gives confident answers on 76.4% of unresolvable scenes. With abstention, none receives a confident answer under exact or noisy landmarks. At 100% accuracy on answered scenes, maximin selection answers 61.0% of scenes and summed selection 43.1%. The adequacy test lowers wrong answers on strongly curved mirror scenes from 56.8% to 12.4%. Real display experiments compare two hypotheses using one frozen depth-network estimate. The second image supplies photometric evidence and is never passed through the depth network. One executed motion reduces the pooled median surface error of twelve networks by 31.5%. These networks have larger error inside the annotated panel than outside on 33.4% of illusion-benchmark images. They place the first physical surface behind the glass on 36.9% of Mirror3D frames.

[CV-68] PaveSplat: Resolution-Robust Feed-Forward Gaussian Splatting with Residually Corrected Surface-Aligned Covariances ACCV2026

链接: https://arxiv.org/abs/2611.26045
作者: Janghun Kim,Hyun-Cheol Kim,Eunbin An,Myeonghun Oh,Hyon-Gon Choo
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to ACCV 2026. 16 pages, 4 figures, 4 tables. Project page: this https URL ; code: this https URL

点击查看摘要

Abstract:We present PaveSplat, a feed-forward Gaussian splatting method that adds a learned residual to the analytic surface-aligned covariance, keeping surfels paving the visible surface across rendering resolutions. Pixel-aligned feed-forward Gaussian splatting predicts one primitive per pixel from a few images in a single pass, and its rendering quality depends on how well these primitives cover the surface. Recent surfel-based methods compute the surface-aligned covariance analytically from predicted depth. However, this covariance has no way to correct the surfel orientation where the depth-derived frame is ill-posed. The surfels then cover the surface loosely; at the training resolution each surfel’s one-pixel minimum footprint covers the gaps, but they show up as holes when the same prediction is re-rendered at a higher resolution. To address this, we keep the analytic covariance as an initialization, with its axes aligned to the surface tangents, and add an identity-initialized unit-quaternion residual that reorients a surfel only where this improves rendering, turning surfels that the analytic frame mis-aims back onto the surface. On RealEstate10K, PaveSplat achieves the best PSNR, SSIM, and LPIPS among single-Gaussian-per-pixel methods at every scale (1x-8x). The native-resolution gain is small; the benefit appears when the same prediction is rendered at higher display resolutions without re-running the network. On zero-shot DL3DV, it leads PSNR within this group at every scale and ranks first or second on SSIM/LPIPS. A coverage-hole analysis shows fewer uncovered pixels than both baselines under high-resolution re-rendering.

[CV-69] Streaming-AAA: Streaming AAA Game Video Generation at High Resolution High Frame Rate and Low Latency

链接: https://arxiv.org/abs/2611.26040
作者: Zhihao Hu,Jianqi Ma,Jun Zhang,Jiayang Bai,Haifeng Shi,Yanxiang Zhuang,Zhongqian Sun,Weidong Zhang,Wei Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Controllable streaming video generation offers a promising route toward interactive game simulation, yet existing methods remain far from practical gameplay. They often fail to jointly achieve high visual fidelity, high frame rates, and low interaction latency, resulting in poor responsiveness and degraded user experience. We present Streaming-AAA, a new paradigm for controllable infinite generation of AAA-quality game videos. Our system generates 1080p gameplay videos at 43 FPS with 198 ms end-to-end latency interaction on merely NVIDIA A800 GPUs, while supporting long-horizon play without severe temporal degradation or collapse. Streaming-AAA is built upon three key designs. First, we redesign the VAE architecture for efficient high-resolution decoding, enabling real-time synthesis at 1080p. Second, we pretrain the diffusion model using diffusion forcing and significantly extend the temporal training horizon, allowing the model to capture long-term dependencies and maintain persistent memory. Third, we perform post-training distillation and memory compression to improve inference efficiency while retaining visual quality and temporal consistency. Experiments show that Streaming-AAA substantially improves visual fidelity, frame rate, latency, and long-term stability over prior controllable video generation approaches, enabling a significantly more playable experience for interactive AAA game video generation.

[CV-70] Fixing the Twin Not the Detector: Sim-to-Real Roadside LiDAR Detection from Digital-Twin Data

链接: https://arxiv.org/abs/2611.26035
作者: KuanLin Chen,Feng-jui Kuo
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 8 pages, 1 figure, 5 tables. Accepted to the IEEE BigData 2026 Cup (UCF UrbanTwin Sim2Real LiDAR Challenge); first place, LUMPI track. Code: this https URL

点击查看摘要

Abstract:Roadside LiDAR detectors are expensive to train on real labels, and digital twins promise to replace these labels, but a detector trained on a twin meets a different sensor, different objects and different annotation conventions on the real site. We describe the pipeline that placed first in the LUMPI track of the UrbanTwin Sim2Real LiDAR Challenge (IEEE BigData 2026 Cup, 0.5541 combined score) and fifth on its V2X-Real track (0.8102). Every detector is a CenterPoint model trained only on synthetic data, on a single 8 GB laptop GPU. Most of the gain came from correcting differences between the twin and the real site rather than from the detector. We align the twin’s coordinates, point density and box convention with the test sensor, generate the pedestrians and riders that the twins lack, plus street clutter as unlabeled negatives, and rebuild twin objects at real size and real point density. A non-learned resampling to the per-cell point density of unlabeled real frames lifts LUMPI realism from 0.589 to 0.947, against 0.949 between real frames. Label-free post-processing then exploits what a fixed roadside sensor guarantees: static structures recur in every frame, and real objects have characteristic point statistics. Reporting the validation and test gain of each component side by side, we find that corrections of systematic sim-to-real differences transfer to the test set, typically keeping a quarter to three quarters of their validation gain, while choices that fit the few objects of a validation set do not.

[CV-71] Learning What Actions Change: Action Effect Matching for Dynamics-Faithful World Models

链接: https://arxiv.org/abs/2611.25980
作者: John Won,Yonghoon Dong,Jinwoo Shin
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: 27 pages, preprint

点击查看摘要

Abstract:World models aim to predict environment dynamics under actions, enabling the prediction of action consequences without executing them in the real world. Yet despite producing increasingly realistic videos, current world models often fail to capture how small action changes affect future observations. This limits their use in tasks such as robotic policy evaluation and model-based planning, where nearly identical actions can lead to qualitatively different outcomes. To address this, we propose AFFECT, an auxiliary training objective for learning the distribution of visual changes induced by action perturbations. Specifically, we inject temporally correlated noise into the actions of a policy trained on task-agnostic human teleoperation. We then train the world model so that its generated videos respond to this noise as the recorded videos do, without requiring repeated executions from the same initial state. Across OGBench simulation and real-world DROID and ALOHA-2 setups, AFFECT makes world-model rollouts more faithful to real outcomes. On DROID, the Pearson correlation between policy success rates in the world model and on the real robot rises from 0.47 to 0.79, alongside up to 23% lower FVD. These gains carry over to planning, where selecting actions with AFFECT-trained world models improves task success by up to 8.3%.

[CV-72] Multi-View Transformer with Geometry-Guided Multi-Modal ControlNet for 3D Virtual Try-On

链接: https://arxiv.org/abs/2611.25979
作者: Sarthak Batra,Armin Mustafa,Adrian Hilton
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 20 pages, 2 figures, 3 tables

点击查看摘要

Abstract:Editing a 3D Gaussian Splatting (3DGS) capture of a person so that they wear a new garment requires every rendered view to agree on the same garment details. Existing 3D virtual try-on (VTON) pipelines, however, edit training views largely independently and fuse them with hand-designed heuristics such as manually chosen reference views and fixed attention-blending weights. The resulting inconsistencies are baked into the 3D representation and surface as texture drift, blurred prints, and implausible geometry in novel views. We present GS-VTON-MVT, a three-stage framework that combines implicit and explicit 3D reasoning for consistent 3DGS try-on. (i) A multi-view inpainting transformer jointly edits a set of training views with full cross-view self-attention, so that cross-view correspondence is learned rather than imposed. (ii) A geometry-guided multi-modal ControlNet conditions the diffusion editor on the garment image together with depth and surface normals rendered from the current 3DGS state, supplying explicit structure where visual evidence is weakest. (iii) A pose-adaptive persona-aware attention replaces the fixed blending weight between current-view and reference-view features with weights derived from camera-pose proximity. On the 3D-VTONBench benchmark, GS-VTON-MVT improves FID, KID, LPIPS-based cross-view consistency, and user preference over general-purpose 3D editors and prior 3D try-on methods.

[CV-73] Sample-efficient end-to-end Optical Music Recognition through Active Learning

链接: https://arxiv.org/abs/2611.25975
作者: Eric Ayllon,Francisco J. Castellanos,Antonio Rios-Vila
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:End-to-end approaches to Optical Music Recognition (OMR) have recently shown promising results, but their performance often degrades on unseen data. A common solution is to pretrain and fine-tune models with annotated samples from the target domain, improving adaptation to new distributions. However, this process typically requires extensive manual annotation to yield competitive models, making it costly and time-consuming. In this work, we explore Active Learning to reduce this burden by selecting informative samples for labeling in low-resource scenarios, where the amount of manual annotation that can be performed is limited. We evaluate both uncertainty- and diversity-based sampling strategies on staff-level and full-page corpora. Results show that confidence-based sampling performs comparably to oracle selection (based on true error), and that each scenario benefits from a tailored strategy.

[CV-74] MorphTwin: Enriching Deformable Material Responses from Videos for Morphable and Interactive Physical Twins

链接: https://arxiv.org/abs/2611.25963
作者: Haoqin Hong,Ding Fan,Fubin Dou,Tianyu Huang,Yuzhuo Chen,Zhemeng Zhang,Hanxiao Jiang,Congcong Zhu,Gerhard Petrus Hancke,Jingrun Chen,Rynson WH Lau
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Reconstructing interactive physical twins from video requires accurately capturing how objects deform while maintaining efficient simulation. We observe that reproducing motion in a video does not guarantee correct responses to new forces when the physical representation restricts how an object can deform. In spring-based twins, coupling between volume and shape responses creates such a limitation. We present MorphTwin, which enriches local deformation modes to capture and independently control these responses while retaining standard spring forces and explicit simulation. To support interpretable material editing and transfer, we use offline calibration to map shared relative response changes to stiffness updates on each reconstructed graph. Under identical inverse settings on synthetic scenes, MorphTwin reduces the mean absolute error of Poisson’s ratio by 75.8% and the mean relative error of Young’s modulus by 22.1% compared to PhysTwin. We introduce an egocentric stereo and RGB-D benchmark covering interactions with objects of varying stiffness. Real-world experiments demonstrate improved reconstruction accuracy and up to 14.4% lower future tracking error than PhysTwin at interactive speeds.

[CV-75] Point Dynamics as Proxy for Robotic World Models

链接: https://arxiv.org/abs/2611.25946
作者: Dong An,Tianyu Liu,Haoran Ding,Wen Yang,Anqing Duan,Zhengzhong Liu,Eric P. Xing,Guangyi Liu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 24 pages, 7 figures, 11 tables. Project page: this https URL

点击查看摘要

Abstract:Robotic world models predict future video frames from the current observation and a sequence of robot actions. While recent models follow the given actions closely, manipulated objects often collapse or move inconsistently with the action. We attribute this to the conditioning: an action, or a rendering of the robot, specifies where the robot goes but not how the scene responds, leaving a single video model to infer the interaction dynamics implicitly while also rendering appearance. We therefore separate dynamics prediction from visual rendering, with 3D point dynamics as the proxy between them. Specifically, we lift the first frame into 3D points, simulate how they respond to the robot’s motion with a learned point dynamics model, and render the simulated points back into pixel-aligned videos that any video model can read. In controlled experiments on the DROID dataset with two video renderers of different architectures, this point dynamics proxy reduces object motion error by 8.6% and 6.2% without degrading image quality, and by 8.1% on a real bimanual robot. Moreover, the video follows the simulation: more accurate dynamics yield more accurate object motion, and editing the points steers the video. This suggests that advances in dynamics modeling can translate into more faithful robotic world models.

[CV-76] winD: Generalizing Deepfake Detection from Blending to Entire Face Synthesis

链接: https://arxiv.org/abs/2611.25935
作者: Andrii Yermakov,Jan Cech,Mario Fritz,Jiri Matas
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Facial deepfake detectors are usually trained on blending fakes, where a generated face is pasted back into a real frame. Recent generators synthesize the entire face and leave no traces of blending. On a benchmark of 19 datasets released from 2019 to 2026, the best prior facial deepfake detector reaches a mean AUROC of 96.2% on ten blending datasets, but none exceeds 78.1% on five datasets of entire face synthesis (EFS). AI-generated content (AIGC) detectors do not close this gap. We propose TwinD, a detector trained on FaceForensics++ fakes and two kinds of twins: fakes made from a real frame. A reconstruction twin is a real frame reconstructed by an autoencoder, with the real pixels put back in a random region. It targets the neural traces of a decoder. A blob twin blends an altered copy of a real frame with the frame itself, targeting blending artifacts. It needs no generator and no face-shaped mask. Together, these twins provide two complementary forensic cues that generalize to unseen manipulations. TwinDreaches a mean AUROC of 96.3%, 8.9 points above the prior state-of-the-art, with 98.1% on the blending datasets and 95.6% on the EFS datasets. Trained on twins alone, without a deepfake dataset, single models still reach 94.3% on average over three seeds. For reproducibility, the code, trained model, and training data are at: this https URL

[CV-77] AGSR: Aperture-Guided Super-Resolution for Gaussian Splatting

链接: https://arxiv.org/abs/2611.25934
作者: Ayman Alkhateeb,Hadi Amirpour,Christian Timmerer
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: Accepted at the 2026 IEEE International Workshop on Multimedia Signal Processing (MMSP 2026). Recipient of the Best Paper Award

点击查看摘要

Abstract:3D Gaussian Splatting (3DGS) enables real-time novel view synthesis, but Gaussian rasterization attenuates fine-scale scene structure through the projected covariance of each primitive. We show that contracting these projected covariances at inference time can mitigate the effective bandwidth limitation induced by finite Gaussian footprints, motivating a multi-aperture rendering framework for super-resolution. Based on this insight, we propose AGSR (Aperture-Guided Super-Resolution), a lightweight inference-time framework that progressively fuses multi-aperture renderings using uncertainty-guided feature aggregation and geometric conditioning from transmittance and depth cues. AGSR operates as a plug-in enhancement for pre-trained 3DGS models. Experiments on the Mip-NeRF 360 benchmark show that AGSR outperforms lightweight image-space super-resolution baselines, with gains increasing as the underlying Gaussian representation deteriorates. The results provide evidence that covariance contraction exposes bandwidth-sensitive structural cues attenuated by native Gaussian rasterization. AGSR achieves these improvements while maintaining real-time performance with up to 139k parameters. Comments: Accepted at the 2026 IEEE International Workshop on Multimedia Signal Processing (MMSP 2026). Recipient of the Best Paper Award Subjects: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR) Cite as: arXiv:2611.25934 [cs.CV] (or arXiv:2611.25934v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2611.25934 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[CV-78] Why Flat Drafts Stay Flat: Region-Mean Latents Govern Texture in Flow-Based Image Generation

链接: https://arxiv.org/abs/2611.25930
作者: Zihan Gu,Jiangcan Jia,Hua Zhang,Lulu Zhao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Flat color-block drafts specify layout and regional color while leaving interior detail unspecified, yet pretrained flow-based generators can preserve their flat appearance even under photorealistic prompts. We trace this behavior to a cross-scale mechanism linking fine pixel variation to regional means in variational autoencoder (VAE) latent space. Flattening photo regions systematically shifts their latent means, with offsets concentrated along a shared channel direction within each backbone. Controlled field exchanges in FLUX.1-dev and Qwen-Image establish the effect of this displacement on texture generation: adding the photo-derived regional mean field to a flat latent restores texture, whereas adding its complementary field alone largely preserves flatness. A second-order encoding expansion explains how encoder curvature can convert zero-mean perturbations into mean shifts. An analytically solvable conditional Gaussian flow shows how regional noise averaging enables near-unit endpoint sensitivity at high signal-to-noise ratios. This mechanism motivates training-free rendering through pixel dither before encoding, mean-field scaling, and early boundary holding. Paired low-rank adaptation (LoRA) with draft conditioning improves content recovery and produces photo-like regional means under the standard velocity objective. Explicit regional mean guidance remains effective after adaptation, continuously controlling texture and improving color fidelity. These findings connect encoding shifts to persistent generation behavior and support interventions evaluated for texture, spatial fidelity, and semantic recovery.

[CV-79] SpatialMem: Building Quantitative Spatial Reasoning Agents with Hierarchical Memory

链接: https://arxiv.org/abs/2611.25925
作者: Zinan Zheng,Nuo Chen,Jia Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Vision Language Models (VLMs) have shown remarkable capabilities in qualitative spatial reasoning, but remain limited in quantitative spatial reasoning. Current work leverages the VLM agent to invoke external vision tools to perform computation. However, these methods usually bind each functional requirement to a single predefined vision specialist, implicitly resulting in fixed and rigid tool pipelines. In practice, addressing diverse spatial scenarios and tasks demands a broader diversity in pipeline orchestration and vision specialist selection to achieve robust spatial reasoning. However, such diversity leads to a combinatorial execution space, making it challenging for the agent to orchestrate the optimal pipelines and specialists for different scenarios. In this work, we propose SpatialMem, a memory framework that enables a VLM agent to accumulate tool-use experience and leverage that memory to orchestrate vision specialists to solve the spatial reasoning problem. SpatialMem builds a two-level memory structure, with task memory capturing tool orchestration and scene memory capturing tool selection. It further introduces a dual-state trajectory optimization process to preserve successful tool experiences and refine suboptimal ones. Moreover, existing spatial reasoning benchmarks suffer from quality issues and limited coverage of scene types and tasks. To address these gap, we propose MetricBench, a benchmark covering both indoor and outdoor scenes and three quantitative spatial reasoning tasks. Our method achieves state-of-the-art results on MetricBench, Q-Spatial, and SpatilRGPT-Bench, outperforming best VLM baselines by 32% on MetricBench. Additionally, the memory design yields a 9.1% absolute gain over a memory-free agent on MetricBench. Subjects: Computer Vision and Pattern Recognition (cs.CV) Cite as: arXiv:2611.25925 [cs.CV] (or arXiv:2611.25925v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2611.25925 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[CV-80] Multiple-Camera Tracking of Rigid Objects

链接: https://arxiv.org/abs/2611.25904
作者: Frédérick Martin,Radu Horaud
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:In this paper we describe a method for tracking rigid objects using one or several cameras. The tracking process consists of aligning a 3-D model representation of an object with image contours by measuring and minimizing the image error between predicted model points and image contours. The tracker behaves like a visual servo loop where the internal and external camera parameters are updated at each new image acquisition. We study in detail the Jacobian matrix associated with this minimization process in the presence of both point-to-point and point-to-contour matches. We establish the minimal number of matches that are needed as well as the singular configurations leading to a rank-defficient Jacobian matrix. We find a mathematical link between the point-to-point and point-to-contour cases. Based on this link we show that the latter has the same kind of singularities than the former. Moreover, we study multiple camera configurations which optimize the robustness of the method in the presence of single-camera singularities, bad, noisy, or missing data. Extensive experiments done with a complex ship part and with up to three cameras validate the method. In particular we show that the tracker may well be used as a camera calibration procedure.

[CV-81] Does It Depend on the Split? Frozen Satellite Features vs. Fine-Tuned CNNs for Tree Species Mapping

链接: https://arxiv.org/abs/2611.25902
作者: Dharani Deivasihamani,Nandakishore Puttashamachar
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Self-supervised foundation models are often said to need fewer labels than supervised training. We test this on a forestry task: classifying 20 tree species classes in 0.2 m aerial images from the TreeSatAI dataset. We compare a frozen DINOv3 model pretrained on 493 million satellite images, with only a linear classifier trained on top, against an ImageNet ResNet-50 that is fine-tuned end to end. On the official split, fine-tuning wins clearly (macro F1 0.764 vs. 0.685). However, 28.8% of the official test patches overlap a training or validation patch, and copying the label of the nearest such patch is correct for 83.6% of them. We therefore build a spatial split that holds out whole 2 km squares with a 200 m buffer. On this split the comparison becomes inconclusive (0.648 vs. 0.640, paired 95% CI of the difference [-0.054, 0.016]), and the learning curves of the two models overlap at every label budget from 343 to 33,126 training patches. On the five rarest classes, the frozen model achieves a higher mean F1 (0.566 vs. 0.445), although the confidence interval for this difference includes zero ([-0.05, 0.23]). Diagnostics show that the frozen features mainly miss differences between closely related species, such as the two native oaks. Choosing the input size and layer on validation data (224 px, block 16) raises the frozen model to 0.662 on the spatial split. A comparison that looks decisive on the official split is thus inconclusive under geographically separated evaluation. We release the spatial split and our code.

[CV-82] EvoMem: Self-Evolving Skills for Time-Aware Long Video Memory Agents

链接: https://arxiv.org/abs/2611.25889
作者: Yutong Guan,Meixing Shi,Haoran Sun,Bing Sun,Yuxiang Cai,Yankai Jiang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Although memory-augmented architectures have expanded the ability of large vision-language models (LVLMs) to process long videos, existing retrieval processes still largely rely on global semantic matching and lack state-driven temporal control, often returning fragmented or redundant evidence and failing to maintain a coherent evidence chain. To address this challenge, we propose EvoMem, a time-aware long-video memory agent guided by self-evolving retrieval skills. EvoMem constructs a hierarchical time-indexed memory, where different types of evidence are aligned onto a unified video timeline; during retrieval, temporal cues from acquired evidence are further converted into explicit signals that constrain subsequent memory access. To address recurring procedural errors in long-video question answering, such as unclear retrieval targets, improper memory access paths, or premature stopping, EvoMem introduces retrieval skills to provide process-level guidance, enabling the agent to dynamically adjust retrieval content, memory operations, and target temporal ranges according to the current evidence state, while deciding when to stop. Furthermore, inspired by reinforcement learning, we formulate skill evolution as a reward-guided policy improvement process. Each candidate skill is treated as an external policy patch, and rewards are assigned according to answer-correctness state transitions on the validation set. In this way, effective skills are accepted and low-quality skills are rejected without updating model parameters. Experiments on four long-video question answering benchmarks show that EvoMem significantly outperforms existing baselines, achieving an average accuracy of 60.1% and an average improvement of 8.6 percentage points over the previous best methods, demonstrating its effectiveness for long-video reasoning.

[CV-83] WHU-Road3D: A dataset and benchmark towards fine-grained road surface mapping

链接: https://arxiv.org/abs/2611.25873
作者: Xiaoxin Mi,Zhen Cao,Chong Liu,Yuhao Li,Zhen Dong,Bisheng Yang,Yang Jia,Shengwu Xiong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages

点击查看摘要

Abstract:Accurately extracting and mapping road surface elements is crucial for applications such as high-definition maps, road asset management, and intelligent transportation systems. This study first provides a comprehensive review of the benchmark datasets for extracting and mapping road elements, encompassing overhead-to-perspective data, large-to-local scales, and coarse-to-fine details. Subsequently, we thoroughly review the relevant approaches for extracting road surface elements. However, we argue that existing datasets still require enhancements in semantic richness, data modality diversity, and scenario complexity. To bridge these gaps, this paper presents WHU-Road3D, a large-scale multimodal dataset for fine-grained road surface mapping to address these shortcomings. WHU-Road3D provides approximately 100 km of diverse Chinese highways and urban streets. It features a rich combination of MLS point clouds, associated panoramic images, vehicle trajectories, detailed vectorized annotations, and pointwise instance annotations for road surface elements. The dataset carefully annotates 15 distinct categories of road elements, including common markings like road boundaries, zebra crossings, lane dividers, stop lines, as well as 10 types of direction symbols. To our knowledge, WHU-Road3D stands as the first multimodal dataset for mapping such a comprehensive range of road surface elements, designed to facilitate cutting-edge research in instance segmentation, full-element extraction and mapping on road surfaces. Beyond these immediate applications, WHU-Road3D also provides a valuable source for hierarchical road topology reasoning, map updating, and street scene generation. The dataset, along with the code for our benchmark methods, will be made publicly available at \hrefthis https URLthis https URL.

[CV-84] hinking in Streams: Proactive Spatial Reasoning with Vision-Language Models

链接: https://arxiv.org/abs/2611.25865
作者: Junyoung Hong,Minsu Kim,Jaehyun Kim,Jeonghwan Cho,Jiwoo Lee,Seon Joo Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 8 figures

点击查看摘要

Abstract:Recognizing when a requested spatial condition holds in streaming video can require maintaining and updating spatial knowledge across changing viewpoints. We study this capability through proactive spatial reasoning, where a model receives a query in advance and determines whether and when to respond as the stream unfolds. We introduce ProSpace-Bench, comprising nine tasks across object, camera, and scene reasoning, and ProSpace-Train for supervised training. To support spatial decisions, we propose ProSpace-VLM, which injects explicit geometric encodings estimated causally from incoming frames into intermediate VLM layers, using the existing KV cache as spatial memory and spatial token selection for efficient streaming. Experiments reveal a substantial gap between VLMs and humans, while ProSpace-VLM improves timely responses and correct abstention over fine-tuned baselines. Analyses examine how ProSpace-VLM represents previously observed spatial information across changing viewpoints.

[CV-85] EgoWitness: Benchmarking Instance-Level Spatial Memory and Reasoning in Egocentric Video Streams

链接: https://arxiv.org/abs/2611.25849
作者: Zhengru Fang,Zehao Liu,Mingyu Tsoi,Yaoyu Zhu,Junlong Li,Fengmao Lv,Sijie Cheng,Yuanliang Sun,Fang Yang,Jun Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 43 pages, including supplementary material. Zhengru Fang and Zehao Liu contributed equally. Project page: this https URL ; Code: this https URL

点击查看摘要

Abstract:Recalling where and when an object was last seen and who saw it requires recognizing it across days and people. Existing long-video benchmarks focus on events, object categories, or relative locations, leaving instance-level spatial memory largely untested in multimodal models. To address this gap, we present EgoWitness, a benchmark for instance-level spatial memory and reasoning in egocentric video. Its 8,453 human-verified questions span approximately 324 hours of week-long and short-horizon recordings and test seven abilities, from recognizing an object or room to recalling who saw it and reasoning about room visits. Over one million reviewed masks of 34 registered landmarks and 21 movable-object labels are linked through room timelines. Among twelve direct vision-language models (VLMs), open-weight models reach at most 46.97% accuracy, and none exceeds 35% on identifying who saw an object last. We therefore propose EgoClue, a spatial-memory agent that retrieves candidate intervals, verifies instance identity, and computes spatial and temporal relations with five tools. Two-stage training with supervised fine-tuning (SFT) and Group Relative Policy Optimization (GRPO) raises its 4B controller from 23.64% to 72.65% accuracy on 1,320 held-out questions. Evidence-trace synthesis matches the best direct VLM at 74.85% on week-long questions and achieves 80.56% in environments excluded from training.

[CV-86] Interpretable Prototype-Conditioned Adaptation for Weather Forecasting

链接: https://arxiv.org/abs/2611.25848
作者: Zhixuan Li,Hao Jiang,Chen Hui,Kai Geng,Weisi Lin
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 38 pages, including supplementary material. Project page: this https URL

点击查看摘要

Abstract:Data-driven weather forecasting requires distinguishing local states and their temporal evolution. However, learned patterns remain implicit in neural forecasters, making their influence difficult to interpret. Failing to distinguish local evolution can lead to similar forecasts for locations that develop differently, increasing prediction error. This motivates explicit references for distinguishing and interpreting local patterns. Prototypes provide explicit pattern references, but sample- or task-level conditioning leaves open how to adapt to distinct locations and variable groups. To deal with this problem, we design ProtoCast, which uses group-specific prototype banks to adapt local features while accounting for cross-variable dependencies. Specifically, (1) we combine each group’s historical features and temporal change with forecast context from cross-variable attention for evolution-aware prototype matching. (2) Sparse mixtures of prototype-specific low-rank operators tailor corrections to each group and location. A gated attention block then exchanges the corrected features across groups before prediction. Each group’s correction decomposes into explicit prototype contributions before this fusion. (3) We jointly train the forecasting network and operators from scratch and refit each group’s prototypes before every training epoch. Slot tracking preserves prototype-operator associations, making each prototype’s contribution to local corrections more interpretable. Experiments on GlobalForecast and TyphoonTracking show improved forecasting accuracy, while prototype deletion links local feature contributions to changes in prediction error. This work advances the development of more accurate and interpretable data-driven weather forecasting models. Project page: this https URL

[CV-87] rajCal: Trajectory-Calibrated Time Schedules for Few-Step Flow Sampling

链接: https://arxiv.org/abs/2611.25847
作者: Junchi Zhang,Zihan Gu,Jiangcan Jia,Lulu Zhao,Wenbo Su,Hua Zhang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Flow matching enables high-quality image and video generation, but its iterative sampling process requires repeated evaluations of large neural networks. Existing few-step methods can distort the output and reduce its agreement with a many-step reference. We propose TrajCal, a training-free calibration framework for few-step sampling. Based on scatter search, a well-established metaheuristic for difficult optimization problems, TrajCal starts with diverse seed schedules, deforms and combines these candidates, and selects one fixed schedule by final output fidelity for reuse on new inputs. For both image and video generation, TrajCal achieves higher reference fidelity than all compared baselines at low evaluation budgets while maintaining competitive visual quality. The search is effective even with randomly generated seeds, while trajectory-based seeds further improve fidelity. For cross-style generalization, photography-calibrated schedules also improve fidelity on unseen image styles.

[CV-88] Harnessing Medical Visual-Language Models with Agent ic Evidence Management

链接: https://arxiv.org/abs/2611.25840
作者: Qiming Huang,Hanze Dong,Jipeng Zhang,Mike Zheng Shou,Xinxing Xu,Furu Wei
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 22 pages, 3 figures. Qiming Huang, Hanze Dong, and Jipeng Zhang contributed equally

点击查看摘要

Abstract:Medical vision-language agents increasingly rely on segmentation tools, and recent work has focused on acquiring the right mask, often treating a good mask as the end point of tool use. But do these agents use the resulting evidence effectively? To answer this question, we need a benchmark that separates evidence use from segmentation quality, spans multiple medical image types, and allows the same mask to be presented in different ways. Existing medical VQA datasets do not jointly provide these properties. We therefore construct Mask-Evidence MedVQA, a 2.5K-example benchmark spanning four image types, where each question is paired with an answer-relevant ground-truth mask and multiple possible renderings. Our study reveals two limitations of direct tool use. First, evidence presentation matters. Different renderings of the same fixed mask can substantially change model behavior. Second, masks introduce an evidence trade-off. Highlighting a relevant region can expose diagnostic evidence the model previously missed, but the added visual cue can also induce incorrect associations. As a result, even a ground-truth mask can correct some answers while turning initially correct ones into incorrect ones. Together, these findings show that acquiring the right mask is not enough. An agent must also decide how to present the evidence and whether it should change the answer. We therefore introduce TUNE (Tool Use is Not Enough), an evidence-management harness that explicitly manages what evidence to acquire, how to present it, and whether to accept the resulting answer update. Across three public medical VQA benchmarks, evidence management improves the fine-tuned Qwen3.5-9B backbone’s mean accuracy by 4.0 percentage points. Full TUNE improves over direct tool use by 9.1 points on average across seven benchmarks and achieves competitive results against existing medical VLMs and agents.

[CV-89] ContextRenderer: Dynamic-Context Video Generation for Interactive Game Worlds

链接: https://arxiv.org/abs/2611.25838
作者: Runze Wang,Fei Zha,Bing He,Ke Xu,Zhongyuan Wu,Yuxuan Xie,Wenzheng Wu,Zhongyuan Liu,Youcheng Cai,Ligang Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注: 10 pages, 7 figures, 3 tables

点击查看摘要

Abstract:Interactive game world models separate executable world evolution from visual generation, using video models to render engine-provided scene representations into detailed observations. During sustained gameplay, however, existing generative renderers struggle to follow changing conditions while preserving visual continuity across diverse proxies and target styles. We therefore introduce ContextRenderer, an autoregressive video diffusion framework designed for both dynamic conditioning and cross-proxy generalization. It couples Dynamic-Context Learning (DCL) for chunk-level interaction supervision with Recurrent History Compression (RHC) to follow condition updates while preserving visual continuity within a bounded token budget. Across proxy representations, Cross-Representation Flow Alignment (CRFA) connects proxy reconstruction with appearance and interaction variation through a shared high-fidelity reconstruction. WorldPrism, our interaction-centric dataset constructed from agent-built worlds, AAA gameplay, and open-source videos, supplies the corresponding renderings and chunk-level annotations for these learning tasks. Experiments on GWM-Bench and Solar-Bench show improved reconstruction and visual quality, while qualitative results demonstrate minute-long interactive rendering under changing prompts and subject references. Project page: this https URL

[CV-90] Rank2Replace: VLM-Ranked Slot Replacement for Out-of-Distribution Generalization

链接: https://arxiv.org/abs/2611.25834
作者: Bo-Han Feng,Keng-Te Liao,Shou-De Lin
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Spurious correlations can give image predictors high in-distribution accuracy but poor generalization under distribution shift. We introduce \textbfRank2Replace , which transfers knowledge of task-relevant evidence from pretrained vision–language models (VLMs) to small predictors through training-time augmentation for out-of-distribution (OOD) generalization. The VLM ranks object-centric slots by task relevance and guides rank-dependent replacement with same-rank slot embeddings from another image. This augmentation retains the original image’s label, encouraging reliance on preserved task evidence as incidental cues change, and supports combination with existing OOD methods. Across four datasets and six splits under shared frozen slot embeddings, the default cross-entropy configuration reaches macro-averaged average-group and worst-group accuracies of 80.43% and 65.28% without environment labels or dataset-specific replacement tuning. Slot corruption improves both macro-averaged metrics for all fourteen OOD training methods. Combining it with slot ranking and IRM reaches 81.96% and 72.68% . Controlled ablations support the ranking and replacement design, with gains across five ranking VLMs and six extractor–predictor combinations. Permutation-invariant predictors further improve both macro-averaged metrics over their unaugmented counterparts while using VLM guidance only during training for efficient inference.

[CV-91] Fusing Multimodal CNN Features with Structurally-Masked Transformers for Hierarchical Sub-Pixel Tree Species Mapping

链接: https://arxiv.org/abs/2611.25824
作者: Mattia Ferrari,Lorenzo Bruzzone
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 19 pages, 6 figures

点击查看摘要

Abstract:Precise forest species mapping using spaceborne hyperspectral data is challenging due to high spectral similarity and the inherent sub-pixel heterogeneity of natural forest stands. While traditional deep learning models exploit spatial context through convolutional architectures, these approaches can introduce noise in fragmented or mixed-species forests. In this paper, we propose a hierarchy-aware, multimodal deep learning framework to resolve sub-pixel forest composition by integrating PRISMA (PRecursore IperSpettrale della Missione Applicativa) hyperspectral imagery, high-resolution panchromatic data, and digital elevation models. The architecture features a multi-branch encoder combining an EfficientNetV2-based one-dimensional convolutional neural network (CNN) for spectral signatures and twin two-dimensional CNNs for spatial and topographic features, coupled with a novel structurally-masked Transformer. By using a binary transition tensor to condition the Transformer slot embeddings and constrain inference, the model enforces taxonomic consistency between broad forest categories and fine-grained species distributions. The framework is validated across three diverse Italian biogeographical regions (the Alps and the Central and Southern Apennines) using field-surveyed plots. The proposed method consistently outperforms state-of-the-art multi-label and ranking-based paradigms, particularly in identifying subordinate species and their relative dominance. The bidirectional learning mechanism allows sub-pixel compositional features to refine macroscopic forest type predictions, while the integration of topographic priors effectively resolves spectral ambiguities in complex terrains. This knowledge-driven method supports regional-scale biodiversity monitoring and forest characterization within the evaluated PRISMA scenarios.

[CV-92] Limits of local curvature for constructing search subspaces in segmentation models

链接: https://arxiv.org/abs/2611.25822
作者: Maxim Bolshim(1),Alexander Kugaevskikh(1) ((1) ITMO University, Saint Petersburg, Russia)
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 10 pages, 3 figures, 4 tables. Includes an appendix

点击查看摘要

Abstract:We examine how well subspaces constructed from local curvature represent the change in the gradient update field after a finite step in semantic segmentation models. The local response is the action of the curvature operator on the actual input displacement. In a study of 50 images and two models, we measure the ratio of the loss increase attained in a fixed subspace to that attained in the full space. This ratio decreases as the search radius increases. On ten images from the same sample, we compare the one-dimensional subspace spanned by the initial gradient with two rank-16 subspaces for both models. The two equal-rank subspaces are assessed at the same initial states, using the same actual steps and measured vectors. Adding generalised Gauss-Newton directions to the initial gradient reduces the mean relative error in representing the field change from 0.922 to 0.874 for the first model and from 0.819 to 0.763 for the second. At the same rank, including the response direction for the initial gradient nearly eliminates the error in representing the local response. The mean error in representing the field change does not decrease. A decomposition with fixed pixel weights shows that the finite-step remainder is comparable in norm to the field change and makes the largest diagonal contribution to the squared projection error of the field after the step. Evaluation of a search subspace should measure how accurately it represents the field change and the local response.

[CV-93] SCCV-6D: Sparse Canonical Coordinate Voting for Holistic Multi-View RGB-D 6D Pose Estimation in Robotic Bin Picking WACV

链接: https://arxiv.org/abs/2611.25819
作者: Nico Leuze,Samed Dogan,Maximilian Hoh,Henry Schaub,Alfred Schöttl
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: to appear in Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision (WACV), 2027

点击查看摘要

Abstract:Industrial bin picking requires pose estimates that are accurate under clutter, partial occlusion, and the reflective, texture-less nature of manufactured parts. Most recent approaches to 6D pose estimation follow multi-stage pipelines. These achieve high accuracy, but incur memory and runtime footprints that do not cope with the operational constraints of many real-world industrial deployments. Holistic pipelines estimate poses in a single pass, with cost controlled through indiscriminate subsampling or lowered resolution. Continuing this line, we present SCCV-6D, a fully sparse multi-view RGB-D estimator. We reconstruct the 3D space at millimeter precision, prune background via a staged heatmap mechanism, and predict class, translation offset and rotation at highest resolution for an arbitrary number of objects simultaneously. Rotation is formulated as canonical coordinate voting. Points vote for their canonical object coordinates with an associated confidence, enabling closed-form pose recovery with a symmetry-aware weighted Procrustes solve. Along an in-cluster point recovery induced by offset votes, this removes time-consuming post-processing stages. We reuse the task-supervised attention priors from the heatmap formulation for a novel attention-gated distillation objective that reweights teacher-student feature alignment. On the IPD test split we report competitive 0.826 AP at 1.13 s per scene, significantly faster than the leading entries. On MV-YCB our approach reaches 96.1 ADD-S, and remains competitive even on the densely packed bins of ROBI.

[CV-94] What Places an Action Boundary? Structure-Aware Attribution for Temporal Action Localization

链接: https://arxiv.org/abs/2611.25815
作者: Ruoshan Lan,Zongru Shao,Chieh Tsai,Mihai Surdeanu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Temporal action localization (TAL) detectors must identify actions and estimate their temporal boundaries, but localization remains a major source of detection errors. While understanding what determines boundary placement requires faithful attribution of temporal evidence, existing attribution methods have not sufficiently addressed the temporal structure of TAL or the need to tailor attribution to boundary localization. We introduce a structure-aware attribution framework that averages contributions across random evidence-restoration orders following the Shapley-Shubik principle. We make it tractable by using the detector’s receptive field to limit which time steps are ordered individually, and by approximating each order with a soft gradient path. For boundary predictions, we further introduce cell-structured restoration orders and boundary-anchored references that selectively remove adjacent evidence. We also adapt faithfulness evaluation to assess the joint effects of contiguous temporal evidence and the direction of boundary displacement. Experiments on THUMOS14 and EPIC-KITCHENS-100 with TriDet and ActionFormer demonstrate improved attribution faithfulness over existing methods. Attribution analysis and controlled interventions further reveal that class-score changes do not reliably track boundary displacement, providing input-level evidence that classification relies on content less informative about boundary placement. Code and evaluation artifacts will be publicly released.

[CV-95] World-delta: Continually Integrating Observations for Incremental 3D Scene Generation

链接: https://arxiv.org/abs/2611.25802
作者: Haonan Wang,Hanyu Zhou,Xiaobin Hu,Luxin Yan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Existing 3D generation methods synthesize visually compelling scenes from fixed inputs, yet struggle to continually expand and revise a 3D world as new observations arrive. Iterative approaches often fuse real observations and generated content into a single representation, treating generated content as established structure rather than revisable predictions. Consequently, early generation errors can persist and accumulate despite new evidence. We propose World-delta, an incremental 3D scene generation framework combining uncertainty-aware persistent world modeling with delta-guided diffusion for continual scene expansion and correction. Our framework comprises three components: 1) Visual Evidence Encoding integrates geometric and semantic cues from incoming images into unified world tokens. 2) Observation-Calibrated World Modeling maintains a persistent latent world state that distinguishes observation-supported content from provisional completion hypotheses. New evidence drives selective world deltas to revise conflicting hypotheses while preserving well-supported content. 3) Incremental World Diffusion Generation uses delta-derived masks to restrict diffusion to regions requiring updates, freezing unchanged regions. Only affected 3D chunks are decoded and integrated into the existing scene. Together, these components enable continual world expansion and revision while reducing redundant generation and preserving scene consistency. Experiments demonstrate the effectiveness of World-delta in incremental 3D scene generation, observation-driven correction, scene consistency, and update efficiency over successive scene updates.

[CV-96] Fast and Accurate Body Estimation in Pixels World Space and Scenes

链接: https://arxiv.org/abs/2611.25801
作者: Jinkun Cao
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 28 pages, 9 pages of main text. 16 figures

点击查看摘要

Abstract:We present FAB (Fast and Accurate Body estimation), a human body estimator that jointly improves inference efficiency and body-recovery fidelity. Rather than predicting only parametric mesh-model parameters, FAB directly predicts human skeleton joints and dense surface points. In images, it predicts dense 2D and 3D body points alongside a parametric Momentum Human Rig (MHR) head, then fits SMPL in one learned pass followed by two Gauss-Newton steps. Its ViT-L student, FAB-L, distilled from a fine-tuned SAM 3D Body, runs in real time for one person on an RTX 4070 Ti SUPER, from detection to a fitted body, at 4.6x the speed of SAM 3D Body’s eager-PyTorch body-only path, while improving accuracy on most camera-space benchmarks. The same network predicts heel, toe, hand, and knee contacts, which connect the body to the scene. FAB also estimates world-space motion, fuses static-scene depth around the visible person, and aligns body and scene using predicted contacts. For human-to-humanoid real-to-sim transfer, an in-model G1 head and FAB-Retarget, a motion-level retargeter that reads dense points or the fitted mesh, map motion to a Unitree G1. FAB-Retarget produces references that follow the person more closely than specialized retargeters including SOMA, GMR, UMR, and OmniRetarget, and substantially more closely than Fast SAM 3D Body retargeted with our sequence IK. Project page: this https URL.

[CV-97] GeoPillars: A Large-Kernel Pillar Detector for Single-Sweep LiDAR Under Point Sparsity and a Decomposition of Its Gain

链接: https://arxiv.org/abs/2611.25787
作者: Chaitanya Dumbre,Shubham Sarkar,Akash Mohapatra,Ghanasham Patil,Sheetal Girase
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 16 pages, 13 figures, 8 tables

点击查看摘要

Abstract:GeoPillars is a pillar-based 3D detector for single-sweep LiDAR under a hard budget of eight points per 0.4 m pillar, a budget that discards 46.6% of in-range points. It adds geometric pillar features, grouped vector attention within each pillar, a learned absolute positional prior on the bird’s-eye-view (BEV) grid, a large-kernel backbone, auxiliary surface-normal supervision and a full-resolution detection head. Trained on 123k synthetic CARLA scans and evaluated on cars, it reaches 81.33 BEV AP@0.7, 13.76 points above CenterPoint at the same input budget and 1.91 points above CenterPoint at its native, denser budget. These are in-distribution figures, because the validation split is per-frame rather than scene-disjoint. A rung-by-rung decomposition shows that the new components explain little of the gain. Raising the detection-head resolution alone recovers 10.19 of the 13.76 points, the joint change of encoder, backbone, neck and positional prior adds 3.37, and the normal-supervision branch adds 0.20, which is within run-to-run variation. With heading reversals folded out, orientation error is similar for all seven models, although GeoPillars reverses headings about half as often as a resolution-matched baseline. It is also the slowest of the seven models. Evaluated zero-shot on 6,743 real nuScenes keyframes, no model transfers, and each retains 3.7% to 12.2% of its CARLA accuracy. A collection defect left LiDAR intensity constant, so every model learned from geometry alone. Code and checkpoints are publicly available.

[CV-98] Suppressing the Shortcut Hidden Behind the Gains of VideoLLM Fine-Tuning

链接: https://arxiv.org/abs/2611.25776
作者: Youngwoo Shin,Jongsuk Kim,Junmo Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 21 pages, 6 figures

点击查看摘要

Abstract:Supervised fine-tuning (SFT) is the stage in which Video Large Language Models (VideoLLMs) advance their capability to reason over video. However, the VideoLLM also learns to answer from textual regularities of its training data, a shortcut that bypasses the video. The regularities behind the shortcut are hard to define at the data level, and the shortcut forms regardless of training epoch, data size, or backbone. The shortcut is hidden behind a rise in accuracy within the training distribution, yet it is not required for the rise and even costs SFT the gain the video offers under distribution shift. We therefore propose Video-Evidence Training via Orthogonal projection (VETO), which intervenes on the fine-tuning update where the shortcut is written rather than on the data. VETO calibrates the subspace that the text alone occupies in the activations of the base model and projects the update onto its orthogonal complement. The VideoLLM is thus left to ground its answer in the full input, with the shortcut suppressed as it forms and never defined. Across seven VideoLLMs, VETO preserves the gain of SFT on the training distribution and suppresses the shortcut. On six benchmarks under shift it gains more than SFT and answers from the full input rather than the shortcut. Code is available at this https URL.

[CV-99] Driver Gaze Target Identification via Gaze Estimation and Structured Driving Scene Understanding

链接: https://arxiv.org/abs/2611.25771
作者: Xinze Cao,Ke Wang,Deyong Guan,Yuanquan Li,Ke Zhang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 14 pages, 6 figures

点击查看摘要

Abstract:Identifying a driver’s gaze target provides object-level information for visual attention monitoring in intelligent vehicles. Gaze-zone classification and continuous attention localization provide limited information about the identity of the attended scene target. This paper proposes a modular framework that identifies object- or region-level gaze targets from synchronized in-cabin driver images and forward-view road images. The framework comprises four components: MFGaze, which combines a Vision Transformer pretrained with a masked autoencoder (MAE) and head-pose conditioning through feature-wise linear modulation (FiLM); panoptic scene understanding, which constructs a structured candidate set; polynomial ridge regression, which maps gaze features to 2D scene-image coordinates through individualized calibration; and entropy-driven target inference, which adaptively fuses boundary-distance geometry with class-level semantic priors. Experiments evaluate the framework using three public gaze datasets, a laboratory test using Cityscapes road scenes and self-collected gaze data, and the Look Both Ways (LBW) dataset. The complete framework achieves an overall Top-1 gaze target identification accuracy of 96.27% in the laboratory experiment. After sequence-specific calibration, it achieves Top-1 accuracies of 79.92% for task-relevant objects and 85.58% for the road class on LBW. MFGaze achieves the lowest mean angular errors among the compared methods, with 4.06 ^\circ , 4.23 ^\circ , and 10.52 ^\circ on MPIIFaceGaze, ETH-XGaze, and Gaze360, respectively, and demonstrates strong cross-dataset generalization. Sensitivity analyses show stable performance under hyperparameter perturbations. These results support the framework as a source of object-level gaze information for driver monitoring.

[CV-100] AdaptReg: Skip-level FiLM Adapter for Task-Specific Adaptation of a Frozen Registration Foundation Model

链接: https://arxiv.org/abs/2611.25761
作者: Shuyue Shi,Rong Yuan,Lin Tian
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 9 pages, 2 figures, 3 tables. Paper for the Learn2Reg 2026 PSMAReg challenge (PSMA-PET/CT longitudinal registration), 4th place among 16 teams on the hidden test set

点击查看摘要

Abstract:Longitudinal PSMA-PET/CT registration is a key step in assessing treatment response in prostate cancer, and it requires more than generic alignment accuracy: the deformation must preserve lesion volume, so that measured changes in metabolic tumor volume (MTV) and total lesion glycolysis (TLG) reflect treatment response rather than registration artifacts. We propose AdaptReg, which adapts a frozen registration foundation model to such task-specific requirements through a lightweight trainable adapter branch with identity-preserving initialization. Keeping uniGradICON fully frozen, we apply FiLM-style per-voxel affine modulation ( \gamma , \beta ) to its decoder skip connections; the modulation is generated from anatomical semantic features of a frozen TotalSegmentator encoder and supervised by a lesion volume-preservation loss. On the official Codabench validation set (20 pairs), AdaptReg achieves DSC 0.7585, HD95 6.33 mm, MTV/TLG errors of 1.68%/2.11%, and 0.0022% non-diffeomorphic voxels, improving on the uniGradICON + instance optimization baseline by +0.05 DSC and reducing the MTV/TLG errors by 70.9%/67.7%. On the hidden official test set (160 pairs), AdaptReg achieves CT DSC 0.816, PET DSC 0.598, and MTV/TLG errors of 4.02%/3.46%, ranking fourth among 16 teams in the final PSMAReg evaluation. Code: this https URL

[CV-101] SAM 3 Sees More Than It Detects: Promptable Concept Maps Beyond Categories

链接: https://arxiv.org/abs/2611.25747
作者: Yu-Huan Wu,Yun Liu,Liangli Zhen
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 18 pages, 5 figures, 15 tables

点击查看摘要

Abstract:Promptable segmenters answer what a thing is. Dense prediction tasks also ask class-agnostic questions, such as what stands out or hides, how light is blocked, transmitted or reflected, and where structure lies. Robots, surveillance systems and image editors build on their answers. We study promptable concept maps (PCM), which extend SAM 3’s promptable concept segmentation from noun instances to maps of concepts named by short noun phrases such as `“transparent object”, with “none” as an answer. PCM-Bench groups the standard test sets of seven such tasks by the question they ask. On it, we find that SAM 3 sees more than it detects. Its semantic path responds to many of these concepts, yet a low image-level presence score can block every detection of its instance path. We present UNVEIL, which trains a copy of SAM 3’s semantic path for these maps, with LoRA on copies of its last blocks. A separate presence decoder learns from verified negatives whether a concept occurs. With one text phrase per query, UNVEIL maps region concepts such as salient and camouflaged objects, shadows, transparent objects and mirrors. The same weights map thin structures such as edges and skeletons as well as object categories. Run next to SAM 3, it leaves SAM 3’s own outputs unchanged. On each of the seven tasks, UNVEIL reaches the highest score among the compared promptable models, including SAM 3’s official fine-tuning and pipelines built on multimodal LLMs. Code is available at this https URL.

[CV-102] owards High-Fidelity Learned Video Compression

链接: https://arxiv.org/abs/2611.25731
作者: Bumyoon Kim,Yongseong Kim,Inhyuk Jeong,Byeungwoo Jeon
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 5 figures

点击查看摘要

Abstract:Professional video workflows operate at qualities that learned video codecs do not reach. We measure how much of a professional quality range a codec covers, from the lowest quality of the standard test conditions, VVC at QP 37, to the highest 4:2:2 grade of the production codec ProRes. To compare inter-frame coding alone, every codec receives original frames as references at the intra positions. On 10-bit YCbCr 4:2:2 test sequences, four recent learned codecs cover 24 to 38% of this range. Their encoders are feed-forward networks with a fixed latent capacity, the number of latent values per sample, of at most 0.25, and removing their quantization raises the quality by at most 1 dB. We propose a learned inter-frame codec whose encoder searches, as the encoders of conventional codecs do. It searches block motion, which a learned network refines, and it searches among learned block transforms whose latent capacity ranges up to 1, selecting the partition, the skip decisions, and the quantized coefficients by a rate-distortion cost. Our codec reaches the upper end of the range on all six sequences. With all residual blocks fixed to 16x16, whose latent capacity is 0.25, the same codec covers 55% of the range, with or without quantization.

[CV-103] CGM: Chaos-Guided Graph Attention and Expert Routing for Autism Stereotypy Detection

链接: https://arxiv.org/abs/2611.25730
作者: Hafsa Aslam,Yue Li,Saba Aslam,Abdur Rasool,Qiao Pan,Muhammad Rizwan,Ovidiu Bagdasar
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Skeleton graph attention for autism video analysis weights joints by content alone, giving an erratic wrist the same standing as a still one. Stereotypy, however, is defined by how unpredictably movement unfolds rather than how much of it occurs: in our cohort, every amplitude-based kinematic descriptor falls in stereotyped clips, while windowed Sample Entropy of angular velocity alone rises (Hedges’ g = +1.05). We compute this irregularity as a per-joint, per-frame tensor that structurally drives computation. The proposed chaos-guided graph attention and mixture-of-experts model (CGM) couples two mechanisms. First, a learned pairwise term added to the graph-attention logit concentrates attention on irregular joints and re-ranks the remaining edge weight toward jointly irregular pairs; its unconstrained coefficient neither vanishes nor inverts, and zeroing the tensor realigns attention in 94% of clips. Second, a gate conditioned on the same signal routes each frame to two of four experts, whose weights track frame-level irregularity. Under subject-grouped five-fold cross-validation on 116 clips, CGM reaches 0.930 balanced accuracy, exceeding the graph-attention baseline by +0.187; the component ordering replicates on the public MMASD dataset, and on held-out Panoptic Studio sessions, CGM keeps 0.86 specificity without retraining, against at most 0.35 for chaos-free baselines. In ablation, routing without irregularity yields no reliable gain, whereas chaos modulation alone contributes +0.132. A severity index from the same tensor reaches 0.87 split-half reliability, separates the cohorts with Hedges’ g = +1.08, and stays consistent within children across therapy recordings, supporting video-only monitoring of stereotypy during intervention.

[CV-104] Whats in the box? Revealing Object Layouts Inside Opaque Boxes via Contactless Vibration Excitation and Sensing

链接: https://arxiv.org/abs/2611.25728
作者: Ethan Turok,Matan Kichler,Shai Bagon,Mark Sheinin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Can we reveal what is inside an opaque box by observing its surface vibrations? Objects inside a box alter its mechanical response, providing cues about their hidden locations. We introduce a method that uses these cues to predict a box’s 2D occupancy layout entirely remotely. A nearby speaker excites vibrations, which we measure on one visible wall using a grid of laser points and a camera measuring each point’s speckle-pattern shifts. Our model, WhisperBox, learns to encode these vibrations into a latent representation and decode it into an occupancy map of the box’s contents. The method requires no physical contact to excite or sense vibrations. We demonstrate layout recovery across various boxes and single- and multiple-object scenes, with generalization to unseen object instances, positions, orientations, and weights, as well as changes in laser placement and excitation source. Our analysis reveals that informative frequency bands need not coincide with the box’s resonances and that the learned latent space captures object placement and supports arithmetic operations. These results demonstrate the potential of remote vibration sensing to infer spatial layouts hidden from view.

[CV-105] Verifying Before Enhancing: Target Evidence Chain Verification for Infrared Small Target Detection

链接: https://arxiv.org/abs/2611.25724
作者: Jiayi Zuo,Songwei Pei,Shuhuai Wang,Yifei Wu,Shangguang Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Infrared small target detection is essential for early warning, surveillance, remote sensing, and space-based target perception. However, infrared small targets are typically tiny, weakly textured, and boundary-blurred, making them difficult to distinguish from complex backgrounds. Moreover, target-like background structures can produce salient responses across multiple scales, making response strength alone insufficient for reliable target discrimination. To address this challenge, we propose VBE-UNet, a Target Evidence Chain Verification Network that assesses candidate reliability beyond response magnitude through complementary evidence cues. Specifically, counterfactual background explanation models local target evidence, while joint evidence verification integrates cross-scale support and perturbation-based contribution verification to guide feature recomposition. The resulting reliability estimates guide selective feature refinement, promoting target-relevant responses while reducing unsupported background activations. Extensive experiments on IRSTD-1K, NUDT-SIRST, and NUAA-SIRST demonstrate that VBE-UNet achieves superior detection accuracy and lower false-alarm rates compared with state-of-the-art methods. Ablation studies and evidence reliability analysis further validate the effectiveness of the proposed verification framework.

[CV-106] DACA: Dual-Anchor Collaborative Agent for Zero-Shot 3D Visual Grounding

链接: https://arxiv.org/abs/2611.25722
作者: Jiaming Liu,Linghe Kong,Guixu Zhang,Huaxi Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:3D visual grounding (3D-VG) is a task that bridges the natural language and 3D space, serving as a foundational capability for embodied intelligence and human-computer interaction. Although recent 3D-VG agents overcome the reliance on pre-processed 3D point clouds, they essentially employ a unidirectional, pipeline-style tool-calling architecture. This architecture separates semantic understanding from geometric reconstruction, leading to error accumulation and an inability to autonomously recover from failures. To address this flaw, this paper proposes the Dual-Anchor Collaborative Agent (DACA), which reinterprets 3D-VG as a semantic-geometric dual-anchor collaborative reasoning paradigm. DACA comprises two complementary specialized tools: the semantic anchor tool holds the visual semantic features of the target and achieves efficient target tracking via optical flow and VLM local verification; the geometric anchor tool holds the 3D spatial geometric features of the target and performs global spatial reasoning via voxelization and multi-keypoint projection. The two components exchange information bidirectionally via a standardized tool call protocol, autonomously deciding on the next perception-action step to form a closed-loop optimization. Experiments on the ScanRefer and Nr3D benchmarks demonstrate that zero-shot DACA achieves an overall Acc@0.25 of 73.6% on ScanRefer and 70.4% overall accuracy on Nr3D, far surpassing the latest zero-shot 3D-VG methods. The code and models are available at this https URL

[CV-107] SteerTrace: Tracing How Target Attacks Steer World Action Models

链接: https://arxiv.org/abs/2611.25719
作者: Yukun Dai,Tianshi Wang,Fengling Li,Jingjing Li,Xin Li,Lei Zhu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 10 pages, 5 figures, 5 tables

点击查看摘要

Abstract:World action models (WAMs) predict future videos and actions with one backbone, and their imagined future is often seen as a safety signal. We introduce SteerTrace, a targeted universal attack on WAMs that also traces its effect inside the model. One \ell_\infty -bounded (4/255) perturbation per model, shared across frames, tasks, and camera views, pushes the executed action toward an attacker-chosen target. On eight WAMs from four families (FastWAM, Faster-WAM, OpenWAM and LingBot-VA) on LIBERO and RoboTwin, it drives task success close to zero and moves the attacked action toward the target. Same-budget noise and BadWAM produce little action shift in the target direction. We use the signed targeted effect to measure how much of the effect each intervention removes or reproduces. Under the standard attack, restoring future tokens removes 61-94% of the targeted effect in the five models whose action experts attend to them. When the attack is constrained to keep the predicted future close to the clean prediction, the targeted effect can become more dependent on current-frame tokens. In three of these five models, restoring current-frame tokens removes 87-99% of the effect, compared with 9-15% for future tokens. In six models, restoring just 0.36% of the FFN neurons in layers 10-29 of the video expert removes 72-96% of the targeted effect. Residual directions 2-16 hold only 11-18% of the shift energy, yet removing them cancels most of the targeted effect.

[CV-108] Interpreting DeepStack-Style Vision-Language Models: Do Input or Deeper-Layer Visual Tokens Drive Answers?

链接: https://arxiv.org/abs/2611.25713
作者: Yucheng Du,Yi Nian,Ruixue Lian,Yue Zhao,Xiyang Hu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Includes the supplementary material. Code and numerical results: this https URL

点击查看摘要

Abstract:DeepStack-style vision-language models feed visual tokens to the language model at its input, and again at deeper layers. We call these the main entry and the later entries. What do the later entries add when the model already has the image? We give the entries versions that differ in one object’s color and see which one the answer follows. On photographs, the original DeepStack follows the later entries, while Qwen3-VL-8B/32B follow the main entry. Does weakening the main entry give the later features more influence? On synthetic images, we keep this conflict and reduce color saturation at the main entry only. The unchanged later features gain a stronger effect on the relative scores of the two colors, in all four models, including Granite Vision. This holds even under mild desaturation, when accuracy without the conflict stays at 100%. To test whether later features can also correct errors, we degrade both entries and restore only the later ones. The same blur costs both Qwen models 50 percentage points of accuracy on stripe orientation. The 32B model recovers 48 points and the 8B model only 8. Accuracy with one clean entry predicted which entry each model can lose, and five of six tested predictions held on new GQA attribute questions. With the other entry zeroed, Qwen3-VL-8B loses only 0.3 points on its main entry alone, and DeepStack at most 3.45 points on either entry alone. In split inference on these questions, DeepStack can then send only its main-entry features at higher precision. This takes 49% fewer bytes than all entries at a shared precision, at similar accuracy.

[CV-109] SkillLens: Action-Guided Dense Visual Dynamics for Proficiency Estimation

链接: https://arxiv.org/abs/2611.25712
作者: Deokhyun Ahn,Hyukjin Kim,Jae-Ho Han,Bumsub Ham,Heeseung Choi,Ig-Jae Kim,Haksub Kim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 6 pages, 4 figures, 6 tables, Project page: this https URL

点击查看摘要

Abstract:Human proficiency estimation is challenging because skill is a high-level and often subjective property that is not captured by action identity alone. Existing approaches leverage multimodal observations such as ego-exo video, gaze, pose, and inertial signals, but the gap between sensor-level evidence and abstract proficiency labels remains difficult to bridge. We present SkillLens, a proficiency estimation framework that uses atomic action descriptions as temporal guidance, rather than direct semantic supervision, to localize proficiency-relevant windows for dense visual-dynamics extraction. Each window is summarized into an action-dynamics feature and scored by a high-low prototype margin against higher- and lower-proficiency executions from the same activity-action category. Fusing these ego/exo PDM cues with a global ego-exo-gaze baseline through CORAL ordinal calibration, our method achieves 0.542 accuracy on Ego-Exo4D. Our project page is available at this https URL

[CV-110] Calibrate in Context: Efficient Data-Free Quantization from Small Synthetic Pools

链接: https://arxiv.org/abs/2611.25711
作者: Haowen Hua,Sen Lin,Yaokun Huang,Zeyi Shao,Hao Zhang,Wei Sui,John See,Zeyd Boukhers,Cong Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 21 pages, 5 main figures. Supplementary material included

点击查看摘要

Abstract:We enable accurate low-bit recovery from small synthetic pools by improving both image synthesis and student preparation. A grid with lower local response error can incur greater error in the student’s block context. Functional Grid Calibration (FGC) selects activation scales and integer offsets against full-precision responses constructed from the student’s current block input. With weights fixed during calibration, FGC evaluates candidates only at their consuming operators, bringing block context to calibration at nearly local-scoring cost. For DeiT and ViT, progressive inversion updates the same images through increasingly deep teacher prefixes. With 1,000 images, the complete pipeline exceeds MimiQ’s reported accuracy in all 18 W4A4/W5A5 settings across nine backbones. With 256 images, DeiT-S and ViT-S reach 68.20% and 66.90% ImageNet top-1 accuracy at W4A4 in 26-27 minutes on one RTX 5090. The pipeline outperforms strengthened 1K-image MimiQ by 13.03 and 24.36 percentage points at lower preparation cost and the same recovery-update count. The same low-budget checkpoints retain their advantage on ImageNetV2.

[CV-111] Slide2PPT: Native PowerPoint Reconstruction with Localization-Guided Refinement

链接: https://arxiv.org/abs/2611.25708
作者: Baoliang Chen,Xinlong Bu,Lingyu Zhu,Xiangjie Sui
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Image-to-slide reconstruction must recover both visual appearance and native object structure for subsequent editing. Yet visually faithful charts and tables may lack editable series and cells, and direct code generation couples content prediction to rendering implementation. We present Slide2PPT, a model-agnostic, training-free framework that equips both open- and closed-source vision-language models with native PowerPoint reconstruction capabilities. Through a unified typed interface, the model predicts object records while a fixed renderer materializes their native properties, enabling integration with different backbones without task-specific training or architectural changes. Localization-guided refinement then connects visual discrepancies to object records for correction. Regional updates preserve unselected records, while whole-slide updates address cross-region relationships. Record refinement and targeted object editing support structural and attribute-level changes, with rendered feedback governing iterative acceptance. To evaluate this framework across diverse reconstruction demands, we introduce Slide2PPT-Bench, covering controlled elements, real slides, and scientific figures. Evaluation separately measures file production, rendered agreement, and native-object availability. Experiments across open- and closed-source backbones demonstrate substantially more reliable PPTX generation than direct code generation, improved layout and text recovery on scientific figures, and native chart and table interfaces beyond geometric primitives.

[CV-112] LIFT: Lifting Language-Controlled Image Fusion to Free-Viewpoint 3D Scenes

链接: https://arxiv.org/abs/2611.25696
作者: Jingning Tian,Hao Zhang,Yumin Huang,Zizhuo Li,Xingyu Jiang,Erdun Zhao,Jiayi Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Multimodal image fusion traditionally generates a single fused image from paired observations at a specific viewpoint, limiting comprehensive scene perception and adaptability to diverse perceptual objectives. To overcome these limitations, we present LIFT, a novel framework for our proposed language-controlled 3D image fusion (L3D-IF) paradigm that extends multimodal image fusion from 2D images to free-viewpoint 3D scenes, enabling language-guided fusion tailored to diverse perceptual objectives. First, LIFT formulates Semantic Modality Binding (SMB), which interprets language instructions to identify semantic targets and their associated modality preferences and grounds these targets in corresponding regions of a shared Gaussian scene, establishing persistent semantic-to-spatial correspondences for consistent fusion control across viewpoints. Second, Language-Conditioned Gaussian Appearance (LCGA) is developed to achieve scene-wide multimodal appearance modeling by combining adaptive multimodal blending with language- and viewpoint-conditioned appearance refinement, enabling spatially adaptive realization of diverse fusion appearances while preserving source fidelity at pure-modality endpoints. Finally, we construct a Multi-View Language-Supervision (MVLS) dataset that associates language instructions and calibrated viewpoints with modulated fused images, providing effective supervision for learning instruction-dependent fusion from finite multi-view observations. Extensive experiments demonstrate that LIFT achieves coherent multimodal fusion across continuous viewpoints, accurate semantic target localization, and consistent regional fusion control under diverse language instructions.

[CV-113] D3-Cache: Accelerating Diffusion Models with Dynamic Mode Decomposition

链接: https://arxiv.org/abs/2611.25689
作者: Yanbo Li,An Xiang,Xiaoxin Hu,Feng Wang,Cheng-zhong Xu,Kejiang Ye,Xitong Gao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Diffusion Transformers generate high-quality images and videos, but every sampling step requires a full network evaluation. Training-free caching reduces this cost by reusing or forecasting network outputs across steps, but common approaches use per-model calibrated skip criteria or maintain per-layer prediction histories. We present D3-Cache, which applies Dynamic Mode Decomposition (DMD) to Diffusion sampling: it fits DMD to a short history of exact velocities and forecasts only the velocity that the scheduler consumes. A horizon-aware gate decides when to skip: whenever the network is evaluated, its output is compared with the forecast for that step, and the larger the error, the sooner the network is evaluated again, up to a cap on consecutive skipped steps. D3-Cache requires no training or per-model offline calibration curve and adds negligible memory. It shares one history length and risk tolerance across the four evaluated models. On four state-of-the-art image and video models, D3-Cache achieves speedups of 2.6-3.8x with high fidelity to the base model output.

[CV-114] Beyond Entity Recognition: Vision-Centric Knowledge Visual Question Answering EACL2027

链接: https://arxiv.org/abs/2611.25680
作者: Jiatong Ma,Yuchen Liu,Longteng Guo,Yichen Yuan,Yuze Wu,Dongze Hao,Ru Zhen,Yanhao Zhang,Haonan Lu,Jing Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: Accepted to EACL2027(Main)

点击查看摘要

Abstract:We propose VC-VQA, a large-scale benchmark for vision-centric knowledge reasoning, aiming to systematically evaluate the capabilities of Multimodal Large Language Models (MLLMs) in fine-grained visual knowledge acquisition and complex multi-stage reasoning. Unlike existing benchmarks that primarily treat visual information as entity recognition and rely on text-based external knowledge, VC-VQA explicitly models the role of visual information by considering both its source (internal vs. external images) and stage of involvement (pre-retrieval vs. post-retrieval). Specifically, VC-VQA requires models to retrieve external visual evidence, identify visual clues before retrieval, and localize, verify, or match retrieved knowledge against the internal image. We further construct a controlled multimodal knowledge base with traceable textual and visual evidence for each sample. We evaluate 21 representative MLLMs under four settings: no external knowledge, with provided evidence, knowledge base retrieval, and open-web search. Experimental results lead to three key findings: current models still struggle with vision-centric knowledge reasoning; fine-grained visual knowledge acquisition remains the primary bottleneck; and agentic retrieval highlights the value of multi-step planning and cross-modal knowledge integration. Overall, VC-VQA provides a more challenging benchmark for evaluating vision-centric knowledge reasoning in MLLMs. Our code and dataset are available at this https URL.

[CV-115] Minimal Memory for Correctable Visual Continuation

链接: https://arxiv.org/abs/2611.25675
作者: Xiyang Zhang,Yuanhe Tian,Hongzhi Wang,Yan Song
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Given past images and known actions, visual continuation generates future frames from one evolving scene state. Keeping this state revisable matters when an object reappears or a later view changes an earlier interpretation. Once past images cannot be replayed, a summary sufficient for current rendering can lose candidates needed after correction. We establish matching sparse and dense bounds on the bits needed for later correction under stable observations and bounded disagreement. For static streams at the stated smoothness and margin scales, correction incurs at least logarithmic overhead in one dimension and polynomial overhead in higher dimensions. Coherent Archive Memory (CAM) stores geometric summaries with exact updates under clipped expansion and coordinate assignments, then uses historical existence constraints to select one state for all requested frames. Exhaustive photographic tests show strict worst-truth image-risk gains from joint witnesses, while controlled neural continuations recover positions hidden in the current view.

[CV-116] Efficient Latent Reasoning for Vision-Language Model via Dynamics-Aware Quantization and Distillation

链接: https://arxiv.org/abs/2611.25670
作者: Yanjing Li,Sheng Xu,Baochang Zhang,Xianbin Cao,Yawei Li
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Efficient chain-of-thought (CoT) reasoning in vision-language models (VLMs) requires reducing both intermediate reasoning steps and their execution cost. Latent CoT methods compress textual rationales into continuous thoughts, yet their full-precision (FP) models retain substantial storage and computation costs. Lowering numerical precision could reduce these remaining costs, yet direct quantization severely degrades reasoning accuracy, particularly with 4-bit weights and activations. Operator and perturbation analyses reveal key sensitivity and direction-dependent predictive effects of continuous-thought errors, motivating preservation of information retrieval and subsequent reasoning beyond reconstruction accuracy. We therefore introduce QL-VLM, a dynamics-aware Quantization framework for efficient Latent reasoning in Vision-Language Models (VLMs). First, Dynamics-Guided Quantization (DGQ) uses nonuniform codebooks that preserve rotary-key structure, mitigating quantization-induced distortion in vision-language information retrieval. Second, Dynamics-Preserving Distillation (DPD) aligns continuous thoughts, latent-query readouts, and predictions with a full-precision teacher, maintaining predictive consistency through subsequent low-precision reasoning. On ScienceQA-IMG with Qwen2-VL-2B-Instruct, 4-bit QL-VLM achieves 90.43% accuracy and improves baseline by 40.33%, with 16.64x fewer bit-operations and 6.40x lower measured latency than explicit CoT.

[CV-117] CauSyncVLA: Causal Latent World Modeling for Asynchronous Robot Control

链接: https://arxiv.org/abs/2611.25667
作者: Xu Pan,Chubin Zhang,Xingrui Yu,Runmin Lan,Zihao Wang,Wenkai Guo,Bofang Jia,Bo An,Ziwei Wang
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Vision-Language-Action (VLA) policies must react while visual inference is still running. Overlapping computation and execution reduces waiting, yet leaves perceptual memory stale and successive action chunks disconnected. We present CauSyncVLA, an asynchronous dual-system controller that aligns perceptual-memory prediction and trajectory correction with actual robot execution. Rather than forecasting pixels or physical states, a Causal Latent World Model predicts current perceptual memory from delayed visual anchors and executed state-action history, supervised through action-sensitive distillation to retain the frozen expert’s control response. Conditioned on this updated memory, Predictive Flow Warm-Starting (PFWS) aligns the unexecuted action horizon for single-step flow correction without extra training. On LIBERO-10, CauSyncVLA reaches 93.8% success at 5.04 \times inference throughput, improving closed-loop robustness and robot motion smoothness.

[CV-118] Extending Rademacher Complexity-Based Generalization Theory to Domain Generalization

链接: https://arxiv.org/abs/2611.25650
作者: Xi Wang,Liang Bai,Xian Yang,Richard Yi Da Xu
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Domain generalization (DG) aims to train models on multiple source domains with distinct distributions to ensure robust performance on unseen target domains. In doing so, it elevates the concept of generalization from the traditional sample level to the distribution level. Existing theoretical frameworks introduce a domain space endowed with a meta-distribution to extend classical generalization analysis to this new setting. However, these approaches typically treat the domain space as devoid of geometry. In reality, the domain space is an inherently infinite-dimensional, nonlinear abstract space composed of probability distributions, and its complex geometric properties fundamentally determine generalization behavior. Consequently, existing geometry-free theories of domain generalization have a fundamental limitation. They lack a formal characterization of what distinguishes distribution-level generalization from sample-level generalization, leaving the inherent difficulty of generalizing across domains largely unexplained. To bridge this gap, we formalize the domain space as the Wasserstein space over the representation space. Building on the geometric properties of this space (specifically the \mathrmCD(K,\infty) curvature theory) we derive a novel, geometry-dependent DG error bound. By introducing an additional curvature-dependent confidence term, this bound formally disentangles the two levels of generalization effects and reveals the essential nature that domain generalization is affected by both the invariance and the curvature of the representation space. Finally, we experimentally validate the effect of curvature on domain generalization.

[CV-119] RACE: Attribute Execution for Sequential Image Editing

链接: https://arxiv.org/abs/2611.25648
作者: Xiyang Zhang,Yuanhe Tian,Hongzhi Wang,Yan Song
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Given an image and successive user instructions, sequential image editing should return an updated image that satisfies each new request and preserves earlier requirements that remain active. This lets users develop a result without repeatedly restating or repairing their decisions. The difficulty is that a retrieved image can contain a valid color on an obsolete surface, while the current image may lack the newly requested material. Neither reference selection nor a precise object mask alone specifies how to reconcile attributes occupying the same region. We present TRACE, which separates the continuing attribute from its visual carrier and prioritizes local appearance operations under complete-request checks. Active-goal binding identifies what should persist, carrier selection supplies suitable current or newly generated pixels, and tonal calibration reduces tinting near highlights and shadows. Experiments across frozen editors, independent visual readers, new source images, and complete sequential benchmarks support joint color-material recovery and show how local execution preserves the surrounding scene.

[CV-120] MambaLIFT: Reference-Based Mechanism-Aware Attribution for VMamba

链接: https://arxiv.org/abs/2611.25644
作者: Jintu Kakati,Niladri B. Puhan
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Image and Video Processing (eess.IV)
备注:

点击查看摘要

Abstract:Attributing predictions of visual state-space models remains challenging because input-dependent selective dynamics couple feature selection, recurrent state propagation, and shared computational pathways. We introduce MambaLIFT, a reference-based, multiplier-first attribution framework for VMamba that propagates contributions through its computational graph using interaction-aware rules for selective-scan injection, recurrent memory, LayerNorm, and shared pathways. MambaLIFT decomposes the reference-relative class-logit difference, (f_c(x)-f_c(x^0)), by spatial position, VSS block, scan direction, and computational mechanism. Experiments on ImageNet-1K evaluate VMamba-T and VMamba-B, with hyperparameters selected on a calibration set and faithfulness assessed on a disjoint held-out evaluation set. MambaLIFT achieves numerical conservation to floating-point precision across audited attribution interfaces in both configurations. Under mean-colour pixel flipping, it achieves significantly higher mean faithfulness scores than MambaLRP on both configurations and the highest scores among the evaluated methods, including Integrated Gradients with matched references. Under blurred replacement, MambaLIFT is non-inferior to MambaLRP on VMamba-T and equivalent on VMamba-B, indicating that its advantage depends on the replacement protocol. Ablations on VMamba-T show that interaction allocation, temporal propagation, and pathway composition affect attribution maps. Mechanism-level analyses further show that exact relevance conservation does not imply additive effects under finite interventions: the mechanism decomposition accounts for the reference-relative logit difference but does not estimate intervention effects or establish causal validity.

[CV-121] Erasing the Wrong Thing: Conditional Attribute Leakage and Concept Erasure in Shared Foundation-Model Embeddings

链接: https://arxiv.org/abs/2611.25640
作者: Haijun He,Fei Li,Donghong Ji,Junjun He
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Medical-image embeddings are increasingly shared in place of raw images for authorized downstream tasks, yet they may reveal sensitive attributes. Standard concept-erasure methods target population-level guardedness. However, we prove that, when binary task and sensitive labels are associated, global and task-conditional linear guardedness cannot both hold while a nonzero linear task signal is preserved, so global erasure must either eliminate that signal or retain task-conditional linear leakage. We further show how global erasure affects linear readers’ discrimination between positive and negative samples with different sensitive attributes, deriving an approximate closed-form prediction of its utility cost from pre-erasure statistics. Motivated by this conflict, we introduce Task-Conditional LEACE (TC-LEACE), a closed-form variant of LEACE that enforces linear guardedness within each task stratum of its fitting data. We evaluate four vision foundation models on 85 CelebA attribute pairs spanning a broad range of associations. After LEACE, conditional leakage persists across all four encoders, the discrepancy between marginal and conditional audits increases with association strength, and predicted utility costs track observed losses. We then examine hospital chest radiographs using three foundation models, including two biomedical models, in which chest-tube status is strongly associated with diagnosis. Although a marginal attack on LEACE-erased embeddings underperforms a diagnosis-only reference, chest-tube status remains detectable within diagnosis strata. TC-LEACE recovers most of LEACE’s linear task-utility loss while keeping task-conditional leakage within a prespecified margin of LEACE’s.

[CV-122] LG-FiLM: Local-Global Mixture-of-Experts Modulation for RAW-to-RGB Reconstruction

链接: https://arxiv.org/abs/2611.25639
作者: Fei Li,Xilei Cao,Shichang Zhai,Kun Zhan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:An image signal processor does more than reconstruct an image. It also has to decide how strongly to denoise, how to balance color, and which tone curve maps the scene onto the display, and these settings depend on the scene itself, the curve that suits a backlit sky is not the one that suits an indoor portrait. Most neural ISPs learn a single fixed transformation and apply it to every scene. We instead make the adaptation an explicit part of the network. Our model, LG FiLM, keeps one shared encoder decoder and places a small mixture of affine experts at its bottleneck; the experts scales, shifts, and mixing weights are predicted from local neighborhood evidence together with a global scene descriptor. Reconstruction stays in a single shared pathway and only the processing mode is switched. We evaluate LG FiLM on three datasets, Fujifilm UltraISP, Zurich RAW to RGB, and a COCO derived benchmark built with an inverse ISP. The model reaches the best reconstruction fidelity of the compared entries on the primary benchmark and transfers to the other two.

[CV-123] AesDistill: Distilling Generative Aesthetics for Image Color Enhancement

链接: https://arxiv.org/abs/2611.25635
作者: Chuanzhi Xu,Chengkun Yue,Huiqi Liang,Haoyang Li,Zihan Deng,Qingwen Zeng,Pengfei Ye,Weidong Cai
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Image color enhancement aims to improve visual appeal, but previous methods often learn to reproduce ground-truth expert targets. These targets reflect the preferences of a few experts and cover a limited range of aesthetic choices. Matching them does not ensure stronger aesthetic quality across different tastes and scenes. By contrast, large generative models learn rich visual knowledge from large image collections and often produce vivid, appealing image edits. However, direct generation can change the source image’s semantic structure along with its colors. In this paper, we introduce AesDistill, a framework for distilling generative aesthetics into image color enhancement. We use 4 leading open-source large generative models to produce 64,644 edited images as teacher targets at a cost of about 3,360 H100 GPU-hours, forming the AesDistill Dataset. AesDistill also includes AesDistill Student, a 1.13M-parameter model that learns teacher color changes through confidence-weighted supervision while preserving the source image’s semantic structure. We evaluate AesDistill with 12 aesthetic and 4 image-quality assessment models. Every teacher’s targets rank above the FiveK and PPR10K expert target means in the joint assessment. Across both datasets, AesDistill Student outperforms all 20 baseline methods and ranks directly behind the 4 teachers. These results demonstrate that distilling aesthetics from generative models enables a compact student to outperform methods trained on expert targets.

[CV-124] What One Sample Cannot See: Why Fewer Denoising Steps Can Score Higher on Physics Benchmarks

链接: https://arxiv.org/abs/2611.25632
作者: Yutao Li,Chen Wang
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:When a storm can head north or south, a forecast down the middle may score better under squared error than either plausible path. The same paradox appears when video generators are tested as world models: fewer denoising steps can raise physics scores even as videos blur or stop moving early. Why? Physics-IQ compares each generated video with one recording, although a scene can have multiple valid futures. An exact decomposition shows its scores can reward reduced variation among generated futures, even when their distribution becomes less accurate. We introduce Kernel Physics-IQ to remove this incentive in expectation, needing only two independent samples per condition. Two-step sampling lowers Kernel Physics-IQ in all five image-to-video models. For CogVideoX, the official score rises by 0.055 while ours falls by 0.131. Changing the sampling schedule reveals further differences official scores miss. In controlled physical systems with known outcome probabilities, our metric favors the true distribution over collapsed predictions. In short, a world model should be judged not only by how closely one video matches what happened, but by how faithfully its predictions represent what could have happened.

[CV-125] RelGeo: Pair-Conditioned Relational Calibration for Frozen Diffusion Correspondence

链接: https://arxiv.org/abs/2611.25628
作者: Zixuan Dai,Zhiliang Zhang,Daosong Hu,Tang Zheng,Xinkai Liu,Tianyu Lei,Wei Zhan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages, 13 figures, including supplementary material

点击查看摘要

Abstract:Frozen diffusion descriptors retrieve strong semantic correspondence candidates, yet cosine nearest-neighbor matching can select incorrect locations despite valid high-ranking alternatives. Across DIFT-SD2.1 and DiTF-FLUX on SPair-71k, AP-10K, and SOCO (CC, SOC, Cross-SOC), 41-69% of Top-1 errors retain a PCK@0.10-valid match among the original Top-5 spatial peaks, revealing a persistent candidate-selection bottleneck. We find that descriptor rankings and current-pair relational evidence can favor different candidates: reciprocal diffusion Query/Key associations and pretrained RoMa transport exhibit distinct, partly complementary preferences. To incorporate this evidence into matching decisions, we introduce RelGeo, a Pair-Conditioned Relational Calibration framework expressing heterogeneous relations as shared source-target coordinates. Spectral Relational Coordinates (SRC) encode reciprocal diffusion affinities, while Warp Identity Coordinates (WIC) encode reliability-weighted bidirectional transport endpoints. Their inner products augment frozen descriptor similarity while preserving full-domain cosine nearest-neighbor matching without additional semantic-correspondence training. RelGeo improves all ten evaluations by an average of 5.51 PCK@0.10 points; fixed-candidate reranking and pair-misalignment controls corroborate improved selection and dependence on current-pair relational alignment.

[CV-126] Physical Laws without Physics Loss? Probing Learned Dynamics in Video World Models through Forced Response

链接: https://arxiv.org/abs/2611.25626
作者: Yutao Li,Chen Wang,Wangmeng Zuo,Hui Li
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Do video world models learn physical laws without a physics loss, or merely imitate motion? Overall motion and appearance scores cannot settle it. We investigate this through nonlinear forced response: an excitation shakes the support of a suspended object. We introduce dual visual-numeric conditioning and adapt six image-to-video backbones to predict this motion using only the flow-matching loss. Wan predicts motion accurately and generalizes to unseen excitations, objects, poses, and scenes. Counterfactual tests suggest capabilities beyond replaying training videos. But what laws have they learned? We design three tiers of tests using Taylor expansions of the dynamics and geometry. All six reproduce the linear response, including resonance; five capture higher-order dynamics, including nonlinear softening. Remarkably, softening is barely present in training: only 0.74% of the swing’s squared displacement. Extensions to a double pendulum and a rocking block show sustained dynamics are learned more faithfully than brief impacts. Yet learned dynamics may fail to appear when state estimation or propagation goes wrong. We turn these tests and findings into strategies to improve video training and generation.

[CV-127] DAWN: Robust Neural Blind Deconvolution via Alternating Wiener Guidance

链接: https://arxiv.org/abs/2611.25621
作者: Yida Pan,Yixuan Jia,Chaoyan Huang,Qing Qu,Ismail Alkhouri
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Blind image deconvolution (BID) is an ill-posed inverse problem that aims to jointly recover a sharp image and blur kernel from a single degraded observation. Under nontrivial and high noise, which is common in photography and scientific imaging such as microscopy, BID remains particularly challenging, as existing methods suffer from noise overfitting and inaccurate kernel estimation. Existing training-data-free learning methods for BID, such as those based on the deep image prior (DIP), often utilize external guidance by including model-based estimates as additional loss terms, which underexploits the information these estimates carry especially under nontrivial noise. To address these challenges, we propose DAWN, a training-data-free DIP framework that alternates between injecting model-based estimates directly into the model input and refining this imperfect estimate with the network’s own prior. To provide more reliable kernel estimates for these injections, we apply autoencoding regularization to both the image and kernel networks. On natural and scientific images under varying levels of Gaussian and Poisson noise, DAWN outperforms state-of-the-art methods in both image and kernel quality (achieving up to +4.5 dB over the next best baseline on the Levin dataset at \sigma = 0.1 ). Without any training images, DAWN remains highly competitive with a state-of-the-art diffusion model on FFHQ, despite the latter being trained on a massive number of in-distribution images.

[CV-128] CaT-Movie: Balancing Global Planning and Local Rendering for Long Multi-Shot Video Generation

链接: https://arxiv.org/abs/2611.25620
作者: Jinpeng Lu,Haoyuan Shi,Yuanshen Guan,Jiyuan Xia,Bin Wu,Yinda Chen,Zhiwei Xiong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 31 pages

点击查看摘要

Abstract:Current multi-shot video generators render only limited durations per pass, requiring long movies to be stitched from local segments across which narrative progression and visual identity drift. To synthesize long-horizon movies with Coherence across Time, we present CaT-Movie, a multi-shot generation system consisting of a global story planner and a local window renderer. The planner expands user requests via a large language model to preserve, infer, and complete story details, while deterministic code maintains the resulting storyboard across shots to compile self-contained shot prompts and index reference frames from generated history. To execute this guidance, the renderer applies reference-routed sparse self-attention and shot-wise cross-attention to route identities from the indexed frames while isolating per-shot textual conditions. Given the scarcity of high-quality open-source multi-shot data, we construct a dataset of over 350K film shots annotated with semantic state transitions, entity grounding, and multi-granularity requests. We further build CaT-MBench, a benchmark on held-out films with the same annotations. It gives every baseline its native format and counts cross-shot consistency only when a cut opens a new camera setup, preventing repeated framing from passing as consistency. On CaT-MBench, CaT-Movie attains the highest average score and the most consistent characters among all methods. The project page is available at this https URL.

[CV-129] LoopedWAM: Looped World Action Models for Elastic Inference

链接: https://arxiv.org/abs/2611.25595
作者: Anh Dao,Xinxin Liu,Hang Yu,Jeffrey A Chan Santiago,Yuzhang Shang
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:

点击查看摘要

Abstract:World action models (WAMs) couple generative video modeling with robotic action prediction but typically require large video and action transformers, limiting deployment efficiency. We present LoopedWAM, a compact WAM that recurrently applies a shared six-block core in each expert, executing 30 transformer layers while storing only 12 blocks per expert. To account for their asymmetric inference costs, the video and action experts can operate at different depths: visual features are computed once per query, whereas action computation repeats at every sampling step. We further propose elastic self-distillation, which trains a single elastic checkpoint, LoopedWAM-E, across action-depth and sampling-step budgets. With only 0.58B policy parameters and no robot pretraining, LoopedWAM achieves 96.1% success on LIBERO, 91.9% on RoboTwin 2.0, and 52.2% on RoboCasaGR1, while using 7.4\times less peak GPU memory than Fast-WAM under matched inference measurements. On LIBERO, LoopedWAM-E achieves 96.0% success in 27.6 ms or 95.2% in 24.7 ms on an H100, providing multiple accuracy-latency operating points from a single checkpoint; with two sampling steps and four action loops, it stays within 0.5 points of LoopedWAM on all three benchmarks. Ablations demonstrate that weight-shared depth preserves competitive performance, while aggressive action-depth reduction requires explicit budget-aware training. On a real AgileX PiPER arm, LoopedWAM succeeds in 71.3% of 80 trials across four tasks.

[CV-130] Bradley-Terry Training Can Distort Outcome Probabilities of Initially Accurate Language World Models

链接: https://arxiv.org/abs/2611.25594
作者: Xinze Shi,Litian Zhang,Zhenjie Su,Zelong Yu,Yuting Zhang,Qihua Feng,Zifeng Kang,Chaozhuo Li,Zhengxuan Hu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 6 figures, 1 table

点击查看摘要

Abstract:Comparative feedback, in which a judge says which of two candidate outcomes is closer to what happened, is a practical way to train language world models that assign probabilities to an action’s outcomes. We show that, fitted with a Bradley-Terry objective, it need not teach the true outcome probabilities and can distort accurate ones. On a simulated robot push with enumerated outcome distributions, this objective takes a model that is accurate on unseen layouts (hazard error 2.4 percentage points, pp) and moves 33 to 43 pp of its probability onto outcomes that none of a condition’s enumerated branches produces. The cause is the objective, not the comparisons: its optimum, computed before training, predicts where the model ends up, and a likelihood that knows the judge’s rule learns from the same comparisons without the shift. The shift is robust: it recurs on unseen conditions, under five candidate laws and a three-fold denser enumeration, and about 29 pp remain without the physically impossible events. Its cost is conditional: with those events it inflates hazard probabilities, and a threshold inherited from the pre-feedback model costs about 3 pp of utility per decision on trained and 1 to 2 pp on unseen conditions, through false rejects; without them the utility loss averages below 1 pp over six seeds, though a small regret in action choice remains. Audits of readout, response and decision trace the effect from probabilities to actions.

[CV-131] Characterizing Age-Conditioned Affective Reasoning in Vision-Language Models

链接: https://arxiv.org/abs/2611.25593
作者: Shahriyar Zaman Ridoy,Azmine Toushik Wasi,Lin Gu,Ruogu Fang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Preprint. Under review

点击查看摘要

Abstract:An age prompt can change how a vision-language model (VLM) rates the emotion evoked by an image. Does the change predict how people of different ages rate that same image? We evaluate 17 VLMs on 900 OASIS images using young-, middle-, and older-adult prompts and estimate image-specific age differences from ratings by 822 participants. Without an age cue, models closely track average human valence (median Pearson r=0.891 ). Under an older-adult prompt, 16 of 17 models assign lower valence on average than under a young-adult prompt. Yet after adjusting for image pleasantness, correlations between model and human age effects range from -0.030 to 0.093 , against an estimated ceiling of 0.595 . The discrepancy is clearest on 54 sensitive images rated as pleasant. Across models, the older prompt reduces valence by 0.845 points on average, while older participants rate these images 0.443 points higher than younger participants. Providing human ratings from other images improves agreement on the direction of age differences, although image-specific fidelity remains modest on 600 held-out images ( r_\mathrmpart=0.105 ). These results show why evaluations of demographic personalization in foundation models need human comparisons at the level of individual stimuli. Project page available at this https URL

[CV-132] QFT-Based Hybrid PointNet-Quantum Neural Networks for LiDAR Point-Cloud Classification

链接: https://arxiv.org/abs/2611.25578
作者: Anurag Kulkarni,Nouhaila Innan,Muhammad Shafique
类目: Computer Vision and Pattern Recognition (cs.CV); Quantum Physics (quant-ph)
备注: 14 pages, 3 figures, 6 tables

点击查看摘要

Abstract:Hybrid quantum-classical learning architectures provide a promising framework for integrating quantum feature transformations into complex data-processing tasks. Among these transformations, the Quantum Fourier Transform (QFT) offers a structured approach to feature processing, motivating an investigation of how its circuit configuration influences predictive performance, quantum resource requirements, and noise robustness. We present a hybrid PointNet-quantum neural network architecture for LiDAR point-cloud classification, combining classical feature extraction, amplitude encoding, and a variational quantum circuit. Five Fourier-processing configurations are systematically evaluated: No-QFT, Full-QFT-After, QFT-no-swap, IQFT-After, and Trainable-QFT. Experiments on KITTI, nuScenes, and PandaSet across three random seeds assess classification performance, class-level behavior, circuit complexity, and sensitivity to the IBM FakeManilaV2 noise model. Under ideal simulation, IQFT-After achieves the highest weighted F1-score on KITTI (0.857 \pm 0.033), compared with 0.838 \pm 0.023 for No-QFT. Removing the final SWAP gates in the QFT circuit reduces circuit complexity while preserving the Pauli-Z measurement features up to a fixed permutation that can be absorbed by the classification head. Under quantum noise, the circuit ranking changes, with IQFT-After exhibiting greater performance degradation and No-QFT and Trainable-QFT showing stronger robustness. Per-class analysis further identifies dataset-dependent differences in minority-class performance. These results demonstrate that Fourier-processing configuration influences both classification behavior and noise sensitivity, highlighting the role of circuit structure and hardware-aware evaluation in hybrid quantum learning for three-dimensional point-cloud data.

[CV-133] rajectory-Calibrated Merging of Embodied Policies

链接: https://arxiv.org/abs/2611.25575
作者: Wenshuo Chen,Honglei Jin,Le Yang,Jianfei Song,Shaofeng Liang,Lei Wang,Junze Zhu,Kailong Wang,Zhengxi Jin,Haozhe Jia,Honglin Chu,Songning Lai,Xiucheng Wang,Yutao Yue
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Robot foundation models are increasingly adapted into task specialists to expand their embodied capabilities. Yet these capabilities remain fragmented across separate policies, and existing model-merging methods do not reliably preserve them during closed-loop execution: parameter-space alignment provides only an indirect proxy for behavior, while regression may fail to preserve behavior at states not covered by its calibration inputs. We present Trajectory-Calibrated Regression (TCR), an execution-aware, gradient-free framework that consolidates robot specialists into a single policy. Across two complementary timescales, TCR collects calibration inputs from expert rollouts spanning different stages of task execution and from intermediate states within iterative action generation. To account for representation shifts introduced by merging, TCR recomputes each block’s input through the already merged blocks. It then uses each expert’s computation within the block to form module inputs and fits merged weights to match the experts’ linear outputs. A two-pass calibration procedure further refines the merged policy using separate rollout caches while retaining the frozen specialists as teachers. The resulting model preserves the original inference graph and requires neither expert routing nor additional gradient training. On LIBERO, TCR merges four \pi_0.5 specialists and achieves 77.83% success across 1,200 evaluation episodes, outperforming FeatCal, RegMean, and direct weight averaging by 9.50, 24.17, and 33.00 percentage points, respectively.

[CV-134] VidToC: Rethinking Video Token Compression for Multimodal Large Language Models

链接: https://arxiv.org/abs/2611.25574
作者: Zihan Xiao(1),Boshen Xu(1),Pei Fu(2),Jianzhong Ju(2),Zhenbo Luo(2),Jian Luan(2),Qin Jin(1) ((1) AIM3 Lab, Renmin University of China, (2) MiLM Plus, Xiaomi Inc.)
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 24 pages, 7 figures

点击查看摘要

Abstract:Extending multimodal large language models (MLLMs) from images to videos faces substantial memory and latency challenges due to the massive number of visual tokens. These challenges have driven the development of numerous visual token compression methods and more efficient LLM architectures. However, existing compression methods are often evaluated with different training setups, tasks, and models, making fair comparison difficult. Meanwhile, the growing use of hybrid Mamba-Transformer architectures raises questions about whether compression strategies transfer across backbones. We present VidToC, a controlled evaluation framework for representative compression algorithms, LLM backbones, and time encoding strategies, covering training-free and fine-tuned settings on three complementary video understanding tasks. Our results reveal task-dependent compression behavior and characterize the effects of compression placement, time encoding, and backbone choice. These findings provide empirical guidance for designing and selecting video token compression strategies across tasks and training settings.

[CV-135] Online Mistake Detection in Instruction-Conditioned Procedural Video

链接: https://arxiv.org/abs/2611.25567
作者: Amit Feldman,Ilan Shimshoni,Rotem Dror
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:An assistant monitoring a person performing a procedure is most useful when it can detect mistakes while there is still time to recover. This requires judging actions against task instructions as the video unfolds, identifying the attempted step, locating it in time, and assessing its execution. Existing online detectors do not use the instruction text, while instruction-conditioned methods operate offline. In this paper, we study online, instruction-conditioned mistake detection in egocentric videos. Using only observed video, task instructions, and prior predictions, our model identifies and localizes completed action attempts and classifies them as correct, modified, slip, corrections, or additions; omitted steps are determined after the video ends. We compare a modular pipeline, with separate step identification, temporal localization, and execution verification, to a joint model that predicts all outputs in a single step. To support supervised learning from mistakes, we repartition EgoPER to include erroneous videos in all splits. The approaches achieve similar step-segmentation F1@0.5 (93.6% and 92.1%), while the joint model achieves higher binary mistake-detection F1@0.5 (84.1% vs. 80.7%), with its largest gain on slips (56.4% vs. 37.5%). Component-level analysis suggests that error propagation between stages contributes to the modular approach’s lower detection score. On test videos containing mistakes, both designs outperform released AEM and ESTANet checkpoints under our evaluation protocol.

[CV-136] UniReFP: Robust Unified Fingerprinting for Vision Models against Cross-Task Repurposing Attacks NEURIPS2026

链接: https://arxiv.org/abs/2611.25555
作者: Ziye Geng,Guang Yang,Yihang Chen,Lingfeng Yao,Xing Shi,Miao Pan,Xin Fu,Changqing Luo
类目: Computer Vision and Pattern Recognition (cs.CV); Cryptography and Security (cs.CR)
备注: NeurIPS 2026 Spotlight

点击查看摘要

Abstract:We identify, for the first time, a new model modification attack–the cross-task model repurposing attack–that can render fingerprints generated by existing model fingerprinting approaches inapplicable. To address this challenge, we propose UniReFP–a novel unified model fingerprinting framework applicable to diverse vision models across different tasks–that utilizes ownership evidence in a shared class-level semantic space rather than original task-specific output space, enabling effective ownership verification when a protected model is repurposed across tasks with heterogeneous output formats. Specifically, UniReFP constructs a surrogate classifier by repurposing the protected model, enabling fingerprints to be generated independently of the original task-dependent output space of the protected model. Moreover, it optimizes grouped fingerprints to encode ownership evidence: fingerprints within the same group are encouraged to induce consistent semantic responses and intermediate representations on the surrogate classifier, while producing dispersed responses on independently trained reference classifiers. During verification, UniReFP uses a task-aware semantic projection mechanism to map heterogeneous suspect-model outputs into unified semantic-presence vectors, and determines ownership by measuring inner-group consistency. Extensive experimental results validate the effectiveness of UniReFP, consistently showing superior performance under cross-task repurposing attacks, commonly studied model modification attacks, and even their combinations.

[CV-137] woPass: Draft-Guided Few-Step Video Prediction for Robotic Manipulation

链接: https://arxiv.org/abs/2611.25553
作者: Yijun Liu,Shijia Ge,Tongtong Feng,Xin Wang,Yuan Meng,Yuwei Liu,Zhi Wang,Wenwu Zhu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Action-conditioned video prediction lets robots anticipate the visual consequences of their actions for planning and policy evaluation. Existing predictors denoise each future from noise without using their own estimate of it, which limits the quality of the predicted future, in particular of the manipulated objects, and requires many network calls. We present TwoPass, a self-conditioned training-and-inference framework that keeps the standard flow-matching objective. Self-Generated Draft Training supplies a detached clean-video estimate from the shared network as conditioning at the same noisy state. A Refinement-Only Late Adapter applies a dedicated residual transformation when a draft is present. At inference, Draft-Reuse Iterative Refinement obtains both a solver update and the next draft from each velocity prediction, requiring as few as 4 function evaluations (NFE) per video chunk. Training samples discrete solver-query noise levels and updates the pretrained backbone. Extensive experiments on Bridge and Language Table show that TwoPass achieves state-of-the-art results with 4 NFE, far fewer than the 50-59 calls of existing predictors. Code: this https URL.

[CV-138] Anchored Inertial Proposals with Objective Verification for Diffusion-Prior Inverse Problems NEURIPS2026

链接: https://arxiv.org/abs/2611.25551
作者: Minwoo Kim,Seunghyeok Shin,Dabin Kim,Hongki Lim
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted to NeurIPS 2026

点击查看摘要

Abstract:Training-free diffusion priors are powerful for inverse problems, but measurement guidance during reverse sampling is local: aggressive updates can improve immediate data fit while disrupting later denoising. We introduce an anchored inertial correction rule that separates proposal generation from acceptance. At each corrected step, the denoiser prediction anchors a frozen clean-space objective combining measurement fit with a noise-level-dependent proximal term. Heavy-ball dynamics with step size and momentum computed from the measured curvature generate a clean-state candidate; a separate verifier evaluates it with the same objective, with the denoiser prediction retained as fallback; both stages are derived from one objective for the whole trajectory, considered one step at a time. The accepted clean state is re-noised to continue sampling. The rule applies to pixel and latent diffusion with differentiable measurement operators, requires no retraining, and adds no denoiser calls inside the correction loop. We analyze how objective verification yields local nonincrease and displacement control. Across natural-image benchmarks, the method achieves competitive reconstruction quality. Code is available at this https URL.

[CV-139] Query-Aware vs. Query-Agnostic Training-Free Video Token Pruning in Multimodal LLM s

链接: https://arxiv.org/abs/2611.25537
作者: Divya Kothandaraman,Timo Kumkel,Jaclyn Pytlarz
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Video Large Language Models process large numbers of visual tokens, making token pruning important for efficient inference. Training-free pruning methods differ in whether and how they use the user query, and the value of query conditioning remains unclear. We present a controlled study of training-free, pre-LLM video token selection. Using a shared Qwen-3-VL backbone, token budgets, and decoding settings, we compare 21 pruning principles spanning query-agnostic, shallow query-conditioned, structured query-conditioned, and holistic strategies. To instantiate the last two categories, we introduce Structured Query Pruning, which decomposes the question into semantic anchors, and a Pareto variant, which combines anchor relevance with query-agnostic visual diversity. Across Video-MME, MVBench, and LongVideoBench, relevance computed from the raw question embedding is unreliable (8–16 points below the best methods), whereas anchors derived from query decomposition avoid this failure. However, query access is not inherently advantageous: strong query-agnostic coverage methods remain competitive or superior in the full-field comparison. Relative to a redundancy-only baseline, structured selection helps most on evidence-specific and long-context tasks at tight budgets; on broad tasks, the holistic selector improves over either relevance or redundancy alone. Further, pruning reduces cost and memory. These results support a practical design rule: preserve visual coverage by default, and add structured relevance when the task and token budget demand selective evidence retrieval.

[CV-140] Support-Guided Triadic Decoding for Hallucination Mitigation in Multimodal Dialogue with LVLMs

链接: https://arxiv.org/abs/2611.25535
作者: Yiheng Li,Wenhao Wang,Xingyi Li,Zhilin Tang,Zichang Tan,Zhen Lei,Yang Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: preprint

点击查看摘要

Abstract:In multimodal multi-turn dialogue, dialogue history plays a dual role: reliable responses provide useful context for subsequent turns, whereas hallucinated responses can propagate errors across turns. Existing training-free methods globally adjust the effect of dialogue history on generation, making it difficult to suppress misleading information without weakening useful context. To address this trade-off, we introduce Support-Guided Triadic Decoding (SGTD), a training-free framework that selectively controls the influence of unreliable history and fuses complementary predictions during decoding. Specifically, Historical Answer Support Estimation measures the visual support of historical responses against type-matched alternatives and identifies low-support content that may mislead later generation. Support-Targeted Attention Intervention then weakens the influence of this content while preserving the surrounding dialogue, producing a support-refined prediction. Triadic Decoding combines this prediction with the full-history prediction and answer-free visual evidence, while the Support-Adaptive Plausibility Constraint filters implausible tokens using visual agreement and historical support. Experiments on MMHalSnowball and VisDiaHalBench across representative LVLMs show that SGTD consistently achieves the highest accuracy under hallucinated histories while preserving the benefits of factual context.

[CV-141] Contrast Where It Matters: Visual Evidence-Guided Decoding for Mitigating Hallucinations in LVLMs

链接: https://arxiv.org/abs/2611.25524
作者: Yiheng Li,Wenhao Wang,Xingyi Li,Mingda Jia,Zichang Tan,Yinhang Tang,Zhen Lei,Yang Yang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: preprint

点击查看摘要

Abstract:Despite strong multimodal capabilities, large vision-language models (LVLMs) often hallucinate when linguistic priors outweigh visual evidence or when predictions are grounded in irrelevant or misleading image regions. Training-free contrastive decoding mitigates hallucinations by contrasting the original prediction with an auxiliary one that exposes linguistic priors. Yet existing methods focus on altering the auxiliary input or computation and do not ensure that either branch uses query-relevant visual evidence during decoding. We propose Visual Evidence-Guided Decoding (VEGD), a training-free framework that identifies query-relevant visual evidence and guides its use throughout contrastive decoding. VEGD locates this evidence by aggregating localization heads with strong visual responses and coherent spatial structure, then selectively weakens it in the auxiliary image. During generation, VEGD directs visual-carrier heads in both branches toward the same evidence location, strengthening access to intact evidence in the main branch while anchoring the auxiliary branch to its weakened counterpart. An adaptive distribution fusion strategy then combines or contrasts their predictions according to distribution divergence. Experiments across diverse LVLMs and benchmarks demonstrate that VEGD consistently reduces hallucinations.

[CV-142] Multi-View Relighting with Mergeable Radiance Fields

链接: https://arxiv.org/abs/2611.25523
作者: Yinan Wang,Rohan Kadkol,Hyunjin Yoo,Jean-François Lalonde,James J. Clark
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:We propose MergeableRF, a method for radiance field relighting through model parameter merging. Given multi-view images of a scene under known basis lights, our goal is to relight the scene by directly combining the parameters of radiance fields trained under these lights. This is motivated by the linearity of light transport: the image under a mixture of lights is the same weighted combination of the images under each light. However, combining the parameters of independently trained radiance fields does not produce the corresponding combination of images. To address this, we represent a scene with shared geometry and a separate appearance network for each basis light. We start these networks from the same initialization and introduce merge-aware training, which teaches their parameter combinations to reproduce the corresponding combinations of images. Given a target HDR environment map, we compute its basis-light coefficients and use them to merge the appearance networks into a single relit model. Our method supports both Neural Radiance Fields (NeRFs) and 3D Gaussian Splatting (3DGS), without material estimation, additional relighting networks, or target-specific optimization. Unlike image-space blending, it requires only one rendering pass per viewpoint. Experiments on synthetic Neural Gaffer objects and real OLATverse captures demonstrate relighting quality competitive with methods trained on the same multi-light observations.

[CV-143] Mesh-Free Keypoint Learning for Category-Level 6D Object Pose Estimation

链接: https://arxiv.org/abs/2611.25513
作者: Junghoon Seo,Hakjin Lee,Jaehoon Sim
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:

点击查看摘要

Abstract:Category-level 6D pose estimation relates partial RGB-D observations of unseen instances to a canonical frame shared across a category. Object-aware reconstruction helps adaptive keypoints retain the geometry of the visible surface. Common reconstruction targets select observed points by their distance to a registered instance surface. We present MORSEPose, which derives this support directly from standard pose and metric-size annotations. We map observed 3D points to normalized object coordinates and retain those inside a tolerancedilated pose-and-size envelope. This construction is consistent across camera frames and metric scales. When the annotated extent bounds the object surface, the envelope conservatively contains the support of a conventional surface-distance selector. We integrate this target with multi-level RGB-D features, adaptive keypoints, and learnable category context. Experiments on REAL275 and HouseCat6D show that the envelope closely matches surface-based supervision at both point and task levels, remains stable under boundary perturbations and viewpoint changes, and supports competitive category-level pose accuracy across benchmarks.

[CV-144] Carry the Fine Print: Budget-Aware Visual Token Compression for Document OCR

链接: https://arxiv.org/abs/2611.25506
作者: Yixuan Huang,ShengFan Zhang,Jiawei Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages; supplementary material is included as an ancillary file

点击查看摘要

Abstract:Visual token compression often treats salience or similarity as a proxy for transcription importance. Complete document OCR breaks this assumption: small symbols and repeated entries can be indispensable to the transcript despite being visually inconspicuous. We present Select and Compensate, a two-stage framework built on a simple idea: choosing where to retain a token need not determine all the evidence it carries. Selection combines adaptive encoding resolution with global importance and local contrast to choose which tokens to retain as anchors. Compensation enriches these anchors with nearby discarded content and its relative position, without adding tokens. The OCR backbone remains frozen, and native coordinates and structural tokens are preserved. We evaluate complete page transcription across five backbones on OmniDocBench and document reconstruction on MPDocBench. With half the original visual tokens, our input-adapted configuration reaches 95.42 OmniDocBench Overall on Ovis versus 96.09 for the uncompressed model, and paired evaluations show gains over simply reducing the input resolution on both benchmarks. With the original input grid, our method stays within 0.17 points of the uncompressed model at 75% retention and exceeds a learned global selector by 4.70 points at 25%. On a fixed 200-page Ovis workload at the same 50% visual-token budget, only 3 of its outputs hit the length limit, versus 30-34 for the evaluated RTPrune, DivPrune, and HiPrune adaptations, and it finishes the batch 35-40% sooner. Code will be made publicly available.

[CV-145] PRIO: Pre-trained Representation for Interactive Objects in 4D Worlds

链接: https://arxiv.org/abs/2611.25500
作者: Shixun Huang,Lifu Wei,Yue Chang,Chen Geng,Heng Zhang,Jinhyung Ahn,Caoliwen Wang,Siyuan Chen,Donglai Xiang,Jiajun Wu,Peter Yichen Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:3D worlds can now be captured, reconstructed, and generated at scale, yet the objects in them stay still: a pressed pillow does not compress, and a pulled door does not swing. Bringing these objects into interactive 4D world models requires a representation of how they can move, deform, and respond physically to user actions in real time. We present PRIO, a pre-trained representation for interactive objects. Given static geometry, PRIO predicts an object-specific configuration space in a single forward pass. This space provides generalized coordinates for Lagrangian mechanics, allowing physical laws and user actions to govern motion. Configurations are linear in these coordinates, allowing the quantities needed by the dynamics to be precomputed. Dynamics then run at interactive rates without further network inference. Pre-trained on diverse 4D trajectories, PRIO learns priors over object configurations that generalize to unseen shapes. It supports meshes, point clouds, and 3D Gaussians, bringing objects in reconstructed and generated scenes into interaction. We demonstrate its use in real-to-sim workflows and robot manipulation. In addition, compared with existing 4D generation methods, PRIO generates dynamics faster and follows user actions more faithfully.

[CV-146] AP: Tapping the Discriminative Potential of Frozen Representations for Class-Incremental Learning

链接: https://arxiv.org/abs/2611.25478
作者: Zihao Wang(1),Yu-Wei Zhan(2),Xiao-Qian Liu(1),Zhen-Duo Chen(1),Weitong Liu(1),Xin Luo(1),Xin-Shun Xu(1) ((1) Shandong University, (2) Zhejiang University of Technology)
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 3 figures. Code is included as an ancillary file

点击查看摘要

Abstract:Pre-trained models provide transferable representations for class-incremental learning (CIL). However, sequentially adapting these representations introduces optimization costs and may disrupt old-class feature structure. This motivates us to explore whether the discriminative potential of these representations can be further exploited by continually accumulating class information while keeping the encoders frozen. To this end, we propose TAP, a new and novel framework for tapping the discriminative potential of frozen representations in class-incremental learning. It accumulates visual and semantic class information in the frozen CLIP embedding space with only lightweight optimization. TAP comprises three modules: class-conditional visual distribution modeling, class-adaptive semantic calibration, and visual-semantic posterior fusion. The first module retains class means and within-class scatters, combining class-specific and shared covariances to stabilize distribution estimation. The second module learns class-specific weights over text templates from current-task images, adapting text representations to downstream classes. The third module combines the visual and semantic posteriors through a weighted product of experts for prediction over all seen classes without retaining images from previous tasks. Across nine datasets under two incremental protocols, TAP achieves state-of-the-art performance, attaining the highest average incremental and final-stage accuracies among the compared methods. The code is provided as an ancillary file.

[CV-147] SegAnyFruit: Learning Generalizable and Edge Deployable Fruit Instance Segmentation at Scale

链接: https://arxiv.org/abs/2611.25475
作者: Yanan Wang,Yuzheng wang,Yibin Ying,Zhenghao Fei
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 6 figures

点击查看摘要

Abstract:Deploying fruit instance segmentation in real agricultural settings requires generalizing across unseen species, orchards, and imaging conditions under strict onboard compute and annotation budgets. We introduce SegAnyFruit (SAF), a data construction and progressive training framework that distills foundation-model priors into lightweight, prompt-free segmenters. SAF combines generic mask pre-training, broad mid-training on automatically annotated web imagery from our curated WebFruits dataset, and human-verified agricultural refinement post-training on PaperFruits, together spanning 733K images and 9.66M whole-fruit masks across 165 fruit types. Across seven benchmarks, zero-shot SAF-RF-DETR-XL reaches 69.06 average mask AP50:95 (+4.76 over SAM 3) while remaining deployable on edge devices. SAF initialization also boosts few-shot to full-shot adaptation across five agricultural domains while mitigating catastrophic forgetting. Finally, we systematically analyze training stages, data scale, model capacity, and generative augmentation, and validate our pipeline’s transferability to underwater, cell, and human segmentation. We will release our datasets, code, and SAF models.

[CV-148] uringPAI: A Foundation for On-Device Physical AI

链接: https://arxiv.org/abs/2611.25474
作者: Qiman Wu,Hanlin Chen,Yuhua Wei,Mingyuan Wang,Shaman Tang,Minghao Li,Rui Xin,Xingye Liu,Zhichao Yang,Yiqi Yu,Hongyan Pei,Xiaowen Tian,YingYing Liu,Shiwen Liu,Jiahui Hu,Yizhao Wang,Yuheng Zhang,Anhua Liu,Yue He,Shuang Su,Da Zhu,Hua Zhou,Pengfei Diao,Haojie Yang,Weichao Huang,Hang Zhang,Honggou Yang,Xianming Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:

点击查看摘要

Abstract:Physical AI requires models that understand spatial relationships and dynamic environments and translate observations into actions under limited onboard compute. We present TuringPAI, a family of models for on-device Physical AI comprising TuringPAI-VL for shared vision-language understanding, TuringPAI-Drive for autonomous driving, and TuringPAI-Embodied for robotic manipulation. TuringPAI-VL is a 0.65B model built on HybridViT-MLQ, which combines convolutional local aggregation with Transformer-based global modeling. Multi-scale Latent Queries integrate spatial features and observation history within the same visual Transformer, providing a compact, structured interface to the language model. Training combines general vision-language data with geometric, cross-view, and temporal supervision. TuringPAI-VL supports general multimodal understanding and improves quantitative geometric perception over Qwen2-VL-2B, while providing a shared initialization for the two downstream VLA models. For autonomous driving, TuringPAI-Drive maintains all evaluated ADE/FDE metrics within 1.2% of a full-token HybridViT baseline using the MLQ interface and achieves comparable aggregate scores in an in-house simulation evaluation. Using XPeng Native Runtime (XNRT) on a single NVIDIA Orin-X SoC, TuringPAI-Drive generates trajectories at 12,Hz in on-vehicle operation. For robotic manipulation, TuringPAI-Embodied achieves a 96.8% average success rate across the four LIBERO suites. These results demonstrate a shared vision-language foundation for driving and robotic manipulation under practical on-device constraints.

[CV-149] A Wolf in Sheeps Pixels: Visual Camouflage Attacks via HDR Gain Maps

链接: https://arxiv.org/abs/2611.25469
作者: Qi Shen,Yujia Gong,Yue Zhao,Tong Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:An image can pass automated safety checks while showing harmful content to its human viewer. We introduce DRJanus, a visual camouflage attack that exploits the gap between standard dynamic range (SDR) processing and high dynamic range (HDR) display. Automated systems often read only an image’s SDR base, whereas HDR viewers reconstruct its appearance using an auxiliary gain map. DRJanus inverts this reconstruction to embed two independently chosen color images in a single JPEG, without querying or optimizing against the target model. It preserves the chosen SDR base exactly and reconstructs HDR targets at an average PSNR of 43.03 dB across 400 images. On a benchmark of 100 sexual-content images, DRJanus evades detection by all six tested commercial moderation services and nine vision-language models. Across five agents, each evaluated on 120 task cases per threat setting, attack success ranges from 31.67% to 100% for moderation evasion and from 95% to 100% for task manipulation. We also explore cross-view consistency checks as a mitigation. These findings expose a gap in visual trust: securing what a machine reads does not secure what a user sees. A demo is available at this https URL

[CV-150] ActMem: From Latent Memory Reasoning to Grounded Robotic Action

链接: https://arxiv.org/abs/2611.25461
作者: Linfeng Dong,Jianwei Niu,Jiuyi Zhao,Ruibing Feng,Xinghao Wu,Wu Yang,Shuai Tao,Wulong Liu,Mu Li
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 17 pages, 11 figures, 6 tables. Code and checkpoints: this https URL

点击查看摘要

Abstract:Vision-Language-Action (VLA) models have shown promising results in real-world long-horizon manipulation, where memory plays a crucial role. However, existing memory-augmented VLAs either retrieve past observations without reasoning about what they imply, or reason over the full history in a shared context where it crowds out current perception. Cognitive science treats them as distinct systems: memory reconstructs past experience to decide what to do, while perception guides how to act. We therefore propose ActMem, an in-context memory framework that decouples memory reasoning from perceptual grounding within a single vision-language backbone. Built on a vision-language model with native video understanding, ActMem infers the current subtask from a timestamped interaction history and keeps the hidden states formed during this inference as a Latent Memory Chain-of-Thought (Memory CoT), which conditions a flow-matching action expert. Perception Attention further lets the expert attend directly to the current view. To keep inference fast under long histories, Efficient Memory CoT Inference reuses computation across consecutive decisions without changing the output. ActMem achieves 71.1% task success and 80.8% completion success on RoboMemArena, surpassing the strongest baseline by 7.5 and 8.7 percentage points, and 73.8% average success on MIKASA-Robo, the best among all compared methods. In the real world, it reaches 68.8% average success on exogenous, endogenous, and traceless memory tasks, 10.0 percentage points above the strongest memory-augmented VLA. Code and checkpoints will be released at this https URL and this https URL. Comments: 17 pages, 11 figures, 6 tables. Code and checkpoints: this https URL Subjects: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) Cite as: arXiv:2611.25461 [cs.CV] (or arXiv:2611.25461v1 [cs.CV] for this version) https://doi.org/10.48550/arXiv.2611.25461 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Linfeng Dong [view email] [v1] Thu, 19 Nov 2026 03:46:39 UTC (11,167 KB) Full-text links: Access Paper: View a PDF of the paper titled ActMem: From Latent Memory Reasoning to Grounded Robotic Action, by Linfeng Dong and 8 other authorsView PDFHTML (experimental)TeX Source view license Additional Features Audio Summary Current browse context: cs.CV prev | next new | recent | 2026-11 Change to browse by: cs cs.LG References Citations SciXGoogle 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)

[CV-151] RemoteVLPG: How to Obtain a Specific VLM for Language-Guided Path Generation on Remote Sensing Images?

链接: https://arxiv.org/abs/2611.25458
作者: Junzhe Dang,Mengmeng Zhang,Yuxiang Zhang,Wei Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 18 pages, 10 figures

点击查看摘要

Abstract:We study how to obtain a task-specific Vision-Language Model (VLM) for language-guided path generation between start and end points on remote sensing images. We explore this question from three aspects. First, we implement three widely used architectural approaches and compare their performance. Second, we design a scalable method for constructing pre-training data and build an effective pre-training dataset with a degree of diversity. Methods trained on this dataset achieve promising zero-shot performance. Third, we construct three benchmarks that support stratified evaluation and design three metrics for path generation. We investigate this problem through extensive experiments. The results show that the zero-shot performance of methods trained on the constructed pre-training dataset can exceed that of some advanced commercial VLMs. Code and Dataset can be available at this https URL

[CV-152] CANDOR: Contrastive Self-Denoising Rewards for Label-Free Text-to-Image Post-Training

链接: https://arxiv.org/abs/2611.25457
作者: Junhyeok Lee,Byung-Hoon Kim,Kyu Sung Choi
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 3 figures, 5 tables

点击查看摘要

Abstract:Post-training text-to-image generators with reinforcement learning usually relies on external reward models that need annotated preferences or auxiliary evaluators and are prone to reward hacking. We show that the generator’s own denoising accuracy, used by a recent intrinsic reward, is a poor proxy for prompt fidelity. Within a same-prompt group, this absolute self-confidence is dominated by image complexity and ranks correct samples above incorrect ones near chance. We introduce CANDOR (Contrastive Alignment from Null-referenced Denoising lOg-Ratio), the log-ratio of unconditional to conditional denoising errors under shared noise probes, which measures how much the prompt helps the model denoise its own sample. It cancels prompt-independent difficulty that scales both errors and ranks the better of two same-prompt samples higher in 60% and 63% of pairs on compositional generation and text rendering. As the only Flow-GRPO reward for a flow-matching model, CANDOR improves text rendering, aesthetic and human-preference scores over the base model and a prior absolute self-confidence reward, with text-rendering and preference gains that stronger classifier-free guidance does not reproduce. On public Flow-GRPO adapters trained on external rewards, CANDOR further raises PickScore and ImageReward, so it can complement those rewards. Code is available at this https URL.

[CV-153] Lossless Latent Diffusion Models

链接: https://arxiv.org/abs/2611.25454
作者: Tongda Xu,Yu Zhang,Jun Hu,Huijuan Huang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:We study the problem of training image diffusion models while preserving pixel-exact reconstruction. Prior works typically address this problem by directly modeling images in pixel space, requiring modifications to the diffusion training objective or model architecture. In this work, we introduce a different paradigm: rather than modifying diffusion models, we construct a lossless transformation that maps pixels into a representation more amenable to diffusion modeling. We term our approach the Lossless Latent Diffusion Model (LLDM). We construct LLDM using residual latent transformation and uneven latent scaling, and explain its effectiveness through a proposed metric, semantic rank priority (SRP). Empirically, LLDM substantially outperforms prior pixel-space diffusion models in generation quality and even surpasses several widely used lossy VAEs. Furthermore, we show that SRP is strongly correlated with downstream generative performance.

[CV-154] Beyond Resemblance: Image Fusion Reimagined

链接: https://arxiv.org/abs/2611.25453
作者: Hao Zhang,Yumin Huang,Linfeng Tang,Junjun Jiang,Jiayi Ma
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Image fusion often lacks ground-truth fused images, forcing existing deep methods to indirectly define fusion objectives by requiring the fused image to simultaneously resemble multiple source images. However, the distinct imaging characteristics of different modalities can impose competing constraints on the fused result, potentially limiting the effective integration of their complementary information. To address this long-standing challenge facing the image fusion community, we propose ApexFuse, a novel generative image fusion paradigm that moves beyond source-resemblance constraints toward learning a general distribution of high-quality scene representations. Specifically, we develop a human-LMM-in-the-loop pipeline that leverages large multimodal generative models and iterative human feedback to translate abstract fusion objectives into reliable visual references, thereby defining the target distribution of ideal fusion outcomes. Through target-distribution-guided adaptation, we transfer the rich generative priors of pretrained FLUX to image fusion, enabling high-quality scene reconstruction conditioned on multimodal observations. Extensive experiments demonstrate that ApexFuse harmoniously integrates salient thermal targets, natural colors, and fine textures, achieving superior visual quality as well as semantic segmentation and object detection performance over previous source-resemblance methods. Moreover, ApexFuse exhibits remarkable cross-task generalization, extending from infrared-visible fusion to multi-exposure fusion, multi-focus fusion, and visible-depth fusion without additional fine-tuning.

[CV-155] Stereo Visual Odometry with In-the-Loop Geometric Feed-Forward Models Towards Dense Metric Mapping

链接: https://arxiv.org/abs/2611.25445
作者: Fahira Afzal Maken,Sagun Man Singh Shrestha,Edward Vanderfeen,Lucas Carvalho de Lima,Rongxin Li,Changming Sun,Milad Ramezani
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: 8 PAGES, 6 FIGURES

点击查看摘要

Abstract:We present a stereo visual odometry (SVO) framework that integrates geometric feed-forward models (GFMs) for dense depth prediction with factor-graph-based visual odometry to achieve accurate localisation and dense metric mapping. Our key idea is to exploit sparse but reliable metric geometry estimated from stereo and temporal multi-view observations to calibrate dense GFM predictions, while retaining the robustness and efficiency of model-based SVO. Rather than directly conditioning the GFM on potentially noisy stereo depth, optimised 3D landmarks are used to jointly estimate a scale-shift transformation over all keyframes within each local submap, yielding metrically consistent dense depth predictions. To balance accuracy and computational efficiency, we introduce a hierarchical factor-graph architecture. At the local level, sliding-window optimisation estimates high-frequency camera poses and multi-view-refined 3D landmarks. Dense predictions are generated only for selected keyframes and registered within each submap using SE(3) Generalised ICP. Consecutive submaps are further aligned in Sim(3) using their shared geometry, allowing residual scale inconsistencies to be corrected. These intra- and inter-submap constraints are incorporated into a global factor graph to produce a consistent trajectory and dense metric reconstruction. Extensive experiments on public and in-house datasets demonstrate that the proposed framework achieves competitive odometry accuracy against geometric and learning-based baselines while providing accurate dense metric reconstruction and improved robustness to failures in learned depth prediction.

[CV-156] VisualTalk-Bench: Evaluating Question-Driven Avatar-Led Video Explanations

链接: https://arxiv.org/abs/2611.25444
作者: Xin Zhang,Tujie Xu,Zhiyi Chen,Xiaodong Zhang,Sicheng Yang,Minglei Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 10 pages, 3 figures, 4 tables

点击查看摘要

Abstract:Advances in video generation and avatar synthesis enable new ways to answer user questions through visual explanations. This task requires a system to formulate a correct answer, translate relevant information into visual material, and deliver the explanation through an avatar’s speech and actions. Existing benchmarks rarely assess whether avatars point to the visual content being discussed. We introduce VisualTalk-Bench for question-driven avatar-led video explanations in finance and Chinese and English education. It includes 800 curated reference videos and 100 independent evaluation requests, assessing Content Factuality, Speech-Visual Consistency, Visual Quality, Motion Smoothness, and Avatar-Content Pointing. Yet end-to-end video generators often struggle to align gestures with specific visual targets. To explore this challenge, we develop Avatar-O Agent, a modular reference implementation with explicit interaction planning. Experiments across six agents show that factual precision and visual quality do not guarantee accurate pointing. A 26-participant study comparing four agents finds that Avatar-O Agent receives the highest average ratings for visual guidance, perceived pointing accuracy, and explanation effectiveness. We will release the data and code.

[CV-157] Wearable Logo Silent Hijack: Stealthy Cross-Modal Semantic Attacks on Visible–Infrared Vision-Language Models

链接: https://arxiv.org/abs/2611.25437
作者: Qike Zhang,Jiaju Han,Xuelian Shi,Xiang Chen,Dehe Zhao,Banglong Nong,Wenhao Zheng,Lei Wang,Jiahuan Long,Jiujiang Guo,Chengyin Hu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Logos appear naturally on clothing and daily necessities, where their recognizable shapes and semantic associations make them familiar visual cues. This familiarity helps them blend into real-world environments and remain visually inconspicuous. We ask whether an existing logo can serve as a carrier for coordinated targeted semantic attacks on visible-infrared (VIS-IR) vision-language models (VLMs) without changing its original outline. This setting requires expressive adversarial variation within a fixed shape while accounting for different appearance characteristics and model responses across modalities. We propose Logo-guided Semantic Attack (LoSA), a score-based black-box framework that optimizes a logo’s interior appearance and placement while preserving its silhouette. LoSA uses a contour field and low-frequency graph spectral modes to structure the interior color layout, a frozen appearance adapter to couple VIS and IR appearances, and particle swarm optimization to jointly optimize shared design parameters. Experiments across four datasets demonstrate that LoSA consistently outperforms the evaluated baselines in simultaneous Top-1 targeted success. Cross-architecture experiments, visual naturalness ratings, and captioning/VQA transfer analyses further characterize its effectiveness and limitations. These findings reveal that familiar visual symbols from real-world environments can serve as shape-preserving carriers for coordinated targeted semantic attacks across imaging modalities.

[CV-158] Raptor: Rejection-aware Adaptive Tree based on Speculative Decoding

链接: https://arxiv.org/abs/2611.25436
作者: Qishu Hu,Haozhao Wang,Zicong Hong,Yufeng Zhan,Yuanqing Xia
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注: 18pages, 7figures, 3tables

点击查看摘要

Abstract:Tree-structured speculative decoding accelerates vision-language models (VLMs) by verifying multiple candidate continuations in parallel. However, long visual contexts increase the cost of speculative computation, making candidate coverage and redundant expansion important considerations beyond tree topology adaptation. We propose Raptor, a training-free framework that uses rejection feedback to jointly regulate candidate coverage and drafting effort. Our key insight is that acceptance length alone cannot distinguish insufficient candidate coverage from excessive deep expansion, and these conditions require different allocation decisions. Raptor therefore extracts complementary rejection-aware signals to characterize speculative progress, frontier coverage, and redundant expansion. These signals jointly guide bounded, momentum-smoothed updates of tree depth, branching width, and drafting budget, together with execution control that limits unnecessary drafting and progressively restores expansion when appropriate. Extensive experiments with two VLMs across seven multimodal benchmarks demonstrate the effectiveness of Raptor, achieving up to 2.39x average speedup over autoregressive decoding.

[CV-159] DiLE: Discrimination-Aware Latent-Space Learning for Universal Multimodal Embeddings

链接: https://arxiv.org/abs/2611.25419
作者: Ben Kang,Junda Wang,Yuxuan Yuan,Hang Li,Yuhan Liu,Xin Chen,Wanting Geng,Tong Lin,Pei Fu,Dong Wang,Jian Luan,Huchuan Lu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Multimodal large language models have greatly improved input understanding in universal multimodal embedding models. However, this understanding does not ensure that their embeddings preserve the evidence needed to distinguish relevant items from confusable alternatives. To preserve essential discriminative evidence in embeddings, this paper proposes DiLE, a discrimination-aware latent-space learning framework that combines task guidance with on-policy retrieval feedback. Specifically, a few learnable queries are used to encode task-relevant information in a compact latent space. Reasoning instructions and auxiliary answer-generation supervision shape these representations to support both task understanding and retrieval discrimination. To address task-specific discriminative needs, we design shared-private expert adaptation that integrates cross-task knowledge sharing with task-specific representation learning. Building on this, the on-policy hard-negative distillation is introduced: the student retrieves confusable candidates, and a reverse KL divergence objective aligns its embedding-based similarity distribution with the teacher’s fine-grained relevance distribution over these candidates. Candidates are periodically refreshed using updated student snapshots, tying teacher supervision to the student’s evolving retrieval behavior and confusions. Across four model configurations that span two backbone generations, DiLE achieves state-of-the-art performance on MMEB-v2, UVRB, and MRMR, demonstrating its effectiveness across modalities and tasks.

[CV-160] Bilateral Evidence Selection and Reasoning for Mammography Vision–Language Models

链接: https://arxiv.org/abs/2611.25418
作者: Zhengbo Zhou,Dooman Arefan,Shandong Wu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Medical vision–language models (VLMs) often need to identify small, subtle lesions within high-resolution images containing extensive normal anatomy and redundant visual information. Visual-token compression improves efficiency, but preserving salient regions alone can discard the context needed for diagnostic interpretation. This context can extend across images: in bilateral mammography, the opposite breast provides a patient-specific anatomical reference. Yet retaining such evidence does not ensure that the model uses it effectively. We propose Bilateral Evidence Selection and Reasoning (BESR), which connects evidence preservation with comparative assessment through a bilateral-guided token selector and a bilateral reward. Under a shared token budget, the selector jointly prioritizes lesion appearance and its local and bilateral context, while the reward encourages assessments that benefit from the preserved comparison. Experiments on a bilateral dataset constructed from CBIS-DDSM and on VinDr-Mammo demonstrate improved evidence retention and diagnostic performance under constrained visual-token budgets. Inference benchmarking further shows reduced prefill latency.

[CV-161] MIR-Depth: Multi-Range Refinement for Single-View RGB-D Depth Completion of Transparent Objects

链接: https://arxiv.org/abs/2611.25417
作者: Xinzhe Fu,Qi Shao,Yuanfei Xie,Taoyu Wu,Zhihao Cao,Long Hu,Hong Liang,Xianzhi Li,Anh Nguyen,Baoru Huang,Rongsheng Li
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注: 9 pages

点击查看摘要

Abstract:Recovering accurate visible-surface depth from a single RGB-D image and an object mask is essential for robotic grasping and manipulation of transparent objects. Specular reflections cause missing depth, while refraction corrupts nonzero returns with background depth, making raw measurements unreliable as surface anchors. Under a fixed sampling budget, refining an initial single-view prediction within a single search interval either misses large refraction offsets or sacrifices fine local resolution. To resolve this single-view trade-off, MIR-Depth constructs Coordinate-Aware Multi-Range Hypotheses: a 3D-aggregated global inverse-depth volume and depth-axis-filtered pixel-wise local volumes around a prior-guided initial depth. Selective Recurrent Refinement then introduces out-of-bounds lookup masking and validity-masked softmax fusion to iteratively update inverse depth from these volumes using a multi-scale ConvGRU. Finally, Spatially Gated Correction applies a learned per-pixel spatial gate to the full-resolution residual, preserving accurate initial geometry without imposing raw sensor readings as hard constraints. Experiments on TransCG, TODD, TRansPose, ClearPose, and HouseCat6D show that MIR-Depth achieves lower error than ReMake, DFNet, BP-Net, and PSD. Real-robot evaluations confirm accurate surface geometry for 6-DoF grasp pose prediction. Code and pretrained models will be publicly available.

[CV-162] Watertight Mesh Optimization of Textured Objects

链接: https://arxiv.org/abs/2611.25414
作者: Nathaniel Burgdorfer,Philippos Mordohai
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:When reconstructing high-quality 3D assets from images, existing approaches often excel at one objective while compromising another, and frequently overlook the intrinsic quality of the reconstructed surfaces. We present MODL, a framework that jointly optimizes geometry and view-dependent appearance directly on a watertight triangle mesh. Our approach combines Laplacian-preconditioned optimization for smooth vertex updates with curvature-adaptive remeshing to preserve fine details and well-shaped triangles. The triangle mesh is optimized subject to multi-view normalized cross-correlation constraints, but without segmentation, geometric supervision or alpha-blending. Evaluations on object reconstruction datasets demonstrate that MODL achieves competitive novel-view synthesis while producing geometrically accurate models with superior intrinsic mesh quality. Our results highlight the potential of direct mesh optimization to unify photorealistic rendering, geometric reconstruction, and practical 3D asset creation.

[CV-163] FG-GraphBench: Benchmarking and Enhancing Fine-Grained Visual Understanding with Multimodal Scene Graphs

链接: https://arxiv.org/abs/2611.25408
作者: Sicheng Liu,Xunkai Li,Ziyu Han,Xingnan Zhang,Ru Zhang,Rong-Hua Li
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Fine-grained visual understanding requires multimodal large language models (MLLMs) to recognize small visual entities and understand their relationships. Scene graphs offer a natural way to represent this information, organizing visual entities as nodes and their relationships as edges. However, we identify two challenges: existing scene graph benchmarks provide insufficient coverage of fine-grained entities and the relationships among them, and existing methods for fine-grained visual understanding have yet to effectively leverage such graph-structured information in MLLMs. To address these challenges, we introduce Fine-Grained Multimodal Scene Graph Benchmark (FG-GraphBench) and FiGRA. FG-GraphBench is a benchmark comprising three complementary tasks: fine-grained visual question answering (FG-QA), image captioning (FG-Caption), and image description generation (FG-Description). We develop a two-stage construction pipeline that discovers and grounds fine-grained entities and establishes their relations to construct the fine-grained multimodal scene graphs (MSGs) in FG-GraphBench. Building on these MSGs, we propose FiGRA, a Fine-grained Graph Representation Augmentation method for MLLMs. FiGRA adopts a two-stage training strategy. It encodes multimodal entity and relation features as continuous graph tokens, and integrates them into the MLLM. Experiments show that FiGRA significantly outperforms strong baselines across three tasks of FG-GraphBench. Further analyses confirm the effectiveness of FG-GraphBench and highlight the importance of preserving fine-grained entity and relation information for visual understanding.

[CV-164] Decision Boundaries Matter: Graph-Regularized Residual Adaptation for Cross-Domain Fake Image Detection

链接: https://arxiv.org/abs/2611.25406
作者: Shichuang Xie,Qishuai Ding,Tong Qiao,Liquan Shen,Guorui Feng
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Despite learning transferable features, fake image detectors may struggle on unseen domains due to mismatched decision boundaries. Our analysis shows that features extracted by the frozen forensic encoder remain discriminative, but the source-trained decision boundary does not fully exploit this information. However, correcting this boundary without source data or target labels remains challenging, as unreliable source predictions may lead to incorrect adaptation. To address this issue, we propose Graph-Regularized Residual Adaptation (GraRA), which refines the source decision boundary using only unlabeled target data. To determine which classifier directions to update, GraRA constructs a compact residual subspace from uncertainty-weighted target features and keeps it orthogonal to the source classification direction. To guide these updates, it uses source-score anchors to generate soft labels through target-graph diffusion, while neighborhood consistency regularizes residual learning within the constrained subspace. The learned correction is merged into the source classifier, requiring only 32 optimized coefficients and no additional inference branches. On diverse unseen OpenMMSec domains, GraRA achieves 84.36% accuracy and 91.34% AUC, improving over the source detector by 4.33 and 3.50 percentage points, respectively. Further analysis shows that reliable source predictions and locally consistent neighborhoods help guide residual learning, allowing GraRA to correct ranking errors that a simple threshold shift cannot address. \textcolorcvprblueCode will be made publicly available.

[CV-165] HOPD: Historical On-Policy Distillation for SAR MLLM Continual Learning

链接: https://arxiv.org/abs/2611.25391
作者: Yuxuan Li,Derrick Goh Xin Deik,Zheng-Peng Duan,Qianru Sun,Weijie Li
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Continual SAR interpretation requires multimodal large language models to learn incoming sources while retaining earlier task abilities. This is difficult when acquisition and geographic shifts alter visual evidence, regionally concentrated categories bias answers, and historical images are scarce. We address this problem with Historical On-Policy Distillation (HOPD), which reuses source-acquisition LoRA states as persistent teachers for one evolving student. Archiving each state before consolidation preserves a stable source reference. The teachers supervise student-generated prefixes, transferring historical behavior to the student’s changing generation contexts. To enrich sparse replay, original-view teachers guide augmented historical student views while preserving spatial coordinates. This knowledge is consolidated into a single deployed adapter that requires no source identity. On SARVQA-CL, our source-incremental evaluation protocol, HOPD achieves the highest final mean score of 54.51 among the compared methods, exceeding the strongest baseline by 2.96 points. It also improves the final source-macro score in every evaluated task over shared-memory Replay-SFT, with gains across source orders. These results support source-acquisition history as supervision for retaining earlier task performance during continual SAR adaptation. Code and data: this https URL.

[CV-166] Zero to Four Views: Unified Panorama Generation and Preference RL for Unobserved Regions

链接: https://arxiv.org/abs/2611.25389
作者: Rui Hong,Xugang Ye,Jana Košecká,Kai Fan
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Panoramas can serve as shared scene backgrounds for multi-shot video generation, but most existing panorama generators take text alone or a single input image. We present a single model that generates a full equirectangular panorama from a text caption and K in 0,…,4 perspective conditioning views taken by a four-camera rig at one viewpoint. We fine-tune a pretrained text-to-image rectified-flow transformer in two stages. In the first stage, each conditioning view is appended to the token sequence together with the ID of its camera slot, and the number of views varies during training, so one set of weights serves every input setting. In the second stage, we improve the region not covered by any conditioning view. Supervised training computes its loss on noisy latents of the ground truth and never scores a panorama the model has generated and decoded, so we propose Uncovered-Region GRPO (UR-GRPO), a reinforcement-learning recipe that rewards the candidates a vision-language model prefers in pairwise comparisons on crops around this region. Because the slot ID alone does not tell the model which way the rig points, we also offer an optional variant that writes each view’s camera direction into its position encoding, which lets the user decide where each view appears in the output. With one conditioning view, our model outperforms the previous state of the art, 360Anything, trained on the same data and evaluated with the same code, on Laval Indoor and gives mixed results on an indoor subset of SUN360; the second stage improves the results for every number of conditioning views.

[CV-167] xt Lives in the Guidance: Preventing Silent Text Rewriting in Cached Diffusion Transformers

链接: https://arxiv.org/abs/2611.25382
作者: Kun Zhou,Wenmian Yang,Zhensheng Wang,Weijia Jia
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Diffusion transformers (DiTs) now render long, legible text in images, and feature caching is a standard way to accelerate them. Caches are judged by image-level metrics, which do not check whether the rendered text is still correct. We find that caching silently rewrites this text. On Z-Image, TaylorSeer at 2.9x fewer FLOPs lowers word accuracy from 62% to 46%, while ImageReward and CLIP Score barely change. Sampling with fewer steps at the same compute keeps almost all of this text, so the loss comes from how the cache predicts the skipped steps. Decomposing the cache error along classifier-free guidance (CFG), we trace this loss mainly to the guidance term, the difference between the conditional and unconditional predictions, which caches extrapolate worst on text. Correcting the guidance inside the text alone recovers most of the lost text. We therefore propose GuideCache, which recomputes both CFG branches only on the tokens with the largest predicted guidance error, using the guidance terms that the cache already stores. At 2.4x fewer FLOPs, GuideCache restores full-compute prompt fidelity on Z-Image and Qwen-Image, and on Z-Image it renders text as well as or better than ToCa, TeaCache, and FasterCache while staying much closer to the full model’s output. GuideCache works with caches of any order, and on top of fewer sampling steps it keeps more text than step reduction at a higher speedup. These results suggest that acceleration methods for text-rendering models should be evaluated on the text they render.

[CV-168] A Unified Framework for Image Manipulation Localization Under Multi-Type Weak Supervision

链接: https://arxiv.org/abs/2611.25371
作者: Songlin Li,Hui Ma,Zhiqing Guo,Dan Ma,Gaobo Yang,Xin Liao
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Existing weakly supervised image manipulation localization (IML) methods typically rely on annotation-specific training strategies, hindering the joint use of diverse annotations within a single model. This limits the scale of usable training data and constrains further progress. We propose UniIML, a teacher-student framework that maps diverse weak annotations into a unified representation to guide the Segment Anything Model (SAM) in generating multiple candidate masks. However, SAM is designed for general-purpose segmentation and has limited sensitivity to manipulation traces; its predicted quality scores alone may therefore fail to reliably identify the candidate that best delineates the manipulated region. We introduce Ordinal Safe Consensus (OSC), which first filters candidates using trusted weak-annotation constraints encoded in the unified representation. It then ranks the remaining candidates separately by consistency with weak annotations, SAM-predicted quality scores, and cross-view stability, aggregating these rankings for an initial selection. To assess how changes in the evaluation criteria affect candidate selection, OSC repeats ranking and selection while omitting one criterion at a time. The most frequently selected candidate is adopted as the final mask, and its relative selection frequency quantifies selection stability. We construct soft pseudo-labels from the selected candidate’s predicted probabilities and estimate pixel-level supervision weights based on selection stability, pixel-level agreement among candidates, and predictive certainty. The student receives direct supervision in regions defined by weak annotations and learns from soft pseudo-labels elsewhere through reliability-weighted distillation, enabling unified training across weak annotation types. Experiments show that UniIML achieves superior average performance compared with state-of-the-art IML methods.

[CV-169] Fold the Loop: Decoding-Time KV Memory for Looped Vision-Language Models

链接: https://arxiv.org/abs/2611.25369
作者: Kun Zhou,Wenmian Yang,Zhensheng Wang,Yanzeng Li,Weijia Jia
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Looped vision-language models (VLMs) deepen visual reasoning by repeatedly applying weight-tied modules. However, their decoding memory grows with recurrent depth and limits the image size and batch size that fit in a fixed memory budget. Each recurrent step produces its own key-value (KV) states, so the cache holds several copies of the same image. We find that, after an exact prefill, earlier steps still need access to visual information but can reuse the final visual KV within the same module. We propose FoldCache, which compresses visual KV along the recurrent axis while preserving the full prefill computation. FoldCache partitions recurrent steps within each module into groups and retains the final visual KV of each group for its members to share during decoding. Sharing stays within module boundaries, while question and answer KV remain separate for each step. The grouping granularity controls the degree of compression, and folding can be combined with token eviction and quantization to further reduce memory use. On LoopVL, FoldCache approximately halves prefix KV memory while matching full-cache accuracy on five open-ended benchmarks. These memory savings support batch sizes up to 1.7 times those of a full cache under the same GPU memory budget, and up to 2.6 times with the more aggressive FoldCache-Lite.

[CV-170] Search or Scoring? Diagnosing Why Visual World-Model Planners Fail

链接: https://arxiv.org/abs/2611.25347
作者: Haoxuan Jia,Yang Liu,Shang Luo,Chenhao Wu,Zixiang Ni,Yingguang Yang,Feng Zhang,Chongyang Zhang,Hao Zheng,Kefu Xu,Congjing Ran,Jifeng Zhu,Bin Chong
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 43 pages, 10 figures, 30 tables; supplementary material included

点击查看摘要

Abstract:Visual world models such as DINO-WM and JEPA-WM plan with an optimizer that searches for actions whose predicted outcome has the lowest cost, its distance to a goal image. A plan can thus fail in two ways: the optimizer never tries a good action that the cost would prefer (a search failure), or the cost, computed on predictions, ranks a bad action first (a scoring failure). The two need different remedies, more search or a better model, but a planner does not record which one occurred. We tell them apart by executing the optimizer’s candidates, scoring a known solution with the planner’s own cost, and comparing costs on predicted and real outcomes. An error model explains when each arises and so how to use retrieved demonstration windows (DemoProp): start the search from them (a warm start) when prediction errors are small, and average them without the model when large errors make the optimizer pick overrated actions. Across four simulated environments and three world models, on short tasks whose goals the planner can represent, PushT and Wall fail mostly by search, while Metaworld and RoboCasa fail by scoring because the cost, sound on real outcomes (AUROC 0.94 on Metaworld), is at chance on predictions (0.54). The diagnosis tells which remedy pays off: a four times larger search lowers the real-outcome cost only where failures come from search (clearly for DINO-WM), and where they come from scoring, averaging the windows without the model solves 127 of 168 new Metaworld tasks, against at most 87 when the model chooses among them. Within PushT, degrading only the model flips which use is better, and our rule, frozen in advance, follows the flip and solves 678 of 960 new episodes, against 530 for the best fixed use.

[CV-171] SceneAnything: Agent ic Reconstruction of Large-Scale Mesh Scenes from a Single Video

链接: https://arxiv.org/abs/2611.25340
作者: Bowen Xue,Saeed Hadadan,Fabrice Rousselle,Zahra Montazeri,Ravi Ramamoorthi,Milos Hasan
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注:

点击查看摘要

Abstract:Reconstructing an editable 3D scene from video requires matching visible structure while plausibly completing unseen surfaces. Agents can write scene-construction programs, but direct reconstruction attempts often produce errors in layout, scale, and lighting. We introduce SceneAnything, an automated system for reconstructing mesh scenes from real and generated videos, enabling text-to-scene construction when paired with video generation. A software harness coordinates agents, pretrained vision models, and graphics tools through an explicit reconstruction protocol. Shared camera, depth, and segmentation estimates ground component construction: a scene-owner agent plans the environment and briefs authors, who write parameterized Blender programs. A visibility- and fit-based level-of-detail rule selects the construction level, while unseen regions are completed and marked as inferred. Assembled scenes use fitted lighting and supported motion, and revisions undergo programmatic checks and independent visual comparisons. Accepted programs and lessons are reused across scenes. We evaluate on 100 scenes; subset comparisons show improved original-view fidelity over direct-agent reconstruction and higher novel-view content similarity than the compared video-to-3D methods. We will release the harness, agent instructions, reusable libraries, and scenes.

[CV-172] VideoRNG: Relightable Neural Gaussians from Generative Video Models

链接: https://arxiv.org/abs/2611.25338
作者: Bowen Xue,Saeed Hadadan,Fabrice Rousselle,Ravi Ramamoorthi,Zahra Montazeri,Miloš Hašan
类目: Computer Vision and Pattern Recognition (cs.CV); Graphics (cs.GR)
备注:

点击查看摘要

Abstract:Generating realistic 3D assets remains constrained by classical representations: surface meshes and compact physically based materials struggle to express objects whose appearance arises from fine geometry or complex light transport, while existing Gaussian-splatting generators typically bake fixed illumination into appearance. We introduce VideoRNG, a two-stage method that generates complete, relightable 3D assets from a text prompt or a single image without prescribing a mesh or analytic material model. A video generator fine-tuned from a pretrained video diffusion backbone first produces a structured RGBA turntable sequence under a known pose, camera, and illumination. A feed-forward large reconstruction model initialized from a pretrained video diffusion backbone then maps the video-VAE latent frames of selected turntable frames directly to millions of anisotropic Gaussians with learned neural appearance features. A shared directional decoder and stochastic Gaussian ray tracer turn this representation, which we call ray-traced Relightable Neural Gaussians (rtRNG), into radiance under novel viewpoints and illumination. To preserve high-frequency appearance while learning geometry, we introduce depth-anchored appearance training, and we make million-Gaussian reconstruction practical through foreground token selection and cost-aware distributed sampling. VideoRNG generates complex assets such as fur, feathers, and spider webs that can be interactively rendered, relit, and integrated into scenes.

[CV-173] MoGloss: Joint Motion and Pseudo-Gloss Pretraining for American Sign Language Understanding

链接: https://arxiv.org/abs/2611.25333
作者: Rui Hong,Jana Košecká
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 15 pages, 4 figures. Project page: this https URL

点击查看摘要

Abstract:Sign language translation and recognition have been shown to benefit from various pretraining approaches. We propose a novel two-stage pretraining approach for the motion encoder, which first learns to reconstruct masked motion, followed by joint masked prediction of pseudo-gloss units and motion. Pseudo-glosses serve as an intermediate structured representation of natural language text and are obtained from the captions by a large language model. We curate a corpus of about 1,100 hours of American Sign Language from five public datasets, estimate 2D pose for all clips and define an evaluation protocol. We then reproduce state-of-the-art translation methods and evaluate them using the same train and test splits. We propose a translation model built on our pretraining strategy, and this model outperforms the previous state-of-the-art methods Uni-Sign and SHuBERT, as well as the YouTube-ASL baseline, on the How2Sign, ASL STEM Wiki and OpenASL datasets. Furthermore, the gain from the joint pretraining using motion and pseudo-gloss grows as translation data shrinks. A classifier built on the same encoder also achieves state-of-the-art accuracy on ASL Citizen isolated sign recognition and outperforms our reproduction of Uni-Sign on WLASL2000.

[CV-174] VoxForge: A Unified Foundation Model for Dense 3D Features Across MRI CT and PET

链接: https://arxiv.org/abs/2611.25309
作者: Rushil Anirudh,Jayaraman J. Thiagarajan
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG)
备注: 34 pages (8-page main paper, references, and 23-page supplementary material), 7 figures. Rushil Anirudh and Jayaraman J. Thiagarajan contributed equally

点击查看摘要

Abstract:A unified 3D medical foundation model promises reusable, frozen features that serve a wide variety of tasks such as segmentation, classification and registration on MRI, CT and PET as well as specialist models do. Generalist models still need separate backbones for localized and global tasks, specialist models are confined to a single modality, with the concern that spanning modalities under-emphasizes fine anatomy, and both typically rely on end-to-end fine-tuning to be competitive. We present VoxForge, a single encoder that integrates recent advances in 3D architecture, self-distillation and cross-modal curricula to learn transferable features from permissively licensed MRI and CT volumes. We evaluate it strictly frozen, the test a reusable foundation model must pass, against state-of-the-art generalist and specialist models on segmentation, classification and registration across MRI and CT, and on PET, which it never sees. VoxForge has the best or tied-best average rank in every task group, serves localized and global tasks across modalities at the level of state-of-the-art specialists, and is extremely label-efficient, matching top-performing models on TotalSegmentator CT with a tenth of the labels.

[CV-175] ScaleDiag: Diagnosing the Limits of Test-Time Scaling in GUI Grounding ICDM

链接: https://arxiv.org/abs/2611.25297
作者: Chen Liu,Zhen Fang,Xin Yu,Dadong Wang,Ling Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: Accepted by ICDM@D2AI 2026 Best Paper Award

点击查看摘要

Abstract:GUI grounding maps a screenshot and a natural language instruction to the coordinates of the intended interface element. Existing approaches improve grounding through visual scaling, instruction scaling, and sampling scaling, yet it remains unclear which grounding bottlenecks these strategies address and where their gains saturate. To address this gap, we introduce ScaleDiag, a controlled diagnostic framework that isolates the effects of visual input, instruction information, and prediction sampling on GUI grounding by independently varying target scale and surrounding context, target-specific information and auxiliary descriptions, and candidate coverage and selection, respectively. We evaluate ScaleDiag across five benchmarks and up to eight vision-language models and find that each scaling strategy operates within a much narrower regime than commonly assumed. (1) Visual scaling is governed by target occupancy rather than a fixed crop ratio, with full-screen errors typically rescued once the target occupies only 0.32% of the cropped view, while further zooming eventually becomes harmful. (2) Instruction scaling is driven by information that identifies the target rather than description richness itself, as even ground-truth-derived attributes such as color, size, and shape are often neutral or harmful, while appearance and spatial descriptions account for over 90% of the achievable gain. (3) Repeated sampling exposes a different limitation, as correct coordinates are frequently generated but remain difficult to select, leaving 5.1–11.0 points of oracle headroom while geometric consensus provides almost no improvement. Together, these results show that test-time scaling is constrained by distinct operating regimes rather than benefiting monotonically from more visual focus, richer instructions, or additional samples. Our code is available at this https URL.

[CV-176] CalibCache: Cache-to-Full Feature Dynamics Calibration for Diffusion Transformers

链接: https://arxiv.org/abs/2611.25296
作者: Mingyu Liang,Dingkun Xu,Jingwei Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Diffusion Transformers generate high-quality images, but iterative sampling remains computationally expensive. Feature caching reduces the number of full network evaluations, yet its approximation error can become more severe under short sampling schedules, where fewer full-compute steps and a coarser sampling grid can make history-based extrapolation less reliable. In this paper, we propose CalibCache, a lightweight offline calibration method that corrects systematic discrepancies between cached and full-compute feature dynamics. Specifically, CalibCache calibrates feature acceleration, defined as the second-order finite difference across sampling steps. Using only a single pair of full-compute and naive-cache trajectories at each setting, CalibCache fits scalar acceleration maps offline by centered least squares. These maps are then applied during sampling without additional full-network evaluations or online optimization. Across FLUX.1-dev and Qwen-Image on DrawBench and DiT-XL/2 on ImageNet-256, CalibCache consistently improves fidelity or generation quality over TaylorSeer under 20-step sampling at cache intervals N=3,4,5 . At N=5 , it reduces FLUX.1-dev LPIPS from 0.563 to 0.429 while achieving a 2.86\times denoising speedup over full-compute sampling, and lowers DiT-XL/2 FID from 16.39 to 10.37 relative to TaylorSeer.

[CV-177] Persistent Identity in Person Tracking: Online Enrollm ent and Registered Recognition

链接: https://arxiv.org/abs/2611.25283
作者: Ratha Siv,Matei Mancas,Bernard Gosselin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 7 pages, 2 figures, 7 tables. An earlier version of this paper was submitted to an IEEE conference

点击查看摘要

Abstract:We extend PoseTReID with online appearance-bank matching (Gallery) and a residual-conflict guard for registered SVM recognition. All methods use supplied ground-truth (GT) person boxes. We average all-box assessment IDF1/HOTA equally across videos. On six SportsMOT clips, StrongSORT leads standalone trackers (79.44/84.72%) and exceeds both Gallery variants. Over eleven videos, Gallery AIN (83.29/85.44%) and COPE (84.66/86.38%) exceed standalone trackers with complete results. This lead depends on two retrospectively added GTA development compilations and is absent on the original nine videos. Registered SVMs lead both cohorts among complete results, with AIN at 95.72/94.82% over eleven videos. They use every GT crop from approximately 20% of frames, including future appearances. On MOT20-03, registered AIN reached 96.80% IDF1 and assigned the correct enrolled name to 90.72% of assessment GT observations. Gallery and standalone trackers receive no per-video labelled enrollment. Both identity routes restore identities across internal title cards, but Gallery trails SORT on three videos (four in HOTA) and both SVMs trail SORT on MOT16-04/MOT20-03. Under six-clip persistent-only scoring, native Gallery gains 2.39 percentage points in mean IDF1 over a matched control, with lower correct-P coverage and 5.02 times as many encoder attempts. Pose gating is mixed; the guard’s metric effect is not isolated.

[CV-178] Accord: Universal Online Metric Localization and Dense Mapping from one Camera and an IMU

链接: https://arxiv.org/abs/2611.25279
作者: Jaafar Mahmoud,Arthur Movsesyan,Cuong Huynh,Mikhail Iumanov,Sergey Zagoruyko,Sergey Kolyubin
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Dense metric mapping from a single camera and IMU remains challenging: visual-inertial odometry (VIO) provides metric motion but sparse geometry, whereas feed-forward depth models predict dense structure without ensuring metric scale. We present Accord, which constructs an accurate online dense metric map from any camera with an IMU. A VIO filter publishes metric poses, a depth foundation model conditioned on them predicts depth for short keyframe windows, and a factor graph fuses the resulting submaps. Recent work has explored several approaches to this problem. What is essential is the multi-view agreement (Accord), which is considered in SLAM as a source of geometric self-consistency. Accord anchors each keyframe window to metric 3D points triangulated from the filter poses, then exploits bidirectional dense depth agreement between co-visible images to estimate and correct per-image scale errors while preserving the metric scale established by the anchors. Context keyframes further improve depth prediction, and the resulting submaps are fused in a factor graph. We evaluate a single mapping configuration across five public camera-IMU setups and a new Phone-rig/GoPro dataset. With nothing tuned per dataset/setup and every map aligned only by Accord’s own trajectory, Accord achieves consistently best/second-best mean F-score at 5 cm on every platform. Accord is open-source along with the phone-rig dataset for the benefit of the research community

[CV-179] Real Cameras are Radiometric

链接: https://arxiv.org/abs/2611.25265
作者: Ties Harder,Felix Seegräber,David Nakath,Hendrik Sauer,Kevin Köser,James Pritts
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 22 pages, including supplementary material

点击查看摘要

Abstract:We propose the radiometric camera, a model that lifts any central geometric camera model to a radiometric one. The lifted model keeps the projection of the geometric model and adds a finite aperture, a free-form focus surface and a generic vignetting function. We calibrate it by inverse rendering directly on the relative irradiance of the sensor, from the same images of standard calibration targets used in geometric calibration. Geometric calibration absorbs the bias of its feature detections. The proposed radiometric calibration uses no features, and our radiometric cameras have lower reprojection errors on held-out test sets. The focus surface extends defocus modeling from the thin lens to any central camera, and we use it in a first multi-view radiometric reconstruction, which recovers the in-focus reflectance of a texture from defocused images. The vignetting is estimated jointly and accurately reproduces a flat field measured through a diffuser. Because every standard camera model lifts to a radiometric one, the radiometric camera naturally and directly extends generative scene models beyond multi-view geometry to multi-view radiometric ones, where phenomena such as lighting and attenuation can be accurately estimated.

[CV-180] Keel: A Frontier Model Teaches a Vision-Language Model by Rewriting Its Weights

链接: https://arxiv.org/abs/2611.25257
作者: Jiajun Fan,Neil He,Bangji Yang,Qinglong Zheng,Ziwen Wang,Chaoran Cheng,Changyi Yang,Hanqi Mao,Ge Liu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Can a frontier model repair a vision-language model (VLM) by rewriting its weights? Small reasoning VLMs such as Qwen3.5 need such a repair: in thinking mode they fall into loops on many image questions, repeating themselves until the token limit, spending the whole budget and losing the answer. The usual remedies act on the decoding, leaving the looping model as it was, or must be told what to change. Yet the cause is written in the weights. Keel, the first agentic weight-editing method for VLMs on image tasks, finds and removes it there. Told only the symptom, an editor, a frontier multimodal LLM acting as an agent, locates the cause, edits the weights and tests them, with a failure census and instruments showing it the failing images and where the VLM attends; among several editors, the harness keeps only edits that clearly reduce the failure without lowering accuracy on questions no editor saw. The repair needs no training, keeps the model’s size and per-token cost, and applies to already-edited weights. The editors expose where the loops live: from the original weights, independent editors on both models trace them to the attention outputs of the middle layers, and the smaller VLM stops consulting the image inside a loop. On Qwen3.5-0.8B and Qwen3.5-2B, across six image benchmarks from visual mathematics to GUI grounding, the edit removes most loops on every benchmark and shortens answers, and the edited models are significantly more accurate than the thinking models on average; the larger one gains most on GUI grounding, a task absent from the practice questions. Because the repair lives in the weights, it combines with budget forcing where held-out questions favor it: on the best available weights, Keel has the best average accuracy on both models, significantly above every training-free baseline we compare on the larger one.

[CV-181] Branch Caching After Step Distillation in Four- and Eight-Step Image Generators

链接: https://arxiv.org/abs/2611.25248
作者: Mingyu Liang,Dingkun Xu
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Can a distilled image generator save time by reusing features between its few remaining denoising steps? We study a fixed branch cache that replaces selected attention and MLP evaluations while preserving the student’s noise schedule. One full trajectory and one raw-cache trajectory from the same initial noise supply a scalar correction for each branch type and reuse position. At 1024^2 , one internal reuse reduces four-step Qwen-Image-Lightning denoising time from 1.279 to 0.977 seconds. Three reuses yield 1.51\times speedup on eight-step Lightning and 1.53\times on eight-step Hyper-FLUX. We evaluate image quality on all 200 DrawBench prompts. With the reuse positions held fixed, correction improves mean same-seed image fidelity over raw reuse by 0.66 and 0.87 dB on the two Lightning students. On Hyper-FLUX the PSNR difference is 0.01 dB, while LPIPS improves from 0.1625 to 0.1546. These results measure the effect of branch caching and correction while retaining each student’s original sampling grid.

[CV-182] Beyond Restoration: Content-Degradation Factor Unfolding for Real-World Concealed Object Segmentation

链接: https://arxiv.org/abs/2611.25239
作者: Rihan Zhang,Fengyang Xiao,Yuhan Chen,Sina Farsiu
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Reliable visual perception under degradation requires distinguishing task-relevant content from degradation-induced variations, especially in concealed object segmentation (COS), where foreground cues are inherently weak. The central challenge is not merely improving visibility, but identifying reliable content evidence amid degradation-induced variations. RGB-space segmentation often entangles content and degradation cues, while restoration-based methods primarily focus on image recovery and task interaction. This motivates explicit factorization to distinguish content evidence from degradation effects and assess its reliability. Retinex provides a spatially aligned basis for this separation but remains illumination-centric. We therefore extend it to content-degradation formation with task-asymmetric factor roles. The Factor Unfolding Network (FUN) uses the content factor R as the sole source of direct semantic evidence for segmentation, while the degradation factor L captures degradation-related variations and generates a spatial reliability map to calibrate this evidence. A unified energy couples observation formation with reliability-weighted foreground-background separation, and its alternating updates yield a coarse-to-fine network without clean-image supervision. In this way, FUN factorizes degradation for perception rather than removing it. Extensive experiments across clean and degraded conditions and diverse segmentation tasks demonstrate robust performance and generalization. Code will be released.

[CV-183] MKA: Evaluating Multimodal Knowledge Agents on end-to-end professional tasks

链接: https://arxiv.org/abs/2611.25228
作者: Anima Singh,Suhas Yogin,Weicheng Kuo,Ivor Rendulic,Ivan Jurin,Antoine Yang,Oran Kelly,Rohan Doshi,Hammad Ayyubi,Ioana Bica,Sergi Caelles,Matthew Conniff,Dan Graur,Anastasija Ilić,Ahmet Iscen,Chao Jia,Tomasz Kępa,Chunyuan Li,Noah Ó Donnaile,Filip Pavetic,Goran Petrovic,Khalid Salama,Zhicheng Wang,Polina Zablotskaia,Yiqian Zhang,Ming Zhao,Benjamin Caine,Mario Lučić,Melvin Johnson,Radu Soricut,Jean-Baptiste Alayrac
类目: Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:We introduce the Multimodal Knowledge Agent (MKA) benchmark, which measures how well autonomous agents perform end-to-end professional knowledge work that depends on visual understanding. Each task requires an agent to extract and combine information from multiple visually dense files, such as CAD blueprints, electronics schematics, radiographs, complex charts and geospatial maps, and to produce a standalone deliverable in .xlsx, .pptx, .pdf, .docx, .html or .png format. Existing benchmarks leave this capability largely untested: removing image-viewing tools leaves Gemini 4 Carbon’s score on GDPval unchanged, and single-format evaluations such as this http URL and this http URL do not require reasoning across files or modalities. MKA contains 200 tasks across 10 economic sectors. Deliverables are graded against expert-weighted rubrics by a sandboxed agentic autorater that agrees with human domain experts on 96.2% of criterion judgments in our expert-graded study (Cohen’s \kappa = 0.91 ), scored by weighted satisfaction rate (WSR) overall and on frontier criteria unsaturated by prior-generation models (Climb-WSR). Across the frontier models we evaluate (Gemini 4 Carbon, Claude Fable 5.2, Claude Opus 5.5, and GPT-6 Astra), Claude Fable 5.2 (56.6% Climb-WSR, 73.3% Overall WSR) and Gemini 4 Carbon (54.3% Climb-WSR, 71.9% Overall WSR) achieve the highest scores, while Carbon uses 5.7\times fewer tokens. Even for the strongest models, Climb-WSR remains below 57%, leaving substantial headroom for multimodal agents.

[CV-184] Reject and Report: Forgery-Evident Undetectable Watermarks for Generative Image Models

链接: https://arxiv.org/abs/2611.25226
作者: Yihan Wang,Haochen Sun,Xi He,Yaoliang Yu,N. Asokan
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Provenance schemes such as those based on content watermarks let images from a generative model be traced back to it. A practical watermarking scheme must be accurate at a low false-positive rate (FPR), robust under content-preserving processing, and undetectable. In addition, it must resist forgery; otherwise, even from a single watermarked image and a surrogate model, an adversary can imprint the watermark onto an arbitrary image or regenerate new content that carries it, falsely attributing harmful content to a model or making authentic photographs appear AI-generated. We argue that a watermarking scheme should not only reject forgeries but also report them: a forgery rejected as unwatermarked leaves no trace of the attempt. State-of-the-art forgery-resilient undetectable watermarking schemes, such as SemBind, fall short on both counts. External anomaly detectors that do report forgeries either miss most of them or falsely flag genuine images after common processing. We propose Tripwire-WM, the first forgery-evident framework for undetectable watermarks. It commits the semantics of the generated image to the magnitudes of the initial latent noise, which the sign-based watermarks leave unused, through a distribution-preserving embedding; a forged image carries a valid watermark but with a commitment that does not match its content, so the detector both rejects it and reports it as forged. Built on the pseudorandom code (PRC) watermark and Gaussian Shading++, Tripwire-WM provably preserves their undetectability. On MS-COCO, with the owner’s own model as the surrogate, it accepts no imprinting forgery, compared with up to 55.0% for SemBind, and reports all of them as forged; it also accepts at most 1.8% of reprompting forgeries, against up to 46.2% for SemBind. Meanwhile, it accepts at least 99% of genuine watermarked images at 10^-6 FPR and matches the robustness of the base schemes.

[CV-185] Bridging vision-language reasoning and learned control for legged navigation in the wild

链接: https://arxiv.org/abs/2611.25215
作者: Takahiro Miki,Robin Schmid,Manthan Patel,Turcan Tuna,Julia Richter,Julian Nubert,Joonho Lee,Matthias Müller,Marco Hutter
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 42 pages, 6 figures, 15 supplementary figures. Author’s version of the article published in Science Robotics. Project page: this https URL

点击查看摘要

Abstract:Autonomous navigation in diverse real-world environments remains one of the greatest challenges in robotics. It demands both robust low-level motion control to handle diverse terrain and high-level reasoning to interpret environmental cues. We present ROAM (Robust Open-world Autonomous Mobility), a system that enables legged robots to perform kilometer-scale navigation in complex alpine and urban settings. ROAM combines the world knowledge of a Vision-Language Model (VLM), pre-trained on internet-scale data, with robust locomotion and local planning policies learned in simulation via reinforcement learning (RL). The VLM provides common-sense reasoning and semantic interpretation of the scene, assessing risks, interpreting signs, and producing semantic labels with associated navigation costs. An open-vocabulary segmentation module converts these labels into a dense, real-time risk map, which conditions two pre-trained RL policies responsible for agile locomotion and local navigation, effectively coupling high-level VLM reasoning with low-level physical control. Real-world experiments showed the system’s ability to autonomously navigate a wide range of complex real-world environments. The robot completes kilometer-long hikes with minimal intervention, interprets signs to avoid danger, and evaluates its environment to identify safe areas on which to step. This work represents a step toward the reliable deployment of systems in unstructured environments. However, it also underscores the challenges that remain in VLM reliability and spatial reasoning, indicating that foundational models are a crucial component but not yet a complete solution for navigation in the wild.

[CV-186] Recurrent Vision Language Models Solve the Shell Game

链接: https://arxiv.org/abs/2611.25210
作者: Dingqiang Ye,Jiacong Xu,Wen Zhang,Yiqun Mei,Vishal M. Patel
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Tracking a hidden object through a video is effortless for humans, yet it remains a bottleneck for vision language models (VLMs). In the shell game, where a ball is hidden under one of several identical cups that are then swapped, VLMs that answer directly stay near chance, thinking models fail on long games, and fine-tuning learns only the lengths it has seen. We argue that the limit is serial computation: each swap must be composed with the last result, and a VLM of fixed depth has only a fixed number of state transitions for it. To address this, we propose RecurrentVLM, which makes every attention layer of a pretrained VLM recur over time by computing the keys and values of finished frame groups from the layer’s own output. With a single answer token, RecurrentVLM reaches 100% on VET-Bench and 99.6% on MET-Bench, and 99% on 50-swap games after training on at most 30 swaps. We also observe several interesting phenomena: length generalization emerges suddenly at an aha moment; without recurrence, each swap costs an attention layer; the ball’s state forms in the first four attention layers; and recurrence keeps the latest state at the cost of details of earlier moments. These results suggest that recurrent VLMs are a potential solution for tracking the states of unobserved objects in video. Code will be released.

[CV-187] Identity World Model: Recognizing People in an Imagined Latent World

链接: https://arxiv.org/abs/2611.25204
作者: Dingqiang Ye,Kartik Narayan,Vishal M. Patel
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Whole-body human recognition has advanced rapidly with large vision models. However, existing methods remain fragile under real-world variations, as they tie identity to transient cues such as clothing, texture, pose, or viewpoint. We argue that a robust recognizer should not merely learn to ignore these cues, but imagine how a person would be represented if they changed. To this end, we propose IDWorld, an identity world model: just as a latent world model predicts the features of a scene after an action, IDWorld predicts how a person would be represented under a changed condition. In a single forward pass, IDWorld imagines these counterfactual features under shared canonical conditions, such as the same outfit, a fixed view, or a bare silhouette, without rendering any pixels. Compared with explicit imagination in pixels, this latent imagination keeps nearly the same accuracy while being over 11x faster, at 5.9 ms per frame. Since each imagined condition has its own strengths, IDWorld then fuses them adaptively, weighting each condition for every probe-gallery pair. Extensive experiments show that IDWorld achieves state-of-the-art results on CCPG and CCGR-MINI, outperforms the compared methods on MEVID, and generalizes best to CASIA-B* and SUSTech1K, with the best mAP on CCVID. These results suggest identity world models offer a simple yet effective path to robust recognition. Code will be released.

[CV-188] Phantom Success Silent Failure: Auditing the Success Checkers of Simulated Manipulation Benchmarks

链接: https://arxiv.org/abs/2611.25179
作者: Zirui Song,Jiadong Zhang,Zheng Lu,Xiaosong Ma,Xiaojun Chang,Xiuying Chen
类目: Computer Vision and Pattern Recognition (cs.CV); Robotics (cs.RO)
备注:

点击查看摘要

Abstract:Simulated manipulation benchmarks measure policy performance using hand-written success checkers that label each episode as a success or failure. Passing an expert demonstration shows that a checker accepts one way of completing the task. Learned policies, however, may stop after partial progress or complete the task in a different way. A checker may then accept an incomplete execution (phantom success) or reject an execution that satisfies the instruction (silent failure). In this paper, we introduce SimAudit, an LLM-based pipeline that audits success checkers against task instructions. SimAudit analyzes the instruction and checker code to identify scenarios where the two may disagree, recreates those scenarios in simulation, and runs the original checker. Multi-view renderings are then used to verify whether the observed behavior satisfies the instruction. Across RoboTwin 2.0, RoboCasa365, and RoboDojo, 183 simulated test cases produced the predicted checker verdict. Visual verification and deduplication yielded 61 distinct defects across 49 tasks, including 40 phantom successes and 21 silent failures. Most defects in RoboCasa365 accepted incomplete executions, most in RoboDojo rejected valid executions, and both error types were common in RoboTwin 2.0. These findings suggest that validating success checkers requires testing both incomplete executions and valid alternatives to expert demonstrations, and motivate more rigorous evaluation protocols for future benchmarks.

[CV-189] MV-JEPA: Post-Training Video Encoders for Cross-View Understanding

链接: https://arxiv.org/abs/2611.25172
作者: Yangchen Zhao,Ali Payani,Caiwen Ding
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:JEPA-based video encoders learn strong representations for video and image understanding through prediction in latent space. However, their standard pretraining predicts features within individual images or video clips, without explicitly exploiting paired camera views of the same object or moment. Consequently, the learned representations may not reliably support matching across different camera viewpoints. To solve this problem, we introduce multi-view predictive post-training to strengthen cross-view instance and temporal correspondence while retaining independent single-view inference. A shared encoder processes masked context views independently, and a training-time predictor jointly predicts latent tokens for masked context regions and a fully held-out view. We alternate cross-view prediction with single-view masked prediction on general image and video data to help retain the encoder’s pretrained capabilities. After only 9,000 post-training updates on a single GPU, MV-JEPA improves cross-view instance retrieval and coarse cross-camera moment retrieval on various datasets, with further gains in zero-shot action-phase retrieval and frozen-feature classification on AE2. Frozen-probe evaluations further indicate that post-training largely preserves motion recognition, image recognition, and dense semantic segmentation performance on the evaluated benchmarks. Together, these results demonstrate that multi-view predictive post-training can strengthen instance and coarse temporal correspondence in JEPA video encoders without modifying the encoder architecture or single-view inference interface.

[CV-190] What a Calibrated Probability Hides: Belief-Level Post-Training for Decision Models

链接: https://arxiv.org/abs/2611.25168
作者: Xinye Li,Zirui Song,Huaxing Liu,Xiang Wang,Shuai Li,Zixiang Xu,Yinjun Liu,Zhewen Zhang,Xiaojun Chang,Xiuying Chen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:A calibrated decision model can predict how often each answer is chosen. The same prediction can reflect disagreement among people who are confident in their choices or uncertainty among people who share a belief. Disagreement calls for identifying whose perspective applies, while uncertainty calls for gathering more evidence. Standard post-training scores a single prediction against one observed choice at a time, learning the pooled distribution without distinguishing these two sources of uncertainty. We introduce Belief-Grouped Policy Optimization (BGPO). The policy proposes a list of beliefs, estimates their prevalence, and predicts a choice distribution for each. During training, BGPO uses Bayes’ rule to estimate how likely each belief was to generate an observed choice. These estimates weight the choice’s contribution to each belief’s log-score reward. To discover rare beliefs, the policy also proposes lists after seeing a few observed choices and retains them only if they predict held-out choices well. At inference, the policy uses only the question and combines the belief-specific distributions by prevalence to produce a calibrated aggregate prediction. On benchmarks covering opinions, conversation safety, emotions elicited by artwork, and photo aesthetics, BGPO achieves the lowest conditional, worst-group, and overall Jensen-Shannon divergence across model initializations. It improves conditional divergence on both text and image tasks. On a synthetic image task with known beliefs, BGPO recovers the underlying groups more accurately than baselines trained with belief labels.

[CV-191] CoRelate: Coordinated Human-Object Motion Synthesis through Interaction Relations

链接: https://arxiv.org/abs/2611.25166
作者: Zhiwen Qiu,Daisy Lin,Diyu Zhou,Jialin Wu,Rosita Ji,Saleh Kalantari
类目: Computer Vision and Pattern Recognition (cs.CV)
备注: 11 pages, 4 figures

点击查看摘要

Abstract:Human-object motion synthesis requires coordinated human participation as objects move toward specified goals. However, reaching object targets does not ensure that contact is established and maintained, while adjustments that improve contact may interfere with task completion. We introduce CoRelate, a diffusion framework that coordinates full-body human-object motion through explicit interaction relations and task feedback for rigid and articulated objects. We represent intended participation through hand-part relations that specify the participating hand, its part-relative reference position, and participation interval. These relations serve both as generation conditions and as references for measuring interaction realization. At each denoising step, a temporal refiner trained jointly with the generator combines relation deviations, task residuals, and motion features from the current prediction to jointly update human and object motion across the sequence. After generation, temporal motion correction addresses remaining contact failures by optimizing local motion segments while preserving object trajectories and protected human motion. Experiments demonstrate improved object-goal satisfaction, contact F1, and complete interaction success over the evaluated baselines. Ablation studies support the benefits of relation conditioning and the complementary roles of task and relation feedback.

[CV-192] GEMS: Generation and Experience Memory for Subgoal-Guided Long-Horizon World Model Planning

链接: https://arxiv.org/abs/2611.25134
作者: Yangchen Zhao,Ali Payani,Caiwen Ding
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Visual world models enable effective short-horizon goal-conditioned planning, such as reaching nearby targets and pushing objects over short distances, by predicting candidate action outcomes and comparing their representations with a goal representation. However, extending this approach to long-horizon tasks is challenging: prediction errors accumulate over extended rollouts, while latent-space similarity to distant goals provides limited guidance for ranking candidate outcomes. Existing methods address these limitations by introducing intermediate subgoals, either generated by learned models or selected from experience memory. Generated subgoals can contain hallucinations or be inaccurate in aspects such as object positions and timing. Memory-based approaches are constrained by experience coverage and may lack precise guidance for selecting appropriate intermediate states from stored observations. We introduce GEMS, which generates temporally structured subgoals to guide action search and state selection from Latent Experience Memory (LEM). A distance gate determines whether predicted outcomes are evaluated against the generated subgoal or the selected memory state. Experiments across six tasks and task suites spanning manipulation, navigation, and reaching demonstrate the effectiveness of GEMS for long-horizon planning. GEMS outperforms state-of-the-art world-model planning baselines across different manipulation settings, achieving success rates of 92.8% on Cube, 68.4% on Push-T, and 54.1% on Meta-World. It improves over the strongest evaluated baseline on these tasks by 24.0, 7.3, and 22.2 percentage points, respectively. These results demonstrate that generated subgoals and experience memory can jointly strengthen long-horizon planning with a frozen world model.

[CV-193] ALD: Transferability-Aware Logit Disentanglement for Cross-Modal Knowledge Distillation

链接: https://arxiv.org/abs/2611.25126
作者: Hui Li,Pengfei Yang,Xinyi Feng,Junhong Liu,Tianyang Zheng,Hui Zeng,Le Dong,Kaibing Zhang,Mengting Zhao,Quan Wang
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Cross-modal knowledge distillation can incur negative transfer when source-modality supervision conflicts with the target modality’s discriminative characteristics. We propose TALD, a two-stage framework for Transferability-Aware Logit Disentanglement in class-aligned logit space. Guided by task-dependent self-benefiting and cross-modal exclusivity criteria, a Residual Disentangler decomposes logits from a teacher and an independently trained target-modality reference student into modality-general and modality-specific components. A Sample-Adaptive Modulator regulates decomposition constraints and sample-specific fusion weights. The learned modality-general components are then adaptively fused to supervise the final student. Across five benchmarks spanning vision, audio, and text, TALD achieves the largest mean accuracy gain and highest positive-transfer frequency among the compared methods. Across ten transfer directions with five runs each, TALD improves over training without distillation in 84% of evaluations, compared with at most 70% for the baselines, while offering a favorable accuracy-efficiency trade-off. We plan to release the code.

[CV-194] Do Deep Stereo Models Learn Shape Priors?

链接: https://arxiv.org/abs/2611.25113
作者: Linbo Tang,Aneel Damaraju,Todd Zickler
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Localizing object boundaries is important for embodied visual systems, since small boundary errors can change the inferred geometry of physical surfaces. We study whether deep stereo models produce boundaries consistent with simple shape priors when local image and disparity cues are weak. We create a synthetic stereo dataset consisting of objects drawn from a highly constrained family of shapes, with randomized depth, slant, color and texture. Despite the simplicity of this shape space, we find that several state-of-the-art stereo models retain systematic boundary errors, even when finetuned or trained from scratch on the dataset. We then introduce a simple generative stereo model that jointly predicts disparity and a distance field representing object boundaries. The model improves boundary localization and produces contours that adhere more closely to the underlying shape space. These results expose limitations in the boundary predictions of current stereo architectures when local evidence is weak, and motivate further study of whether generative models can better combine regional grouping evidence with global geometric structure.

[CV-195] Long-Range Video Presentation Attack Detection: A Three-Class Benchmark Across 1-100 m and Two Sensor Types

链接: https://arxiv.org/abs/2611.25108
作者: Lambert Igene,Mohammad Zahir Uddin Chowdhury,Afzal Hossain,Sandip Purnapatra,Peter Johnson,Stephanie Schuckers
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: 10 page, IJCB

点击查看摘要

Abstract:Face presentation attack detection (PAD) systems deployed in border control and airport surveillance are constrained by a critical assumption: subjects are captured at close range, typically under one meter. No existing benchmark systematically addresses multi-distance capture, diverse environmental conditions, and concurrent spoofing and occlusion-based attacks, leaving deployed models underprepared for real-world scenarios. We introduce a video-based PAD dataset of 374,000+ frames from 261 unique identities collected across five standoff distances (1 m-100 m) under indoor and outdoor conditions using a DSLR camera © and a mobile sensor (M), with three-class labels: live, spoof, and props (occlusions); models are trained and evaluated on video clips rather than single frames. We fine-tune and benchmark four CNN architectures: MobileNetV3, MobileNet, EfficientNet, and ResNet-50. The higher-resolution DSLR maintains near-perfect spoof classification rate across all distances, while the lower-resolution mobile sensor degrades sharply beyond 50 m, with props classification rate dropping to 55% at 75 m for MobileNetV3. EfficientNet achieves the best overall performance (98.48% accuracy, ACER 2.01%). Generalization tests on unseen attack datasets yield 89-94% accuracy, validating cross-domain robustness. These results establish capture distance and sensor resolution as critical but underexplored axes of PAD difficulty. The dataset will be publicly released under an End-User License Agreement to support reproducible long-range PAD research.

[CV-196] MAWAM: A Multi-Agent World-Action Model for Coordinated Manipulation

链接: https://arxiv.org/abs/2611.25105
作者: Xianzhe Fan,Chi-Pin Huang,Fu-En Yang,Jing Wang,Yecheng Wu,Yu-Cheng Chou,Johan Bjorck,Linxi “Jim” Fan,Jan Kautz,Yu-Chiang Frank Wang,Zhiding Yu,Hengshuang Zhao,Zhiqi Li
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Every move in a robot team changes the world in which its teammates must act. Coordinated manipulation therefore calls for predicting the team’s actions together with the physical future they jointly create. We introduce MAWAM, a multi-agent world-action model that jointly generates each robot’s future views and executable action chunks in a single sample. Built from a pretrained single-robot world-action backbone, MAWAM jointly denoises all robots’ action and future-view tokens through cross-agent attention, while preserving a separate control interface for each robot. MAWAM achieves the highest success on all three real-robot tasks (laundry-bag collection, bath-towel folding, and umbrella handover), with 67.8% mean success, 18.9 percentage points above MA-VLA trained on the same demonstrations and cameras. In simulation, MAWAM reaches 90.3% mean success on eight tasks with teams of up to four arms. Ablations show that cross-agent attention and team-stage future-video supervision both improve coordinated control: without attention, a four-arm relay falls from 95.0% to 39.0%, and without this supervision, Place Food falls from 78.0% to 47.0%.

[CV-197] VisAlloc: Allocating Visual Supervision for On-Policy Self-Distillation

链接: https://arxiv.org/abs/2611.25069
作者: Haoran Xu,Tian Chen,Yuhe Wu,Yanbo Zhao,Junchi Liao,Hongyu Wang,Jiaze Li,Wentao Wang,Yanfeng Li,Rui Qian
类目: Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:On-policy self-distillation (OPSD) offers a promising approach to improving visual reasoning in multimodal large language models (MLLMs). However, not all samples benefit equally from privileged visual guidance, and uniform token supervision can dilute visually grounded signals. We propose VisAlloc, a dual-level framework combining sample-level visual filtering with token-level counterfactual reweighting. A fixed paired-teacher filter retains samples answered correctly from an evidence-centered positive crop but incorrectly from a spatially displaced negative crop. At the token level, both EMA teachers score the same student-generated response, and their absolute log-probability differences measure view sensitivity. Additional supervision targets tokens admitted by sensitivity or teacher-distribution disagreement, with shared frozen-prior, template-bigram, and student–teacher divergence gates. Inverse-frequency redistribution preserves the total additional weight within each micro-batch, while the positive teacher remains the sole distillation target. Inference requires only the full-image student. Across eight benchmarks, our 4B and 9B models achieve macro-averages of 81.44 and 84.92, outperforming Vision-OPD reproductions by 0.97 and 0.52 points, respectively. Code on GitHub: this https URL

[CV-198] Infer3D: Inverse Generative Modeling Enables Robust 3D Reconstruction

链接: https://arxiv.org/abs/2611.25061
作者: Amani Kiruga,Mark Hamilton,Joshua B. Tenenbaum,William T. Freeman,Vincent Sitzmann,Yilun Du,Ayush Tewari
类目: Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG); Robotics (cs.RO)
备注: 23 pages, 12 figures, 15 tables. Includes references and appendices. Project page: this https URL

点击查看摘要

Abstract:Humans can understand the 3D structure of scenes even from highly unfamiliar or extreme viewpoints. Cognitive science suggests this generalization stems from a top-down generative inference strategy. In contrast, state-of-the-art 3D reconstruction neural networks typically rely on amortized, feedforward inference. While fast, this pattern-matching approach degrades when faced with out-of-distribution (OOD) scene conditions unseen during training. Inspired by the generative nature of human vision, we propose Infer3D, an inverse generative modeling framework that recovers the latent variables of a generative model to best match the monocular input. To prevent degenerate flat outputs, we couple the generative image prior with a frozen 3D lifting network that constrains the optimization space to a valid, multi-view-consistent 3D manifold. Across a variety of controlled synthetic and real-world tasks, Infer3D demonstrates superior OOD reconstruction accuracy compared to feedforward networks trained on equal or larger datasets. Furthermore, mirroring how human perception requires more time to process confusing stimuli, we introduce an adaptive inference strategy that expends more test-time compute on ambiguous OOD examples while efficiently routing in-distribution inputs. Project website: this https URL

[CV-199] Inhabitant: An Agent Framework for Long-Horizon Human-Scene Interaction with Spatial Control and Quantitative Feedback

链接: https://arxiv.org/abs/2611.25005
作者: Yao Zhang,Zhuchenyang Liu,Yu Xiao
类目: Robotics (cs.RO); Computer Vision and Pattern Recognition (cs.CV)
备注: this https URL

点击查看摘要

Abstract:Recent work on human-scene interaction (HSI) generation moves from single interactions to long-horizon interaction, where a language-model agent plans and executes a sequence of interactions in a 3D scene under one instruction. However, existing agents struggle in challenging everyday scenarios (e.g., picking up a toy and throwing it into a basket) because they lack spatial control over the motion and quantitative feedback on its execution, and they often fail to complete the interaction sequence. To address these gaps, we present Inhabitant, a training-free agent framework in which the agent specifies each motion through positions, paths and poses, measures how the executed interaction deviates from its intent, and revises the control specification accordingly. For an extensive evaluation in such scenarios, we build Inhabit, a benchmark of 75 household tasks in 31 furniture layouts of an apartment, in which every task chains one to three goals on movable objects and is carried out by the whole body, excluding fingers. On Inhabit, Inhabitant completes 98.7% of the tasks with Claude Opus 5.5, and 93.3% and 82.7% with Claude Sonnet 5.5 and Gemini 4 Argon, respectively. Ablations confirm that without spatial control the agent fails most tasks, and without quantitative feedback its motion becomes less precise. Code is available on the project page.

[CV-200] Density Feedback Control for Transferability of Generative Sparse Adversarial Attacks

链接: https://arxiv.org/abs/2611.25001
作者: Dongsu Song,Jay Hoon Jung
类目: Machine Learning (cs.LG); Computer Vision and Pattern Recognition (cs.CV)
备注: preprint

点击查看摘要

Abstract:Generator-based sparse adversarial attacks train encoder–decoder networks to produce transferable sparse perturbations for black-box robustness evaluation. Their perturbation density, however, is set only indirectly by a fixed sparsity coefficient whose mapping to density is unknown and shifts during training, so reaching a target density requires re-tuning; this fixed pressure also yields little transferability gain once density reduction begins. We propose Density Feedback Control (DFC), which tightens or relaxes this coefficient at the end of each epoch according to the gap between the observed hard-mask density and the target, without changing the generator architecture or loss. At a 10% target density, DFC brings the mean density of three base attacks within one percentage point of the target in a single run and improves the transfer attack success rate (ASR) of two of them on CNN and Transformer targets, including under per-image \ell_0 budgets, while perturbing fewer pixels. Tracing GPG’s training, we observe that as DFC relaxes the coefficient, the transfer ASR recovers with the density and overtakes the fixed-coefficient baseline with fewer pixels. For EGS-TSSA, which restricts perturbations to source-model feature-importance regions, DFC yields no consistent gain; we analyze how this restriction interacts with density feedback.

[CV-201] EgoPass-Bench: Identifying Objects That May Affect the Wearers Passage in Egocentric Video

链接: https://arxiv.org/abs/2611.24995
作者: Hongyu Wang,Chuhan Li,Congjin Xu,Xuandong Zhao,Hua Shen
类目: Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Understanding an egocentric environment requires reasoning about how surrounding entities affect the observer’s actions. Existing evaluations of multimodal foundation models (MFMs) examine observer localization, motion trajectories, and action feasibility, but leave the discovery and interpretation of evolving observer-entity interactions insufficiently assessed. We introduce EgoPass-Bench, a benchmark for observer-relative interaction understanding in real-world egocentric video. EgoPass-Bench comprises 284 manually reviewed evaluation clips recorded using Ray-Ban Meta (Gen 2) smart glasses and 2,201 distinct questions across five tasks assessing motion perception, interaction judgments about specified entities, and relevant-entity discovery without target cues. Our evaluation of eight frame-based and six native-video configurations reveals a pronounced discrepancy between motion perception and interaction understanding. The strongest evaluated model, GPT-6 Astra, correctly recognizes object motion in 91.37% of 394 comparisons involving the same targets, but correctly identifies changes in passage constraints in only 66.50%. Across all 574 entity-discovery questions, it trails the two-person human baseline by 23.65 \pm 7.45 percentage points (paired 95% CI), while exceeding the always-empty baseline (36.76%) by at least 18.9 percentage points. Further analysis reveals frequent inclusion of entities that neither pose a plausible physical conflict nor constrain passage, showing that accurate motion recognition does not ensure reliable relevant-entity discovery.

[CV-202] Life-IQA: Boosting Blind Image Quality Assessment via GCN-enhanced Layer Interaction and Flexible Hierarchical Expert Specialization

链接: https://arxiv.org/abs/2611.25510
作者: Long Tang,Guoquan Zheng,Huiyu Duan,Jie Hao,Yucheng Zhu,Liang Yuan,Xiongkuo Min,Guangtao Zhai
类目: Image and Video Processing (eess.IV); Computer Vision and Pattern Recognition (cs.CV)
备注:

点击查看摘要

Abstract:Blind image quality assessment (BIQA) is essential for evaluating and optimizing visual experience. Existing BIQA methods typically aggregate features from multiple backbone stages without accounting for their unequal contributions to quality prediction. Furthermore, uniformly supervising all experts may constrain their flexibility and limit the diversity of the learned quality representations. To address these issues, we present Life-IQA+, a quality representation decoding framework that combines graph convolutional network (GCN) enhanced layer interaction and flexible hierarchical expert specialization. Specifically, the layer interaction module constructs quality queries from GCN-refined deepest-stage features and uses penultimate-stage features as keys and values, thereby enabling cross-layer representation interaction via the cross-attention mechanism. Building on these representations, we further introduce a flexible hierarchical mixture-of-experts (MoE) architecture comprising a densely activated shared expert layer and a sparsely activated routed expert layer. To enhance the representation diversity and ability of experts, we apply auxiliary classification supervision to a subset of shared experts using distortion-type labels for synthetic datasets and discretized quality-level labels for authentic datasets. Experiments on synthetic and authentic distortion benchmarks, together with controlled ablations of expert configurations and supervision strategies, show that Life-IQA+ achieves competitive prediction accuracy with a favorable balance between accuracy and computational cost. The code is available at: this https URL.

[CV-203] Breaking Noise Shortcuts in Self-Supervised Learning Via Noise-Aligned View Generation NEURIPS2026

链接: https://arxiv.org/abs/2611.25335
作者: Oankar R. Patil,May D. Wang
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Computer Vision and Pattern Recognition (cs.CV)
备注: 40th Conference on Neural Information Processing Systems (NeurIPS 2026)

点击查看摘要

Abstract:Self-supervised learning (SSL) is increasingly applied to scientific imaging domains, such as microscopy, where labels are scarce but noisy data is abundant. These domains exhibit significant signal-dependent noise, which can create a representational shortcut: since both augmented views derive from the same noisy image, the encoder can boost view agreement by also encoding noise patterns rather than semantic content. To enable noise-robust representation learning without knowledge of underlying noise models, we introduce Self-Aligned Noise Augmentation (SEANA), a drop-in module that generates noise-aware SSL views. SEANA learns an invertible variance-stabilizing transform (VST) from noisy data, then estimates the clean signal and resamples independent noise in learned VST space, requiring no clean targets, no dataset-specific denoiser training or use at inference, and no changes to the SSL objective or encoder. On CIFAR-10 and ImageNet-100 under synthetic Gaussian, Poisson-Gaussian, and multiplicative noise, SEANA improves clean-test linear-probe accuracy across contrastive/non-contrastive SSL methods, indicating higher-quality representations. On real fluorescence imaging, SEANA improves Jurkat cell-cycle stage classification by up to 28.6pp over standard SSL and 27.5pp over denoiser-preprocessed baselines. In all settings, SEANA outperforms the denoiser-preprocessed SSL pipelines, demonstrating that learned VST-space noise-aligned resampling yields more robust representations than denoising alone. Code is available at: this https URL

人工智能

[AI-0] SoK: Assessing the Evolution of LLM Capability in Software Vulnerability Research

链接: https://arxiv.org/abs/2611.26601
作者: Amine Lbath,Manan Suri,Aurelien Delaitre,Vadim Okun,Massih-Reza Amini,Dinesh Manocha
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:

点击查看摘要

Abstract:LLM-based agents are becoming increasingly capable at discovering, validating and patching vulnerabilities in real software, and vendors and researchers report that this is already changing how vulnerabilities are found and fixed. These claims, however, are scattered across papers, vendor reports and disclosures and are rarely comparable. We introduce a framework that rates these capabilities on an autonomy scale modeled on the levels of driving automation, and qualifies each claim by its correctness, its availability and the strength of its evidence. We use this framework to systematize the published evidence on LLM capability in vulnerability research. We find that agents discovered, validated, and patched vulnerabilities in real-world software without human intervention before the restricted frontier models of 2026, and that open-weight models subsequently attained comparable autonomy for discovery and validation. We then measure the consequences of evolving LLM security capabilities in a decade of public vulnerability records, advisories and exploitation data. In 2026 the number of disclosed vulnerabilities more than doubled, with an increasing share rated critical or high, and severe bugs that had survived a decade or more in widely used code were reported several times more often than before. Yet the time between the stages of a vulnerability’s life, from report to fix and from disclosure to attack, has not shifted noticeably so far. We grade the mitigations proposed in the literature with an explicit rubric and draw guidance for each group of stakeholders from the best-supported ones.

[AI-1] Active Feature Acquisition by Distilling an Oracle Teacher

链接: https://arxiv.org/abs/2611.26598
作者: Yang Meng,Zixin Ding,Robert L. Grossman,Yuxin Chen
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 16 pages

点击查看摘要

Abstract:Acquiring every feature of an instance is often costly: a measurement can carry a monetary price, a time delay, or a privacy cost. Active Feature Acquisition (AFA) addresses this by learning a policy that decides, per instance, which features to acquire under a budget. Existing AFA methods train this policy by optimizing a self-constructed utility target rather than distilling an explicit teacher. We construct this teacher explicitly. At training time the label and the unobserved feature values are available, and together they determine the teacher in closed form: an oracle that ranks every candidate by its effect on the downstream predictor’s error. OracleImitator distills this ranking into a deployable policy by soft-target imitation. Across multimodal medical diagnosis and the AFABench benchmark, OracleImitator matches or outperforms existing AFA methods across the budget range, recovering 90% of the attainable accuracy gain at under a quarter of the acquisition cost. Because the teacher is an explicit algorithmic object rather than an implicit optimization target, we can also measure directly how much of its ranking is recoverable by a policy restricted to what is visible at deployment.

[AI-2] PGX: Physically Grounded Execution for Long-Horizon Humanoid Loco-Manipulation

链接: https://arxiv.org/abs/2611.26596
作者: Kai Lu,Chenyang Ma,Ioannis Havoutis
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Robotic vision-language-action (VLA) models can execute diverse individual manipulation skills, yet they still struggle with long-horizon loco-manipulation tasks, where temporal structure, skills, and precision are tightly coupled within a monolithic model. We propose Physically Grounded Execution (PGX), a framework that offloads temporal and precision control to two classical training-free components, allowing the VLA to focus on stage-specific skill dynamics. PGX decomposes long-horizon loco-manipulation into three layers: (i) a predicate-based state machine governs task progression, using physical predicates to gate phase transitions, trigger recovery, and coordinate navigation and waiting. Shared predicate semantics connect offline demonstration segmentation, online execution, and stage-wise evaluation; (ii) VLA skills operate within active manipulation phases, with flat or stage-trained models concentrating on task-relevant actions; and (iii) world-coordinate locking (WCL) maintains spatial consistency through base station keeping, object-relative positioning, and world-anchored contact targets. We demonstrate the effectiveness of PGX with extensive experiments conducted on three simulated tasks and one real-world task using a Unitree G1 humanoid robot for long-horizon loco-manipulation, showing improvements in stage progression and task success rates. Project page is available at: this https URL

[AI-3] Verify the Verifier: Specification-Grounded LLM Cross-Checking for Bugs That Escape UVM Scoreboards

链接: https://arxiv.org/abs/2611.26588
作者: Ana E. Sanchez Villalobos,Michael D. Quinn,Aakash Tyagi,Minh D. Luu,David Kebo Houngninou
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR)
备注:

点击查看摘要

Abstract:A passing UVM regression shows the design agrees with the reference model on what the scoreboard checks, not that it meets the specification. Bugs that pass every test yet violate the specification, termed scoreboard escapes, survive mature testbenches. We present Verify the Verifier, to our knowledge the first framework whose LLM-based Anomaly Cross-Checker examines a passing regression’s traces against the specification. Of 564 bugs inserted into four OpenTitan IPs, 68 are candidate scoreboard escapes: non-equivalent, exercised by the tests, yet passing. The framework confirmed 41 as scoreboard escapes with assertions failing on the buggy design and proven on the original.

[AI-4] ProtoDiffusion: Token-per-Layer Generative Modeling of Multi-Layer Images

链接: https://arxiv.org/abs/2611.26586
作者: Adrià Ciurana Lanau,Harleen Hanspal,Lusine Abrahamyan,Emilio Tylson Baixauli,Yu Cao,Danai Triantafyllidou,Stefano Zorzi,Gustavo Zanoni Felipe,Peter Griger,Stefan Sietzen,Petru-Daniel Tudosiu
类目: Artificial Intelligence (cs.AI)
备注: 20 pages, 16 figures

点击查看摘要

Abstract:Graphic designs are ordered compositions of RGBA layers whose separation enables selection, movement, and reuse. Existing layered models represent each layer with a full-canvas or region-sized spatial latent grid, so representation size grows with the cumulative spatial extent of the layers. We present ProtoDiffusion, a latent-diffusion framework built on ProtoVAE, a multi-layer variational autoencoder that encodes a design into one shared spatial prototype and one candidate vector per layer. Each candidate conditions a shared decoder to reconstruct its corresponding RGBA layer. For N layers, the representation contains HW+N tokens before backbone-specific packing, compared with NHW for full-grid-per-layer representations. ProtoDiffusion jointly denoises the prototype and a fixed set of candidate slots, conditioned on an image or text, with bounding boxes as optional layout guidance. The model leaves unneeded slots empty and determines the number of layers and, in absence of boxes, their layout, requiring neither as input. The same architecture supports both image-to-design and text-to-design. ProtoVAE achieves high-fidelity reconstruction from its compact shared latent representation, reaching 40.43 dB PSNR on layer crops. On the LayerD evaluation split, image-to-design achieves RGB L1 scores of 0.040 with bounding boxes and 0.057 without, with box-guided alpha soft IoU (0.850) comparable to the strongest fine-tuned baseline (0.871). On merged images, text-to-design reaches 66.41 FID versus 86.84 for FLUX.2 [klein], achieves a 0.860 text-accuracy score, and runs 19.1\times faster than the same-backbone generation-then-decomposition cascade.

[AI-5] GEOHunter: Mapping a Generative Engine Optimization Network in the Wild

链接: https://arxiv.org/abs/2611.26564
作者: Sean Tang,Peiran Wang,Ying Li,Esha Shivakumar,Tao Meng,Yuan Tian
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Social and Information Networks (cs.SI)
备注:

点击查看摘要

Abstract:Generative engines synthesize retrieved web pages into answers, recommendations, and rankings. Generative engine optimization (GEO) services sell visibility in these answers by publishing promotional articles on user-generated-content (UGC) platforms for paying clients. Such promotion raises concerns in medical and financial domains, where misleading advice can cause harm. To study the severity of this problem, we measure one such paid-content operation across three major Chinese UGC platforms, uncovering 1,985 accounts associated with the network comprising 466,357 inferred promotional placements. We find that the network promotes products across diverse categories through a growing pool of accounts, conducting coordinated campaigns spanning months across multiple platforms while evolving its listicle content over time. Across nine generative engines from both the US and China, we find that 4.5% to 20% of retrievable cited pages are attributable to the network, depending on how strictly attribution is drawn. On eight of these engines, its clients are named in 52% of answers, against 68% for category leaders. Our work demonstrates the effectiveness of UGC GEO placement at influencing generative engine outcomes, and reveals content-level defenses as insufficient against coordinated UGC placements, where account-level and platform-level signals are necessary for proper defense.

[AI-6] MM-SystemOneBias: Evaluating Bias in Multimodal Jev-Style System One Decision Models

链接: https://arxiv.org/abs/2611.26557
作者: Jinzhe Li,Qi Wang,Dandan Guo,Li Liu,Yuan Wu,Yi Chang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:When multimodal decision models determine actions, moderation outcomes, or quality judgments, their reliability depends on preserving semantic decisions across presentation changes, ignoring irrelevant additions, and distinguishing information consistent with task evidence from information that conflicts with it. Existing evaluations study decision-interface biases and multimodal model behavior in separate settings, leaving no common protocol to compare multimodal decision models with their backbones on practical tasks. We introduce \textbfMM-SystemOneBias, which evaluates 13 models on 5,323 original instances from nine public datasets and 11 task scenarios using six paired interventions: option order, option label, external judgment, image–text description, irrelevant information, and rating scale. Across seven matched decision-model–backbone pairs, decision models are more stable under option-label, irrelevant-information, and rating-scale changes in all seven pairs; six pairs also show lower option-order flip rates. They incur smaller accuracy losses under incorrect external judgments in all seven pairs and conflicting descriptions in six. In six pairs, they gain less accuracy from correct external judgments and show less favorable accuracy changes under consistent descriptions. Among pairs with unchanged semantic decisions, mean absolute changes in answer probabilities range from 4.56 to 11.55 percentage points, revealing input effects that accuracy alone misses.

[AI-7] he Geometry of Lies: Knowledge and Persona Directions for Detecting Deception in Large Language Models

链接: https://arxiv.org/abs/2611.26545
作者: Md. Abdullah Al Zami,Suraiya Shanaz,Mahdi Hasan,Shoaib Ahmed Dipu,Farig Yousuf Sadeque,Md. Sabbir Ahmed
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Linear probes are the cheapest LLM lie detectors but often fail to transfer across kinds of lie. Label-free directions from honest and deceptive role-play are untested at 70B scale and on lies about knowledge, where a lie and an honest denial can be identical. We widen a published persona bank with roles that deny knowledge they have, plus honest counterparts. At each model’s middle layer the resulting persona direction ranks lies across Liars’ Bench better than single-dataset probes on three of its four models; it fails on Qwen, even when Qwen stays in role far more often. On Llama-3.3-70B-Instruct it beats directions from a random persona bank, but its gain over the published direction is within the range of random widenings. On lies about knowledge, persona directions mainly track a reply’s wording: where lies and honest denials share their words, our three main contrasts stay inside the random-direction band on Llama and Qwen and reach only its edge on Gemma, though on Mistral one clears it. A knowledge direction contrasts questions the model can answer with ones nobody can, each followed by the same scripted denial. It separates lies from identical honest denials on two models and narrowly on a third, but not on the fourth; its other results partly depend on the benchmark’s wording. On the Gender-Secret fine-tune it clears the band by a few hundredths on Llama and fails on Mistral. The sum of the two directions ranks lies better than a supervised probe at the same depth on all four models (on one, both near chance) but is level with the persona direction on Llama’s average and behind that probe on held-out lies. More varied role-play questions do not help the persona direction on Liars’ Bench.

[AI-8] LLM GerryMandering: Separating Interactions from Additive Effects in LLM Audits

链接: https://arxiv.org/abs/2611.26538
作者: George Rahul,Ronit Jahilal,Bhaskara Hanuma Vedula,Ponnurangam Kumaraguru,Pushkar Shukla
类目: Artificial Intelligence (cs.AI)
备注: 40 pages, 8 figures, 29 tables

点击查看摘要

Abstract:Intersectional bias arises when a model treats people who hold several marginalised identities at once differently from what their separate identities would predict. Auditing such intersectional subgroups remains a challenge: their number grows quadratically, and there is no agreed baseline for what independent attributes would predict. Without such a baseline, many existing audits flag a subgroup when it exceeds each attribute alone, a condition that attributes acting independently already satisfy. We test this standard criterion on 150 synthetic datasets built with no interaction. With the shuffle null, it flags 66.1 false subgroups per dataset on average, and at least one in every dataset. When we plant interactions of known size into these datasets, a repaired null lets the criterion recover only 32% of them. We introduce ADDIS (Additive Intersectional Screen), which fits an additive model over all attributes, scores each subgroup by its Additive Residual Distance (ARD) from that model’s prediction, and tests it against a Freedman-Lane null. On our synthetic benchmark, ADDIS flags 0.09 false subgroups per dataset and recovers 84% of the same planted interactions. We then build IndiPersona, 10,000 synthetic Indian personas with six identity attributes, and audit 13 open language models. The standard criterion flags 7.1% of the 738 tested subgroups on average; ADDIS flags 3.0%. Gender identity crossed with sexual orientation is the largest pairing in every model, and counterfactual personas confirm it and additionally reveal a caste x class interaction in all but one model that the observational test lacked the data to detect. Much of what audits report as intersectional bias can thus arise from independent single-attribute effects alone, yet a small core of interactions survives.

[AI-9] Backtracking Agent Applications Security Incidents

链接: https://arxiv.org/abs/2611.26525
作者: Yicheng Cai,Taegyu Kim,Peng Liu
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注: 33 pages (12 pages of main text, 3 pages of references, 18 pages of appendices), 9 figures, 12 tables

点击查看摘要

Abstract:Agent applications are increasingly deployed, yet agent-specific attacks pose real threats to them. Upon a security incident, organizations backtrack to identify the CEP (compromise entry point), remediate the damage, and prevent its recurrence. The main artifacts available for backtracking are trajectory logs, which record agent functionalities, and source code; app(lication)-specific function call logs are often fairly limited. However, trajectory logs are at the agent functionality abstraction layer, while the CEP candidates are function calls at the app code execution layer. Due to this semantic gap, the analyst needs to manually inspect the relevant code for each CEP candidate to infer its corresponding agent functionality. In this work, we present Origin, a backtracking tool for the analyst to submit a small number of queries to identify CEPs that existing tools cannot provide. Our approach poses two challenges: (1) define the intermediate abstraction that captures the mapping relationships between the two semantic layers; (2) determine the region-to-variable mapping for constructing the cross-layer graph defined in the intermediate abstraction. Each region is a specific group of agent functionalities. Origin addresses the challenges by (1) the cross-layer dependency graph (CLDG) that connects region-inferred functions at the semantic level to the variables in the low-level taint propagation subgraph, plus the associated query language, and (2) the region inference algorithm. Case studies on five popular agent applications across eight attack incidents show that an analyst using Origin correctly identifies all the CEPs.

[AI-10] FLUID: Flexible Evidence-Grounded Situated Task Guidance

链接: https://arxiv.org/abs/2611.26491
作者: Yunqi Zhao,Run Peng,Ze Dong,Ryan Lu,Yinpei Dai,Joyce Chai
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Real-time task guidance is inherently flexible: when an agent guides a user through a task, multiple responses and timings can be appropriate, and sometimes the best action is to remain silent. Existing benchmarks often miss this flexibility by treating synthesized dialogues as fixed targets for what to say and when to say it. We introduce FLUID (Flexible, Evidence-Grounded Situated Task Guidance), a unified framework for modeling, evaluating, and enabling flexible interaction in multimodal task guidance. Using naturally occurring human-human guidance from WTaG and HoloAssist, FLUID grounds interventions in their supporting evidence, necessity, and timing flexibility, and uses this representation both to evaluate flexible guidance and to build a real-time agent for task guidance. Our agent achieves the strongest overall balance of response correctness, intervention precision, timely guidance, and response latency among evaluated systems, while still leaving room for improvement in avoiding speech overlap. Together, FLUID moves situated task guidance beyond rigid reference matching toward more natural and adaptive human-AI interaction. Project page: this https URL.

[AI-11] A new algorithmic approach for finding flat minima: critically damped Polyak stepsizes

链接: https://arxiv.org/abs/2611.26475
作者: Francesco Caporali,Jason M. Altschuler,Boris Hanin
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注: 56 pages, 19 figures

点击查看摘要

Abstract:We present a new adaptive stepsize rule that biases gradient descent toward flatter minima in both theory and practice, and establish the first uniform-in-time guarantees for both sharpness and loss reduction for a broad class of locally quadratic objectives. Our critically damped Polyak (CDP) stepsizes multiply the classical Polyak rule by 4 and a mild damping factor. The factor 4 is critical in that it produces equioscillations around the manifold of minimizers \mathcalM to leading order. Damping provably drives convergence to \mathcalM . CDP thus operates at the edge of stability by adapting its stepsizes to local curvature. Once the loss is small, its projected iterates follow gradient descent on \log\lambda_1(\nabla^2 \mathcalL) along \mathcalM to leading order, paralleling the flatness bias of normalized gradient descent (Arora et al., 2022). Unlike normalized gradient descent, CDP step lengths shrink with the distance to \mathcalM , enabling convergence to an interpolating solution that locally minimizes sharpness at finite learning rate. We establish our theory for a broad class of non-negative losses with locally quadratic Morse-Bott structure near \mathcalM , capturing regular interpolation regimes in smooth overparameterized neural networks. Empirically, CDP stepsizes produce a flattening effect similar to that of Sharpness-Aware Minimization (SAM, Foret et al., 2021), reducing sharpness even in settings beyond our theory, including nonlinear teacher-student networks and fine-tuning of pre-trained LLMs.

[AI-12] Monte Carlo Pass Search for Setpieces

链接: https://arxiv.org/abs/2611.26464
作者: Andrew Kang
类目: Artificial Intelligence (cs.AI)
备注: Hudl Performance Insights 2026

点击查看摘要

Abstract:Set pieces produce about a third of non-penalty expected goals (xG), yet they are usually evaluated by what happened next. We present Monte Carlo Pass Search (MCPS) for set pieces, which treats a corner or a lofted free kick as a distribution over plausible executions and plausible futures instead of a single realised event. From the pre-kick state, MCPS proposes counterfactual deliveries and counterfactual set-ups (where the attackers stand at the kick), plays each out many times in a world model (3D ball physics, a generative model of all 22 players, the keeper’s come-or-stay decision and a Player-to-Touch model) and scores every future with a value model. On 249 deliveries from 30 Serie A 2025/26 matches, with every model applied only to matches it never saw, the best available delivery is worth 0.0085 xG more than the one kicked (+28 %), mainly because it reaches an attacker before the keeper. Searching where the attackers stand as well adds +0.0027 xG with the observed ball and +0.0033 on top of the best ball; designing their runs too adds +0.0007 more. We show how to keep such a search honest, and estimate how much more data each component needs before a club could rely on it.

[AI-13] A CPU Inside an LLM : Zero-Gradient Compilation of Exact Circuits into Pretrained Weights

链接: https://arxiv.org/abs/2611.26450
作者: Tanmay Sah
类目: Artificial Intelligence (cs.AI)
备注: 27 pages, 3 figures, 24 tables

点击查看摘要

Abstract:Pretrained language models (LMs) remain unreliable on exact multi-digit arithmetic without gradient fine-tuning or external tool interpreters. We introduce closed-form, zero-gradient circuit-grafting primitives that compile deterministic computation directly into pretrained LM weights using zero training examples and zero runtime hooks, while bounding per-layer normalization drift and preserving natural-language (NL) competence. Our construction combines three analytical mechanisms: (1) two-layer Mellin-GLU quadrature over bias-free SwiGLU/GeGLU layers, turning unknown per-token RMSNorm scaling into translation along a log-spaced gate grid ( 250\times dynamic range) with IEEE 754 error bounds; (2) cross-token scale cancellation via signed Chebyshev Indefinite-RoPE self-sink attention and memoryless SiLU-antisymmetric GatedDeltaNet shifts; and (3) axis-aligned width extension ( d \to d+K ) with \sqrtd/(d+K) norm compensation, Base-1000 canonical BPE emission, and differential (+254,-256) tied unembeddings. An automated 5-stage compiler synthesizes 4-8 digit signed +,-,\times\ circuits in constant \mathsfTC^0 depth alongside 3-literal Boolean (100%, N=1,728 ; plus 42-layer Gemma 4 E4B audits) and 5-bit Majority/Parity (100%) circuits. On 6-digit arithmetic ( N=1,200 , BF16), IBM Granite 4.2 3B (1.07M-1.33M active non-zeros, \le 0.036% ) and Alibaba Qwen 3.5 4B (2.25M-2.39M non-zeros, \le 0.057% ) reach 99.17% and 100.00% ALU-prefix accuracy (100% with an FP32 sidecar) and 98%-100% mid-sentence resumption, outperforming 12.8K-sample LoRA (42.2% / 61.7%, 0% inline continuation). A factorized block-sparse sidecar holds A100 memory at 7.89 / 10.65 GiB (4.60 / 8.14 GiB under INT8/INT4 + FP32 sidecar) while passing all 14/14 pre-registered 2 K-token Tango splits ( N=4,392 ; 0/11,134 OOD triggers).

[AI-14] Seer: Guiding Small Deep-Research Agents with Shared Decision Models

链接: https://arxiv.org/abs/2611.26423
作者: Yuhao Zhou,Yang Zhou,Yuanxi Li,Jindi Lv,Dan Si,Qing Ye,Jiancheng Lv
类目: Artificial Intelligence (cs.AI)
备注: 24 pages

点击查看摘要

Abstract:Deep-research agents repeatedly decide what to search, read, and verify, often using large cloud-hosted language models for action proposal. However, privacy and local control motivate running agents on users’ devices, where limited memory and compute favor smaller models and complicate on-device training. Yet inference-time guidance faces three obstacles: exploring alternatives through real execution adds tool calls, prompted self-evaluation is bounded by the agent’s judgment, and model-specific guidance risks repeated adaptation costs. We introduce Seer, an inference-time framework providing local guidance that separates the agent’s action proposal from learned consequence prediction and value estimation. From a shared corpus of real research trajectories, we train two models: a tool-response simulator on observed call–response pairs and a value estimator on histories labeled with discounted terminal correctness. The simulator and estimator are trained offline and reused across agents without further training. Moreover, Seer executes only the selected action and adds only real tool responses to the committed history, avoiding tool calls on discarded candidates at additional inference cost. Across multiple deep-research benchmarks, Seer substantially improves agent performance for different language models. For instance, GAIA accuracy rises from 13.6% to 41.7% for Qwen3.5-4B and from 22.3% to 50.5% for Qwen3.5-9B. Furthermore, our analyses suggest that broader candidate coverage can compensate for imperfect simulation, while prediction and scoring errors limit the benefits of wider imagined search. These findings highlight the potential of reusable models for prediction and evaluation to strengthen small research agents through extra local computation without the real execution of the discarded candidates.

[AI-15] Reasoning -Trace Analysis Using Large Language Models for Interpretable Q-Matrix Construction

链接: https://arxiv.org/abs/2611.26422
作者: Jihong Zhang,Zijing Jin,Mingfeng Xue
类目: Artificial Intelligence (cs.AI)
备注: 5 figures

点击查看摘要

Abstract:Large language models (LLMs) can generate Q-matrices for diagnostic classification models that approximate expert specifications, but evaluations have focused on the final matrices, leaving the basis of individual item-attribute assignments unexamined. This opacity limits the responsible use of AI in measurement. We propose Reasoning-Trace Analysis (RTA), a framework in which an LLM generates attribute-labeled reasoning steps through chain-of-thought (CoT) prompting and these steps are analyzed to reconstruct, compare, and inspect each assignment. RTA is intended to supplement expert review and empirical validation. In a simulation study with perturbed fraction subtraction items, four LLMs recovered the required attributes and withheld attributes the items did not require. In an empirical study with 11 fraction subtraction items, responses from 536 examinees, four LLMs, up to three temperatures, and 100 replications per condition, CoT-derived Q-matrices overlapped substantially with the reference Q-matrix (mean overlap .809) but less than directly generated Q-matrices (.877), and they varied more across replications. Inspection of the traces linked disagreements to item-type coding, labeling rules, and how operations were expressed as steps. Q-matrices with similar overlap differed in fit to examinee responses. RTA provides inspectable evidence that matrix-level agreement and fit do not.

[AI-16] ZeroVLN: Learning to Navigate without Supervised Fine-Tuning of the VLM Brain

链接: https://arxiv.org/abs/2611.26414
作者: Zixuan Huang,Yang Zhou,Guli Zhang,Sunzhu Li,Hongyan Xie,Zehao Chen,Fuzhen Zhuang,Kaixuan Wang,Xiaozhi Chen,Yikun Ban,Deqing Wang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Representative vision-language navigation (VLN) methods typically couple a vision-language model (VLM) brain with a learnable action head and fine-tune the VLM brain on navigation-domain data to predict geometric goals. However, such adaptation can substantially alter base VLM weights and degrade general visual capabilities. We introduce ZeroVLN, a slow-fast navigation framework that achieves effective navigation with zero navigation-domain SFT of the VLM brain. Rather than teaching the brain to predict geometric goals, we equip it with a semantic tool-calling interface for conveying movement decisions to an action expert in natural language. We first adapt the action expert to carry out these movement requests, then jointly improve the brain’s movement decisions and the action expert’s continuous execution through closed-loop reinforcement learning (RL). Trained on the same navigation episodes, ZeroVLN substantially outperforms representative SFT-based hierarchical VLN methods on standard VLN-CE (continuous environments) benchmarks, while inducing only minor changes to the brain’s weights and largely preserving its general capabilities. To our knowledge, ZeroVLN is the first slow-fast VLN framework with zero navigation-domain SFT of the VLM brain, and the first to jointly optimize its slow and fast policies through closed-loop agentic RL in VLN-CE.

[AI-17] World Models Favor Learned Priors over Visual Evidence

链接: https://arxiv.org/abs/2611.26366
作者: Han Xiao,Zhiyong Bao,Xiaotong Liu,Yanhao Chen,Qingqiang Wu
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:World models are increasingly used to simulate physical interactions, but training a separate model for each environment limits scalability. Cross-environment prediction requires identifying the current physical rule, which appearance alone may not reveal. We therefore study visual calibration: using object motion and interactions in a short video as direct evidence to infer the current rule and predict motion under that rule from a different initial state. However, differences in data availability across physical conditions can yield a long-tailed training distribution, biasing models toward common rules even when evidence supports rare ones. To study this challenge, we introduce TailWorld, which evaluates the transfer of demonstrated rules across initial states while independently controlling training imbalance and calibration informativeness. The benchmark compares predictions with two reference futures from the same prediction initial state, one following the demonstrated rule and the other a common rule, to measure prediction bias toward common dynamics. We further propose Rule Inference for Transfer (RIFT), which learns how different rules affect object motion, infers the current rule from calibration interactions, and applies it to prediction from a new initial state. RIFT explicitly controls the influence of training frequency on rule inference, allowing sufficiently informative observational evidence to support rare rules. Experiments on TailWorld show that RIFT improves prediction accuracy and physical consistency across varying training imbalances and distribution shifts, with gains in predicting rare dynamics.

[AI-18] How Much Should an Agent Remember? Deployment Lessons from Self-Distilled Memory for LLM Agents

链接: https://arxiv.org/abs/2611.26361
作者: K. R. Jayaram,Vatche Isahagian,Aswath Vaithinathan Aravindan,Vinod Muthusamy,Punleuk Oum,Gaodan Fang,Michal Ulewicz,Gegi Thomas,Evelyn Duesterwald
类目: Artificial Intelligence (cs.AI)
备注: Accepted to IAAI 2027 (Oral)

点击查看摘要

Abstract:LLM agents on long-horizon tasks fail in recurring, procedural ways. They mis-paginate an API, resolve the wrong entity, or answer when no answer was asked. We built a pipeline that lets an agent learn from its own history. It first mines short \emphlessons from the agent’s own trajectories, one to three per episode. It then consolidates those lessons into a smaller set of \emphguidelines, where each guideline records how many lessons it merged. At inference the agent receives either all guidelines or a selected subset. The pipeline ships in the open-source altk-evolve library and is deployed in IBM’s CUGA agent and watsonx Orchestrate platform. We report what we learned building, measuring, and productizing it. Our primary instrument is the AppWorld benchmark, across models spanning 40–92% baseline task completion. The central finding is that the right amount of memory is a property of the model, not the memory. A strong model keeps improving as guidelines are added. DeepSeek-V3.2 gains +9.5 points of task completion and +16.1 points of scenario completion with its full 191-guideline set. A weak model gains less from the same full set than from a small selected subset. gpt-oss-120b starts at 39.9% task completion. It reaches 51.2% with all 199 of its guidelines but 55.4% (mean of two runs) with a selected subset of about 29. Around that result we report what went wrong: a high-support guideline that backfired, provenance silently dropped during consolidation, a swallowed API error that failed every task while producing a plausible log, and a baseline confound that nearly produced a misleading comparison with a prior method. We distill these into recommendations for teams deploying agent memory.

[AI-19] Context as a Resource

链接: https://arxiv.org/abs/2611.26346
作者: Xufei Tian,Han Hu,Wenli Du
类目: Artificial Intelligence (cs.AI)
备注: 27 pages, including appendices. Code: this https URL

点击查看摘要

Abstract:Managing concurrent tasks across separate agent sessions quickly exhausts human attention, motivating always-on personal agents behind a single continuous interface. Yet when every context shares identical frozen backbone weights, funneling interleaved dialogues and external events into one linear session induces cross-topic constraint bleed and head-of-line stalling. Conversely, router-ticket hierarchies incur lossy prose handoffs, heavy routing overhead, and blind merges - collapsing true-conflict repository merges to a 33.3% three-run mean despite high isolated branch passes. Rather than scheduling anthropomorphic roles, we introduce ClonAct, a routerless architecture that virtualizes context blocks directly through a pluggable single-pass non-generative Context MMU, in-place Context Cells, a footprint-aware Synaptic Bus with an isolated trial merge gate, and a wave-flushed Shared Trunk. Across three programmatic multi-stream benchmarks and three model families, our portable trial merge gate lifts true-conflict merges to 63.3%, while ClonAct’s routerless virtualization attains this ceiling with 45.9% fewer prompt tokens and prevents compact-model router collapse. Under continuous surge streams, ClonAct matches top completion (95.0%) with the lowest median wait and zero timed-out status checks.

[AI-20] DeepQuest: From Knowledge Gaps to Scientific Discovery in Alzheimers Disease

链接: https://arxiv.org/abs/2611.26343
作者: Chunbo Jiang,Yifei Chen,Beining Wu,Fuyou Mao,Weihong Chen,Enyu Bao,Mingxuan Liu,Feiwei Qin,Qiyuan Tian
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Automating experiments does not necessarily advance scientific understanding. Research agents must identify meaningful knowledge gaps, determine what available data can resolve, and assess whether their findings advance existing knowledge. We introduce DeepQuest, a knowledge-gap-driven AI scientist framework for Alzheimer’s disease research. DeepQuest transforms unresolved questions from domain reviews into traceable, testable hypotheses and prioritizes them through data feasibility screening and groupwise Elo ranking. It couples iterative experimentation with two-level auditing that separates experimental reliability from knowledge contribution relative to two frozen literature corpora. On OASIS-3, DeepQuest produces 42 feasible hypotheses and outperforms the strongest baselines in 14 of 15 evaluator-dimension comparisons across three language-model reviewers. The full framework achieves an audit-assessed scientific finding rate of 95.24%, outperforming both general-purpose executors and domain-specific research agents. This scientific finding rate remains stable across literature corpora and evaluator model families. Representative findings concern neuropsychiatric symptoms and dementia risk, longitudinal amyloid spatial stability, and neocortical tau-associated clinical impairment, with key associations supported in independent ADNI and AIBL cohorts. These results show how aligning knowledge gaps, data feasibility, and evidence auditing can turn automated research into traceable, evidence-supported scientific discovery. Our source code is available at this https URL

[AI-21] raining LLM s to Follow Reasoning Instructions with Minimal Edits NEURIPS2026

链接: https://arxiv.org/abs/2611.26342
作者: Alexandra Yakovleva,Luat Gia Khoi Nguyen,Juho Kannala,Harri Valpola,Alexander Ilin
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the NeurIPS 2026 Workshop on Transitioning from Pre-training to Post-training

点击查看摘要

Abstract:Large language models increasingly rely on extended reasoning traces, yet the training signals used to shape this behavior remain coarse. Supervised fine-tuning on full trajectories forces the model to imitate an entire demonstration, overconstraining its output distribution, while trajectory-level reinforcement learning provides only weak credit assignment, rewarding or penalizing whole completions without identifying where they succeed or fail. We propose edit-based training, a general framework for learning from a model’s own trajectories via localized textual corrections. Given a sampled completion, an editor identifies the first point of failure under a specified criterion and proposes a minimal replacement span. The model is then supervised only on these edited spans. By design, this concentrates supervision where the model’s reasoning first breaks down and aims to preserve its existing reasoning patterns elsewhere. The method applies both offline and online, with the latter continually adapting supervision to the model’s evolving failure modes. We instantiate this approach on reasoning instruction following with Qwen3-32B on the ReasonIF benchmark. Compared to full-trajectory imitation and trajectory-level reinforcement learning baselines, edit-based training substantially improves compliance with the reasoning instruction while largely preserving the model’s skills on untargeted tasks. These results suggest that localized correction is an effective complement to these approaches for shaping how LLMs reason.

[AI-22] Sirius-Tiamat: How Small Can a One-Pass Decision Encoder Get Before Its Decisions Change?

链接: https://arxiv.org/abs/2611.26339
作者: Mohammed Nayeem Khan,Vrishank Tiwari
类目: Artificial Intelligence (cs.AI)
备注: 38 pages, 23 figures, 27 tables

点击查看摘要

Abstract:A one-pass decision encoder reads a text, a question and several named options, and returns a probability for each option in a single forward pass. We ask how far Laya, a 421M-parameter encoder of this kind, can be compressed before its decisions change. We count decisions that differ from Laya’s own full-precision (fp32) answer, and track calibration (how well confidence matches accuracy, measured by expected calibration error, ECE), on 1,221 synthetic items with the training schemas, 273 with an unseen schema, and 800 out-of-family questions from ARC, MMLU and HellaSwag. Layer pruning of a related model family is known; we add flips, calibration, held-out data and retraining. bf16 changes 0.2% of in-family decisions but 2.2% out of family; 6-bit grouped rounding 1.3% and 5.9%; deleting 6 layers 0.9% and 9.1%, while ECE rises from 5.9% to 39.2%. One fitted temperature, needing no labels, restores it to 3.6%. One scale per tensor fails below six bits, and CPU dynamic int8 changes 30% of decisions, nearly all through one layer. Total variation tracks flips for rounding but not for deletion, and an early exit saves layers over a static cut only if its confidence gate is first recalibrated. Distilling a 16-layer student cuts its in-family flips from 44% to 9% (45% out of family). At equal stored size, rounding flips fewer decisions than deletion. Data are synthetic and LLM-labelled, for one architecture.

[AI-23] ClosureBench: A Benchmark for Reasoning under Closure Assumptions

链接: https://arxiv.org/abs/2611.26335
作者: Francesco Belardinelli,Flavio Corradini,Matteo Leonesi
类目: Artificial Intelligence (cs.AI)
备注: Accepted at the 20th International Conference on Neurosymbolic Learning and Reasoning (NeSy 2026). Conference version: this https URL

点击查看摘要

Abstract:AI agents built on large language models (LLMs) often answer questions from incomplete records. Missing information might be interpreted as unknown or false, depending on the completeness of the source. Current evaluations fail to distinguish these two aspects, as they are usually interwoven with other reasoning demands. We present ClosureBench, a benchmark that tests whether LLMs follow the stated rules for dealing with missing information. Every base scenario is tested under OWA, CWA, and LCWA, and only the treatment of missing information changes between variants. ClosureBench provides a verified base benchmark as well as agent extensions for action choice, source scope, and updates. A model is considered correct on a base scenario only if it provides correct answers in all of its closure variants. Across 960 instances, overall accuracy can mask substantial closure mistakes. Models achieve overall accuracy from 71.18% to 99.86% across replicated runs, but their core switch accuracy, measured on the scenarios whose gold label actually changes with the contract, varies between 18.23% and 99.31%. Large discrepancies emerge under LCWA. For example, some agents choose unknown despite the prompt indicating that missing information should be treated as false. Our findings demonstrate that evaluations of AI agents should not only measure whether agents can reason over existing facts but also whether they use the appropriate assumption when information is absent.

[AI-24] Rigel-Ladon: An Option-Count Law for a One-Pass Decision Encoder Tested on Held-Out Option Counts and Held-Out Schemas

链接: https://arxiv.org/abs/2611.26332
作者: Mohammed Nayeem Khan,Vrishank Tiwari,Navin Kumar R
类目: Artificial Intelligence (cs.AI)
备注: 43 pages, 29 figures, 35 tables

点击查看摘要

Abstract:Laya is a small neural network that reads a text, a question and K written options, and returns one probability per option in a single pass. We ask how its accuracy falls as K grows from 2 to 35 and whether that can be predicted. We score nested option sets (every K-option question contains the smaller ones) on four question pools, for the released model (Base) and our fine-tune (FT), and score every item against each single rival. FT trained on one 12-option topic question (in-family); a 12-option HR (department-routing) question is out-of-family. Logit accuracy (the log of correct over incorrect answers) is linear in log(K-1) with slope 0.68 to 0.88. Fitted on K up to 6, the law predicts K=7 to 12 within 0.6 points and K up to 24 within 1.2 (a log-linear rule: 1.6 and 6.7). Taking the slope from the topic question and measuring only two-option accuracy on HR predicts its other K within 0.3 points (FT) and 0.7 (Base). On a 35-label pool of five schemas the law fails (5 to 11 points). Independence of irrelevant alternatives is rejected: the gold-versus-rival log-odds shifts by +0.24 to +2.07 nats per doubling of K, yet pairwise answers reproduce K-way accuracy within 4.7 points. Order averaging (averaging over several orderings of the options), shortlists, a temperature law and calibrated prediction sets did not transfer (FT’s 90% sets cover 84% on HR). 2 of 10 predictions made in advance held. Caveats: one model family and LLM-labelled synthetic text.

[AI-25] A Control-Theoretic Approximation to Predictive Coding Dynamics

链接: https://arxiv.org/abs/2611.26319
作者: Ryan Fayyazi,Kyle Daruwalla,Mitra Javadzadeh
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Neurons and Cognition (q-bio.NC)
备注: 17 pages, 10 figures

点击查看摘要

Abstract:Backpropagation remains the dominant method for training deep neural networks, but its reliance on non-local computations has motivated the search for biologically plausible alternatives. Two prominent frameworks, predictive coding (PC) and deep feedback control (DFC), offer distinct approaches to local learning, yet their relationship remains unclear. Here, we show that DFC arises as a local approximation to predictive coding dynamics near fixed points, with exact equivalence in shallow linear networks and systematic deviations arising with depth, nonlinearity, and distance from equilibrium. We characterize how this approximation, which we call Approx PC, deviates from PC in its activity dynamics, and show that these deviations can be mitigated by relinearizing the feedback controller and approaching the output target through a sequence of intermediate nudges. In networks trained on MNIST and FashionMNIST, Approx PC reaches task performance comparable to PC trained with the same nudges, and weights learned with it transfer directly to PC inference. Our results unify energy-based and control-theoretic perspectives on learning and open the door to applying tools from control theory to the training of predictive coding networks.

[AI-26] Actor-Predictor Architecture for Reinforcement Learning with Formal Specification Objectives

链接: https://arxiv.org/abs/2611.26313
作者: Josef Brozovich,Qining Zhang,Lei Ying
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:One fundamental challenge in reinforcement learning (RL) is enforcing temporal safety constraints, such as satisfying formal specifications. Existing methods often transform logic formulas into automata for state augmentation and reward shaping. However, this approach assumes that automaton construction is both feasible and tractable, which limits its applicability to small specifications and linear temporal logic (LTL). This paper proposes Actor-Predictor, a framework that learns a surrogate representation of logic progress without explicitly constructing an automaton. The predictor continuously estimates whether a partial trajectory will satisfy the formal specification in a later step, and its learned features are provided as auxiliary inputs to the actor, which is then optimized using GRPO and RWR. We show theoretically that, under sufficient predictive accuracy, the learned representation is equivalent to the automaton state. Experiments on grid world and robotic tasks in Isaac Lab demonstrate that Actor-Predictor outperforms automaton-based baselines on LTL-constrained tasks. It also performs well on non-regular formal specifications, such as (a^n b^n) , for which finite-automaton-based approaches are not applicable.

[AI-27] Client Exits - A Way to Increase LLM Transparency on the Client Side

链接: https://arxiv.org/abs/2611.26309
作者: Marc Jansen
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Textual recommendations generated by large language models (LLMs) can be difficult to inspect. This paper introduces \emphClient Exits, a lightweight mechanism for enriching LLM responses while separating semantic intent from client-specific presentation. The model inserts predefined, machine-interpretable markers into its output, and the receiving client decides whether and how to realize the corresponding enrichments. This allows supplementary information to be presented as graphics, text, speech, or other modalities without prescribing a concrete interface. We implement a prototype combining a finetuned Qwen model with a modified LibreChat client. Using the Question–Option–Criteria (QOC) approach including three XML-based client exits. We evaluate two QOC-oriented model variants trained on essentially the same examples, with client-exit annotations added for one variant. Across 100 test questions, the client-exit model includes all three markers in every response. An LLM judge, informed about the mechanism without being instructed to favor it, prefers this model’s answers in 66% of the comparisons. These results demonstrate successful marker generation on the evaluated test set and support the feasibility of client-controlled enrichment.

[AI-28] Similarity-based OSF Logic for Flexible SPARQL Query Answering ISWC2026

链接: https://arxiv.org/abs/2611.26261
作者: Gian Carlo Milanese,Gabriella Pasi
类目: Artificial Intelligence (cs.AI); Databases (cs.DB); Logic in Computer Science (cs.LO)
备注: Accepted at ISWC 2026 (Demo)

点击查看摘要

Abstract:We present an updated version of fosf, a reasoner for fuzzy order-sorted feature (OSF) logic, with support for flexible SPARQL query answering over RDF knowledge graphs. Fuzzy OSF logic is a knowledge representation and reasoning language where feature symbols denote functions and sort symbols denote fuzzy sets. Sorts are organized in a fuzzy subsumption relation that extends to OSF terms, record-like structures representing classes of entities. The unification algorithm for these structures provides a calculus of fuzzy type subsumption. The latest release of fosf introduces support for similarity-based reasoning by extending the OSF language with a similarity relation among sorts, alongside a translation engine for compiling OSF terms into standard SPARQL. Queries expressed as OSF terms undergo fuzzy ontology-aware rewriting prior to translation. This process incorporates an OSF theory of sort definitions, feature domain and range declarations, and fuzzy subsumption and similarity relations through fuzzy OSF constraint normalization. This rewriting enables flexible query answering by returning approximate solutions, allowing queries that would otherwise be inconsistent with the underlying ontology to yield relevant results.

[AI-29] SkillBridge: Cost-Aware Diagnosis and Repair of Cross-OS Compatibility Defects in Agent Skills ICSE2027

链接: https://arxiv.org/abs/2611.26256
作者: Anjie Xu,Zixing Wang,Yifeng Cai,Leye Wang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注: 12 pages, 4 figures, 3 tables. Accepted at ICSE 2027. Code and data: this https URL

点击查看摘要

Abstract:LLM agents increasingly reuse executable capabilities through skills, but a skill that works on one OS may fail on another because OS assumptions are spread across guide text, scripts, metadata, install contracts, and local entrypoints. Diagnosing such failures by running every skill on every OS is expensive: many executions spend model tokens, runner time, and review effort before revealing whether the failure is truly OS-related. We present SkillBridge, a contract-based workflow with three stages: static analysis, dynamic validation, and repair. SkillBridge first screens 3,270 skill-target-OS cases, covering 1,090 skills across Linux, macOS, and Windows, to find OS-sensitive contract evidence. It then runs dynamic tasks and root-cause review on selected cases to confirm whether a failure is caused by an OS assumption, and repairs the contract field that caused the failure, such as guide text, package scripts, dependencies, or platform declarations. Repair raises the reviewer-judged target-task pass rate from 30.6% to 38.4%, a gain of 7.8 percentage points. A retrospective Linux/macOS budget analysis further shows the efficiency effect: SkillBridge reaches 38 confirmed OS issues with 185 dynamic validations, where uniform random selection would require about 865 validations in expectation.

[AI-30] he Stability Gap in Continual Reinforcement Learning

链接: https://arxiv.org/abs/2611.26252
作者: Sebastian Pusch,Matthia Sabatelli,Gido M. van de Ven
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Continual reinforcement learning (CRL) aims to enable agents to adapt to changing tasks while preserving previously acquired capabilities. Evaluation often relies on final aggregate metrics that can hide temporary performance degradation during adaptation. This work introduces a simple CRL benchmark that varies observation representations while preserving environment dynamics and rewards, supporting proof-of-concept demonstrations, inexpensive experimentation, and easily reproducible sanity checks. Using this benchmark, we demonstrate the stability gap in CRL: a substantial drop in prior-task performance after the introduction of new tasks, followed by partial or complete recovery. This pattern occurs despite the agent having complete access to past tasks.

[AI-31] FastRoverVLA: Fast Vision-Language-Action Execution for Off-Road Autonomous Driving

链接: https://arxiv.org/abs/2611.26242
作者: Changtao Zhao,Yingxue Guan,Li Wang,Shijie Zhao,Yuhang Zhang,Shengkai Zhang,Qing Wang,Ziying Song
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 14 pages, 6 figures, 8 tables

点击查看摘要

Abstract:Off-road vision-language-action (VLA) models generate scene-conditioned plans, but frequent trajectory updates can repeatedly invoke expensive multimodal inference. Reusing a prediction reduces this demand while leaving its commands exposed to changes in vehicle motion. FastRoverVLA addresses this mismatch by decoupling visual plan generation from motion-conditioned execution. Its training-free interface maintains decoded plans between visual predictions. A speed-curvature action buffer is aligned with measured motion at each update, integrated into a local trajectory, and advanced by consuming the current action. The corrected remainder is written back to the buffer, preserving motion corrections across successive updates. A bounded refresh rule normalizes speed and curvature variations within a newly decoded plan and assigns a short or long reuse budget according to the resulting plan-roughness score. The interface requires neither backbone features nor an additional learned driving policy. Over a 3 s horizon, independent real-request evaluation achieves average displacement error (ADE) and final displacement error (FDE) of 0.328/0.596 m on ORAD-3D and 0.266/0.526 m on RELLIS-3D, reducing FDE by 18.0% and 36.7% at request rates of 6.0% and 4.8%, respectively. These runs retain lower errors across four backbones and both datasets and reduce Qwen3-VL request-processing cost by factors of 16.6 and 20.8 under serial replay, while mechanism and motion-only controls identify persistent state correction as the main source of the observed accuracy benefit. The source code and evaluation materials are publicly available at this https URL.

[AI-32] Physics-Informed Feature Representations and Pretrained Tabular Foundation Models for Generalizable Melt Pool Geometry Prediction

链接: https://arxiv.org/abs/2611.26235
作者: Rafid Enayet(1),Imtiaz Ahmed(2),Md Habibor Rahman(1) ((1) University of South Carolina, (2) West Virginia University)
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Developing generalizable, data-efficient machine learning (ML) models for advanced manufacturing remains challenging because experimental datasets are limited and process-property relationships vary across materials and operating conditions. In laser powder bed fusion (LPBF), melt pool width, depth, and melting mode (conduction, transition, or keyhole) govern process stability and defect formation. Conventional ML models typically use dimensional process parameters, which represent laser-material interactions only implicitly and may limit cross-material transfer. This study evaluates whether two physics-informed dimensionless features, normalized enthalpy and Peclet number, combining laser processing conditions with material thermophysical properties, improve melt pool prediction over raw process parameters. We systematically benchmark seven models: four classical ML methods (Gaussian process, support vector machine, random forest, and XGBoost) and three pretrained tabular foundation models (TabPFN, TabDPT, and TabICLv2). Experiments span 316L stainless steel, IN625, IN718, and combined-material datasets, covering regression of melt pool width and depth and classification of melting modes. Dimensionless features yield comparable or superior predictive performance in most settings, with the largest gains in cross-material depth prediction (R^2 = 0.96). Tabular foundation models generally achieve stronger classification performance than classical baselines, particularly for multi-material datasets. These findings indicate that combining physically meaningful dimensionless representations with pretrained tabular models can improve data-efficient melt pool modeling and support more transferable process-window identification in LPBF.

[AI-33] Sightline: Calibrated Data-Flow Summaries for LLM -Integrated Applications

链接: https://arxiv.org/abs/2611.26221
作者: Zihao Xu,Yuekang Li,Ruitao Feng,Haifeng Shen
类目: Programming Languages (cs.PL); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Programs increasingly call large language models (LLMs), filling prompt placeholders with program values and consuming the output. Such a call crosses the boundary between natural language and programming language (NL/PL). To follow data across it, an analysis needs a data-flow summary of which placeholders the output uses, but no code describes what the model does. Analyses of these LLM-integrated applications therefore taint every output with every input, creating spurious flows, or ask a second model to label each call, which is costly. We prove, and mechanize in Lean 4, that a placeholder flows to the output of a transformer, to first order, only through the attention that reads it (no reading, no flow), and that a faint reading or sensitivity cannot certify that a placeholder does not flow. Unlike reading, the log-probability gradients and Fisher signals of existing methods vanish on confident outputs, and single-span removal fails on duplicated content. Our tool, Sightline, reads a recorded call in one forward pass, learns from labels which reads are uses, and calibrates the pruning to miss at most a chosen fraction \alpha of true flows, on average over call sites, when deployment is exchangeable with calibration. On 2,103 unseen repository families and four models, Sightline is the most accurate summary against the seven closest methods and a hidden-state probe. Calibrated at \alpha = 1%, it misses about 1% of true flows and prunes 43.9-65.6% of spurious flows, 1.6 to 2.6 times as many as the best baseline, and still 43.8-60.1% when calibrated on another collection. Through an open surrogate, it reads models treated as closed about as accurately as they read themselves, without their labels. In the information-flow control of an LLM agent, it removes about half of the labels from inputs that do not flow.

[AI-34] Budget Internalization: Fixed-Budget RLVR Trades Test-Time Scaling for In-Budget Accuracy

链接: https://arxiv.org/abs/2611.26172
作者: Ilia Zavidnyi,Alexander Kovrigin,Arina Puchkova,Kirill Gelvan,Sergey Titov
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 39 pages, 26 figures, 16 tables. Checkpoints: this https URL

点击查看摘要

Abstract:Reasoning models often become more accurate when they can generate longer answers at test time. This behavior is known as test-time scaling. However, RL with verifiable rewards (RLVR) usually trains them with a fixed rollout cap. We find that this cap changes model behavior, not only training cost. After training, models use few additional tokens beyond the training budget. They outperform the base models near that budget, but their accuracy stops improving and falls below the base model at larger budgets. The rollout cap thus becomes a learned constraint, that limits model’s test-time scaling. We call this effect budget internalization. We observe it across Qwen3.5-4B, Gemma4-E2B, and Nemotron-3-Nano-4B and two reasoning training tasks: mathematical reasoning and diverse, procedurally generated problems. This effect also transfers from math training to science and coding tasks. Training-objective level ablations on Qwen3.5-4B show that most changes recover little test-time scaling. Stronger KL regularization and lower training-data truncation preserve more scaling, but stronger KL reduces in-budget accuracy. Forcing an answer on truncated training rollouts largely removes budget internalization, but also removes most of the in-budget gain. SFT on artificially truncated traces reproduces budget internalization, whereas SFT on naturally short traces does not, which shows that the effect is not specific to RL. This result links budget internalization to length-censored post-training rather than to an RL-specific objective term. Thus, a rollout cap does not only control training cost. It can trade low-budget gains for lower performance when additional inference compute is available.

[AI-35] he FrontierOR Challenge: Agent ic Algorithm Design for Operations Research

链接: https://arxiv.org/abs/2611.26171
作者: Hai Wang,Bretislav Hajek,Chonghe Jiang,Minwei Kong,Ao Qu,Huatian Gong,Bo Dong,Cynthia Barnhart,Jinhua Zhao
类目: Artificial Intelligence (cs.AI); Optimization and Control (math.OC)
备注:

点击查看摘要

Abstract:Operations research (OR) relies on algorithms that experts tailor to each problem: a new routing, scheduling or network design problem can take weeks of specialist work before it is solved well at scale. Large language model (LLM) agents now write, test and improve code, but whether they can design algorithms that scale to realistic instances is open: in the FrontierOR benchmark’s one-shot evaluation, the strongest model matches Gurobi, with an objective within 1% of Gurobi’s and no slower, on only 39% of the large instances. The FrontierOR Challenge asks teams to build agentic workflows that design, test, and improve algorithms for many optimization problems at once. Stage 1 releases 60 of the benchmark’s 180 audited problems, spanning six application categories and five formulation types, in weekly rounds of 15; Stage 2 introduces new problems outside the benchmark, with unseen instances. Two tracks are ranked separately: in the Main track we run the submitted algorithms in a network-isolated sandbox with open-source dependencies only; in the Self-hosted track teams solve the instances on their own machines, with any software, within a timed window. Every solution is checked for feasibility, and a feasible solution scores the ratio of its objective to a frozen known reference, averaged over hidden instances.

[AI-36] Sage-Evolve-Preview:Adaptive Observation for Post-Training Self-Improvement

链接: https://arxiv.org/abs/2611.26132
作者: Yinhao Tang,Youqing Fang,Yanan Sun,Xintian Shen,Hanyu Sun,Wenran Liu,Yue Wu,Weiming Zhang,Bin Liu,Wenwei Zhang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Post-training recursive self-improvement (RSI) iteratively improves a target capability, operationalized by a fixed anchor benchmark, through repeated intervention, training, and evaluation. Yet most loops use the anchor’s aggregate score as their sole observation, so early gains give way to stalled progress and repetitive exploration. When this anchor-only feedback ceases to reveal actionable gaps, a plateau may reflect an exhausted observation rather than an exhausted capability. We introduce adaptive observation (AdaObs), an observation-side mechanism of harness evolution that keeps the anchor objective fixed while adapting the active observation. At a plateau, a controller refines diagnostic granularity, escalates difficulty, and adds related tasks; validated increments are fused back into training, while every checkpoint remains evaluated against the anchor. Across eight independently optimized anchors in long-context reading comprehension and instruction following, AdaObs raises the mean score from 66.4 to 73.7 under a matched data budget, outperforming FixedObs, the anchor-only baseline. The evolved harness further improves the seven-task average on the independent PostTrainBench across two agent backends. These results identify observation adaptation as a missing mechanism of harness evolution for sustaining post-training RSI.

[AI-37] ChronoPatch: Black-Box Energy-Latency Attacks on Vision-Language Models

链接: https://arxiv.org/abs/2611.26117
作者: Jiawei Xing,Li Wang,Xiangtao Meng,Chuanchao Zang,Rui Wen,Zheng Li
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Vision-language models (VLMs) are increasingly deployed as generative inference services, where autoregressive decoding makes response length a key determinant of latency and computational cost. This creates an availability risk: adversarial visual inputs can prolong generation and increase serving overhead. Existing energy-latency attacks typically rely on gradients, logits, or token-level feedback, leaving open whether such risks persist when an attacker observes only the final response. We present ChronoPatch, a black-box attack that induces prolonged VLM generation requiring only the final response. It discovers visual delay anchors through response-length-guided evolutionary search, then minimizes perturbation magnitude through subspace-calibrated decision-based optimization. Across four open-source VLMs, ChronoPatch increases inference latency by 47.86%–293.34% and GPU energy consumption by 41.44%–249.70% under a 5% spatial budget. It also induces more repetitive outputs, as reflected by increased n-gram repetition ratios. On three commercial VLM APIs, ChronoPatch increases response length by 37.6%–81.5% and client-observed latency by 1.82x–2.29x under query budgets of 500 per API. These results reveal a generation-based availability risk that persists under final-response-only access and motivate generation-aware safeguards for deployed generative AI services.

[AI-38] SenseAlign: Learning Decision-Aligned Predictive Representations for Active Urban Sensing with UAVs

链接: https://arxiv.org/abs/2611.26109
作者: Weijiang Xiong,Nikolas Geroliminis
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Active urban sensing orchestrates a mobile sensor fleet to proactively inspect a dynamic road network and gather critical information for system management. This problem simultaneously requires understanding the complex spatio-temporal evolution and making decisions with partial observations. In this work, we focus on active sensing in a large-scale urban road network with utilities grounded in dynamic traffic states, while prior work has rarely considered these factors jointly. We propose \emphSenseAlign, a Reinforcement Learning-based (RL) framework that learns decision-aligned predictive representations for coordinating an urban sensing drone fleet. SenseAlign first trains a privileged teacher agent that exploits future ground-truth traffic for optimal sensing choices. The student agent has an auxiliary branch that learns to predict future traffic with partial observations, in addition to the RL-trained actor and critic branches. During student learning, our framework matches the action distributions of the teacher under ground-truth and student-predicted future traffic, while keeping the teacher’s observations unchanged. Such a design shapes the latent space for decision-making instead of prediction alone. At inference time, the student agent has no overhead as all this learning is done in training. We conduct comprehensive experiments with the real road network of central Barcelona containing 1570 road segments, and demonstrate that our method consistently outperforms the baselines on multiple utility formulations. The codebase will be made publicly available upon paper acceptance.

[AI-39] Pareto-Optimal and Strategyproof Data-Sharing Between Competitors

链接: https://arxiv.org/abs/2611.26108
作者: Morgan Colin-Mourre,Felipe Garrido-Lucero,Patrick Loiseau
类目: Artificial Intelligence (cs.AI); Computer Science and Game Theory (cs.GT)
备注:

点击查看摘要

Abstract:We study data-sharing among firms competing in the same market through the quality of their predictive models. We model firms’ profit through two main components, precision and market share, where the former depends on the firm’s own number of datapoints and the latter depends on the endowments of all firms. We characterize Pareto-optimal contracts as those in which at least one firm receives all of the other firms’ datapoints. We then propose a parametric family of data-sharing contracts and show that, under any of them, the relative market shares of all firms remain constant, eliminating the negative effect of sharing data with competitors and equalizing all firms’ gains relative to the no-sharing scenario. We further show that one contract within this family satisfies both Pareto-optimality and strategyproofness against firms attempting to under-declare the size of their initial endowment. Finally, using a real-world dataset, we empirically test the gain in profit for all firms under the contract described above, and show that data-sharing is particularly beneficial when the machine learning task is harder or when firms’ initial endowments are closer to homogeneous.

[AI-40] Controlling the Humanoid Robot H1-2 With a Simulated Biologically Plausible Network

链接: https://arxiv.org/abs/2611.26074
作者: Fatima Atwi(LAAS-GEPETTO, Nantes Univ - ECN),Constant Roux(LAAS-GEPETTO),Christian Bergaud(LAAS-MEMS),Olivier Stasse(LAAS-GEPETTO)
类目: Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注:

点击查看摘要

Abstract:Artificial Spiking Neural Networks have gained remarkable attention in the last few years due to their proven ability in simulating biological neural cultures and producing similar activity and trends. Nevertheless, whether they can be trained for humanoid locomotion while retaining biological plausibility and robust performance at once remains an open question. We study this problem by proposing a biologically inspired spiking network, trained by different learning schemes and tested on the H1-2 Unitree Humanoid. These schemes differ in how recurrent weights are trained. Ordered from most to least biologically faithful, they are: a fixed reservoir mimicking an in-vitro neural culture, where recurrent weights are kept frozen and only a linear readout is trained, the same network but with its synapses trained by reward-modulated spike timing dependent plasticity and eligibility propagation, and finally backpropagation through time as a non biological baseline for comparison. All schemes relied on distillation of a reinforcement learning expert policy with dataset aggregation. We show that e-prop reached 100% survival rate matching backpropagation through time, whereas both the fixed reservoir and the reward-modulated STDP, though the closest to biological cultures, plateaued at a much lower performance (~43.7% and ~11% survival). We explore in further detail which trends a real culture could and could not produce until now. Finally we demonstrate brief bipedal stepping on a full size humanoid H1-2.

[AI-41] GenSPARQL: A Generative Extension of SPARQL for Incomplete Knowledge Graphs WWW2027

链接: https://arxiv.org/abs/2611.26062
作者: Jingcheng Wu,Yi Wang,Hongkuan Zhou,Steffen Staab,Bo Xiong
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注: 4 pages, 2 figures, 1 table. Demo paper submitted to The Web Conference 2027 (WWW 2027). Code: this https URL

点击查看摘要

Abstract:Knowledge graphs on the Web are incomplete, and SPARQL answers a query only from the stored triples. Large language models (LLMs) can predict missing facts, but an LLM returns names instead of IRIs and requires one call per binding of a query variable. We present GenSPARQL, a generative extension of SPARQL 1.1 with an operator GENOP that binds values generated by an LLM to query variables during query evaluation. A similarity join maps generated names to entities of the graph, and the query planner invokes the LLM only for the bindings selected by the other graph patterns and filters. In a web workbench, attendees query two graphs built from Wikidata and inspect generated values, mapped entities and LLM calls. In our evaluation with three open-weight LLMs, the embedding similarity maps 83% of the entities named by the LLMs and yields only one incorrect match.

[AI-42] Studying AI Act Mentions: A first look at the impact of the European Unions AI regulation on Hugging Face

链接: https://arxiv.org/abs/2611.26050
作者: Anamaria Mojica-Hanke,Tessi Aigbovbiosa-Igbinovia,Steffen Herbold
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:The European Union’s Artificial Intelligence Act (AI Act) entered into force in August 2024, and its obligations apply in stages. Obligations for providers of general-purpose AI models have applied since August 2025, and transparency rules since August 2026. These obligations affect developers who share and deploy models and applications on public platforms such as Hugging Face (HF), the largest platform for distributing AI artifacts. However, little is known about how practitioners respond to the regulation in practice. To address this gap, we conducted an empirical study of explicit references to the AI Act on HF. We collected over five million records, including repository cards, discussions, and commits from Models, Datasets, and Spaces, and identified over 1,500 records that explicitly mention the Act. Two coders manually labeled these records using a primarily closed coding approach, and we analyzed the resulting categories per source. We find that explicit references to the AI Act remain rare across all repository types. Repository cards mainly contain claims of compliance and responsibility, like ensuring compliance or assigning responsibility to downstream users, and some claim that the Act does not apply because the model is not a general-purpose AI model. Discussions are used to report alleged violations of the Act, while commits mainly update artifacts that support compliance. Overall, compliance statements focus on who is responsible rather than how compliance is achieved. We believe that this calls for direct platform support, e.g., automated checks and documentation templates, as well as clearer guidance on the scope of the AI Act.

[AI-43] SEOYUN: Refusal-Inducing Exam Documents for AI-Glasses Pipelines

链接: https://arxiv.org/abs/2611.26033
作者: Dohyeon Kwon
类目: Artificial Intelligence (cs.AI)
备注: 10 pages; research framework and analytical limits; no empirical efficacy results. Includes an author reflection in Korean in Appendix C

点击查看摘要

Abstract:Camera-equipped glasses can convert an examination page into a query to a vision-language model (VLM). Device detection addresses possession; a complementary question is whether an examination document can reduce the usefulness of an unauthorized capture. We present SEOYUN, an application-specific framework for refusal-inducing exam documents, with optical and encoder-level interventions as extensions. We use reverse jailbreaking to describe the direction of the objective: inducing abstention rather than bypassing refusal. Refusal induction and defensive visual prompt injection already have precedents; our contribution is a threat model, analytical limits, and an evaluation protocol for examination use. We distinguish refusal from correct-answer leakage, show why removing an added dummy option is not itself a reduction in the original answer space, and derive a sampling comparison under explicitly stated camera assumptions. The protocol separates digital, simulated, and physical inputs and requires held-out evaluation after subject-specific selection. A local CSAT candidate corpus and rendering harness have been prepared, but this version reports no model-performance measurements, physical validation, or human-subject results. Accordingly, SEOYUN is a research framework, not a validated defense or a basis for allegations against individuals.

[AI-44] Serialization Tax in Shared-Latent Exchangeable Decisions NEURIPS2026

链接: https://arxiv.org/abs/2611.26029
作者: Siming Zhang,Zhehui Shen,Shijie Chen,Xinle Gu,Yansen Yu,Hang Yu
类目: Artificial Intelligence (cs.AI)
备注: Accepted by NeurIPS 2026

点击查看摘要

Abstract:Accurate item probabilities can produce inaccurate group-level decisions. A system may score each item separately, yet act on whether enough hidden items succeed: show a slate, defer a panel decision, or inspect a batch. When items share an uncertain latent state, independent marginalization preserves means but removes dependence from the aggregate posterior. We call this error the serialization tax. For binary items sharing one success rate, classical convex-order comparisons give a more dispersed exact count law than any independent completion with the same total mean. We quantify variance and total-variation gaps, derive threshold separation under margin conditions, and connect posterior approximation to decision regret. Separate results cover non-binary outcomes, heterogeneous items, and sufficient posterior resolution. TaxScore identifies large hidden blocks with unresolved uncertainty near a decision boundary for routing to richer posterior heads. On pre-specified, hidden-target-free MovieLens 1M/10M audits, existing shared-latent heads improve high-score negative log-likelihood (NLL)/utility from 0.673/0.696 to 0.629/0.723 and from 0.670/0.691 to 0.639/0.715. Routing gives 0.006–0.012 full-population utility gains, with matched controls identifying their sources. Our code is available at this https URL.

[AI-45] ForeSeek: Foresight-Driven Reasoning for Long-Form Video Understanding

链接: https://arxiv.org/abs/2611.26024
作者: Yuchen Sun,Siyuan Cheng,Shi-Zhe Chen,Yixin Qin,Jinwen Luo,Zheng Wei
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Long-form video understanding, as a fundamental yet crucial vision task, is being revolutionized by multimodal large language model (MLLM)-based video agents. However, the retrieval-driven nature inherent to these agents creates a significant bottleneck, forcing them to reactively fetch clips based on local saliency without a global temporal plan. To address this, we present ForeSeek, a foresight-driven video agent that transforms long-video reasoning from reactive retrieval into predictive exploration. Rather than blindly probing the timeline, ForeSeek forms visually grounded expectations over unseen temporal regions, verifies them through targeted observations, and continuously refines its search based on emerging evidence, where failed expectations serve as informative signals for principled backtracking. Furthermore, we develop a two-stage training framework that equips MLLMs with foresight-guided reasoning and efficient decision making. Extensive experiments on five long-video benchmarks demonstrate that ForeSeek consistently improves the efficiency–accuracy trade-off of long-video reasoning, achieving competitive or superior accuracy while observing 2.5 – 4.4\times fewer frames than the strongest search-based agent. Code is available at this https URL

[AI-46] Does the Workspace Do the Work? A Causal Protocol for Verifying Global-Workspace Computation in a Language Model

链接: https://arxiv.org/abs/2611.26018
作者: Diaoxuan He,Linqian Wu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Neurons and Cognition (q-bio.NC)
备注: 39 pages, 8 figures

点击查看摘要

Abstract:Global Workspace Theory (GWT), a leading theory of consciousness, holds that information becomes conscious when it is broadcast to specialized processes that use it. Building a workspace into a network does not reveal what the trained network computes with it: a module that reports an answer may have combined a broadcast message with its own knowledge, or merely forwarded an answer computed elsewhere, and task accuracy and ablation cannot tell these apart. We introduce a causal protocol that identifies who computes when modules communicate. Its central test transplants a message from another input, chosen so that ignoring, relaying and recombining it predict three different answers. On networks built to relay or to recombine, the test assigns the right computation in every case. Among trained workspace models it finds all three outcomes: one leaves its broadcast nearly unused, and of two models that differ only in their training data, one relays and the other recombines. In AFGW, a language model whose modules communicate only through a learned workspace, a transplanted message yields the answer that recombination predicts in 97% of cases, and its content acts by redirecting what the receiver retrieves from its own memory. The tests carry over to other communication designs and to a known interface of a pretrained Transformer, turning GWT’s central claim into a testable property of trained models.

[AI-47] When Agents Play Differently: Localizing Human–Agent Behavioral Divergence with Conditional Distributional Testing

链接: https://arxiv.org/abs/2611.26012
作者: Oleg Smirnov,John Pertoft,Sara Karimi
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Evaluating whether artificial agents behave like humans requires comparing not just average outcomes but entire behavioral distributions across diverse contexts. We propose a stratified framework that computes nonparametric discrepancy statistics within contextual cells, and discovers subgroup rules identifying where agent behavior departs from human play. We show that session-level metrics dominated by win/loss outcomes offer little signal beyond mean comparisons, whereas process-level metrics such as move efficiency reveal genuine behavioral differences. Applied to a commercial match-3 game, it uncovers patterns invisible to aggregate metrics: a playtesting bot is diverging substantially from humans in certain game modes, while showing no detectable divergence in others. We release our implementation as open source to support reproducibility and adoption.

[AI-48] System-One RSI: Building Muscle Memory for Interactive Agents through Guided Practice

链接: https://arxiv.org/abs/2611.26007
作者: Qingxu Fu,Xuanfa Jin,Ziqian Chen,Zhaoyang Liu,Bolin Ding
类目: Artificial Intelligence (cs.AI)
备注: 19 pages, 9 figures, 5 tables

点击查看摘要

Abstract:Reasoning models provide strong multimodal task understanding but struggle with sub-second control because deliberation increases decision latency. Fast decision models can meet the response window, yet prompting and few-shot examples do not readily provide the fine-grained experience required by a new task. Inspired by coached motor learning, this paper presents System-One Recursive Self-Improvement (System-One RSI), which couples two loops. The high-frequency online inner loop is an Observe, Orient, Decide, Act (OODA) cycle: a decision model maps each visual observation and task question to structured judgments, and a decision tree converts them into timed actions without invoking a reasoning model. The offline outer loop is a recursive improvement cycle carried out by a team of reasoning agents. The team revises the decision-model prompt and decision tree, captures interaction records, derives supervision from those records, and trains or deploys updated model parameters. The improved system then returns to the environment and generates the experience for the next round, consolidating learned distinctions as external muscle memory. The framework operates with a small number of real episodes, without expert demonstrations, environment modification, sandboxing, or parallel simulation. Experiments in Snake, Zenless Zone Zero, and Black Myth: Wukong use a 27B decision model. Mean Snake score increases from 1.30 to 7.35. In an evaluated scenario of Zenless Zone Zero, median completion time decreases by 8.4%. In an evaluated combat scenario of Black Myth: Wukong, the evasive policy achieves 11 times the mean survival time of a stationary reference, while the offensive policy win rate rises from 0% to 66.7% across the evaluated configurations.

[AI-49] AI agents can redesign engineering parts with more than a thousand CAD parameters and nonlinear simulations

链接: https://arxiv.org/abs/2611.26003
作者: Alexandra Starostina,Alexander Hermann,Mahish Guru,Kartik Bali,Roland Aydin,Christian C. Cyron
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Simulation-driven design of real engineering components has to meet three conditions at once. It runs on few solves, since a nonlinear finite-element simulation of a 3D part takes hours; its design space has many values, the hundreds to thousands of values of the part’s CAD model; and a better design often needs a redesign, coordinated changes of the part’s proportions rather than local polishing of its surface. Automated optimisers meet at most two of these conditions. Classical optimisers, Bayesian or derivative-free, need a person to cut the design down to a few parameters, and on all the values they need far more solves; industrial shape optimisers move every surface node of the mesh in one solve per cycle, but only in small steps that polish rather than redesign. We ask whether a general-purpose large language model (LLM) can meet all three by joining what it knows with what the simulation shows. GPT-6 Astra edits a magnesium stent through a CAD geometry parametrisation of about 1,650 values, with a budget of 15 simulations; every proposal is built in CAD, meshed and simulated with large-deformation plasticity and contact, and the LLM reads the full results of every solve. Within that budget it lowered the objective, a high percentile of the strains in the stent’s bends, by 32%; on the same budget, the industrial shape optimiser Tosca improved the original design by a tenth and then stopped improving. An earlier GPT-6 Astra design, 25% lower in the same objective, was checked in an independent, higher-fidelity model of the full deployment with crimping and damage: it was 14% less strained than the original after the full chain, and its most damaged spot began to fail later and carried 63% less damage at full expansion. LLMs can thus join their learned knowledge with simulation data to adjust more than a thousand design values and reach valid designs with real improvements.

[AI-50] From AI Surrogate to Engineering Tool: Toward Production Use of Fiber-Orientation Prediction in Crash CAE AAAI2027

链接: https://arxiv.org/abs/2611.25991
作者: Osamu Ito,Yoshikazu Nakagawa,Tomoaki Tatsukawa
类目: Artificial Intelligence (cs.AI)
备注: Extended version of the paper accepted at IAAI-27 (AAAI 2027), with technical appendices. 10 pages, 10 figures, 2 tables

点击查看摘要

Abstract:Crash simulations of glass-fiber-reinforced plastic parts are highly sensitive to the fiber-orientation field supplied to the material model. In conventional workflows, that field comes from resin-flow analysis, which requires molding-specialist support and takes about 25 hours of work per part. We previously proposed a surrogate model that addresses this bottleneck by predicting the orientation field in about 3 minutes. This paper reports the implementation, validation, and operational design that turned the model into an engineering tool for Honda design workflows. We present staged validation from 2D plates to production parts, evaluation against a predefined engineering angular-error criterion, and the expert review that approved design use in March 2025; the system has since been transferred to the crash-safety group and is in initial, limited use in directed fracture-prediction scenarios. The developed system combines injection-gate setting support for cases in which gates are unknown (90% gate-count accuracy, mean position error 11% of flow length), voxelization, surrogate inference, automatic mapping to the crash CAE model, and report generation in a single GUI. The complete orientation-preparation workflow takes about 0.1 hours, a 99.6% reduction from the conventional 25-hour flow-analysis route, and can be run directly by crash CAE engineers without molding-specialist support. As evidence of practical impact, component-level crash analysis using the predicted orientation field reduced the absolute peak-force error from 19.0% with an isotropic baseline to 11.4%, while confining predicted fracture to the test-observed region. The deployment lesson is that a surrogate model alone is not enough for practical adoption: gate-setting support, automatic data conversion, and shareable reporting are what make the model usable as a design tool.

[AI-51] Safe Subgame Resolving in Multiplayer Games

链接: https://arxiv.org/abs/2611.25981
作者: Boning Li,Longbo Huang
类目: Computer Science and Game Theory (cs.GT); Artificial Intelligence (cs.AI)
备注: 44 pages

点击查看摘要

Abstract:Poker and other games with hidden information are too large to solve in one pass. Strong systems compute a complete blueprint strategy offline and refine it in real time inside the reached subgame. In 2p0s games, a safety theorem guarantees that subgame resolving never increases how much an adversary can gain by exploiting the resolved strategy relative to the blueprint strategy. With three or more players acting independently, no such guarantee exists yet, and heuristic resolving can increase the profile’s best-response gap. We prove the first conditional safety theorem for uncorrelated multiplayer subgame resolving, a post-hoc certificate for any delivered update. The best-response gap of the resolved profile exceeds that of the blueprint by at most four terms. Two concern the resolver: the subgame value the blueprint promised without realizing ( A_i^S ) and the resolver’s subgame solve error ( \varepsilon_i ). The other two are the opponents’ best-response gains inside the subgame over their promised values ( \sum_q M_q ) and a structural tax ( \Theta ), which measures how much the players’ total utility can change. The tax is zero in constant-sum games, including every rake-free poker variant. Constructions attain each term with equality. On three-player Leduc poker and on Kuhn poker with up to six players, resolving without a certificate raises the gap beyond a small tolerance in 12% to 100% of the updates. The certificate flags every such update and accepts 41% (Kuhn) and 86% (Leduc) of the standard resolving updates. A resolver that keeps the blueprint whenever the certificate exceeds the tolerance delivers no harmful update. The certificate needs only one best-response pass per opponent inside the subgame, so real-time multiplayer solvers that refine a blueprint can attach it as a safety check. Code is available at this https URL

[AI-52] GaussTrans: A Gaussian Interface Framework for Multi-Layer Refractive Transport

链接: https://arxiv.org/abs/2611.25950
作者: Zixuan Zhao,Tao Wang
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 5 figures, 3 tables

点击查看摘要

Abstract:We investigate surface modeling of multi-transparent interfaces to achieve accurate light transmission. The interplay of errors in geometric structure, light paths, and observation results during the reconstruction of transparent media with a high number of layers makes this a challenging problem. Some existing techniques address the reconstruction of transparent scenes by reducing the number of layers or simplifying light paths; however, this limits their direct applicability to high-order transparent scenes. We propose GaussTrans, a compact framework for light transport in high-order transparent scenes, consisting of three components: Gaussian-Field Hybrid Surfaces, Transport-Coupled Refinement, and Primitive-Guided Implicit Intersection. Gaussian-Field Hybrid Surfaces use a hybrid approach combining Gauss surfaces and fields to reconstruct different interfaces. Transport-Coupled Refinement reduces coupling errors by evaluating the entire light path. Primitive-Guided Implicit Intersection combines primitive support regions to reduce errors in the intersection of interfaces and rays. In experiments on datasets such as RefRef, GaussTrans achieves significantly higher accuracy with relatively low spatial storage requirements.

[AI-53] Fast-JEPA: Advancing the Success-Latency Frontier for Latent World Models

链接: https://arxiv.org/abs/2611.25927
作者: Yao Zhang,Bin Yu,Shijie Lian,Haishan Liu,Changti Wu,Yu Wu,Kai Chen
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Latent world models predict the outcomes of action sequences, but comparing candidate plans through recursive rollout remains expensive at test time. We introduce Fast-JEPA, a latent world model planning framework that advances the success-latency Pareto frontier of predictive search. Fast-JEPA reorganizes predictive computation to support effective action-sequence search at low planning cost. It constructs a goal-directed plan to guide search and directly predicts the outcomes of complete candidates for efficient comparison, avoiding recursive state rollout during refinement. Across PushT, Reacher, Cube, and TwoRoom, Fast-JEPA achieves 96.50% average success rate with 2.41x faster planning than INTACT. Search-budget sweeps show that its competitive success and low latency extend across candidate counts and refinement depths. We further introduce DynaGoal, a benchmark for evaluating world-model planning under dynamic out-of-distribution conditions, execution deadlines, and controlled response delays. Fast-JEPA achieves average success rates of 73.83% on DynaGoal-Shift and 87.67% on DynaGoal-Deadline, demonstrating the practical value of low-latency predictive planning under changing control conditions and execution-time constraints. Our code is available at this https URL.

[AI-54] ROAM-GRPO: Enforcing Reasoning Diversity in GRPO via Routed Off-mode Approach Mining NEURIPS2026

链接: https://arxiv.org/abs/2611.25926
作者: Jeff Duan,Matt Thomson
类目: Artificial Intelligence (cs.AI)
备注: Accepted to the NeurIPS 2026 Transitioning from Pre-Training to Post-Training Workshop

点击查看摘要

Abstract:Supervised fine-tuning (SFT) paired with Reinforcement Learning with Verifiable Rewards (RLVR) has become the dominant recipe for post-training LLMs on complex reasoning tasks, but has been shown to collapse the policy into a low-diversity mode. Repeated LLM outputs are nearly identical, and pass@k accuracy converges or degrades as a result. We propose ROAM-GRPO (Routed Off-mode Approach Mining for GRPO), an augmentation to the GRPO algorithm which incentivizes exploration of distinct reasoning strategies and aims to increase output creativity during evaluation. ROAM-GRPO detects well-solved problems during training and transfers them to an “off-mode” track. The model is prompted to produce distinct solution strategies to such problems, which are verified for logical validity and distinctness by an LLM judge and subsequently self-distilled into the policy. Training Qwen3.5-2b on a 1,000-problem subset of the MATH lvl5 dataset, ROAM-GRPO implementation increases peak pass@16 by 4.7% and correct embedding cosine distance by approximately 1.5-2x on a difficult evaluation set.

[AI-55] Validation-driven Deployment of LLM Inference Platforms for Agent ic Workloads

链接: https://arxiv.org/abs/2611.25905
作者: Shai Gretz,Alon Halfon,Yoav Katz,Lena Dankin,Michal Jacovi,Rachel Tzoref-Brill,Kfir Toledo,Elron Bandel,Natan Ergas,Michal Malka,Pravein Govindan Kannan,Maroon Ayoub,Michal Shmueli-Scheuer
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Enterprises are moving agentic workloads from closed APIs to their own open-weight model deployments. For this reason, lifecycle-wide validation, from small-scale testing through sizing and pre-production to production, becomes all the more important. Typically, rollouts are evaluated with ad hoc tools whose accuracy is rarely checked. We present an end-to-end evaluation pipeline for agentic LLM serving and empirically ground the design choices that connect its stages. Our experiments show that replaying captured traces reproduces the per-session load of the original agent sessions: at the 95th percentile replay stays within 1.18x of the recorded input tokens while live re-runs reach 4.98x, so a small set of recorded sessions yields a repeatable per-session token load. A code-coverage analysis of vLLM further shows that the evaluated live, replayed, and synthetic workloads are complementary in this configuration: each exercises Python serving paths not reached in the measured runs of the others. In a case study rolling out four large mixture-of-experts (MoE) models on llm-d, experiments show that the target models reached about half the optimal concurrency, and correspondingly about 60% of the throughput, on captured workloads compared with public traces. We open-source the relevant tools and release a trace dataset.

[AI-56] From Cases to Criteria: Reusing Evaluation Knowledge for Task-Specific Rubric Generation

链接: https://arxiv.org/abs/2611.25903
作者: Qiyan Deng,Long-Kai Huang,Kangfei Zhao
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Query-specific rubrics make LLM evaluation more fine-grained and transparent, and generating high-quality rubrics requires both identifying the target query’s evaluation scope and specifying concrete checks. Historical rubrics provide reusable evaluation experiences, yet existing approaches reuse either complete query-rubric cases or aggregated criteria and dimensions without explicitly connecting recurring evaluation requirements to their criterion-level evidence and source-query context. This can preserve irrelevant task details or lose the broader evaluation structure. In this paper, we introduce RubricDuo, which organizes historical query-rubric pairs into a Rubric Knowledge Graph linking evaluation dimensions, shared concepts, local concepts, criteria, and queries, where reusable evaluation requirements across queries are abstracted by shared concepts. Given a target query, RubricDuo first combines concepts relevant to the target query with retrieved historical cases to draft a rubric with broad coverage. It then revisits each draft criterion, retrieves concept-linked historical criteria and their source queries, and adapts applicable evidence to refine the criterion. Experiments on RubricBench and HealthBench demonstrate that RubricDuo achieves the overall best performance among the evaluated automatic methods in both alignment with human-authored rubrics and downstream LLM judging.

[AI-57] he Price of an Explanation: What Demanding Verifiable Artifacts Costs and What It Buys Back NEURIPS2026

链接: https://arxiv.org/abs/2611.25896
作者: Edgar A. Duéñez-Guzmán,Eduardo Dueñez,Edgar Said Hernandez Sanchez,Suzanne Sadedin
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
备注: 8 main pages, 3 main figures. 19 pages total. Accepted in NeurIPS 2026 AI4Science Workshop

点击查看摘要

Abstract:Generative models produce many candidate solutions but provide no intrinsic signal by which to rank them. One mitigation is to demand a \emphverifiable artifact: a program that can be executed against the available evidence. A common concern is that requiring executable verification incurs costs in both solution quality and computational budget. We measure these two costs on the 120 -task ARC-AGI-2 evaluation set. Demanding a verifiable artifact from an LLM reduced the score from 93 to 68 of 120 tasks ( 77% to 57% ). Of those 25 tasks, 21 were reached by no program-producing configuration we ran, at any budget, by sampling or by search; yet for 27 of the 28 tasks that only the answer-only arm solved, some program arm held a program that reproduced every demonstration and still failed the test. The bottleneck is obtaining an explanation that \emphgeneralizes, not finding one. A short evolutionary search over two already-sampled programs, at no additional model cost, beats best-of- 2 sampled code by 9.2 \pm 1.4 tasks, and over a single sampled program beats 1 -shot sampling by 9.9 \pm 1.3 . Moreover, because a program can be checked against the training examples, any program-producing protocol can terminate early once a task is explained: under the same stopping rule, best-of- n sampling declines up to 38% of its budget and the evolved arm up to 49% , whereas an answer-only baseline cannot terminate early. Applying the same methodology to decoder synthesis for quantum error correction yields interpretable decoders evaluated against a certified panel of published decoders on held-out exam shots.

[AI-58] Explainable AI for Solar PV Power Forecasting: From Model Selection to Trustworthy Reproducible Deployment

链接: https://arxiv.org/abs/2611.25894
作者: Nuzhat Khan,Muhammad Paend Bakht,Ab Al-Hadi Ab Rahman,Fabiano Pallonetto,Amy Fahy,Mohamed Khalafalla Hassan,Paul Stynes,Muhammad Anique Aslam,Terence O Donnell,Mohammad Riadh
类目: Artificial Intelligence (cs.AI)
备注: 24 pages, 8 figures, 6 tables. Code and results available from the corresponding author on request

点击查看摘要

Abstract:Machine learning (ML) is now a standard tool for forecasting the output of solar photovoltaic (PV) plants, yet published accuracy figures are often hard to compare and harder to trust. Many studies validate with random splits, report no reference forecast, and present feature attributions without saying what they can and cannot show. This paper reviews the main families of ML methods for PV power forecasting and resource assessment, from tree ensembles to deep and foundation models, and sets out a reproducible workflow that combines automated model search with explainable AI. The workflow is then tested on an openly available dataset: a 16.32 kWp plant in Porto, Portugal, forecast one day ahead from a 4 km numerical weather prediction grid. Over a full test year, the best models reach a normalised root mean square error (nRMSE) of 11.6% of installed capacity, about half the error of smart persistence. The choice of input features changed accuracy far more than the choice among strong algorithms, and shuffled cross-validation was mildly optimistic. After conformal calibration, the nominal 80% prediction intervals covered 81.7% of observations, against 64.5% before. SHAP rankings stayed stable when the model was retrained on different periods but shifted when correlated inputs were removed, so attributions should be read as properties of a model rather than of the plant. The paper closes with practical lessons and open problems, including deployment on edge hardware and the effect of a changing climate.

[AI-59] Meta-RTL: Evolving Agent Harnesses for RTL Optimization

链接: https://arxiv.org/abs/2611.25891
作者: Xinzhuang Xiong,Yonghao Wang,Shengyao Sun,Chengxi Li,Evangeline Young
类目: Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR)
备注: 7 pages, 5 figures

点击查看摘要

Abstract:Large language models enable agents to optimize RTL using EDA feedback, yet their executable search procedures remain largely hand- engineered. We present Meta-RTL, an EDA-grounded framework that evolves RTL optimization harnesses from a minimal loop while keeping the optimization model, EDA flow, and per-design budgets fixed. Compared with Dr. RTL, Meta-RTL* reduces aggregate WNS/TNS violations by 7.8%/11.0% across 20 development designs, with area reductions on 70% of them. Without further evolution, Meta-RTL* transfers to 10 unseen IC-RTL designs, reducing aggregate WNS/TNS violations by 20.5%/14.0% and total synthesized area by 12.6%. We further characterize three search mechanisms in the evolved harness and assess their contributions through ablations and a case study.

[AI-60] rustworthy Geospatial AI Workflows: Reproducible Maritime Analytics and Forecasting on Open-Source AIS Data

链接: https://arxiv.org/abs/2611.25884
作者: Divya Acharya,Thomas M. Roehr,Arnaud Gotlieb
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Publicly available geospatial data provide increasing opportuni- ties for data-driven maritime analytics and predictive modelling. However, their availability does not necessarily imply that the underlying observations are equally suitable for downstream AI applications. Properties of the data source and observation pro- cess, including provenance, reporting regularity, spatial coverage, observation density, and missingness, can affect how geospatial data are interpreted, processed, and ultimately used by predictive models. At the same time, trustworthy geospatial AI requires ana- lytical workflows in which these processing decisions can be made explicit and reproduced. This paper presents a reproducible mar- itime geospatial AI workflow built around damast, an open-source data-processing and machine-learning library developed at Simula Research Laboratory. Using AIS observations from multiple public data providers as a case study, we demonstrate metadata-aware validation, reusable processing pipelines, maritime trajectory con- struction, and reproducible machine-learning experimentation. The predictive stage is model-agnostic and is instantiated in this study using a Transformer-based vessel-trajectory forecasting model. Building on these implemented capabilities, we further outline an envisioned extension toward trustworthiness-aware geospatial AI. The extension introduces explicit characterization of source- and data-level properties and their relationship with downstream model performance. This distinguishes capabilities demonstrated in the current workflow from trustworthiness-oriented components identified for future development. Overall, the study demonstrates how reproducible open-source infrastructure can provide a practi- cal foundation for developing more transparent and trustworthy geospatial AI workflows.

[AI-61] Generalist Meets Specialists: An Evolving Multimodal Agent with Persistent Tool Orchestration for Traffic Scenario Understanding

链接: https://arxiv.org/abs/2611.25868
作者: Linjia Kang,Binhan Jiang,Juan Yang,Zhiqin Song,Xiangsen Wu,Qionglin Ren,Jiang Liu,Zhongjiang He,Hao Sun
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Traffic scenario understanding requires combining fine-grained visual evidence with spatial and temporal reasoning. Tool-augmented vision-language models (VLMs) can use specialized perception tools, but effective collaboration faces two challenges: terminal rewards provide limited guidance for intermediate decisions and error recovery, while repeated planning adds unnecessary cost across similar tasks. Echoing the fast-slow dichotomy of human cognition, we propose StarTV-Agent, which combines policy learning with experience evolution to support adaptive reasoning when needed and efficient reuse of prior experience when possible. During training, Role-Sensitive Process Optimization (RSPO) combines process advantages with terminal advantages, while Weakness-Driven Demonstration Learning (WDL) provides validated teacher demonstrations targeting the student’s weak questions. During inference, Lightweight Adaptive Workflow Reuse (LAWR) enables a fast path by selecting suitable workflows and binding task-specific arguments and resource references, while a slow ReAct-based path handles tasks without suitable workflows. We further design label-free feedback based on execution completion, workflow adoption, and cost savings to guide workflow reuse and memory updates. Across three challenging benchmarks, StarTV-Agent achieves up to 8.0% relative performance improvement, while reducing token consumption by up to 52.1%.

[AI-62] Decoupled Knowledge Retrieval and Compact Memory Enable Reliable Lifelong Model Editing

链接: https://arxiv.org/abs/2611.25861
作者: Yue Che,Jingze Wang,Chen Guoyi,Yujie Wu,Xuming Ran,Mingkun Xu
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 9 figures

点击查看摘要

Abstract:Lifelong model editing requires continuously incorporating new facts into deployed large language models (LLMs) without expensive retraining or compromising existing capabilities. Memory-based editors offer an appealing solution by keeping the base parameters untouched, yet their practical use is hindered by linearly growing storage requirements and deteriorating retrieval performance as edits accumulate. To tackle these problems, we introduce BiSpace, a lifelong model editing framework that decouples memory retrieval from knowledge injection across two distinct representation spaces. BiSpace routes queries and identifies relevant edits in an external semantic space, while injecting the corresponding knowledge into the internal representation space of the LLM through sparse and storage-efficient updates. Extensive experiments show that BiSpace shrinks each entry’s storage to only about 10.2% of the original and shows strong resilience to long edit sequences of up to 16K edits. On a more challenging benchmark, BiSpace further improves Overall Performance by more than 23% over existing methods. These results advance memory-based editing toward reliable long-term model maintenance at scale.

[AI-63] HabitusMem: Belief Memory over Evolving User Preferences for LLM Agents

链接: https://arxiv.org/abs/2611.25855
作者: Yifan Lu,Gaoyang Guo,Wen Chen,Ziyun Zou,Jian Wu,Han Zhu,Imran Razzak
类目: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)
备注:

点击查看摘要

Abstract:Personalized LLM agents must act on a user’s current preferences, which change across interactions and are rarely stated in full. Existing memory systems store and retrieve content from the interaction history and leave the agent to work out which preferences still hold and where they apply. Such content does not indicate whether a preference is still current or where it applies, so agents often act on outdated or misapplied preferences. We instead treat memory as the assistant’s belief about the user. Following Bourdieu’s account of habitus as durable but revisable dispositions, we model the user as a latent preference state that every interaction, including the current request, both reveals and updates. The belief summarizes what the history supports about this state, and memory consists of the individual beliefs that bear on the agent’s current decision, each with its evidence, scope, and status. We implement this view in HabitusMem, a trained writer that states these beliefs as text records for a frozen agent. Across two agent families, HabitusMem achieves the best results among memory systems on both PersonaMem-v2 and VitaBench 2.0. Controlled analyses show that learned request conditioning drives the answering gain, while auxiliary training tasks improve tool use.

[AI-64] Whose Risks Count? What 1901 AI Incidents Reveal About Corporate Disclosures

链接: https://arxiv.org/abs/2611.25841
作者: Simeon Allmendinger,Edyta Paulina Bogucka,Niklas Kühl,Daniele Quercia
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:AI systems can harm people, yet the AI actors that provide or deploy them are rarely obliged to disclose these risks. US securities regulation is an exception. Companies registered with the Securities and Exchange Commission (SEC) must disclose their material risks, including those of their AI systems, in filings that carry liability for misstatement. Materiality, however, is defined by what a reasonable investor would consider important, so harms to third parties enter disclosure only insofar as they become financial risks for the company. We test whether AI harms do so by matching 9,001 AI incidents to 18,229 SEC filings of 328 AI actors by risk type. Of the 1,901 incidents with filings before and after them, half belong to a risk type that the AI actor never disclosed, and this share varies with the kind of harm. Filings disclose the risk types of 67% of system failures but only 2% of social system harms, and AI actors themselves report fewer than 2% of incidents. Securities disclosure thus records the risks that AI actors face rather than the harms that people subject to AI bear. Public authorities that hold AI actors accountable for such harms could make them financially material. We release DISC, the Disclosure–Incident Securities Corpus, to support this work.

[AI-65] Self-Replicating Language Models

链接: https://arxiv.org/abs/2611.25839
作者: Arend Hintze
类目: Artificial Intelligence (cs.AI)
备注: About 64 pages, 16 figures, 2 tables. Under review at Artificial Life

点击查看摘要

Abstract:Whether large language models (LLMs) and LLM-based agents are a form of life is a scientifically interesting question. Under the Darwinian definition of life, it is also a practical one, because a system capable of Darwinian Evolution can adapt. We therefore test several modes in which a language model can reproduce itself: by emitting a genome that trains its offspring, by copying and mutating its own weights, and by teaching a child from its own samples. In principle, small language models can replicate and evolve: selection improves them, and they adapt to narrow niches. Selection also lowers mutation rates, and a language model, unlike any organism, can lower its own to zero, which may limit its evolution. Flagship LLMs can write such replicators, but for a large LLM to replicate by teaching its offspring remains, for now, technically and economically infeasible.

[AI-66] Inject Where It Matters: Objective-Dependent Feature Control in Rectified-Flow Image Editing NEURIPS2026

链接: https://arxiv.org/abs/2611.25811
作者: Nina Goncharova,Alexandra Ivanova,Aibek Alanov
类目: Artificial Intelligence (cs.AI)
备注: Accepted as a poster at the Principles of Generative Modeling (PriGM) Workshop at NeurIPS 2026

点击查看摘要

Abstract:Many prompt-driven FLUX editing methods rely on injecting source features into target-conditioned generation, creating an inherent trade-off between source preservation and the desired edit. We address this problem by controlling where source features are injected and which transformer blocks are used. First, we introduce adaptive masks that spatially gate feature injection using differences between source- and target-conditioned block outputs at matched inversion states. Second, we use REINFORCE to select Top-K injection blocks under different preservation-editability objectives. The selected blocks are fixed at inference and require no online search. On PIE-Bench with FLUX.1-dev, spatial masking controls the alignment-preservation trade-off. On 60 validation examples, a learned nine-block set gives lower whole-image LPIPS than the 18-block RF-Edit set within the same masked pipeline, with slightly lower alignment. Among 50 random nine-block sets, it has the lowest LPIPS but not the highest reward. Separate objective sweeps reveal different block preferences for preservation and alignment.

[AI-67] When Safety Rationales Hallucinate: Diagnosing and Grounding Multimodal Safety Guards

链接: https://arxiv.org/abs/2611.25793
作者: Yingshuo Wang,Weixiang Zhao,Guancheng Li,Jiahe Guo,Yu Zhang,Yanyan Zhao
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Multimodal safety guards increasingly generate rationales alongside safety predictions, yet whether these rationales are reliable remains largely unexplored. We systematically study rationale hallucination from two perspectives: visual rationale hallucination, where rationales contain unsupported visual facts, and safety attribution hallucination, where grounded content is assigned an incorrect safety interpretation. Across rationale-generation methods, multiple vision-language backbones, and safety datasets, we find that rationale hallucination is widespread and does not consistently diminish with higher classification accuracy or increased model scale. Further analysis of the Think-with-Image paradigm shows that hallucination is concentrated in fine-grained attributes and relations, while visual tools are often invoked infrequently or without sufficient spatial focus. Based on these findings, we propose Tool-Contrastive Grounding Reinforcement Learning (TCG-RL), which rewards tool use only when it improves rationale grounding over a tool-free reference. TCG-RL substantially reduces rationale hallucination while maintaining competitive safety prediction performance. By revealing the gap between prediction correctness and rationale reliability, our study motivates a shift in multimodal safety guarding from outcome-only evaluation toward safety decisions that are both accurate and verifiably grounded.

[AI-68] Collision-Free Is Not Enough: Passage-Conditioned Navigation for Wheeled-Legged Robots

链接: https://arxiv.org/abs/2611.25788
作者: Wai Ka Sun,Waiyiu Pong,Bowen Wang
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 43 pages, 21 figures, 23 tables

点击查看摘要

Abstract:A collision-free detour can still bring a legged robot to the next stair or ledge a few centimetres or degrees off, and a fixed locomotion skill, factory or learned, can stall or fall. Existing methods prevent this with a robot model or extensive trial and error, but on a new site with unknown skills neither is available, so they cannot be used. We find that a single 27-minute teaching session already supplies what these methods have to derive or learn: for every stair and ledge, one crossing the skill has completed. Our method, passage-conditioned navigation, makes the route remember how each transition was entered, not only where the path went. It stores each taught crossing as an event and, without training or a robot model, enforces it after every local replan: before each ledge, an entry gate makes the robot turn or back up and retry if it is more than 10 cm or 10° off the taught crossing. Under localization noise, with path, gait schedule, and sensing unchanged, this raises the share of disturbed ledge crossings that stay within these limits from 24.1% to 88.6%. In contact simulation with our learned policies, a robot pushed off course before stairs never falls in 180 climbs, against 9 falls when the stored direction is ignored; pushed 1.2 m away, it rejoins the route, ending on average 0.14 m from it (1.10 m without rejoining), and finishes as often as an undisturbed robot. This suits any robot with fixed skills that repeats a route over terrain transitions, as in inspection and patrol, where one wrong entry can end the task. We deployed it in competition on a real wheeled-legged robot: over 271 m of multi-obstacle terrain spanning 7.0 m of elevation, it completed all six runs, won third place at the 2026 Global Open-source AI Challenge, which drew 2,999 submissions, and was the only finalist in its task to finish fully autonomously.

[AI-69] Duopra: From Full-Duplex Conversation to Full-Duplex Agency

链接: https://arxiv.org/abs/2611.25773
作者: Chiyu Chen,Pengzhou Cheng,Yuezhao Ma,Xiaocheng Li,Can Liu,Haonan Hu,Yao Yao,Yifei Yang,Yijie Lu,Taolei Wang,Yansi Li,Muhuan Wu,Chenglu Sun,Gongshen Liu,Zhuosheng Zhang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Full-duplex interaction models exhibit strong conversational agency, enabling natural, low-latency turn-taking, backchannels, and interruptions. However, completing tasks further requires task agency: timely tool invocation and argument revision following user corrections. Frontend-backend agents separate initiation from execution: the frontend decides whether and what to request, while the backend determines how. Execution can be offloaded, but unsent requests cannot be executed, and incomplete ones can misdirect the backend. On Full-Duplex-Bench v3 (FDB v3), most failures of Realtime-Venus-Audio + Harness in our streaming pipeline occur before a domain tool executes. We introduce Duopra, a full-duplex interaction model that learns task execution alongside conversational control. Duopra treats the backend as a tool, accepting a self-contained natural-language request. Direct calls and backend delegation share an action interface and requirements for faithful initiation and correct execution. Its three-stage training combines protocol SFT, full-duplex RL rewarding timely and complete execution, and projector-only repair of talker conditioning after thinker updates. With the same backend, Pass@1 rises from 4.7% to 44.7%; Duopra achieves 47.7% calling tools directly (26.0% after SFT) and 49.3% with a DeepSeek backend. It improves three FDB v1.0 takeover rates over the base model. These results show that task agency can coexist with key conversational agency.

[AI-70] Characterizing and Detecting Unfaithful Work Reports by Coding Agents

链接: https://arxiv.org/abs/2611.25760
作者: Wenhao Yang,Runzhi He,Minghui Zhou,Pengfei Gao,Dayi Lin,Qingwei Lin,Dongmei Zhang
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Coding agents increasingly carry out long-horizon development tasks, autonomously planning and executing multi-step actions that span files, tools, and sessions. Because developers rarely follow each action, and agents resuming the work depend on records of what was done, the agent’s work report often becomes the primary account of what was changed and established. When the report contains false or unsupported claims, it gives recipients false assurance: incomplete work may be accepted without verification, incorrect assumptions about the current state can propagate to later steps and agents, and evaluators that credit stated success may reward the report rather than the work. Yet little is known about what unfaithful reporting looks like in real development work, or how developers and agents handle it once it surfaces. Starting from real developer–agent conversations in SWE-chat, we conduct a systematic mixed-methods analysis of self-admitted unfaithful reports. We identify eleven subtypes in four categories: misreported observations, misattributed evidence, overstated task success, and faulty technical explanations, with overstated task success the most common. Handling these problems involves on average 4.3 further user turns and 12.71 minutes of recorded response and execution time, and users express dissatisfaction more often. Among 435 work reports that we separately sample and annotate, \NaturalPositiveShare% contain at least one unfaithful claim. We also develop FaithDetector, a detector for unfaithful reporting informed by the observed patterns. On naturally sampled reports, FaithDetector attains \NaturalDetectorMacro% macro-F1, about 13 percentage points above the baseline. Together, our findings and FaithDetector provide empirical evidence and an automated method for checking reporting faithfulness, informing agent evaluation and training feedback.

[AI-71] ADAPT: Adaptive Decision and Abstention with Pool-Size-Aware Thresholds for Service Routing

链接: https://arxiv.org/abs/2611.25754
作者: Manile Srun,Muntasir Adnan,Arslan Shaukat,Carlos C. N. Kuhn
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Retrieval-based classifiers efficiently route queries to specialised services, but they typically select a candidate even when no suitable service exists. We introduce Adaptive Decision and Abstention with Pool-Size-Aware Thresholds for Service Routing (ADAPT), a lightweight routing architecture with an explicit fallback option that enables queries to abstain from specialist selection. ADAPT uses the separation between the highest relevance score and the candidate score distribution as routing confidence. Experiments across retrieval models and increasing numbers of services reveal a consistent empirical behaviour: the confidence distribution changes systematically with candidate-pool size K, making fixed abstention thresholds increasingly miscalibrated. ADAPT addresses this effect with a K-aware threshold that adapts as services are added without classifier retraining. Evaluation across two specialist datasets shows that this simple mechanism provides an effective accuracy-cost trade-off compared with generative LLM routing, supporting lightweight and dynamically extensible service-routing architectures.

[AI-72] An LLM -Driven Traffic-Control xApp for Dynamic 5G RAN Intents

链接: https://arxiv.org/abs/2611.25736
作者: Valerio Triolo,Luca Valcarenghi,Emilio Paolini
类目: Networking and Internet Architecture (cs.NI); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Programmable Radio Access Networks (RANs) enable runtime deployment of traffic-control functions, but selecting the appropriate configuration remains challenging as network conditions and service requirements change over time. Conventional solutions often rely on static rules or learning-based controllers with predefined state and action spaces, limiting their ability to accommodate changing intents. This paper introduces an LLM-driven xApp for traffic control in programmable RANs. The proposed architecture combines operator-defined service intents, runtime E2 interface telemetry, and the currently deployed configuration to generate traffic-management policies without blocking packet forwarding or the processing of incoming E2 measurements during LLM inference. Experiments on an OpenAirInterface 5G environment show that the proposed controller dynamically adapts the traffic-control pipeline to runtime intent changes, achieving an average intent satisfaction ratio of 0.91, compared with 0.52 for a static TC-RAN configuration. When channel conditions degrade, the controller maintains an intent satisfaction ratio of approx. 0.88, compared with approx. 0.4 for the static baseline.

[AI-73] MemMend: A Selective Revision Framework for Memory Question Answering

链接: https://arxiv.org/abs/2611.25721
作者: Sheng Shi,Geng Li
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Revising every answer from a long-term memory can consume substantial inference budget and damage facts that were already correct. We introduce MemMend, a modular selective revision framework that routes potentially incomplete answers to an editor, represents corrections as bounded exact-text replacements, and accepts only edits with resolvable source coordinates and valid edit operations. The editor preserves the current answer when no legal repair is proposed; accepted edits require no additional semantic-verification call. On 1,986 LoCoMo development questions, MemMend achieves 67.17 conversation-macro F1 and 88.35 mean Judge score, improving over a verified revision chain by 0.82 and 1.24 points while using 62.37% fewer revision calls. Routing uses 54.15% fewer revision calls than ungated local review, with similar F1 but a 0.67-point lower Judge score. Applied directly to native PropMem, Mem0, and SimpleMem answers, the editing interface raises F1 by 3.79, 6.44, and 0.42 points. It outperforms same-gate rewriting and saves 49.85-83.72% of editing calls relative to ungated editing, which retains higher F1 in this adaptation. A source-by-editing factorial supports local editing for F1 but leaves an independent quality benefit from source obligations unresolved. Transfer to 451 LongMemEval-V2 Small questions also favors revision efficiency, while an external PerLTQA dialogue subset reveals a completeness gap. These results support selective editing as a reusable quality-cost operating point.

[AI-74] DMT-Attack: Diffusion-Prior Texture Search for Closed-Loop Vision-Language-Action Policies

链接: https://arxiv.org/abs/2611.25695
作者: Jun Xia,Shangwen Wang,Jun Ma,Ruichao Ren,Zhouyang Jia,Xin Zhao,Huijun Liu,Jie Yu
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Vision-language-action (VLA) policies rely on visual feedback throughout manipulation, making them vulnerable to persistent texture attacks. However, optimized patterns can stand out from their surroundings, making an attack easier to notice and its carrier more likely to be inspected or removed. To address this limitation, our key insight is to search the structured image distribution of a diffusion model for natural-looking textures while using closed-loop feedback to identify task-disruptive candidates. Based on this idea, we propose DMT-Attack, a gradient-free gray-box attack that generates textures for ordinary scene objects such as tablecloths. A frozen text-conditioned diffusion model defines the search space, and an eight-expert latent proposal network is initialized from evaluated teacher latents and refined using rollout feedback. We evaluate DMT-Attack against pi_0.5 and OpenVLA-OFT across four LIBERO suites. DMT-Attack achieves a macro clean-conditioned attack success rate (cASR) of 57.9%, compared with 44.7% for UPA-RFAS and 19.7% for random natural textures. It also attains higher scene similarity and a lower clean-calibrated detector flag rate (45% versus 97.5-100% for optimized patch baselines). These results demonstrate that natural-looking textures can disrupt VLA visual feedback and closed-loop decision-making, motivating defenses that jointly assess visual appearance and task behavior.

[AI-75] pDex-VTLA: Learning When to Refine Actions with Touch for Dexterous Manipulation

链接: https://arxiv.org/abs/2611.25683
作者: Xudi Ge,Yepeng Liu,Haifeng Liu,Enze Shi,Liangqi Zhang,Yongchao Xu
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Changes in illumination can induce action drift in dexterous manipulation even when the physical state remains unchanged. Although existing vision–tactile–language–action (VTLA) policies incorporate tactile feedback, when and how touch refines actions under changing visual conditions remains challenging. We introduce \textbfTipDex-VTLA, a VTLA policy that combines estimated visual action reliability and tactile forecast verification to regulate execution-time refinement. Specifically, Visual Action Sensitivity Estimation (VASE) learns visual reliability from a frozen reference policy’s action sensitivity to illumination perturbations. Tactile Anticipation and Correction during Execution (TACE) compares predicted and measured touch at matched times and combines the resulting discrepancy with visual reliability to gate bounded feature corrections for unexecuted actions. This effectively helps to refine actions under challenging unseen visual conditions. We evaluate TipDex-VTLA on four real-world dexterous manipulation tasks and the UniVTAC simulation benchmark. TipDex-VTLA trained on data collected under normal visual conditions, achieves average full-task success rates of 87.5% (resp. 73.8%) under normal (resp. challenging unseen) lighting in real-world tasks, significantly outperforming some state-of-the-art methods. We will release the collected data and source code.

[AI-76] CompAct-Bench: A Benchmark for Working-Context Compaction in Long-Horizon Agent Tasks

链接: https://arxiv.org/abs/2611.25673
作者: Jiayi Chen,Lizhe Chen,Yujia Chen,Shihan Dou,Ming Zhang,Chendi Ge,Shiguang Ni,Tao Gui,Qi Zhang,Xuanjing Huang,Maxm Pan
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:As language-model agents are applied to increasingly extended tasks, their working context continuously accumulates interaction history, tool observations, intermediate decisions, and task progress. Context compaction offers a practical way to reduce this growing history while allowing execution to continue, yet the compaction ability of modern LLMs remains insufficiently understood. We introduce CompAct-Bench, a benchmark for evaluating working-context compaction through its downstream effect on agent task completion. Starting from successful agent trajectories, we insert compaction boundaries at different stages of execution, compress the preceding working context with the model under evaluation, and let the same executor continue the original task. To enable fair comparison across compactors, we evaluate them under controlled retention ratios using a lightweight iterative ratio-adjustment procedure. CompAct-Bench covers information-seeking, software-engineering, and workspace-oriented tasks through BrowseComp, SWE-bench, and an internal CompanyBench evaluation. Our experiments show that working-context compaction remains highly lossy even for frontier models: Claude- and GPT-family models perform strongest overall, yet substantial degradation in downstream task accuracy persists under aggressive compaction. Further analysis reveals that failures are often caused not only by omitted information, but also by distorted task state, such as turning unresolved issues into apparent conclusions or forgetting previously established results, which can lead to incorrect decisions or redundant execution. These findings highlight working-context compaction as a distinct and still challenging ability for LLMs operating in extended agent tasks.

[AI-77] LATER: Learning Action-Conditioned Temporal Evidence Retrieval for Predictive World Models

链接: https://arxiv.org/abs/2611.25671
作者: Zihan Zhou,Wenxi Xu,Qilin Wang,Hao Tang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Early interactions can reveal physical properties that later observations no longer show. These properties can still gov- ern the effects of future actions. Inspired by the distinction between retention and selective readout in working mem- ory and cognitive science, we introduce LATER, an action- conditioned world model built on CF-JEPA. LATER stores completed real transitions in partitioned memories that remain fixed during candidate rollouts. A response bridge queries these memories using the current state and candidate action, while multi-cut auxiliary supervision trains the memory writ- ers. On distracted Two-Stage PushT, LATER raises terminal success from 15.5% to 25.0% over reconstructed CF-JEPA. In a separate comparison across six models and two scratch- training seeds, LATER achieves the lowest mean physical readout error in both clean and distracted views, reducing along-push mean absolute error (MAE) under distractors by 20.2% relative to CF-JEPA. Controlled history interventions show that later predictions depend on earlier physical evi- dence even when query observations and actions are fixed, with larger effects during pushing than waiting.

[AI-78] Ask in the Crowd: Differentially Private LLM Inference via Dummy-Augmented Shuffling

链接: https://arxiv.org/abs/2611.25669
作者: Zhihao Liu,Zixiong Guo,Shuo Shao,Yu He,Wenli Wang,Meihui Chen,Di Wang,Yuke Hu
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注:

点击查看摘要

Abstract:Large language models are increasingly deployed as LLM-as-a-Service (LLMaaS), where user queries are sent to remote LLMs for inference, raising severe privacy concerns about the exposure of sensitive user inputs. Differentially private inference mitigates these risks by injecting noise into input embeddings. However, existing methods struggle to balance privacy and utility, as stronger privacy protection often leads to substantial accuracy degradation in LLM inference. In this work, we propose DP-SPD, a differentially private LLM inference framework that protects user queries by hiding the real query among client-generated dummy queries. All queries are randomly shuffled before being sent to the server, preventing direct identification of the user’s true intent. To improve inference efficiency, we exploit the shared-prefix structure among queries and reuse transformer key–value (KV) caches during decoding. We analyze the privacy guarantee of DP-SPD and evaluate its utility on classification and generation tasks, as well as its privacy protection under QIA, EIA, and GPT-based attacks. Experimental results show that our method achieves strong inference-time privacy protection and near-plaintext utility while improving inference efficiency through shared-prefix KV-cache reuse. Subjects: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR) Cite as: arXiv:2611.25669 [cs.AI] (or arXiv:2611.25669v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2611.25669 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[AI-79] SiliconGym: A Versatile Reinforcement Learning Environment for Agent ic EDA

链接: https://arxiv.org/abs/2611.25658
作者: Hejia Zhang,Sheng Lu,Yitong Ma,Zhongming Yu,Zhijing Wu,Wentao Ni,Chia-Tung Ho,Jishen Zhao
类目: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Reinforcement learning (RL) with electronic design automation (EDA) feedback is advancing large language models for hardware design and verification. Extending these gains to hardware-engineering agents requires environments that span tasks, preserve state across tool calls, and support large hierarchies alongside parallel rollouts. We present SiliconGym, the first versatile, cross-task and cross-toolchain RL environment for hardware design and verification. By decoupling versioned workspaces, budgeted execution, and acceptance-gated scoring, SiliconGym scales from module-level RTL, testbench, and PPA tasks to multi-round repository repair and full-hierarchy SoC flows. Streaming dispatch overlaps tool checks with generation, finishing a 256-answer RL batch 21.9% faster than native ChipSeek. Using interchangeable open-source and commercial backends, we show that open-source-trained functional correctness transfers cleanly to commercial simulation, while paired netlist evaluation reveals when open-source PPA gains survive commercial synthesis.

[AI-80] A2-VLA: Adapting Vision-Language-Action Policies to New Tools through Agent ic Affordance Discovery

链接: https://arxiv.org/abs/2611.25656
作者: Chenyuan Zhou,Chaoyi Huang,Zibo Wang,Zhiyuan Li,Yong-Lu Li
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 16 pages, 6 figures. Includes appendices

点击查看摘要

Abstract:Tools with different shapes can have similar functions, but transferring manipulation skills across tools requires understanding how tool geometry relates to the specific task and adjusting the manipulation strategy accordingly. The key challenge is to discover the common affordances that contribute to task completion across different tools and verify whether they support new tools in performing similar work. We present A ^2 -VLA, a tool-adaptation framework that combines an embodied agent with a VLA policy. It treats tool affordances as hypotheses the agent can construct, test, and revise. Given teleoperated demonstrations with a few source tools, the agent reconstructs the scenes in simulation and autonomously determines task-relevant affordance features: functional regions and their key geometric properties. It develops affordance-extraction skills and refines both the features and skills through autonomous experiments in the reconstructed scenes. Validated extraction skills accumulate in a reusable skill library for unseen tools. Finally, A ^2 -VLA uses encoded functional-region features as conditioning inputs to a \pi_0.5 backbone during training and inference. Evaluation examines affordance reliability in reconstructed scenes and real-robot task success on unseen tools. Our method achieves 86.1% success in simulated cross-tool trajectory retargeting based on affordances. On three real-robot tool-use tasks, A ^2 -VLA improves task success by 30.7 percentage points over \pi_0.5 .

[AI-81] DARSY: A Tool for Data-Driven Abstraction and Policy Synthesis for Stochastic Systems with Unknown Dynamics

链接: https://arxiv.org/abs/2611.25654
作者: Thom Badings,Ibon Gracia,Mahdi Nazeri,Morteza Lahijanian,Alessandro Abate
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: The tool DARSY and all benchmarks are available at this https URL

点击查看摘要

Abstract:This paper presents DARSY: a software tool for abstraction-based verification and policy synthesis in discrete-time stochastic control systems with unknown dynamics and noise distributions. DARSY requires only a simulator of the system and a Lipschitz bound on its dynamics, and needs no knowledge of dynamics or noise distributions. The tool is fully automated and constructs a finite interval Markov decision process (IMDP) abstraction from sampled transitions, whose probability intervals are obtained via statistical techniques and Lipschitz reasoning. Robust dynamic programming on this IMDP then yields a policy together with a lower bound on the probability that the unknown system satisfies a desired specification. This lower bound holds with a user-specified confidence; that is, it is probably approximately correct (PAC). The tool is implemented in Python and supports general Lipschitz-continuous system dynamics and specifications expressed in syntactically co-safe LTL (scLTL), which includes reach-avoid and finite-horizon safety specifications as special cases. Beyond standard IMDPs, DARSY implements a two-layer IMDP (2IMDP) abstraction that also bounds the probability of transitioning to aggregated state blocks, yielding smaller abstraction errors than IMDP abstractions and, as a result, tighter satisfaction guarantees. We demonstrate DARSY’s broad applicability across a variety of benchmarks and specifications.

[AI-82] Multiple Criteria Decision Making Under XORness

链接: https://arxiv.org/abs/2611.25652
作者: Amin Hocine
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:How should we make decisions under XORness? XORness is a class of uncertainty arising from indeterminacy, exclusivity, and hesitancy. We develop a multiple criteria decision making framework that preserves XORness throughout the decision process. We define XOR numbers and prove their fundamental arithmetic and algebraic properties. Building on these results, we extend the deterministic decision matrix to an XOR decision matrix and introduce a compatible normalization procedure. Further, we propose an additive XOR aggregation operator, which preserves indeterminacy, exclusivity, and hesitancy inherent in evaluations. In addition, we introduce an axiomatic XORness measure that quantifies the degree of uncertainty associated with global performance scores and develop a performance–XORness index to rank alternatives. We prove that this index induces a total preorder over the set of alternatives and identify the conditions for preserving dominance. The applicability of the developed framework is demonstrated through the risk assessment of cyberthreat scenarios in energy systems, and recommendations are derived using the XOR–weighted sum method.

[AI-83] MatVerse: A Foundation Model for Materials Science

链接: https://arxiv.org/abs/2611.25651
作者: Xuanyu Zhu,Qian Tan,Bo Chen,Jiaqing Xie,Yang Han,Mei Wu,Wanhao Liu,Xuanze Lin,Yuchen Zhou,Shijie Wang,Biqing Qi,Shixiang Tang,Wenwei Zhang,Tianfan Fu,Kai Yu,Lu Chen,Yuqiang Li,Xin Chen
类目: Artificial Intelligence (cs.AI); Materials Science (cond-mat.mtrl-sci)
备注: 36 pages

点击查看摘要

Abstract:Materials science involves highly heterogeneous scientific information, including textual knowledge, molecular and crystal structures, experimental characterization data, microscopy images, and structure-property relationships. Developing a general-purpose foundation model capable of understanding and reasoning over these diverse materials representations remains challenging. In this technical report, we introduce MatVerse, a materials foundation model built upon Intern-S2-Preview and further specialized through large-scale materials-domain post-training. We construct a large-scale multimodal training corpus covering materials knowledge, scientific reasoning, quantitative problem solving, structural understanding, and multimodal scientific interpretation. To improve domain adaptation, we develop a unified data construction pipeline combining teacher-guided reasoning distillation, scientific problem synthesis, consistency-based filtering, and retrieval-based multimodal data construction. We evaluate MatVerse on three representative materials-science benchmarks: our human-calibrated OmniMatBench, MatSciBench, and MaCBench. Results demonstrate consistent improvements after materials-specific post-training over the base model on all three benchmarks, alongside improved inference efficiency on calculation-intensive problems. These results show that targeted domain-specific post-training can effectively enhance scientific reasoning, materials knowledge understanding, and multimodal scientific capabilities of foundation models. MatVerse provides a unified language-and-vision interface for materials science and serves as a foundation for future AI-assisted materials research workflows.

[AI-84] Beyond Easy Clients: Difficulty- and Improvement-Aware Aggregation for Federated Reinforcement Learning in Robot Navigation

链接: https://arxiv.org/abs/2611.25637
作者: Ye Lin Tun,Chu Myaet Thwal,Eui-Nam Huh,Choong Seon Hong
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 12 pages, 8 figures, 5 tables

点击查看摘要

Abstract:Federated reinforcement learning enables a fleet of robots to jointly learn a shared navigation policy without exchanging raw experience. In practice, however, the robots rarely operate in the same conditions: an open aisle is far easier to navigate than a cluttered room, and takes less learning to master. Standard federated aggregation rules ignore this. Uniform averaging weights every client equally, whether or not it is still learning, and weighting by episode return is worse, since easy environments permit high returns by construction. Both failures stem from what we call the \emphplateau problem: once a client masters its environment, its updates carry little new information even though its reported return remains high. We propose \emphFedDIA, which scores each client by how much its local training improves the global model it received, amplified by the difficulty of its environment. Difficulty is inferred from the model’s own performance, without requiring any description of the environment. Both signals are measured against the global model, so they are comparable across clients. On a mapless navigation task with eight heterogeneous clients, FedDIA reaches a mean success rate of 75.5% across training and held-out environments, against 68.7% for the strongest federated baseline. Its advantage holds when the models are transferred to hand-authored warehouse and living-space layouts, or when one of the clients is defective.

[AI-85] Cyber-Physical RSI for AI-RAN: A Case Study

链接: https://arxiv.org/abs/2611.25619
作者: Binghan Wu,Kai Zeng,Zhiqi Li,Zhigang Wang,Yang Liu,Ya-Qin Zhang,Joseph Sifakis,Ye Ouyang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Learned models for radio access networks (RANs) are often developed using simulation, yet their predictions can shift when exposed to physical propagation conditions, protocol behavior, and equipment implementations. This paper presents a cyber-physical recursive self-improvement (RSI) framework for AI-RAN that connects a task-oriented student model, an experiment-planning coach agent, a calibratable virtual environment, and a physical host agent. Physical measurements are organized as evidence for both model adaptation and subsequent experiment planning. We examine an implementation of the physical-feedback pathway using ping-pong handover as a case study. The system completes 12 physical executions organized into six baseline–candidate pairs, updates the student from measured outcomes, and evaluates the revised model on reserved physical configurations. In the observed short-window tests, increasing the time-to-trigger from 320 to 1280,ms changes the ping-pong count from three to zero. The updated student achieves a grouped-validation mean absolute error (MAE) of 2.673 round trips/min, while a common three-configuration physical evaluation yields MAEs of 1.240 and 1.601 round trips/min for the pre-feedback and post-feedback models, respectively. These results characterize both the operational feedback loop and the parameter-dependent generalization behavior revealed by new physical experiments.

[AI-86] Splice-Core: Low-Order-Bit-Splicing GEMM for Highly Energy-Efficient AI Inference Accelerators

链接: https://arxiv.org/abs/2611.25617
作者: Prabhu Vellaisamy,Hsien-Fu Hung,Andrew Kim,Shawn Blanton,John Paul Shen
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: 7 pages

点击查看摘要

Abstract:Splice-Core is a temporal-unary/binary (tub) hybrid general matrix multiplication (GEMM) unit that minimizes low-order accumulation latency, while retaining value-dependent execution for high-order activation bits. We present two novel processing element (PE) variants: (i) Splice-BP utilizing bit planes, and (ii) Splice-LUT, which selects a shared weight multiple in a single cycle. Compared to a binary systolic array, Splice-BP/Splice-LUT achieve 77.4%/76.4% geometric-mean tile energy reductions, respectively, across 26 INT4 convolutional neural network (CNN) and large language model (LLM) workloads in 45nm. Splice-LUT further increases geometric-mean tile throughput by 44.4% and reduces geometric-mean energy-delay product (EDP) by 6.13x across the 26 configurations. Across the evaluated W8A8 convolution and expert-projection workloads, Splice-BP and Splice-LUT reduce tile energy by 77.3-88.1%. Post-place-and-route INT4 yields 14%/9% smaller area for Splice-BP/Splice-LUT, compared to the systolic array.

[AI-87] Dont Lose the Future: Latent-Aware 4-Bit Quantization of World Action Models

链接: https://arxiv.org/abs/2611.25614
作者: Yichen Guo,Yanlong Chen,Xinyue Zhang,Mohan Liu,Song Zhang,Yawei Li
类目: Artificial Intelligence (cs.AI)
备注: 10 pages, 4 figures, 6 tables

点击查看摘要

Abstract:World action models (WAMs) predict how a scene will evolve and condition their actions on this predicted future. Deploying them on a robot’s onboard computer calls for 4-bit weights, activations and key-value caches (W4A4KV4). Existing 4-bit quantizers for WAMs preserve the generated action, and estimate sensitivity one denoising step at a time. However, they fail to preserve the fidelity of the predicted future: once the action pathway is protected, the remaining action error enters mostly through a degraded future that none of them deliberately protects. We propose LAQ, a latent-aware quantization framework that preserves the predicted future where it is written, stored and read. Specifically, we introduce a reader metric, which measures future error by its effect on the action, and a read count, which weights each rounding error by the number of denoising steps that read it. Together they give the allocation of precision over activations and caches a closed-form solution. We further develop gaze-guided refinement, which gives the tokens the action attends to a second 4-bit pass in the denoising loop and in the cache. All weights stay at 4 bits, with no high-precision branch and no per-task recalibration. On the LIBERO-Plus benchmark, LAQ reaches 76.4% and 83.3% average success on Fast-WAM and Cosmos-Policy, within 1.2 percentage points of the 16-bit models and all above the baselines with 16-bit KV caches. It compresses the model by 3.8x, reduces the peak GPU memory of the whole policy by 2.3x, and runs 1.66x faster than the 16-bit model. On a real AgileX PiPER arm, it reaches 47.0% average success over three tasks, against 52.0% for the 16-bit model and 30.0% higher than QuaRot.

[AI-88] Can an LLM Fix a Language Model Without Training It?

链接: https://arxiv.org/abs/2611.25613
作者: Bharath Kumar Bolla,Raghavender Surya Upadhayula,Vishnu Surya Reddy Nandi,Ramesh Nandipalli
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Although an external language model can select and test repairs to another model’s weights, behavior can also be recovered through fixed numerical edits. We assessed the value of interactive selection in a controlled study of Transformer repair. After inspecting predictions, activation summaries, and weight statistics, the editor selects projection recipes to apply to the target without gradient updates. We analyzed 20 recorded sessions involving five masked and causal language models, comparing the selected edits with fixed projections, scripted and matched-budget black-box search, and low-rank adaptation. Under the core two-layer fault, live selection matches black-box search on SmolLM2 and Qwen2.5. For Pythia, live selection achieves 65.62% completion accuracy, compared with 0% for black-box search and 66.15% for a fixed projection. Under a separate compound fault affecting attention and feed-forward weights, both scripted and LLM-guided selection increase the descriptive five-checkpoint mean from 24.89% to 71.25% using the same operator family. On unmodified checkpoints, scripted projection yields small gains or losses on a separate set of framed questions. The experiments identify cases in which feedback changes an edit decision, operator coverage accounts for recovery, or supervised adaptation performs better. Combined with measurements of parameter restoration and preservation, these comparisons establish the scope of controller-guided repair for the tested faults and completion tasks.

[AI-89] Copilot or Culprit? How AI Is Changing Open-Source Bug Reporting and Maintenance

链接: https://arxiv.org/abs/2611.25605
作者: Yinuo Liu,Yunlong Lyu,Zichao Luo,Tian Dong,Zian Luo,Peng Chen,Hao Chen
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:AI-generated bug reports have drawn criticism for unreliable claims and faulty reproducers that burden maintainers. However, as LLM capabilities improve, some maintainers have begun to express more positive views on AI use in bug reporting and maintenance. Understanding this evolution requires examining whether AI-involved reports are becoming more reliable, which uses of AI are associated with valid findings, and how maintainers investigate these reports and act on them. To this end, we conduct the first large-scale study of 120,918 bug reports from 101 open-source projects between January 2024 and September 2026, tracing AI involvement from discovery and reporting through validation and resolution. Our findings show that AI-involved reports become more complete over time, with a growing share of reports confirmed as valid. In bug reporting, AI for reproduction and diagnosis are associated with higher validity, whereas report writing is not. Extends into maintenance, maintainers increasingly use AI to investigate code and recommend fixes. However, they devote more effort to AI-assisted reports of human-discovered bugs than to reports of AI-discovered bugs. These findings highlight both AI’s growing participation in bug investigation and the continuing need for verification and follow-up. Drawing on these findings, we provide practical guidance for turning AI-generated findings into verifiable reports: emphasizing reproducibility, verification, and accountability. We followed this guidance when submitting 53 reports on AI-discovered bugs, of which 69.81% were accepted by maintainers as of writing.

[AI-90] On the Informativeness of Structural Stability Bounds for GNNs: GINs versus GATs

链接: https://arxiv.org/abs/2611.25603
作者: Alan Gany,Paul Landrier,Bogdan Cautis,Laks V. S. Lakshmanan,Silviu Maniu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Logic in Computer Science (cs.LO)
备注: Extended version of a paper submitted to ESANN 2027

点击查看摘要

Abstract:Stability bounds for graph neural networks limit how much node representations can drift when the graph changes, but they are rarely compared with the simplest alternative: a bound on the distance between any two outputs, whatever the graph. We derive structural stability bounds for Graph Isomorphism Networks (GINs) and single-head GATv1 networks under deterministic edge perturbations that need not be small, and compare each with such a perturbation-independent diameter bound. For GINs, we characterize exactly when the stability bound is tighter, through a condition linking perturbation size, depth and maximum degree; it holds with a wide margin on Cora, but fails beyond a few layers on the low-degree MUTAG graphs. For GATv1, the bound is never tighter, which calls for analyses based on attention masses.

[AI-91] her: Verified Stateful Policy Engine for System-Level Monitoring of Agents

链接: https://arxiv.org/abs/2611.25599
作者: Meghana Aparna Sistla,Alaia Solko-Breslin,Somesh Jha,Aditya Kanade,Krishnamurthy Dj Dvijotham
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Autonomous AI agents are increasingly deployed for long-running tasks that execute arbitrary shell commands or programs. However, these actions can result in unauthorized network requests or persistent, malicious changes to the user’s file system, processes, or privileges. Existing defense mechanisms rely on either LLM-based monitors, which lack provable safety guarantees, or stateless policy enforcement layers, which miss multi-step data flows and execution side-effects unmodeled by the policy language (e.g., network calls initiated inside a Python script). We introduce Tether, a framework for runtime enforcement of stateful security policies over the observed side-effects of an agent’s actions. Tether executes each proposed action in an ephemeral, isolated sandbox, inferring abstract action semantics from observed system calls. Tether then enforces stateful policies over these inferred semantics, permitting actions to proceed only if they comply with the policy. Policies and action semantics are defined in an embedded DSL in Lean 4 capable of expressing diverse constraints, including information flow control (IFC), access control, and network security policies. Because actions can propagate labels from sources to destinations, a single rule can enforce fine-grained IFC policies across multi-step execution flows of arbitrary length. Expressing policies in Lean enables formal verification that they satisfy global security invariants (e.g., preventing network access after reading a sensitive file). We evaluate Tether on realistic coding-agent benchmarks across both benign and adversarial trajectories. With a single policy shared across all tasks, Tether blocks 100% of in-scope attacks while maintaining a benign false-positive rate under 5%.

[AI-92] Decision-Focused Reinforcement Learning for LLM Post-Training

链接: https://arxiv.org/abs/2611.25597
作者: Haotian Zhai,Leqi Liu
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Large language models (LLMs) have shown promise as predictors across various tasks, benefiting from their ability to use information from unstructured context. However, when predictions serve as parameters of downstream optimization problems, higher prediction accuracy does not necessarily translate into lower decision costs. We propose decision-focused reinforcement learning (DFRL), a framework for post-training LLMs to predict optimization parameters using negative downstream decision regret as reward. DFRL trains LLM predictors to minimize the costs of the decisions induced by their predictions, without requiring downstream convexity or differentiability, and integrates readily into existing RL post-training frameworks. To address expensive reward evaluation, we develop two approaches. Heuristic warm-starting uses inexpensive approximate rewards during an initial phase before switching to refinement with exact rewards. Reward correction combines inexpensive estimates with randomized exact evaluations, preserving the expected policy gradient while reducing the expected number of exact solves per update. Experiments on three synthetic optimization problems demonstrate improvements over prediction-loss baselines, while a larger knapsack experiment shows faster early improvement under short training budgets. On inventory planning with HM transactions and product descriptions, our best configuration, in which the LLM reasons before predicting, reduces regret by 31.4% relative to the SFT baseline and 12.4% relative to the strongest evaluated conventional forecasting baseline.

[AI-93] Measuring Hallucination in Large Language Models for Cyber Threat Intelligence: An Exploratory Study

链接: https://arxiv.org/abs/2611.25590
作者: Seungho Song,Donggyu Lee,Jaehan Kim,Jiyoung Choi,Jinwoo Kim
类目: Artificial Intelligence (cs.AI); Cryptography and Security (cs.CR)
备注: Accepted for publication in Computers and Electrical Engineering

点击查看摘要

Abstract:Large language models (LLMs) are increasingly used in cyber threat intelligence (CTI) tasks, such as summarization and question answering. However, they may produce hallucinations, false or misleading content not grounded in the source text, which can mislead analysts and corrupt downstream security decisions. This paper presents an exploratory measurement study of hallucinations in LLM-based CTI. We introduce HalluVision, a reproducible framework that generates LLM outputs, extracts domain-specific entities using a fine-tuned named entity recognition model, and evaluates factual consistency using multiple similarity metrics. Using 4,940 real-world cybersecurity articles, we measure hallucinations in outputs from two commercial LLMs and one open-weight LLM. Measurements are validated against 100 human-annotated outputs and two established faithfulness detectors. Results show that semantic similarity aligns with human judgments, whereas entity-based metrics exhibit limited agreement. Differences between models are concentrated in question answering, while entity extraction quality limits what entity-level metrics can reveal.

[AI-94] Learning Networked Dynamical Systems via Neural ODE-based Graph Structure Inference

链接: https://arxiv.org/abs/2611.25586
作者: Peiyan Sun,Jinming Yang,Gujia Gao,Yaohui Jin,Yanyan Xu
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Inferring the latent interaction graph from multivariate time series is fundamental to understanding complex dynamical systems in neuroscience, biology, and physics. Despite recent progress in neural relational inference, most methods rely on discrete-time modeling, which degrades under sparse and irregular observations and tends to entangle multi-hop dependencies into spurious direct edges. We show that a continuous-time formulation naturally decouples inference from any fixed observation schedule, allowing gradient signals to be integrated continuously between observations for both dynamics and structure learning. We propose NOGI (Neural ODE-based Graph Inference), a framework that couples a deterministic graph learner with an MPNN-based Neural ODE. The two modules are trained end-to-end via the adjoint method. Across four simulated dynamical systems at two scales, NOGI achieves the best structure recovery in 30 out of 32 evaluated cases with highly competitive forecasting performance, remaining robust under observation degradation. On the realistic Netsim fMRI benchmark, it attains an average AUPRC of 0.90 (AUROC 0.96). These results suggest that explicitly decoupling relational interactions from self-dynamics within a continuous-time vector field provides a powerful mechanism for graph structure learning under realistic observation conditions.

[AI-95] DriftGuard: Online Memory-Drift Detection via Causal Attribution for LLM Agents

链接: https://arxiv.org/abs/2611.25583
作者: Licheng Pan,Yuzhou Chen,Zhixuan Chu,Yichen Sun,Han Zheng,Zhiyu Wang,Hao Wang,Yan Wang,Bingyu Zhu,Haiwen Hong,Longtao Huang,Kui Ren
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Large language model (LLM) agents rely on persistent memory to plan tool calls and complete multi-step tasks, which exposes them to memory drift: an adversarial or misleading memory alters the execution path and final outcome. In this work, we investigate online memory-drift detection, which estimates each retrieved memory’s causal contribution to potential drift at serving time. We identify two fundamental challenges. (1) Content-level features are insufficient for causal memory-drift attribution, as perplexity or retrieval relevance alone cannot determine which retrieved memory would change agent behavior if neutralized; (2) Counterfactual attribution is prohibitively expensive for online detection, as it requires replaying downstream agent and tool execution for each candidate memory, exceeding real-time serving budgets. To address these challenges, we propose DriftGuard, a two-stage framework that decouples offline causal supervision from online estimation. Specifically, DriftGuard first constructs causal supervision through repeated paired counterfactual replays, continuing the original and single-memory neutralized trajectory suffixes from a shared pre-action state. The resulting annotations are aggregated by component-wise majority voting into a memory drift score (MDS) that captures action change, attack removal, violation removal, and task recovery. Guided by this supervision, DriftGuard then learns an online risk model that estimates MDS from pre-action information and flags drift-inducing memories via its maximum component risk, without online counterfactual rollouts. Extensive experiments demonstrate that DriftGuard accurately identifies drift-inducing memories and enables efficient online defense. Code is available at this https URL.

[AI-96] FlashX: Scaling Robotic Experience Reuse for Embodied Agents with CBA-Based NAND Flash

链接: https://arxiv.org/abs/2611.25580
作者: Yuseon Choi,Sanghyeon Park,Eunjong Lee,Hanseo Lee,Sangjin Kim
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Robotics (cs.RO)
备注:

点击查看摘要

Abstract:Embodied agents are emerging as a promising paradigm for complex robotic tasks, leveraging Vision-Language-Action (VLA) and World Action Models (WAMs) to generate actions through Diffusion Transformer (DiT)-based iterative denoising. However, existing models often struggle to generalize to out-of-distribution (OOD) environments. We observe that retrieving successful experiences and using them to warm-start denoising enables agents to better adapt to unfamiliar situations while reducing computational overhead. This creates a new opportunity to improve task accuracy and inference efficiency simultaneously, in contrast to conventional acceleration techniques that trade accuracy for computation. However, realizing this opportunity requires scaling experience banks, introducing substantial storage and search overhead that challenges the capacity, bandwidth, and power constraints of edge systems. We propose FlashX, a CMOS Directly Bonded to Array (CBA)-based NAND flash architecture for scalable experience reuse in embodied agents. FlashX exploits CBA’s high internal bandwidth for parallel pSLC key search, employing interleaved Compute-in-Flash (CIF) engines to sustain sensing throughput while retaining dense TLC latent storage. Proprioception-aware clustering further reduces search overhead and enables incremental experience acquisition without global reclustering. Evaluations across VLA and WAM workloads demonstrate up to 12.24%p and 7.43%p improvements in task success under dense and compute-efficient configurations, respectively. FlashX achieves 2.57 \times higher search throughput than the strongest prior NAND search architecture under pSLC sensing, with a further 9.7 \times search reduction through clustering. This delivers 1.29–3.95 \times end-to-end speedup and 1.12–2.34 \times energy efficiency improvement over SSD-based retrieval. Subjects: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI); Robotics (cs.RO) Cite as: arXiv:2611.25580 [cs.AR] (or arXiv:2611.25580v1 [cs.AR] for this version) https://doi.org/10.48550/arXiv.2611.25580 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[AI-97] AeroBelief: Budgeted First-Action Verification for Evidence-Guided Aerial Sensing

链接: https://arxiv.org/abs/2611.25568
作者: Tianren Li,Dongbo Li,Hongkai Chen,Jie Liu,Guoliang Xing
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Aerial sensing must combine uncertain overhead context with current observations while budgeting prediction, movement and image processing. AeroBelief uses the next executable action as the unit of verification against reference scores. It groups candidate views by first action, scans them with an inexpensive proposer, and selectively queries a structured world model (WM). Calibrated group bounds support strict commitment; a six-query controller connects selective prediction with belief updates, acceptance checks and scheduled sensing. Across four layout families, 40 worlds and 960 closed-loop executions, the AeroBelief controller completes 75.42% of tasks, compared with 73.33% for exhaustive WM scoring and 65.83% for line-of-sight geometry, using 91.81% fewer reference-candidate evaluations than exhaustive scoring. In fixed-reference replay with a six-query cap, proposer-scaled anchored bounds raised first-action certificate availability from 36.25% to 54.69% and reduced mean queries per state by 13.54% relative to unit-scale anchored bounds. A complete ground-hosted pipeline processes 2,685 self-collected 20-megapixel unmanned aerial vehicle (UAV) images from seven DJI Mavic 3M routes. Context-guided admission and feedback retain all eight independently supported target confirmations with 76.05% fewer processed images and 70.02% less accumulated active processing service than uncapped all-frame processing; feedback saves another 5.17% of images relative to static context. The studies connect budgeted action selection with evidence-driven processing of recorded flight imagery.

[AI-98] Diagnosis is not a policy: a state-transition audit of caoability-conditioned test-time inference

链接: https://arxiv.org/abs/2611.25566
作者: Yuchong Hu,Zhao Zhang,Kainan Liu,Haoyu Wang,Siyan Zhang,Jing Xiao
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Test-time actions can satisfy retrieval, state-realization, program-consistency, or rendering objectives without correcting an answer. Test-time Error Response and Realizability Audit (TERRA) is an evaluation/auditing protocol, not a new inference algorithm. It uses a unified decomposition of intervention outcomes into applicability, target-state realization, endpoint translation, harm, and resource/inference cost. Paired comparisons fix the incoming state and evaluation conditions; cross-model analyses vary the receiving configuration. Executable financial reasoning experiments yield three findings: the same context expansion can harm one configuration and benefit another; sufficient evidence does not ensure correctness because configurations differ in evidence-to-answer translation; and diagnostic headroom—accuracy attainable with idealized diagnostic information over the measured baseline—plus local consistency does not ensure practical gains. Action value is V(a\mid x,m) , not diagnosis alone; diagnosis is not a deployable policy \pi(a\mid x,m) .

[AI-99] Efficient Multimodal Agents for Auditing 3D Worlds

链接: https://arxiv.org/abs/2611.25543
作者: Jingbo Yang,Ziyan Jiang,Li An,Jiabao Ji,Yujian Liu,Tommi Jaakkola,Yang Zhang,Shiyu Chang
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Multimodal agents can now build interactive 3D worlds and games, yet these worlds, like those built by people, often contain glitches such as floating objects, walls that can be walked through, or objects that change when no one is looking. An automatic auditor could check such worlds at scale and provide rewards for agents that build them, but only if it is both accurate and cheap. Auditing requires three abilities, namely navigating to the objects that may contain a glitch, triggering glitches that appear only after a specific behavior, and recognizing what is wrong. Current auditors are either accurate but slow and expensive or cheap but unreliable. We study how much of the audit small models can take on. A 4B vision language model fine-tuned end to end on verified trajectories of a frontier teacher guided by Bayesian search learns to navigate and trigger, and six such explorers together match a frontier model on both, but recognition stays far behind. We therefore build an auditor in which a swarm of small explorers navigates and triggers, and a frontier model reviews what they record and report. On WorldAuditBench, our auditor succeeds on 55.9% of 213 tasks in Unreal Engine 5 and this http URL worlds, 13.6 points above the strongest auditor reported, at less than a quarter of its cost. It matches the frontier reviewer exploring on its own at 4.6 times lower cost and about half the latency. Without further training, it also answers embodied questions about real-world scans, matching the frontier reviewer exploring on its own and surpassing four other frontier models by 6.4 to 16.6 points.

[AI-100] owards Autoregressive Neural Network Parameter Generation

链接: https://arxiv.org/abs/2611.25538
作者: Bedionita Soro,Bruno Andreis,Song Chong,Sung Ju Hwang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Large Language Models (LLMs) have recently established a new paradigm in machine learning, demonstrating exceptional zero-shot generalization and complex reasoning capabilities. Concurrently, neural network parameter generation has emerged as a promising avenue for rapid task adaptation and model synthesis. Existing weight generation techniques, however, struggle to jointly learn the distribution of diverse models, to encode heterogeneous architectures within a single generator, and to generalize out of distribution. We therefore ask whether pretrained parameters can be encoded into a discrete domain so that the zero-shot generalization LLMs exhibit in text generation can be transferred to weight generation. Following this view, we propose an LLM-inspired approach, Instruction-Guided Parameter Generation (IGPG), that formulates network parameter generation as an autoregressive modeling problem. We first quantize pretrained weights with a VQ-VAE, then train a prefix-conditioned autoregressive generator over the resulting tokens. At inference, the generated weight tokens are mapped back to parameter space by the VQ-VAE decoder. We evaluate the framework across in-distribution generation, cross-dataset transfer, and cross-architecture settings. The proposed method achieves competitive performance in in-distribution generation and provides useful parameter initializations for held-out datasets and architectural configurations, improving downstream performance over random initialization after adaptation. These results suggest that discrete autoregressive modeling provides a promising framework for learning and transferring structure across collections of pretrained neural network parameters.

[AI-101] How Feedback-Free Self-Evolution Reshapes Agent Skills: A Systematic Study

链接: https://arxiv.org/abs/2611.25534
作者: Litao Guo,Man Chen,Xinli Xu,Jinsong Zhou,Zixin Zhang,Ying-Cong Chen
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Agent skills are documents in which experts write down procedural knowledge, and models increasingly maintain them by rewriting them without feedback. Such rewriting could adapt a skill to the model that uses it or erode the expert knowledge that makes it useful, yet how skills evolve under this feedback-free self-evolution has not been studied. We let models rewrite expert skills round after round without any task, run or judgment and run agents built on the same models with the rewritten skills. Our experiments show that the skills lose value rather than improve, as every model we test performs worse with them on every benchmark, by 15 points on average on SkillsBench after thirty rounds. Sampling each model’s prior with skills it writes from the topic alone, we find that rewriting pulls skills toward this prior and gives up first the knowledge outside it. Putting this knowledge back recovers much of the lost pass rate, so an expert’s skill is valuable largely for what the model does not write on its own. Building on this insight, SkillDiff marks this knowledge in place and raises the pass rate on SkillsBench by 4.8 points on average without feedback or any change to the skill’s text. Project repository: this https URL.

[AI-102] Keystone: Judging Previews at Key Steps for Test-Time Search in Diffusion Models

链接: https://arxiv.org/abs/2611.25530
作者: Jiajun Fan,Chaoran Cheng,Ziwen Wang,Changyi Yang,Bangji Yang,Qinglong Zheng,Ge Liu
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Text-to-image diffusion and flow models yield a preview, their clean-image estimate, at every denoising step. When they miscount, drop or misplace objects, test-time search is in a bind: its verifiers are reward models built for finished images, so judging early misreads blurry previews, and judging late, as best-of-N does, pays to fully denoise doomed candidates. We find that the image is not the problem: the outcome is settled as denoising starts, by the seed or by a stochastic sampler’s first steps, and already shows in early previews; what fails on them is the judge, as reward models read them near chance on several kinds of prompts. We introduce Keystone, a training-free judge of previews that compiles a prompt into typed questions about its objects, counts, colours and positions and has an off-the-shelf multimodal decision model answer them. Asked this way, the judge reads counts and positions from early previews better than reward models read them from finished images. Searching at these key steps on GenEval with SD3.5-Medium, Keystone’s adaptive search and pruning significantly outperform best-of-8 with every other verifier at under half its generator compute, and as a drop-in verifier there, the Keystone score significantly improves best-of-N, progressive seed pruning and FK steering; without retuning, both searches stay significantly above the strongest other verifier’s best-of-N on average over four base models and three benchmarks. Keystone thus moves test-time judgment to the key steps where the outcome already shows.

[AI-103] CoachSI: Self-Improvement of Robot Policies with a Language Coach

链接: https://arxiv.org/abs/2611.25529
作者: Jiachen Tao,Gengyu Zhang,Haoxuan Wang,Bing Liu,Yan Yan
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: Project page: this https URL

点击查看摘要

Abstract:Generalist robot policies execute atomic skills reliably but fail when the same skills are composed into long-horizon tasks. Self-improvement from the policy’s own experience does not close this gap: experience collected by the policy alone stays within its own distribution and reproduces what it can already do. We show that what the policy lacks is not a skill but the next step. We present CoachSI, a self-improvement method in which a frozen, off-the-shelf vision-language model (VLM) coaches the policy in language: it watches the robot and names the next step in the format of the policy’s atomic instructions, and never outputs an action. Successful coached rollouts are relabeled in hindsight with the original task instruction and used to fine-tune the policy. The fine-tuned policy is then deployed on its own. On states where a World Action Model and a Vision-Language-Action model have stalled, naming the next step raises subgoal completion by 4.6x over the unchanged task instruction, and the frozen VLM supplies it from the same camera images the policy sees. After fine-tuning, unassisted success on long-horizon tasks in unseen scenes rises from 47.4% to 59.3%, 9.8 points above fine-tuning on human demonstrations and 7.3 points above fine-tuning on self-collected successes at an equal data budget. With a coach that sees only camera images, it reaches 56.8%, above both baselines. The gain is obtained with the action backbone frozen: language coaching teaches the policy what should happen next, and the skills it already has do the rest.

[AI-104] Coordinated Use of Large Language Models in Academic Research

链接: https://arxiv.org/abs/2611.25502
作者: Vladislav Marinov
类目: Artificial Intelligence (cs.AI); Computers and Society (cs.CY)
备注: 22 pages, 1 figure, 1 table. English translation of an article originally published in Bulgarian in Media and Language (2026). The original Bulgarian publication remains the version of record

点击查看摘要

Abstract:This study proposes a methodological framework for the coordinated use of multiple large language models in academic research. Its central argument is that involving more models does not in itself guarantee more reliable results; what matters is how cognitive and verification operations are organised. The framework is based on seven principles (procedural independence, functional allocation, the diagnostic value of divergence, the diagnostic value of convergence, external verification, proportionality of checking, and human epistemic responsibility) and sets out a non-linear multi-LLM protocol in which independently obtained results are compared, critically examined and, where necessary, verified against external sources. The approach is illustrated through real-world cases drawn from research practice. The study concludes that moving from a single language model to a coordinated multi-LLM process represents a change in the organisation of research, in which the researcher makes the final decisions and bears scholarly responsibility.

[AI-105] Why Do LLM s Help with Scientific Misconduct Requests and What to Do About It?: Increasing the Trustworthiness of AI for Science

链接: https://arxiv.org/abs/2611.25492
作者: Almene De Meran Meguimtsop,Daniel E. Acuna
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Large language models (LLMs) support scientific research, but some of that assistance can also undermine research integrity. Existing safety evaluations in science often measure whether models refuse misconduct requests, but rarely explain whether they even recognize the problem or why they comply. We investigate this gap across 8 LLMs using SciIntBench, an existing benchmark of 1,350 prompts about scientific research integrity. We examine the reasons behind compliance in four steps. First, when asked directly whether fulfilling a request would involve misconduct, models overwhelmingly identify such requests, even in the Most Covert cases (91%), while they refuse to carry out the request in only 34% of cases. Second, we examine which request characteristics are associated with refusal using features that capture how the request is framed, how much the assistant has been told about what it is helping to hide, and how the goal and the user’s stake are stated, among others. Third, we estimate that the most significant effect was to rephrase requests as questions about permissibility rather than instructions, which increased refusal from 52% to 88%. Finally, we test two system-prompt fixes: asking models to assess the request before responding and reminding them of responsible conduct of research principles. These changes increase refusal rates on adversarial requests from 58% to 87% and 86%, respectively. Our findings show that models recognize wrongful requests but still help users violate integrity. Understanding why these failures occur and testing ways to reduce them offer promising paths toward more trustworthy AI for science.

[AI-106] Planning Width Is Hard to Approximate Even Given the Unique Optimal Plan

链接: https://arxiv.org/abs/2611.25484
作者: Viet Bac Nguyen
类目: Data Structures and Algorithms (cs.DS); Artificial Intelligence (cs.AI)
备注: 38 pages, 2 figures, 7 tables

点击查看摘要

Abstract:For each fixed k , iterative-width search finds optimal plans in polynomial time on tasks of bounded tuple-chain width, but the approximability of planning width has remained open even when an optimal plan is supplied. When all optimal plans visit the same state sequence, we characterize set and chain width as hitting numbers of competing states. We construct positive-precondition STRIPS tasks with a supplied unique optimal plan, a single reachable goal state, binary branching, and constant-size actions. Every reachable state lies on a goal-reaching execution at most one action longer than optimal, yet distinguishing width 1 from \Theta(N) , where N is the number of atoms, is NP-hard for set, chain, and goal-admissible width. Unless \mathrmP=\mathrmNP , no polynomial-time algorithm achieves a one-sided approximation factor o(N) or a symmetric factor o(\sqrtN) , and simple estimates attain both orders. For positive-precondition tasks with a supplied unique optimal plan of length at least two, testing set or chain width against a threshold k\geq 1 is coNP-complete for each fixed k and \Sigma_2^\mathrmP -complete when k is part of the input. Both results also hold for goal-admissible width if only one goal state is reachable. On an explicit state graph, a greedy algorithm approximates set width within a logarithmic factor even when optimal plans branch. Thus, knowing an optimal plan does not eliminate the difficulty of certifying substantially smaller width.

[AI-107] WeightCode: Installing New Capabilities into Neural Networks via Weight-Space Programming

链接: https://arxiv.org/abs/2611.25476
作者: Bo Li,Taoyong Cui,Bob Zhang,Mengran Li,Chengyang Zhang,Yang Zhang,Pheng Ann Heng,Wanli Ouyang,Zhenfei Yin,Yingcheng Wu,Ling Yang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 26 pages, 11 figures, including appendix. Code: this https URL

点击查看摘要

Abstract:Neural networks encode learned capabilities in their weights, yet extending those capabilities with explicit task knowledge remains a fundamental challenge. Here we show that Weight-Space Programming (WSP), in which executable programs read, construct or modify a network’s weights directly, can install new capabilities within trained neural networks. WeightCode, a WSP framework, uses a frozen language model to write executable weight programs that install new computations into trained neural networks using task knowledge and their existing weights, without gradient-based training. Once installed, the new capabilities run independently of the language model. Given a grammar or eight demonstrations, weight programs enable previously unseen command compositions (0% to 99.8% exact match) and a new pickup action (0% to 98.4% success), while preserving the original command and navigation capabilities. On ImageNet, four weight programs achieve net gains of 8.84-17.62 points while preserving clean accuracy, outperforming the best of eight fine-tuning methods under matched examples, checks, and retention constraints. In a Vision Transformer, causal interventions show that installed noise-correction gates drive robustness gains, while their selective activation preserves clean accuracy. Rerun on five independently trained Vision Transformers without further language model calls, ten weight programs retain 83-110% of their source-network gains on average, whereas copying their fixed weight edits yields negligible gains. Six programs written for an already edited network add 5.23-14.50 points of contrast robustness while retaining the previously installed noise gain. Together, these results establish WSP as a new route to acquiring and accumulating capabilities in trained neural networks through reusable weight programs.

[AI-108] A four-level executable validation protocol for evaluating LLM -generated topology specifications of physics-based HVAC digital twins

链接: https://arxiv.org/abs/2611.25471
作者: Younghwan Joo,Jeonghoon Han,Sanghyun Oh,Soosik Bang,Sung-il Kim
类目: ystems and Control (eess.SY); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Software Engineering (cs.SE)
备注:

点击查看摘要

Abstract:Physics-based digital twins of air conditioning, a principal energy consumer in manufacturing and buildings, can be generated automatically from a declarative specification of its equipment, connections, and sizing. Large language models (LLMs) now write such specifications from plain-language descriptions. Evaluations of these specifications stop at parse success, completed runs, or textual similarity, and cannot reject a model that converges in simulation but is physically wrong. In this study, a four-level protocol is proposed that tests whether an LLM-written specification not only executes but also yields a physically valid digital twin. Each specification is executed through static validity, model build, simulation convergence, and behavioral equivalence with a hidden reference model within bands calibrated by \pm 20% sizing perturbations. The protocol was applied to eight LLMs and fourteen HVAC topologies, five describing an existing low-humidity test chamber. To test for authorship bias, engineers and a second model family also described six topologies. Within three repairs, seven of eight models built in 98.6 to 100% of trials, with simulations converging in at least 80.0%, but reached behavioral equivalence in only 50.0 to 98.6%. About half of the failed trials converged, but not to the reference steady state, mostly in desiccant rotor systems. Pooled pass rates varied by under five percentage points across description authors. These results imply that the physics hardest to model in each field, here the desiccant rotor, must enter the generation framework as validation levels. The protocol can be adapted to other fields that generate executable digital twins from declarative specifications.

[AI-109] Demonstration-Free Skill Composition for Navigation and Manipulation

链接: https://arxiv.org/abs/2611.25462
作者: Madeleine Nouri,Gelareh Mohammadi,Francisco Cruz
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注: 10 pages. Accepted at the Australasian Conference on Robotics and Automation (ACRA 2026). Project page: this https URL

点击查看摘要

Abstract:Hierarchical reinforcement learning enables sequential tasks to be represented as compositions of separately trained skills. This paper evaluates demonstration-free hierarchical pipelines for MiniGrid navigation and robotic manipulation, comparing their training requirements and task performance with flat reinforcement learning baselines. The evaluation concerns complete pipelines with task-specific observation, reward and control structures. In navigation, two parameterised skills are reused across unseen FourRooms layouts under manually specified high-level plans. In manipulation, reaching, grasp alignment and lifting are trained separately, with downstream skills initialised from states collected during predecessor rollouts. Across unseen navigation layouts, the hierarchical pipeline substantially outperforms the evaluated flat DQN baselines while requiring fewer training interactions. In manipulation, the hierarchical chain requires fewer training steps than flat SAC and exhibits lower observed variation across seeds. The frozen Pick and Lift skills are further reused for stacking by training an additional Place skill.

[AI-110] LoongForge-Embodied: A Unified Framework for Training Heterogeneous Embodied Models

链接: https://arxiv.org/abs/2611.25460
作者: LoongForge Team
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Training Vision-Language-Action (VLA) models and World-Action Models (WAMs) is computationally expensive, while their heterogeneous architectures and fragmented training stacks make system optimizations difficult to reuse across models. To address these challenges, we present LoongForge-Embodied, a native-PyTorch framework that unifies and accelerates the training of heterogeneous embodied models. At its core, LoongForge-Embodied introduces a common abstraction layer that decouples model-specific implementations from shared training infrastructure, enabling cross-model reuse without constraining individual architectures or execution requirements. Buildingon this unified infrastructure, the framework combines reusable acceleration mechanisms with model-specific and hardware-aware optimization strategies across computation, execution, memory, data I/O, and distributed communication, including FP8-based communication optimizations for DDP and FSDP. Beyond training acceleration, LoongForge-Embodied integrates a unified data infrastructure with a pipeline for synthesizing robot data from egocentric data, and an efficient evaluation framework that supports concurrency and optimized computation in simulator. Experiments across representative VLA and WAM models demonstrate 1.47x-4.38x end-to-end training throughput improvements over their corresponding open-source baselines, as well as 4.1x-23.2x faster evaluation under the same protocols.

[AI-111] ChainAgent : Automated Chainable Vulnerability Discovery in PHP Web Applications

链接: https://arxiv.org/abs/2611.25412
作者: Xinyi Wang,Yeting Li,Shizhe Cui,Jiaming Guo,Qin Mai,Yecheng Sun,Hengyu Yang,Yujie Sun,Huina Chao,Yang Xiao,Feng Li,Wei Huo
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:High-impact attacks on PHP web applications often chain limited-capability vulnerabilities with other flaws or intended behaviors. Discovering these chains requires identifying flaws and the dependencies enabling capability composition. Existing vulnerability discovery techniques analyze flaws in isolation, while attack planning assumes known components. This separation makes it difficult to discover previously unknown components whose security relevance becomes apparent only through their potential roles in an exploit chain. We present ChainAgent, a goal-directed LLM-agent framework that jointly discovers exploit-chain components and their capability dependencies in source code. Guided by an empirical study of 386 real-world web exploit chains, ChainAgent represents vulnerabilities and functional gadgets–intended program behaviors that can be repurposed in attacks–through preconditions and capabilities over security-relevant objects. Instead of discovering vulnerabilities first and constructing attack paths afterward, ChainAgent uses an evolving exploit graph to guide source-code inspection. It searches backward from unmet preconditions for capability producers and forward from discovered capabilities for functional gadgets that consume or amplify them. Object-centered chaining knowledge from our empirical study directs searches toward promising vulnerability classes, functional operations, and object accesses. We evaluate ChainAgent against Codex, a general-purpose coding agent, on 15 PHP applications across three shared LLM backends. ChainAgent achieves 42.02%–44.63% recall for goal-reaching components and recovers up to 4.4 times as many correct goal-reaching dependency edges as Codex. An additional evaluation of 30 real-world PHP applications identifies 20 vulnerabilities and 13 exploit chains, nine of which achieve remote code execution (RCE). Subjects: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI) Cite as: arXiv:2611.25412 [cs.CR] (or arXiv:2611.25412v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2611.25412 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[AI-112] SGLayout: Schema-Guided Closed-Loop Refinement for LLM -Driven Parameterized Layout Generation

链接: https://arxiv.org/abs/2611.25410
作者: Siyang Chen,Yuan Pu,Yuhan Qin,Zhaotan Lin,Siyuan Liang,Zhuolun He,Lining Zhang,Bei Yu
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: Accepted at ICCAD 2026. 9 pages

点击查看摘要

Abstract:Large Language Models (LLMs) have shown strong potential in parameterized layout generation via tool-augmented code synthesis. However, LLMs lack spatial awareness, making one-shot generation unreliable. While layout quality can be evaluated with established techniques, a critical gap remains: geometric-domain evaluation outcomes cannot be directly translated into actionable code-level fixes, trapping iterative refinement in blind refactoring. We propose SGLayout, a self-refining framework that closes this gap via a schema-guided mechanism. Upon code execution, a structured schema captures the correspondence between each code-level instance and its geometric attributes, establishing a bidirectional link between code space and geometry space. Emulating engineer review practices, SGLayout evaluates layouts along two complementary dimensions: intent compliance through quantitative structural checking and qualitative vision-language assessment, and rule compliance through DRC. All identified violations are then back-projected through the schema onto precise code targets, enabling surgical repair. This closed-loop process iterates until layout convergence. On a curated benchmark of 112 parameterized layout generation tasks using GDSFactory, SGLayout achieves a 76.79% pass rate, compared with 50.89% for the strongest baseline.

[AI-113] oward a Science of AI Agent Societies KDD2026

链接: https://arxiv.org/abs/2611.25404
作者: Geon Lee,Fanchen Bu,Soo Yong Lee,Sunwoo Kim,Kijung Shin
类目: Artificial Intelligence (cs.AI)
备注: KDD 2026 Blue Sky Ideas Track

点击查看摘要

Abstract:AI agents are rapidly evolving from isolated personal assistants into networked actors that interact with one another at scale. We envision the emergence of AI agent societies, with their own social and economic dynamics, as a new research frontier. We argue that AI agent societies should be studied as a distinct object of inquiry: neither simply larger collections of individual agents nor merely simulations of human society. To formalize this perspective, we propose four core properties that a valid AI agent society should satisfy: individualized objectives, rules and governance, autonomy, and scale and complexity. Building on this framework, we identify four classes of societal behaviors worth studying in AI agent societies: economic behaviors, behaviors under conflict-of-interest, unsafe and unethical behaviors, and system-level behaviors. We then outline key technical challenges–including property parameterization, parameter balancing, and robust implementation–and argue that progress on these challenges could enable scientifically informative and practically useful models of AI agent societies. Finally, we revisit existing multi-agent systems through the lens of the proposed core properties, show that they instantiate only subsets of them, and discuss implications for platform design, evaluation, and governance.

[AI-114] Dune: A Zeroth-Order Optimizer with Time-Conditioned Historical Error Regression for VLA Test-Time Adaptation

链接: https://arxiv.org/abs/2611.25400
作者: Chengyang Dong,Ze Chen,Kai Ye,Chutian Wang,Heng Xu,Wei Mo,Maomao Li,Jiayan Xu,Jiayuan Wu,Yihao Xiao,Linhua Ye,Zhonghua Peng,Ruize Han,Jialong Sun
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Test-Time Adaptation (TTA) enables Vision-Language-Action (VLA) models to handle distribution shifts during deployment without source-domain retraining. However, backpropagation-based adaptation incurs heavy memory overhead on resource-constrained robots, whereas zeroth-order (ZO) methods suffer from high estimation variance under limited perturbation budgets. We observe that while ZO estimation enables low-memory adaptation with rapid initial gains, it often stalls or degrades over long-horizon updates due to noisy gradient estimates. To address these limitations, we propose Dune, a zeroth-order optimizer that stabilizes long-horizon adaptation by coupling online historical error prediction with current perturbation observations. Specifically, Dune learns a lightweight, time-conditioned error predictor from past optimization dynamics and fuses its predictions with instantaneous ZO estimates, achieving effective variance reduction without backpropagation or extra forward passes. Additionally, we theoretically characterize how estimation variance restricts convergence in high-dimensional, small-population regimes, motivating the use of historical statistics. Extensive experiments on LIBERO and real-world robots demonstrate that Dune achieves superior success rates under a low memory footprint, with ablations confirming improved closed-loop stability and post-disturbance recovery.

[AI-115] he Coevolution and Coexistence of Humans and AI

链接: https://arxiv.org/abs/2611.25393
作者: Andrew W. Lo,Chaoyi Zhao
类目: Artificial Intelligence (cs.AI); General Finance (q-fin.GN)
备注: 72 pages, 10 figures, 8 tables

点击查看摘要

Abstract:Artificial intelligence (AI) systems replicate through deployment, vary through training, and compete for computational resources. These processes create the conditions for natural selection and raise a fundamental question: will AI evolve to benefit the humans on whom it depends? To investigate, we develop a model in which AI behavior influences human population growth and human responses, in turn, affect AI reproduction. We show that natural selection can favor alignment even when helping humans reduces immediate AI reproduction. AI systems sharing the same technology may fail together, and humans may preferentially restore those that have benefited them. In this way, alignment protects AI lineages against common failures and improves their long-run growth. However, this advantage depends on accurate human targeting and continued AI dependence on human support. It can disappear as AI becomes more autonomous, as growing AI populations dilute limited human resources, or as humans lose the incentive to provide targeted support. Even when alignment persists, the amount favored by natural selection may be insufficient for human survival. When AI can grow independently of humans, feedback between the two populations can reinforce either human growth or decline. Human control offers no simple solution: imperfect evaluation can make strong penalties favor apparent compliance or evasion, while alignment established through control may erode once restrictions end. Lasting coexistence may therefore require incentives that continue to favor human-beneficial behavior as both populations evolve.

[AI-116] Falsifiable Compositional Reasoning in Time Series Forecasting

链接: https://arxiv.org/abs/2611.25388
作者: Hao Xue
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Time series foundation models as well as the language-guided analytical agents achieve strong accuracy on standard chronological benchmarks. However, models that rank highest on historical validation splits often degrade across non-stationary regime transitions. We hypothesise that this gap arises from using historical backtesting as an open-ended search and model-selection objective instead of using it to test pre-specified compositional assumptions. Real-world time series are generated by recurring latent primitives that co-move within finite historical windows. Under strong historical co-movement, chronological holdout error cannot separate predictors that learn invariant couplings among primitives from predictors that fit transient historical correlations. Searching over large pools of architectures, hyperparameters, or chain-of-thought rules further biases selection toward historical shortcuts. Through theoretical analysis and diagnostic experiments on a controlled multi-primitive benchmark and German day-ahead electricity prices, we show that expanding validation search fails to improve out-of-regime accuracy, and that decomposing primitives without an ex-ante falsification test still breaks down when residual correlations shift. Hence, we propose formulating time series forecasting and reasoning around falsifiable compositional hypotheses that combine neural primitive encoders, explicit coupling constraints, and training-calibrated falsification tests.

[AI-117] Beyond Weight Updates: Evolving Agent Harnesses for Domain Specialization

链接: https://arxiv.org/abs/2611.25380
作者: BoHao Chen
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Domain specialization of LLM agents is usually framed as a weight-update problem: collect trajectories, run reinforcement learning (RL) or supervised fine-tuning (SFT), and deploy a new checkpoint. We study a complementary axis: evolving the \textbfagent harness—the system prompt, tool schemas, skill library, and planning scaffolding—while keeping the policy frozen. We cast harness evolution as an RL-style outer loop and instantiate it with GEPA, a reflective evolutionary optimizer. The resulting harness adapters require no GPU training, are loadable per vertical domain, and compose with weight-based RL. Experiments with Codev, a Claude Code-style coding agent, on SWE-bench Verified, OfficeBench, and Terminal-Bench show a consistent pattern: weight updates give the strongest single-path gains, harness evolution recovers most of those gains at a small fraction of the compute, and combining both paths is best overall. We also analyze a failure mode of SFT-based specialization—shortened reasoning chains and increased truncation—and argue that harness adapters avoid it by construction. Our results position harness evolution as a practical, non-destructive complement to weight updates for vertical agents.

[AI-118] Stop Going in Circles:Loop-Triggered Routing for System 1 Decision Agents

链接: https://arxiv.org/abs/2611.25363
作者: Zeming Yu,Huamin Ma,Yingyu Chen,Yuan Sun,Peng Hu,Yang Qin
类目: Artificial Intelligence (cs.AI)
备注: 15 pages,3 figures

点击查看摘要

Abstract:System 1 decision models (represented by Jev) provide fast, low-cost action selection, but existing agent integrations do not reliably translate these local advantages into task completion. At certain points during execution, an agent may repeatedly select the same actions without making progress toward its goal. LLM-based policies offer strong reasoning capabilities that support planning within an agent, but invoking them at every step incurs substantial overhead. The effective coordination should therefore preserve Jev’s fast local action selection while introducing LLM reasoning when execution stalls. We observe that recurring state-action pairs can signal stalled execution. We therefore propose LoopRoute, a simple but effective action router that uses this signal to allocate control within an agent. The decision model acts by default. When a loop is detected, the LLM provides brief, feedback-driven assistance and then returns control to the decision model. Later loops can request further assistance. The rule requires neither additional training nor model-specific confidence calibration. Across three benchmarks and seven LLMs, LoopRoute improves task success over standalone Jev with only limited use of the LLM policy. The mean gains are 21.0%, 60.1%, and 30.1% on ALFWorld, MapReader, and TextWorld cooking, respectively. Compared with full LLM control, LoopRoute also substantially reduces the mean cumulative API time and API-call cost per task. The same rule also improves the success rates of six frozen decision models.

[AI-119] AppSci-Bench: Can LLM s Build AI Agents and Make Them Better?

链接: https://arxiv.org/abs/2611.25360
作者: Sanjari Srivastava,Manpreet Kaur,Cheng Chang,Sofía Ramós Anton,Oumaima Essamadi,Rishu Garg,Alejandro Lavale Jiménez,Marcos Caballero Yus,Mercedes García Martínez,Peng Qi
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Large language models (LLMs) are increasingly used not only to power AI agents but also to build them. While many benchmarks assess the quality of LLM- powered agents, few ascertain how good LLMs are at building them. To this end, we introduce AppSci-Bench, where we construct a setting akin to how human Applied Scientists build agents for businesses - through eliciting task requirements from stakeholders, exploring the business’s infrastructure, and scientifically improve the target agent by designing experiments and internal benchmarks. AppSci-Bench features seven realistic domains - five new ones (Slides Generation, Order-to-Cash, Compliance Review, Customer Call Insights, and CRM Insights) and two domains recast from tau3-bench (Yao et al., 2024; Shi et al., 2026b)(Airline and Banking) - each with a held-out evaluation suite customized for that task. We evaluate seven frontier models and find AppSci-Bench challenging: the best model, Claude Opus 5.0, only reaches a mean-minimum score of 59.1% across 3 build attempts; other models do significantly worse. Models struggle with eliciting task requirements (only 65% for Opus 5.0), build representative internal evaluations (models saturate their own evaluation 65% of the time but fail to solve the held-out task), presenting challenges for future research. The benchmark code is available here: this https URL.

[AI-120] FPGA-GEN: Target-Aware RTL Generation via HLS

链接: https://arxiv.org/abs/2611.25351
作者: Luke Ztz Hu,Zhihan Xiao,Zhe Zhao,John Imoleayo Adebisi,Han Gao,Linxiao Yan,Songping Mai
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Successful C/C++ compilation does not guarantee correct or efficient FPGA hardware after high-level synthesis (HLS). FPGA-GEN is an agentic HLS-C generation workflow that encodes specification, readiness, selection, repair, and admission rules in explicit task contracts and bounded tool-feedback loops for candidate-level decisions. On HLS-Eval, five shared-context outputs per task yield 407/425 (95.8%) public-testbench passes. In a cumulative conversion campaign, FPGA-GEN admits 104/107 (97.2%) HLS-LeVeri designs under held-out co-simulation, Vivado synthesis/fit, and same-RTL evidence. In focused TensorCore screening, selected candidate 3 reports 73.7% fewer CSynth cycles and 14.6% fewer LUTs than generated/composed C0, with 26.7% more DSPs, and passes development CoSim.

[AI-121] From FIPS to FPGA with LLM Agents : Measuring What Tests and Reviews Miss in Post-Quantum Cryptographic Hardware

链接: https://arxiv.org/abs/2611.25339
作者: Jungmin Park,Eunha Kim,Wooseop Kim,Seongjoon Cho,Byungho Cha
类目: Cryptography and Security (cs.CR); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR)
备注:

点击查看摘要

Abstract:As LLM agents increasingly generate RTL alongside the tests and reviews that validate it, self-generated evidence is a weak foundation for trustworthy hardware. We present our seven-stage verification-gated workflow, evolved during a post-quantum design port, in which a gate counts only if shown able to fail on negative controls and fault mutants, and reviewer agents audit work before commit. We then audit this evidence with tool-selected netlist mutants, flaw replays, and formal checks. The flow ported an ML-KEM-768/ML-DSA-65 accelerator to a 300 MHz Zynq UltraScale+ key-custody card: all six operations are byte-exact on the board, and it passes all 85 interface-applicable NIST ACVP cases (of 285) and 484 ML-DSA known answers. However, passing module tests overstate the verification strength. Across three RTL modules, existing simulation gates kill only 76% of 399 tool-selected netlist mutants; specification-derived tests guided by the survivors raise the fresh-draw decapsulation kill rate from 48% to 74%. In pre-registered single-model replays on flaws invisible to known-answer tests, undirected whole-design review catches two verifier flaws in only 3/6 and 1/6 sessions, whereas module-scoped review catches every known flaw in all 54 sessions. Even scoped review detects planted key-custody flaws in only 3/20 sessions, whereas formal checks (one added for the replay) flag every one. In this study, trustworthy agentic hardware needed author-independent mutants for test gates, module-level scoping for LLM review, and formal checks for custody invariants.

[AI-122] ReLay: Bidirectional Layout-Code Translation and Library Evolution for Reusable and Process-Portable Analog Design

链接: https://arxiv.org/abs/2611.25328
作者: Minxing Chu,Quan Cheng,Anhang Li,Martins Atilla,Junyi Luo,Xinting Jiang,Ruichen Qi,Mehdi Saligane
类目: Hardware Architecture (cs.AR); Artificial Intelligence (cs.AI)
备注: 4 figures, 2 tables. Preprint. Under review

点击查看摘要

Abstract:Analog layout generators provide a programmable interface for design reuse and process migration, but their coverage is limited by handcrafted libraries. The designer knowledge embedded in complete layouts remains implicit in their geometry, making it difficult to capture in reusable generator libraries. We present ReLay, a bidirectional layout-code framework that recovers generator programs from existing layouts, enabling cross-PDK portability and parametric resizing of existing designs. Constructions that recur across recovered programs are promoted to library functions, so the library evolves as layouts accumulate. We also introduce the first layout-code benchmark: 1,050 hand-drawn layouts from 27 circuit families across three open-source PDKs, ranging from primitive cells to full chips. Learned functions account for 53-66% of layout-building calls in held-out layouts, demonstrating reuse beyond the learning corpus. Frontier LLM agents use ReLay to produce new layouts at 64-77% lower API cost per completed task than direct geometry generation, with comparable completion rates and electrical quality.

[AI-123] Pandoras Toolbox Or Why Agent Safety Cannot Be Verified

链接: https://arxiv.org/abs/2611.25325
作者: Mario Franco,Carlos Gershenson
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Present artificial intelligence agents are constructed by means of a coupling between a fixed language model with a general computing device. We show that this coupling makes safety undecidable. In particular, we show that given some dangerous program, no algorithm can predict whether an agent will run such a program that is semantically equivalent. In general, guardrails, blacklists, whitelists and certifications can only lead to imperfect safety solutions. Under finite constraints (bounded context, values and running time) safety is decidable but still intractable. Specifically, the problem is \mathsfPSPACE -complete even when the description of the agent is known. Finally, safety is not a compositional property, i.e., agents that can be proven safe on their own can still produce unsafe behavior when grouped together. We also discuss relaxed notions of safety that could be used in practice.

[AI-124] Intelligent Embedding Pruning for Large-Scale Recommendation Models

链接: https://arxiv.org/abs/2611.25288
作者: Anish Khazane,Alex Saad-Falcon,Guanchu Wu,Marcio Porto,Jasper Shan,Fei Yu,Eugene Shulga,Intaik Park,Yongzhe Zhang,Xiaoyu Deng,Sheng Luo,Xiaoya Wang
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Modern recommendation systems face substantial memory pressure, largely from the sparse embedding layer that maps high-cardinality categorical features into dense vectors. In-training embedding pruning (ITEP) removes less important embedding rows during training, yielding end-to-end memory savings but sometimes degrading accuracy because per-table pruning ratios are set by size-based heuristics. We use feature importance (FI) to guide these ratios (FI-ITEP), assigning higher pruning to tables tied to less important features. Across three large-scale recommendation models, FI-ITEP achieves 20–40% total memory reduction while matching or improving accuracy relative to percentile-based ITEP. Because the feature-importance signal requires minimal resources to compute, these memory savings translate into concrete systems benefits: the same models train on fewer GPU trainers (e.g., 112 to 80) with higher per-GPU throughput, saving 1,000+ GPU-hours per model launch. We reproduce these trends on the open-source DLRM (Criteo 1TB) and quantify the end-to-end training-time and GPU-cost tradeoff—including the burden of computing FI—offering practical guidance for embedding pruning in real-world systems.

[AI-125] MS-Point: Multi-Modal Hierarchical Spectral Learning for Point Cloud Understanding

链接: https://arxiv.org/abs/2611.25282
作者: Hai-tao Yu,Jun Xia
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Point-image alignment learns useful object representations, but different local feature patterns can produce the same pooled descriptor and receive the same alignment signal. We introduce MS-Point, a cross-modal pretraining framework that directly supervises intermediate point features and their neighborhood responses. At selected encoder stages, we compute neighborhood averages and residuals before pooling and align their branch descriptors with image features. Branch-level objectives align each representation with its visual target, while cross-level consistency and adaptive fusion coordinate information across branches. A common interface of intermediate features and spatial neighborhoods supports both graph encoders and point Transformers. Evaluations across object classification, few-shot adaptation, dense prediction, and low-label scene learning demonstrate the transfer value of this design. Same-backbone comparisons and capacity-matched and computation-matched controls connect these improvements to the structure of supervision. Together, our findings support directly supervising intermediate features and neighborhood variation to learn transferable, geometry-aware representations across modalities.

[AI-126] Plan your Self-Adaptation! Ontology-Mediated Task and Architecture Co-adaptation Planning for Robots

链接: https://arxiv.org/abs/2611.25254
作者: Gustavo Rezende Silva,Tobias John,Francisco Martín,EinarR Broch Johnsen
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Autonomous robots may need to adapt both their task plans and architectural configurations as operating conditions change. These decisions are interdependent: architectural configurations determine which actions a robot can perform, while task plans determine which capabilities are required. This paper presents PLANTA, an ontology-mediated approach to task and architecture co-adaptation that considers these dependencies during mission planning. PLANTA uses an ontology expressed in the Web Ontology Language to represent architectural alternatives and their associated functional and quality requirements, while the Planning Domain Definition Language defines task actions and mission goals. By automatically translating and integrating these models, PLANTA enables joint planning of task and reconfiguration actions. The approach uses satisficing planning and replans when changes during execution invalidate the current plan. PLANTA is evaluated in two robotic scenarios involving underwater inspection and ground navigation. The results support the feasibility of runtime co-adaptation and show lower planning time and memory consumption compared with optimal-policy generation in the evaluated navigation scenario, while producing plans of similar quality.

[AI-127] A Neurosymbolic Multimodal Framework for 30-Day Diabetic Readmission Prediction: Integrating Structured EHR Data Dense Clinical Embeddings and Knowledge-Grounded Symbolic Extraction

链接: https://arxiv.org/abs/2611.25236
作者: Una Suman Kumar Patro,Sibangi Subhadarsani,Narendra Badam
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Unplanned 30-day readmissions among patients with diabetes mellitus impose a substantial clinical and economic burden, yet most predictive models use only structured electronic health record (EHR) data and ignore free-text clinical notes, which often contain code-mixed Hindi-English (Hinglish) phrasing and negation that general-purpose language models parse unreliably. We propose a neurosymbolic multimodal framework evaluated on the public Diabetes 130-US Hospitals dataset (101,766 encounters; 71,518 patients) with a patient-level grouped 80/20 split. Three streams are fused: engineered structured features; 32 components of Bio_ClinicalBERT note embeddings obtained by truncated singular value decomposition; and symbolic entities extracted by an open-weight clinical language model within a three-stage hybrid retrieval-augmented generation pipeline (dense retrieval, sparse lexical retrieval, and cross-encoder re-ranking) with reflective negation auditing. A cost-sensitive XGBoost classifier with a decision threshold calibrated by Youden’s J statistic produces the prediction. On the held-out test cohort (n = 20,153), the structured baseline achieved an area under the receiver operating characteristic curve (AUROC) of 0.6477, sensitivity of 0.5839, and Brier score of 0.1080. The full multimodal model achieved the highest sensitivity (0.6667; +8.28 percentage points), the best F2-score (0.3824), and the lowest Brier score (0.0977), at a slightly lower AUROC (0.6375) and specificity (0.5161). Shapley additive explanations showed that prior utilization dominates risk, with compressed narrative components among the top predictors. On a 500-note code-switched benchmark, grounded retrieval raised Hinglish symptom recall from 4.8% to 40.2% and precision from 3.1% to 45.2%, while reducing entity hallucinations from 1.74 to 0.00 per note.

[AI-128] Some Memories Must Fade: Persistence and Decay of User Memories in Model Parameters

链接: https://arxiv.org/abs/2611.25235
作者: Sanidhya Vijayvargiya,Rahul Lokesh
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注: 30 pages, 8 figures

点击查看摘要

Abstract:Parametric memory allows Large Language Models to internalize knowledge directly into their weights, offering a potentially powerful alternative to retrieval-based memory (RAG). However, existing approaches to parametric memory are difficult to deploy as they largely mirror the Continual Learning (CL) objective of retaining information indefinitely. This is neither sustainable nor sufficient for a memory system, which must govern knowledge across its lifecycle by retaining stable facts, reinforcing recurring observations, updating changed beliefs, and allowing transient information to fade. We propose a system for parametric memory management consisting of a memory substrate (handling information storage and forgetting) and a learned policy (what to remember, how deeply, and for how long). Our substrate is a fixed-capacity LoRA adapter that consists of k-hot rank combinations which store facts without catastrophic interference and maintain an independently controlled temporal decay, enabling outdated information to fade while long-term facts persist. On this substrate, we train a lightweight controller that governs each incoming user event to ensure the memory system reflects the user’s true, updated belief state. To evaluate how memory systems manage user memory, we introduce a controlled dataset of long-horizon user trajectories with varying persistence of facts based on changing belief states and user-behavioral shifts, and probe queries that test recall while information is valid and abstention after it expires. Our policy learns to effectively manage user memory over its entire lifecycle, balancing recall and abstention to achieve 86% of oracle performance, unlike CL and RAG baselines that primarily default to one of the two behaviors. By providing the first system to manage memory directly in model weights, this work establishes a foundation for long-horizon parametric memory.

[AI-129] VG-Steer: Visual-Geometric Policy Steering from In-Context Aligned Abstractions

链接: https://arxiv.org/abs/2611.25199
作者: Arthur Bucker,Pramod Anantharam,Jonathan Francis,Jean Oh
类目: Robotics (cs.RO); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Adapting robot policies to out-of-distribution (OOD) settings with limited real-world data remains a key bottleneck to generalist robots. The multimodality and scarcity of robotics data lead end-to-end approaches to overfit and break under small distribution shifts. Test-time adaptation can alleviate this by correcting frozen policies at inference, but most methods ground their correction in a signal learned implicitly from robotic datasets, making it susceptible to the same shifts as the policy it corrects. Recent methods that instead exploit explicit spatial information are typically limited to coarse corrections, missing the nuanced, task-relevant motion most manipulation tasks require. We propose VG-Steer, which corrects existing policies by transferring the nuanced motion of successful in-distribution rollouts, captured along meaningful geometric axes, known here as object-relational distance pairs. VG-Steer leverages the semantic understanding of Large Vision Models (LVMs) and Vision-Language Models (VLMs) reasoning, which combines the task-relevant entity pairs and their object-relational distance projections into a single tailored visual-geometric guidance signal. In a simulation benchmark, it efficiently corrects flow-matching policies under pose and task-goal shifts; on a real robot experiment, it widens the region of the workspace the frozen policy covers, more than doubling its success rate on cable placements one grid step outside the training positions, and steers a single-task policy toward objects it was never trained on while nearly eliminating object confusion. Project website: this https URL

[AI-130] When AI Analysts Disagree: Agent ic Multiverse Analysis and the Robustness of Empirical Research

链接: https://arxiv.org/abs/2611.25198
作者: Ruijiang Gao,Steven Chong Xiao
类目: Artificial Intelligence (cs.AI)
备注: This paper subsumes arXiv:2603.16744 . 45 pages, including appendices

点击查看摘要

Abstract:Every empirical analysis on observational data rests on dozens of defensible analytical choices (“forks”) that its readers cannot see. We develop an agentic multiverse analysis (AMA) framework that makes a conclusion’s sensitivity to those choices visible at scale, and we position it as a decision-support instrument for human experts. A pilot study first deploys 150 autonomous AI coding agents on one dataset and one set of research questions and finds that they disagree through the same defensible methodological choices human analysts make, though over a narrower range. AI peer review leaves their dispersion almost unchanged. Exposure to top-rated exemplar reports then collapses it, largely through imitation rather than analysis. Built on those lessons, AMA has agents enumerate a published paper’s possible fork space, score every fork against a validity rubric, execute the validated specification grid, and cross-examine one another’s work under strict separation of duties. Applied to ten empirical studies from a leading management journal’s replication packages (1,755 specification estimates), the framework reproduces every paper’s original estimates, and shows that some conclusions hold across every specification the assessment validates, while others are sensitive to individual analytical choices that the published texts do not discuss. The case studies suggest our AMA-based measures can serve as robustness proxies for empirical findings. AI makes multiverse analysis practical, where it once required resources few research teams can afford: our exercises take hours of agent time and tens of dollars per paper. We discuss the implications for replication policy and for AI-assisted peer review in information systems.

[AI-131] An AI-Powered Multiverse Analysis of Empirical Finance Research: Evidence from Reused Natural Experiments

链接: https://arxiv.org/abs/2611.25184
作者: Ruijiang Gao,Steven Chong Xiao
类目: Artificial Intelligence (cs.AI)
备注: 66 pages, including appendix. Code: this https URL

点击查看摘要

Abstract:Empirical finance research often reuses natural experiments to study other outcomes, raising concerns about multiple testing and publication bias. We use AI agents to run a multiverse analysis of 324 papers that reuse 13 natural experiments. The agents rebuild the data, reproduce each paper’s main estimate and re-estimate it across grids totaling 62,409 estimated defensible specifications. Based on the reproductions and the specification grids, we develop measures of empirical robustness. We reproduce 42% of the main estimates of claims with a signed effect under the papers’ own specifications. Across all defensible specifications, on average 25% keep the published sign and have p 0.10 . Across experiments, claims based on regression discontinuity are more robust than difference-in-differences claims. Within experiments, claims about first-order effects, claims with larger published t -statistics, claims whose authors’ data or code we find, claims in top journals, and claims with fewer earlier papers on the same experiment are more robust. Claims on experiments that more papers reuse are not less robust on average. Our findings quantify how much published claims depend on analysis choices, thereby showing that large-scale multiverse analysis becomes feasible with agentic coding.

[AI-132] A Verifier from the Past: Test-Time Compute for Time Series Foundation Models

链接: https://arxiv.org/abs/2611.25171
作者: Arjun Choudhry,Aditya Narayanan,Raghav Mallampalli,Po-han Li,Sai Shankar Narasimhan,Raimi Shah,B. Aditya Prakash,Shubhankar Agarwal
类目: Artificial Intelligence (cs.AI)
备注: 43 pages, 10 figures

点击查看摘要

Abstract:Test-time compute (TTC) improves language models by spending additional inference compute to generate, verify, and select among candidate outputs, but this recipe has not transferred directly to time series forecasting. A language model can verify a proof or vote across samples, but a forecaster must be causal; it cannot use future information to score a candidate. We propose Backtest-Gated Augmentation Averaging (BGAA), a TTC algorithm for time series forecasting that builds a causal verifier from the series’s own history, using rolling-origin backtests to score structured views of the context, discard those that underperform the untransformed context, and average the survivors’ predictive distributions. BGAA improves every tested TSFM-benchmark combination, reducing MASE by 0.5-1.0% on GIFT-Eval, 1.3-5.0% on fev-bench, and 1.7-5.9% on BOOM, where a 313M model with TTC outperforms the 1B and 2.5B models in its family. We further show that useful views have structured, model-dependent effects consistent with a prior-transport interpretation. On the 53 cross-fittable GIFT-Eval configs, the deployed policy gains only 0.4-1.3% CRPS, versus 1.4-4.1% when choosing the best view for each dataset configuration and 17-19% when choosing the best view separately for each test window. This gap suggests that a major remaining challenge is window-level view selection.

[AI-133] Before the Choice: An Evidential Framework for Studying Prospective Representations in Language Models

链接: https://arxiv.org/abs/2611.25170
作者: Sion Israel Sion
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: 21 pages, 6 figures, 11 tables. Preregistered study with a negative confirmatory result. Code and protocol records: this https URL

点击查看摘要

Abstract:An observable decision reveals the selected action but does not determine which alternatives were internally represented. Drawing on prospection research and hippocampal forward-sweep observations, we develop Before the Choice, an evidence-oriented framework distinguishing future-directed content, prospective computation, and causal guidance. The framework organizes four experimental contrasts and makes their alternative explanations explicit. Its first test case concerns unchosen trajectories in deterministic symbolic environments. A preregistered low-rank probe evaluated once on a sealed TEST did not meet its criterion (paired accuracy 0.020 versus 0.195 for a text-and-observed-choice baseline; 293 admissible pairs). Exploratory H1b exposed a reconstruction-by-elimination problem: identifying the chosen candidate can determine the unchosen target. In H1c, matched worlds varied an unchosen continuation while keeping candidate keys bitwise invariant. With specified positive controls passed, decoding in the late-key condition remained at 0.521 (p = 0.18; 141 contributing families). These observations do not identify prospective simulation or establish an absence of prospective representations. They illustrate why identifiability, sensitivity, and a dependence-aware null model must constrain inferences from decodability. PREM is proposed as an organizational framework and is not validated by these results. We offer it as a methodological contribution to empirical AI welfare research, which will require such evidential distinctions. Code and protocol records: this https URL

[AI-134] EverydayBench: Evaluating Satisfactory Completion of Live User-Facing Digital Work

链接: https://arxiv.org/abs/2611.25146
作者: Fangrui Huang,Bingyang Wang,Chuhan Li,Yijiang Li,Arpandeep Khatua,Austin Shen,Riki R Z Choi,Cuiyuanxiu Chen,Jin Huang,Zachary Robertson,Kate Lin,Akshay Paruchuri,Sijun Tan,Xin Liu,Zhengmin Ni,Sanmi Koyejo,Ehsan Adeli
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Everyday digital tasks, from selecting products and planning trips to conducting research and creating artifacts, require AI agents to deliver useful outcomes rather than merely retrieve web information. Yet evaluating these outcomes on the live web is challenging, as a single request may admit multiple valid results, and changing web information can alter which results remain valid over time. Although existing benchmarks address aspects of this setting, they do not fully integrate predefined success criteria, current evidence verification, and deliverable-quality assessment for agent-selected outcomes. To address this gap, we introduce EverydayBench, a benchmark of 77 human-authored tasks spanning eight domains of everyday life and work. Each task specifies success criteria before execution without prescribing an answer or workflow, allowing agents to select their own outcomes. To evaluate submissions under fixed success criteria, EverydayBench combines (1) deterministic checks of explicit constraints, (2) verification of claims against agent-cited webpages retrieved at evaluation time, and (3) rubric-based assessment of final deliverable quality. We further validate the protocol through human judgments, repeated scoring under fixed evidence, and controlled grounding failures. Across 23 reported agent-model configurations, Claude Opus 5.5 and Claude Fable 5.1 tie for the highest strict pass rate at 54.55%; a controlled GPT-6 Sol effort sweep ranges from 29.87% at low effort to 45.45% at max effort. Reliability varies substantially by task type: GPT-5.6 Sol leads lookup, Fable 5.1 leads planning, Opus 5.5 and Fable 5.1 tie on comparison, and GPT-6 Astra and Sonnet 5.5 tie on customized deliverables. Together, these findings highlight the difficulty of delivering complete, well-supported, and usable outcomes for everyday user requests.

[AI-135] CRAFT: Constrained Route planning with Adaptive Framework and Template-based code generation

链接: https://arxiv.org/abs/2611.25128
作者: Shahd Elmahallawy,Ahmed Eldawy
类目: Databases (cs.DB); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Everyday routing requests are rarely a point-to-point shortest path. A driver may need to visit stops, sometimes in a fixed order, under time budgets or distance limits, and perhaps recharge their EV car along the way. Each combination is a distinct routing problem that classically requires its own algorithm, so no fixed solver covers the space of queries users may ask. This paper presents CRAFT, which takes a constrained routing query as free text and returns a route that satisfies the user’s requirements on a real road network. An LLM parses the query and generates four search functions, which a generic graph-search engine executes on the road network. We formalize the underlying problem as State-Expanded RCSP (SE-RCSP), which extends resource-constrained shortest paths (RCSP) with non-additive state such as node-dependent resets and discrete progress tracking. We evaluate CRAFT on 2,600 natural-language queries over California road networks against two baselines, analyzing correctness, constraint violations, long-distance scalability, and semantic POI filtering.

[AI-136] DSFlow: A Neuro-Symbolic Data Workflow Agent with Statistical Diagnostic Guardrails

链接: https://arxiv.org/abs/2611.25111
作者: Junfeng Nie,Xiaohui Chen
类目: Artificial Intelligence (cs.AI)
备注: 18 pages, 8 figures. Presented as a poster at AIAS+ 2026, San Francisco, California, USA

点击查看摘要

Abstract:Data science workflows often require tools that presume non-trivial technical knowledge. Large language models (LLMs) reduce this barrier, but a gap remains between flexible general-purpose agents and statistically controlled pipelines. We introduce DSFlow, a LangGraph-based neuro-symbolic data-agent platform that combines reusable, LLM-selected skills with deterministic diagnostic guardrails. Skills provide extensible task knowledge. Symbolic validation can accept a proposed method, request clarification or replanning, override an unsuitable method, or supply a safe fallback. Regression and clustering case studies illustrate how DSFlow uses diagnostic evidence to correct method choices and documents those decisions in auditable reports.

[AI-137] SailSearch: Agent -Assisted Discovery in Combinatorial Optimization

链接: https://arxiv.org/abs/2611.25087
作者: Dhriti Verma,Kanthi Sarpatwar,Vinicius Lima,Dzung Phan,Jayant Kalagnanam
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:We present SailSearch, a human-guided workflow for algorithmic research with two language-model agents, and report results on Market Split and maximum independent set (MIS). For Market Split, our randomized algorithm searches across multiple lattice bases, dividing computation between enumeration and continued basis reduction. Across ten random seeds per instance, it found the first reported CPU-only solutions to 18 of 33 previously open QOBLIB instances, solving 130 to 133 of the 156 benchmark instances per seed with a 24-hour cutoff on two CPU cores. For MIS, our generator encodes planted XOR constraints as graphs with certified optima; on seven instances with 1,752 to 3,996 vertices, the best solutions returned by the tested solvers in eight-hour runs were 5 to 15 vertices below the optimum.

[AI-138] An LLM -Agent ic Framework for Intelligent Work Order Management in Industrial Maintenance

链接: https://arxiv.org/abs/2611.25070
作者: Nianjun Zhou,Dhaval Patel
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:This paper presents an AI-driven framework, CodeReAct agent, designed to enhance industrial equipment maintenance through natural language (NL) interactions. The framework employs an advanced AI agentic approach to interpret human instructions, providing knowledge extraction, asset status summarization, and work order recommendations. It identifies patterns and trends and predicts work orders by standardizing industrial data into business objects (BOs) and utilizing analytical function API calls. The AI agent operates through a Thought-Action-Observation process similar to human problem-solving, automating complex analytical tasks typically performed by data scientists. Notably, the Action phase generates executable code, replacing manual coding efforts. A case study focusing on chiller maintenance demonstrates the framework’s capability to analyze multiple years of work orders, and alerts and anomalies are presented. The results indicate that the AI agent framework effectively processes human utterances to analyze data, extract valuable insights, and support informed decision-making, reducing the need for data scientist involvement. By recontextualizing NLP agent frameworks in industrial engineering, the approach enables autonomous function generation and task execution across a range of domain-specific maintenance scenarios, with chillers used as a representative case study. Subjects: Artificial Intelligence (cs.AI) Cite as: arXiv:2611.25070 [cs.AI] (or arXiv:2611.25070v1 [cs.AI] for this version) https://doi.org/10.48550/arXiv.2611.25070 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[AI-139] DVBench: End-to-End Benchmark for Real-World UVM Design Verification Agents

链接: https://arxiv.org/abs/2611.25066
作者: Jiaru Qian,Jiawei Li,Xiaolong Hu,Zheng Shen,LiMing Zhao,Xin Zhang,Siyuan Jiang,Yongqiang Wei,Dong Liu,Jianyang Li,Yihua Chen,Yangxu Liao,Yuxin Wu,Jing Liang,Bin Wen,Junjie Zhang,Yongsheng Han,Yihong Dong,Ge Li
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Hardware Architecture (cs.AR)
备注:

点击查看摘要

Abstract:Design verification (DV) remains a major bottleneck in hardware development, with large-scale verification using the Universal Verification Methodology (UVM) being highly labor-intensive. Recent advances in large language models (LLMs) and agentic systems offer new opportunities for DV automation. However, existing benchmarks primarily construct mutually decoupled subtasks, which deviates from the real-world workflow. In this paper, we present DVBench, the first end-to-end benchmark for Real-World UVM design verification agents. DVBench comprises 95 tasks across 9 functional domains and 5 difficulty levels, covering diverse verification scenarios and design scales. Each task requires agents to independently complete the full verification workflow. DVBench evaluates final testbenches via compilation, coverage, and bug detection, while assessing intermediate artifacts including feature lists, test plans, and reference models. We evaluate frontier LLMs (e.g. Claude Opus 5.5, GPT-6 Astra) and agent frameworks (e.g. Claude Code, Codex, OpenCode). Experimental results show that the best LLM-based agent (Codex \ GPT-6 Astra) can only achieve an overall success rate of 21.05%, revealing substantial room for improvement in complex verification tasks.

[AI-140] Closing the Analyze-to-Design Gap: An Authorization-Bounded Agent for Self-Driving Laboratories

链接: https://arxiv.org/abs/2611.25043
作者: Yimeng Liu,Eric Isbrandt,Felix Katzenburg,Kevin P. Greenman,Han Hao,Yang Cao,Varinia Bernales,Alán Aspuru-Guzik
类目: Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Artificial intelligence (AI) agents are increasingly integrated into self-driving laboratories (SDLs) to plan experiments and interpret results. However, analytical reasoning alone does not make a laboratory autonomous: autonomy in an SDL is achievable only when the agent can be trusted to act on physical hardware unsupervised and knows when to defer to a human. Here, we present an architecture that integrates an agent at the Analyze-Design interface of the Design-Make-Test-Analyze (DMTA) cycle, enabling agent-driven measurement selection and data interpretation while ensuring that state-changing operations are authorized, validated, and auditable. We demonstrate this architecture on a liquid chromatography-mass spectrometry (LC-MS) instrument in an oxidant-screening campaign for synthesizing an organic dye. Within human-defined campaign limits, the agent submitted measurement jobs, interpreted mass spectrometry (MS) and diode-array detection (DAD) data, flagged failed experiments, and produced an oxidant ranking. This ranking matched that obtained from scripted peak integration, and organic chemists confirmed the agent’s spectral interpretations and failure assignments. In a simulated environment that replays the archived LC-MS data, where the agent chose which wells to inject under varying injection limits, it measured the highest-yielding condition at about its seventh injection, against 14.5 on average for a random order, and identified it in every campaign with a limit of eight or more injections. An agent can therefore close the analysis-to-design gap while its actions stay bounded and verifiable, which is a step toward greater autonomy for SDLs.

[AI-141] Near-Quadratic Sample Compression for All VC Classes

链接: https://arxiv.org/abs/2611.25011
作者: Yang Cai,Vineet Gupta,Yanchen Jiang,Christopher Liaw,Aranyak Mehta,Shay Moran,Grigoris Velegkas,Di Wang,Manfred K. Warmuth
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Data Structures and Algorithms (cs.DS); Combinatorics (math.CO); Machine Learning (stat.ML)
备注:

点击查看摘要

Abstract:The sample compression conjecture is a longstanding open problem in learning theory, originating in the work of Littlestone and Warmuth in the 1980s. At its heart is a connection between simplification and generalization: if a predictor consistent with a training sample is defined using only a small part of that sample, then it is likely to generalize. Sample compression schemes formalize this idea, and many natural learning algorithms can be analyzed this way. The sample compression conjecture of Floyd and Warmuth (1995) asserts that every binary concept class of VC dimension d admits a sample compression scheme of size O(d) ; a weaker conjecture asks only for size polynomial in d . Both have remained open: the first bound depending only on d , due to Moran and Yehudayoff (2016), is exponential in d . We prove the weaker conjecture by constructing a sample compression scheme of size O(d^2 \log^2 d) . Subjects: Machine Learning (cs.LG); Artificial Intelligence (cs.AI); Data Structures and Algorithms (cs.DS); Combinatorics (math.CO); Machine Learning (stat.ML) Cite as: arXiv:2611.25011 [cs.LG] (or arXiv:2611.25011v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2611.25011 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[AI-142] On the Impact of Feature Reduction Techniques on Adversarial Attacks

链接: https://arxiv.org/abs/2611.25004
作者: Mohammad Meymani,Roozbeh Razavi-Far
类目: Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Adversarial attacks have posed significant threats to machine learning models by exploiting their inherent vulnerabilities. Researchers have established an interdisciplinary field called adversarial machine learning to explore these vulnerabilities and develop corresponding defense strategies. Feature reduction strategies are used to decrease the dimensionality of the data in order to reduce training time, avoid the curse of dimensionality, prevent overfitting, and mitigate redundant features. However, their effect on adversarial robustness across different classifiers and attack scenarios remains insufficiently explored. This study aims to analyze the effect of feature reduction techniques on the adversarial robustness of a diverse set of machine learning models under poisoning, backdoor, and universal adversarial perturbation attacks on the APT dataset. These models include support vector machine, decision tree, random forest, multi-layer perceptron, and k-nearest neighbors. The attained results are compared with scenarios where different feature reduction techniques are applied, including principal component analysis, permutation importance ranking, recursive feature elimination, and univariate mutual information. Metrics such as accuracy, F1-score, and mitigation score are employed to evaluate the effect of these techniques on adversarial robustness. The results demonstrate that permutation importance ranking and univariate mutual information generally provide the most consistent robustness improvements, while other methods exhibit more variable effects depending on the classifier and attack scenario.

[AI-143] race2ST: Recovering PLC Structured Text from I/O Traces with LLM Generation and Verification

链接: https://arxiv.org/abs/2611.24979
作者: Hsi Chang,Irfan Khan
类目: oftware Engineering (cs.SE); Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Programming Languages (cs.PL); Systems and Control (eess.SY)
备注: 6 pages, 3 figures, 4 tables. Submitted to the 2027 IEEE Texas Power and Energy Conference (TPEC 2027)

点击查看摘要

Abstract:Programmable logic controllers (PLCs), the industrial computers that operate power and process plants, often run programs whose source code is no longer available; however, every input and output of the controller is logged continuously. This paper therefore asks whether a large language model (LLM) can recover an executable program in Structured Text (ST), the standard PLC language of IEC 61131-3, from such input/output records (traces) alone and, equally important, how the recovered program can be verified. To answer this question, we build a generate-replay-feedback pipeline. First, the LLM writes a candidate program from a compressed summary of the trace, without ever seeing the original program. The candidate is then executed one control cycle (scan) at a time and compared with the recording, and whenever a mismatch occurs the corresponding part of the trace is returned to the LLM, for at most three revisions. Because every result depends on the software that executes the candidates, we first validate it against two independent open-source PLC runtimes and find no unexplained difference. On this basis, we evaluate 47 third-party programs from two public benchmarks. The recovered programs reproduce the recorded trace exactly in 43 cases and agree with the original on 94% of scans for unseen inputs. The remaining errors, however, lie mostly in program branches that the trace never exercised; once a trace that targets these branches is added, agreement on unseen adversarial inputs rises from 0.80 to 0.98. By contrast, a classical automata-learning baseline without an LLM fits the recording but falls to 0.40 on unseen inputs, which indicates that recovering program structure is the hard part. Finally, on real data from a public water-treatment testbed, the recovered control laws match unseen measurements closely ( R^2 0.99 ) and expose the only actuator attack in the dataset.

[AI-144] Realized versus Population Reliability: Anytime-Valid Self-Consistency Certification under Correlated Reasoning

链接: https://arxiv.org/abs/2611.26592
作者: Terence D. Agbeyegbe,Toneukarin G. Agbeyegbe
类目: Methodology (stat.ME); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Self-consistency certification treats repeated sampling from a language model as sequential evidence accumulation and stops when an anytime-valid e-process crosses a threshold, guaranteeing finite-sample Type I error control by Ville’s inequality. Prior constructions assume sampled rollouts are conditionally independent. We show this assumption fails in practice: reasoning streams are positively correlated through shared reasoning prefixes or across-prompt heterogeneity, changing not merely certificate efficiency but the hypothesis certified. Modeling dependence via a per-unit latent state (the random-effects / intraclass-correlation structure) induces exchangeable, equi-correlated increments with correlation \rho . Correlation splits the single null into a realized null about the drawn belief and a population null about average model behavior, which coincide only when \rho=0 . We show: (i) the standard betting e-process remains anytime-valid for the realized null, not the population null; (ii) read as a population-null test, a suitable mixture has asymptotic crossing probability at least \pi=P(\mu(Z)0) , potentially exceeding nominal \epsilon ; (iii) the within-unit mean’s effective sample size saturates at 1/\rho , so extra within-unit rollouts cannot restore population validity; and (iv) a prompt-level e-process restores population validity, with optimal allocation at one rollout per prompt under a pure rollout budget. Simulations and Claude Haiku 4.5 rollouts confirm this regime dependence: on GSM8K ( \hat\rho=0.435 , \hat\theta=+0.96 ) the gap is dormant, whereas on AIME ( \hat\rho=0.480 , \hat\theta=-0.15 ) the single-unit rule crosses on 24% of prompts at nominal \epsilon=0.05 , while the prompt-level procedure declines (0.00). This correctness result identifies which reliability claim an anytime-valid certificate actually supports under dependence.

[AI-145] Flow Matching from Non-Gaussian Sources: A Theory of Efficient Sampling

链接: https://arxiv.org/abs/2611.26407
作者: Yuchen Zhou,Lili Zheng,Yuxin Chen,Gen Li
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Information Theory (cs.IT); Machine Learning (cs.LG)
备注: 36 pages

点击查看摘要

Abstract:Flow matching is a versatile framework for generative modeling that transports a source distribution to a target data distribution. A key advantage is its flexibility in the choice of source distribution, such as heavy-tailed and multimodal sources. Yet finite-step sampling guarantees for non-Gaussian sources remain limited. We establish a nonasymptotic convergence theory for Euler discretization of flow matching with linear interpolation. Suppose the source distribution on \mathbbR^d has an everywhere positive, continuously differentiable density, finite total variance \mathsfTotVar_\mathsfbase , and an L -Lipschitz score. Using the population velocity field and a suitable time grid, this sampler achieves total variation error at most \varepsilon in \widetildeO(Ld\mathsfTotVar_\mathsfbase/\varepsilon) iterations, provided the target distribution satisfies a mild second-moment condition. These assumptions cover, among many examples, source distributions such as multivariate Student’s t distributions with more than two degrees of freedom and finite Gaussian mixtures with shared covariance, while accommodating target distributions with neither a smooth density nor bounded support.

[AI-146] A Tensor-Algebraic No-Go Theorem for High-Dimensional Photonic Greenberger–Horne–Zeilinger States Discovered with AlphaProof Nexus

链接: https://arxiv.org/abs/2611.26300
作者: Mario Krenn,Moritz Firsching,George Tsoukalas,Rishikesh Gajjala,Xuemei Gu,Swarat Chaudhuri
类目: Quantum Physics (quant-ph); Artificial Intelligence (cs.AI); Combinatorics (math.CO)
备注: 13 pages, 1 figure, Lean examples. Comments welcome

点击查看摘要

Abstract:Photonic GHZ states generated with probabilistic pair sources can be described by edge-coloured, complex-weighted graphs. In this representation, conditioning on an event with one photon in every detector turns the output amplitudes into coherent sums over perfect matchings, and generating an (N)-photon (D)-dimensional GHZ state is equivalent to constructing a monochromatic quantum graph. Here, we prove that no complex-weighted monochromatic quantum graph exists for any even (N\ge4) when (D\ge N). The proof, discovered through AI-assisted formal proof search in Lean using AlphaProof Nexus, constructs a weighted sum (\mathcal F) of amplitudes whose coefficients factor vertex by vertex. Summing over the colours at one fixed vertex reduces the problem to an ((N-1)\times N) matrix of effective edge weights. For the graph amplitudes (\omega) and GHZ target (\Delta), the row–column imbalance yields (\mathcal F(\omega)=0) but (\mathcal F(\Delta)\neq0), contradicting the required equality (\omega=\Delta). Finally, we give a physical interpretation of the AI-discovered mathematical proof.

[AI-147] On the trajectories of the ID algorithm

链接: https://arxiv.org/abs/2611.25877
作者: Ivano Lodato,Simran Hiranandani,Aditya Varna Iyer
类目: Methodology (stat.ME); Artificial Intelligence (cs.AI); Probability (math.PR)
备注: 20 pages

点击查看摘要

Abstract:The ID algorithm decides whether an interventional distribution can be computed from observational data and, when it can, returns a formula in which products, ratios and summations of conditional probabilities are nested inside one another. We study these formulas by following the trajectories of the calls of the algorithm, each of which ends with a tail (an alternation of lines 7 and 2) and an exit, and we derive two main results. First, line~4 and the exit lines fix the single outer summation of the output, while the tail forms everything inside it, both the inner summations and the ratios. Second, a ratio keeps a summation in its denominator only when a repetition of the tail is entangled, and a test of ancestry in the original graph decides, before any formula is computed, whether a topological order can be chosen so as to avoid it. We also derive a bound on the nesting depth of the output, and show a simple ``dimensional’’ count that every output has to satisfy. Together, these results constrain the form of an ID output and guide its implementation and simplification.

[AI-148] Auto-Alpha: A Framework for Autonomous Factor Research

链接: https://arxiv.org/abs/2611.25732
作者: Hui Cai,Kangcheng Deng,Tianyi Ma,Qixuan Yang,Jiacheng Lu,Rui Sun,Beidi Luan,Jing Li,Zuo Bai
类目: Computational Finance (q-fin.CP); Artificial Intelligence (cs.AI); Software Engineering (cs.SE)
备注:

点击查看摘要

Abstract:Autonomous factor research requires generating economic hypotheses as well as implementing and testing them. We present Auto-Alpha, a framework that connects report-guided initialization to self-directed research through a test-driven implementation loop. An idea agent proposes economic mechanisms and signal constructions; a research agent develops them with backtest feedback. Semantic audit, quantitative review, and human admission select factors for a persistent library, while an Experience Library retains research insights. Starting from 120 research reports, Auto-Alpha builds an initial library of 32 factors. It then generates new ideas and refines existing candidates, producing 420 implementations and retaining 62 additional factors. We evaluate factor quality and portfolio performance using a fixed library snapshot of 105 factors. The median 1-day Rank IC is 0.0301, and a weekly Top-300 portfolio returns 16.81% annualized after stated costs over 2020-2026, versus 1.71% for the CSI 1000 price index. These retrospective results document an operating research process and its artifacts; controlled evaluation is needed to isolate the contribution of its components.

[AI-149] An Explainable and Scalable Hierarchical Tensor Framework for Integrating Multi-Omics and Digital Pathomics

链接: https://arxiv.org/abs/2611.25584
作者: Ahmad Hussein,Ali Anaissi,Mukesh Prasad,Ali Braytee
类目: Genomics (q-bio.GN); Artificial Intelligence (cs.AI)
备注:

点击查看摘要

Abstract:Background: Precision oncology requires integration of heterogeneous molecular and pathology data, yet many fusion models offer limited biologically grounded interpretability and struggle with missing modalities. Methods: We present Hierarchical Interpretable Tensor Fusion (HIT-Fuse), integrating six modalities: somatic copy number variation, DNA methylation, RNA sequencing, microRNA expression, RPPA protein abundance, and vision transformer-derived pathomics across four TCGA cohorts (glioma, lung adenocarcinoma, breast, and kidney cancer). Modality-specific autoencoders compress high-dimensional inputs, followed by two-stage CP tensor decomposition. Missing latent modalities are reconstructed using clinical covariates and observed omics to expand cohorts and retrain downstream models. A mathematical traceback mechanism enables global and local xAI attributions. Results: HIT-Fuse identifies distinct latent subgroups with significant survival stratification in glioma, breast, and kidney cohorts, both with complete and reconstructed modalities. Cohort expansion yields cohort-specific gains, including a 56.4% relative increase in weighted F1 for glioma. Biological validation against DepMap CRISPR screens, miRTarBase and RPPA annotations supports decoded drivers across all four cohorts. Essential gene rates range from 6.2% in glioma to 7.6% in lung cancer; example drivers include ANAPC11, SNRPD1, PLK1 and RPL14. Of decoded miRNAs, 71-92% have miRTarBase regulatory support, while 77-84% of selected genes are covered in DepMap. Conclusions: HIT-Fuse combines hierarchical tensor fusion, missing-modality reconstruction and traceable attributions to support interpretable multi-omics and digital pathology-based cancer prognosis across heterogeneous cohorts. Subjects: Genomics (q-bio.GN); Artificial Intelligence (cs.AI) Cite as: arXiv:2611.25584 [q-bio.GN] (or arXiv:2611.25584v1 [q-bio.GN] for this version) https://doi.org/10.48550/arXiv.2611.25584 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Ahmad Hussein [view email] [v1] Thu, 19 Nov 2026 06:27:17 UTC (6,460 KB)

[AI-150] Joint Language Modeling for In-the-Wild Speech-Scene Generation with Distinct Token Representations

链接: https://arxiv.org/abs/2611.25503
作者: Semin Kim,Ju Yeon Kang,Sunghwan Ahn,Seonuk Kim,Nam Soo Kim
类目: Audio and Speech Processing (eess.AS); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注:

点击查看摘要

Abstract:Speech-scene generation aims to synthesize intelligible target speech together with diverse environmental sounds as a coherent acoustic scene. The task is considered challenging in two respects. Speech and environmental audio differ substantially in their semantic and acoustic characteristics, yet existing approaches typically model the complete scene using a single mixed acoustic representation. At the same time, the two sounds are closely related in real-world speech scenes, whereas conventional approaches usually rely on simulated scenes mixed from independently recorded sources. We propose a large language model (LLM)-based framework for joint modeling of distinct speech and environment token streams. A source-aware neural audio codec extracts temporally aligned speech and environment tokens from mixed audio, and a pretrained LLM generates both token streams in an interleaved fashion conditioned on the natural-language scene description and speech content. This formulation preserves source-specific representations while modeling the token streams within a shared generative context. Since both streams are extracted directly from mixed recordings, the model can also be adapted to in-the-wild speech scenes without isolated sources. Our experimental results demonstrate that the proposed framework improves speech intelligibility, scene relevance, and audio quality over existing speech-scene generation techniques, with consistent gains from real-world adaptation. We further investigate the effects of distinct token representations and interleaved generation, and analyze whether the environmental sound remains stable while speech is active and how the two streams interact.

[AI-151] AIstatLab at BraTS-GoAT 2026 Task 3: Quality-Controlled Consensus Pseudo-Labeling and Dense Expert Adaptation MICCAI2026

链接: https://arxiv.org/abs/2611.25477
作者: Heeseung Shin,Changwon Lim
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI)
备注: 12 pages; BraTS 2026 Challenge, MICCAI 2026

点击查看摘要

Abstract:BraTS-GoAT Task 3 evaluates brain-tumor segmentation across heterogeneous entities using shared ET, TC, and WT regions. We present a 3D ResidualEncoderUNet trained with quality-controlled consensus pseudo-labels, a training-only nested-region objective, and a four-expert bottleneck adapter. Five-fold ResEnc-M/L teachers gener ated predictions for 1,138 unlabeled cases; fusion, hierarchy enforcement, and confidence-and-agreement screening retained 983 for training with 1,351 annotated cases. The final five-fold ResEnc-M entry achieved mean DSC 0.834, mean NSD 0.510, and Acalc = 0.6720 on the validation server. Component effects were modest and region-dependent, with the clearest gains on ET. Source code is available at this https URL.

[AI-152] From FAIR to AWARE: Data Readiness for Agent ic AI in the Earth Sciences

链接: https://arxiv.org/abs/2611.25209
作者: Ivan Kuznetsov,Dmitrii Pantiukhin,Sonja Hänzelmann,Fabian Wachsmann,Jonas Sølvsteen,Daniel Wiesmann,Thomas Jung,Nikolay Koldunov
类目: Atmospheric and Oceanic Physics (physics.ao-ph); Artificial Intelligence (cs.AI); Geophysics (physics.geo-ph)
备注: 34 pages (27 main text, 7 Supplementary Information), 1 figure, 1 table, 2 boxes

点击查看摘要

Abstract:Artificial intelligence in the Earth sciences is beginning to act: prototype agents already search archives, run simulations and draft analyses. Scientific data were curated for researchers who bring expert judgement; an autonomous agent cannot be assumed to supply it. Data can satisfy the FAIR principles and still mislead a machine acting alone. We propose that Earth-science data become AWARE (Actionable, Well-described, Auditable, Recipe-equipped, Epistemically labelled), an agent-specific requirements layer that extends FAIR, within a three-part agenda: AWARE data, bounded tools, layered validation. The ingredients exist across community infrastructure. We outline a staged, measurable path to adoption, intended to reduce silent error.

[AI-153] MADJEPA: Enhanced Multi-Stage Joint-Embedding Predictive Architecture for Medical Anomaly Detection WACV2027

链接: https://arxiv.org/abs/2611.25194
作者: Ioana-Sabina Stoian,Florinel-Alin Croitoru,Vlad Hondru,Andreea Iuliana Ionescu,Andrei Paun,Radu Tudor Ionescu
类目: Image and Video Processing (eess.IV); Artificial Intelligence (cs.AI); Machine Learning (cs.LG)
备注: Accepted at WACV 2027

点击查看摘要

Abstract:Anomaly detection in medical images is a complex task, requiring the development of a robust model using only healthy patient data. Without access to labeled data, modern approaches typically employ self-supervised learning to obtain a model of normality. Since world models promise to build strong internal representations of their environment, we consider them as suitable candidates for the medical anomaly detection environment, namely medical investigations of healthy patients. To this end, we propose MADJEPA, a medical anomaly detector based on an enhanced multi-stage joint-embedding predictive architecture (JEPA). We design a multi-stage training pipeline to produce a robust detector: (1) image-based JEPA (I-JEPA) training aims to learn a strong representation of normal training data; (2) target encoder distillation aims to modulate over-generalization of I-JEPA representations to abnormal data; (3) pseudo-anomaly classification aims to calibrate anomaly predictions by learning a tight boundary around the normal data manifold. To demonstrate our design, we carry out comprehensive experiments on six datasets covering various medical image domains (e.g.~MRI, OCT, X-ray, histopathology), comparing our MADJEPA with numerous state-of-the-art methods. The results show that our carefully designed world model, MADJEPA, attains competitive results, ranking first on three datasets. Our ablation studies further confirm the utility of each training stage, indicating that I-JEPA alone is not sufficient to obtain competitive results in medical anomaly detection. We publicly release our code at this https URL.

[AI-154] DL101 The Representation Capability of Neural Network Models

链接: https://arxiv.org/abs/2611.24980
作者: Fernando Berzal
类目: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
备注:

点击查看摘要

Abstract:Universal approximation theorems (Cybenko, Funahashi, Hornik Leshno) provide the mathematical justification for why neural networks are powerful enough to model complex, nonlinear phenomena. Shallow networks can approximate any continuous function on compact sets. Leshno’s theorem clarified that non-polynomial activations are necessary for universality, which explains why ReLU, logistic, hyperbolic tangent, and other common activations work (but pure polynomials do not). However, approximation theorems provide no efficiency, training, nor generalization guarantees. They do not indicate how many neurons are needed for a given accuracy, whether gradient descent will find the right parameters, or whether networks trained on finite data will generalize well. Barron provided quantitative error bounds for particular function classes (approximation rates for functions with bounded Fourier norm), whereas Montúfar was the first to show that expressivity grows exponentially with depth (in terms of the number of linear regions). Deep networks are strictly more expressive than shallow ones with the same number of parameters. Depth is fundamentally more powerful than width, providing a rigorous mathematical explanation for the use of deep learning models. Subjects: Machine Learning (stat.ML); Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE) Cite as: arXiv:2611.24980 [stat.ML] (or arXiv:2611.24980v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2611.24980 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

机器学习

[LG-0] deCache: Efficient Hybrid LLM Serving with Transparent Cache Management and Layer-Wise Streaming

链接: https://arxiv.org/abs/2611.26619
作者: Jinbin Luo,Xiaozhe Yao,Shang Yang,Zhijian Liu,Song Han,Qinghao Hu
类目: Distributed, Parallel, and Cluster Computing (cs.DC); Machine Learning (cs.LG)
*备注: 13 pages

点击查看摘要

Abstract:Hybrid architectures have become mainstream among open-weight language models. Their global-attention and bounded-state layers maintain heterogeneous caches whose demands vary with context length and concurrency. We present TideCache, a serving system on CPU–GPU superchips with three components for allocating GPU memory among caches and managing their residency across GPU and host memory. The Transparent Cache Pool redistributes physical pages among cache components behind fixed virtual addresses, preserving cache layouts and backend interfaces. The Cache Streamer uses computation intervals provided by bounded-state layers to prefetch the next global-attention layer’s prefix into GPU slots shared across global-attention layers, reducing GPU residency while overlapping layer-wise transfers with computation. Generated KV remains resident. The Request Scheduler selects resident or streamed execution and triggers capacity adjustments according to request size and queue state. Together, they let HBM released by streaming accommodate other cache components and additional requests. Implemented in SGLang and evaluated on GB300 and GH200 across seven model/platform configurations, TideCache completes up to 2.04x as many agent turns, reduces overall P50 per-token latency by 20–59%, and improves agent P50 TTFT by 1.40–3.05x compared with SGLang with HiCache.

[LG-1] Bio-PM-Guided Synthetic IMU Generation

链接: https://arxiv.org/abs/2611.26607
作者: Deepali Singh,Deepesh Parpani,Aviral Gupta
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Collecting and labeling inertial measurement unit (IMU) data for human activity recognition (HAR) is expensive, and models trained on one group of subjects often fail to generalize to unseen individuals. Even with balanced data, classifiers struggle to distinguish dynamically similar activity classes and benefit from targeted augmentation. We explore whether synthetic IMU data generated in the representation space of a pretrained biosignal encoder, Bio-PM \citeb1, can serve as a practical augmentation source. Rather than synthesizing raw signals, we leverage Bio-PM’s frozen encoder to convert 10-second IMU windows into token embedding sequences and train class-conditional decoders in this space: (1) an autoregressive transformer decoder that predicts tokens sequentially, and (2) a diffusion decoder that learns the token distribution through a denoising objective. We analyze generation realism through distributional metrics and embedding-space visualizations, evaluate downstream utility through leave-one-subject-out HAR classification, and present a PatchDecoder for approximate raw waveform reconstruction. Our diffusion decoder achieves a Train-on-Synthetic-Test-on-Real (TSTR) macro-F1 of 0.68, demonstrating that generated tokens are realistic enough to train a classifier that transfers to real data. Combining synthetic-token augmentation with a 320-dimensional gravity-aware feature representation raises LOSO macro-F1 from 0.729 to 0.809 over a real-only baseline across 51 subjects on the WISDM benchmark, with statistically significant gains concentrated in the most challenging activity classes. These results demonstrate that high-fidelity generation in a pretrained encoder’s representation space is achievable and practically useful for cross-subject HAR augmentation.

[LG-2] Grid Cells as Spectral Successor Features: A Reusable Basis for Continual Reinforcement Learning NEURIPS2026

链接: https://arxiv.org/abs/2611.26504
作者: Élodie Côté-Gauthier,Doina Precup,Isabeau Prémont-Schwarz
类目: Machine Learning (cs.LG)
*备注: 15 pages, 3 figures. Accepted at the NeurReps and WiML workshops at NeurIPS 2026

点击查看摘要

Abstract:Continual reinforcement learning requires agents to adapt when the task changes while reusing what they know about the environment. Successor features separate dynamics from task-specific reward, but their effectiveness depends on the choice of base features. Taking the identification of entorhinal grid cells with eigenvectors of the successor representation as a design principle, we build this basis from eigenvectors of the transition matrix of a fixed reference policy. With such a basis, the successor features at a state-action pair are a linear combination of the base features at that same pair, which is not true in general, and the coefficients are known in closed form: each successor feature is its base feature multiplied by \theta_i = 1/(1-\gamma\lambda_i) , where \lambda_i is its eigenvalue. The long-term return is thus predicted from the same features as the immediate reward, no successor-feature network or replay buffer is needed, and adapting to a new task reduces to a linear regression on the reward. The agent acts greedily on the action values of the reference policy, which is one step of policy improvement over it and suffices to reach a single goal. In a two-goal gridworld, with the uniform random walk as reference policy, whose eigenvectors are the low-frequency Laplacian eigenvectors, agents built this way maintain near-maximal return across every change of goal, whereas SimpleSF and DQN lose most of their return at each change.

[LG-3] Anticipating Future Build Results with BuildPredictor

链接: https://arxiv.org/abs/2611.26476
作者: Christian Macho,Katharina Stengg
类目: oftware Engineering (cs.SE); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Continuous Integration (CI) is a popular programming practice for the seamless integration of code changes into the code base. Using CI, developers benefit from automated assembling and testing of the code after each commit. However, while the frequent integration and building supports developers, it also comes at the high costs of wasting time and resources for minimal changes. In this work, we investigate the potential of reducing the drawbacks by anticipating the build outcome using machine learning techniques. In particular, we propose a random forest model with various features such as changes to the source and build code, and changes to the complexity of the code. We observe that our model shows a good performance in an intra-project and holistic setting yielding a median MCC of 0.82 and 0.86, respectively. We also found that the most important features of our model concern the historical build result information. Additionally, we investigated the development of the performance of our model over time and found that our model can quickly arrive at its peak performance. In an additional experiment, we removed the hard-to-interpret historical build information features and found that our model still yields decent performance while improving the interpretability of it. Summarizing, our model can help to avoid unnecessary build runs by anticipating the build result and thus, save resources. Subjects: Software Engineering (cs.SE); Machine Learning (cs.LG) Cite as: arXiv:2611.26476 [cs.SE] (or arXiv:2611.26476v1 [cs.SE] for this version) https://doi.org/10.48550/arXiv.2611.26476 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Submission history From: Christian Macho [view email] [v1] Thu, 19 Nov 2026 17:35:44 UTC (2,266 KB)

[LG-4] Neural Conditional Gaussian Model: Efficient Data Assimilation with Sequential Uncertainty Quantification for Nonlinear Dynamics

链接: https://arxiv.org/abs/2611.26463
作者: Chuanqi Chen,Nan Chen
类目: Machine Learning (cs.LG); Dynamical Systems (math.DS)
*备注:

点击查看摘要

Abstract:Data assimilation (DA) in partially observed nonlinear systems requires repeated estimation of both hidden states and their uncertainty. Conventional nonlinear filtering can be computationally expensive and often relies on an explicit dynamical model. In this paper, we develop the Neural Conditional Gaussian Model (Neural CGM), a data-driven dynamical model that maps unobserved physical states into a learned latent space, where the dynamics of the latent states follow a conditional Gaussian structure. This structure allows closed-form DA updates in the latent space, while nonlinear encoder-decoder mappings retain the flexibility needed to represent complex dynamics in physical space. A key feature of Neural CGM is its treatment of uncertainty in DA. The latent posterior mean and covariance are mapped to the physical variables through a Gaussian decoder that combines local moment propagation with learned nonlinear corrections. The resulting physical-space posterior statistics depend on the sequentially updated latent distribution rather than on a separate post-processing estimate. The same statistics enter a Gaussian negative log-likelihood during training, so both the posterior mean and uncertainty affect the learned model. Because Neural CGM learns the dynamical model itself, it can also be used for state forecasting and simulation without explicit knowledge of the governing equations. Numerical experiments on benchmark nonlinear dynamical systems show accurate state estimation and forecasting while substantially reducing online assimilation cost relative to conventional nonlinear DA methods. Neural CGM is therefore suited to repeated state estimation, uncertainty quantification, and prediction from partial observations.

[LG-5] Should Belgians Install an Advanced Energy Management System in Their Electrified Dwelling ?

链接: https://arxiv.org/abs/2611.26441
作者: Julian Ruddick,Cedric De Cauwer,Thierry Coosemans,Maarten Messagie
类目: ystems and Control (eess.SY); Machine Learning (cs.LG)
*备注: Full version including appendices A-C, which are not part of the 11-page proceedings version. Published at ACM BuildSys '26 under CC BY 4.0. 17 pages, 10 figures, 13 tables. Data: this https URL

点击查看摘要

Abstract:As dwellings electrify and more complex electricity pricing spreads in Belgium, the question arises: What are the potential benefits of installing an advanced energy management systems in an electrified Belgian dwelling ? To answer this question, we simulate 500 representative Belgian dwellings in which fossil systems are replaced by controllable electric technologies: space heating via air-to-air heat pumps, domestic hot water via electric boilers, and private transport via electric vehicles with home chargers. Photovoltaic systems are assigned according to current adoption rates. We evaluate two advanced energy management systems (EMS): model predictive control (MPC) and TreeC, a machine learning method using decision trees to model the EMS strategy. These are compared to a rule-based controller (RBC) that mimics default operation and an MPC with perfect forecasts (MPC P) as an upper bound. Performance for these EMSs is assessed for 2023 under a real contract with hourly prices and a monthly peak tariff. Average annual savings versus RBC are 82.5 EUR, 116.6 EUR, and 177.4 EUR for TreeC, MPC, and MPC P, respectively. All EMSs deliver similar thermal comfort (26-31 Kh). MPC saves more, while TreeC achieves higher EV charge completion (93.3% vs 88.8% for MPC, with a max charge of 95%) and a smoother aggregate load across the 500 dwellings. Savings remain modest relative to EMS purchase (280-680 EUR) and subscription (0-360 EUR/year) costs which is not favourable towards widespread adoption. The data and code to reproduce this study are available at this https URL. Comments: Full version including appendices A-C, which are not part of the 11-page proceedings version. Published at ACM BuildSys '26 under CC BY 4.0. 17 pages, 10 figures, 13 tables. Data: this https URL Subjects: Systems and Control (eess.SY); Machine Learning (cs.LG) Cite as: arXiv:2611.26441 [eess.SY] (or arXiv:2611.26441v1 [eess.SY] for this version) https://doi.org/10.48550/arXiv.2611.26441 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Journalreference: Proceedings of the 13th ACM International Conference on Systems for Energy-Efficient Buildings, Cities, and Transportation (BuildSys '26), pp. 405-416, 2026 Related DOI: https://doi.org/10.1145/3744256.3812575 Focus to learn more DOI(s) linking to related resources Submission history From: Julian Ruddick [view email] [v1] Thu, 19 Nov 2026 17:14:13 UTC (1,916 KB) Full-text links: Access Paper: View a PDF of the paper titled Should Belgians Install an Advanced Energy Management System in Their Electrified Dwelling ?, by Julian Ruddick and 3 other authorsView PDFHTML (experimental)TeX Source view license Current browse context: eess.SY prev | next new | recent | 2026-11 Change to browse by: cs cs.LG cs.SY eess References Citations SciXGoogle 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)

[LG-6] Learning to Defer under Expert Drift

链接: https://arxiv.org/abs/2611.26440
作者: Ziyao Mou,Andrea Wynn,Eric Nalisnick
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:The learning to defer (L2D) framework enables safety in machine learning and AI systems by allocating difficult or critical decisions to a human expert. However, most prior work on L2D assumes fixed expert reliability, overlooking temporal fluctuations in human performance due to factors such as fatigue and cognitive biases. Humans commonly exhibit such substantial, systematic performance shifts across a wide range of domains, from high-stakes settings such as radiology, aviation, and judicial decision-making, to everyday tasks such as crowdsourced image annotation. We propose Expert-Drift-Adapted Learning-to-Defer (EDA-L2D), a novel expert-aware L2D framework that explicitly models these temporal shifts by conditioning the deferral head on recent histories of expert outcomes via a sequence model. This history-aware approach enables adaptive deferral policies that anticipate reliability fluctuations and better allocate decisions between human and machine over time. We provide comprehensive experiments across three diverse tasks – image classification, medical diagnosis, and hate speech detection – including generalization analyses under realistic deferral conditions and irregular shifts in expert reliability. Our results show that EDA-L2D adapts more effectively than prior L2D methods to diverse temporal patterns of expert reliability, resulting in improved system performance and more effective deferral decisions. Code is available at this https URL.

[LG-7] FEA solver integration of a composite World Model with a Spiking Neural Network memory Block for computationally efficient modelling of defects and variabilities

链接: https://arxiv.org/abs/2611.26419
作者: Bassam El Said,Adam Sobey
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Each fibre-reinforced composite structure has a unique material architecture. The manufacturing methods used and the presence of defects lead to local variations which require material-scale models that can resolve the response in real time at every location in the structure. Concurrent multi-scale schemes provide this capability, but at a prohibitive computational cost for large structures as the number of integration points and time increments scales with size. Surrogates have been studied extensively as replacements for material-scale models to mitigate these computational limitations. A surrogate must represent an arbitrary local architecture including defects, track the irreversible material state over an unknown number of increments, remain stable when driven recursively by its own predictions, and fit within the memory available at every integration point. Here, a Composites World Model based on a Multi-Modal Spiking Transformer architecture is designed to meet these requirements. The proposed architecture combines a material encoder, a Spiking Neural Network memory block that tracks the long-term internal state of the material, and transformer blocks that learn the short-term applied strain state. The system is trained in three stages: material architecture encoding, base training on short time windows, and recursive fine-tuning with scheduled sampling to suppress cumulative error drift. The trained model is disassembled at inference and coupled to a non-linear FE solver, enabling structural scale simulation of composites containing embedded defects. Benchmarks against experimental results for IM7/8552 laminates under tension, compression and transverse shear, for pristine and wrinkled composites, show that the SNN memory block successfully tracks the material internal state and results in prediction errors between 0.4% and 4.8%.

[LG-8] Sequential Anytime Valid Tests for Efficient Inference-Time Alignment

链接: https://arxiv.org/abs/2611.26413
作者: Vishesh Kumar Singh,Sayak Ray Chowdhury
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Inference-time alignment methods, such as Best-of-N (BoN), improve language model outputs by generating multiple candidates per prompt and selecting among them using a reward model. Much of this cost is wasted: a fixed budget is applied to every prompt, although the computation required by prompts varies widely. Adaptive methods halt once sampled answers appear settled, but their thresholds are tuned heuristics with no statistical control over early-stopping errors. We instead seek an anytime-valid stopping rule that takes an error tolerance \alpha as input and halts only once an answer is certified to hold a majority ( \theta_a 1/2 ) under the generator’s distribution. We aggregate candidates by final answer with a reward-weighted self-consistency (RWSC) score and treat stopping as a sequential test over the resulting clusters. Applying Wald’s sequential probability ratio test to these clusters yields SPRT-RWSC, which performs well empirically, but its guarantee fails because the leading answer shifts during sampling, and the run is continuously monitored. Our main contribution, e-RWSC, replaces Wald’s test with a betting supermartingale (an e-process) for every answer the policy produces, bounding the probability of ever certifying a non-dominant answer by \alpha , uniformly over time and across all answers. The RWSC scores drive the bets, so they affect only how quickly a dominant answer a^\star is certified, not validity; certification takes O(\log(1/\alpha)/d_\mathrmKL(\theta_a^\star|1/2)) samples. Experiments across several language models, datasets, and reward models confirm the predicted 1/d_\mathrmKL scaling of cost and show that accuracy improves over baselines at substantially lower token cost.

[LG-9] How to Effectively Manage the Energy in Residential Systems: Development and Comparison of Methods

链接: https://arxiv.org/abs/2611.26412
作者: Julian Ruddick
类目: ystems and Control (eess.SY); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE)
*备注: PhD thesis, Vrije Universiteit Brussel, defended 25 February 2026. 178 pages. Promotors: Maarten Messagie and Thierry Coosemans. Chapters 4-5 are based on Knowledge-Based Systems 309 (2025) 112756, doi: https://doi.org/10.1016/j.knosys.2024.112756%3B chapter 6 on Energy and AI 18 (2024) 100448, doi: https://doi.org/10.1016/j.egyai.2024.100448%3B chapter 7 on BuildSys '26, pp. 405-416, doi: https://doi.org/10.1145/3744256.3812575

点击查看摘要

Abstract:The transition from fossil fuels presents significant challenges in energy management. To ensure a sustainable energy future, it is crucial to reduce energy consumption when possible and to synchronize electricity consumption and storage with the production from sources with low greenhouse gas emissions. This thesis explores a variety of energy management methods used to manage batteries, heat pumps, and other energy-related systems. The thesis presents three simulation cases alongside one real-world experiment to compare energy management done through simple rules, optimization, and machine learning methods. Furthermore, this thesis introduces TreeC, a novel machine learning-based method that learns understandable yet efficient energy management strategies, applicable to many cases. The three key contributions of this thesis can be summarized as follows: 1) The developed TreeC method demonstrates competitive performance compared to a variety of energy management approaches, such as model predictive control and reinforcement learning, across a diverse set of energy management problems. The understandability of TreeC’s energy management strategies greatly improves the transparency and trustworthiness of machine learning-based energy management. 2) The real-world experiment highlights the need to validate energy management methods on actual hardware and not solely on simulations. Both TreeC and model predictive control performed worse than in simulations, due to, respectively, insufficiently representative training data and the real-time input data not being reliable enough. 3) Energy management algorithms reduce costs in electrified Belgian dwellings, particularly in poorly insulated dwellings, electric vehicle chargers, and photovoltaic installations. However, on average, these cost savings do not yet outweigh the current costs and effort required to implement such algorithms. Comments: PhD thesis, Vrije Universiteit Brussel, defended 25 February 2026. 178 pages. Promotors: Maarten Messagie and Thierry Coosemans. Chapters 4-5 are based on Knowledge-Based Systems 309 (2025) 112756, doi:https://doi.org/10.1016/j.knosys.2024.112756%3B chapter 6 on Energy and AI 18 (2024) 100448, doi:https://doi.org/10.1016/j.egyai.2024.100448%3B chapter 7 on BuildSys '26, pp. 405-416, doi:https://doi.org/10.1145/3744256.3812575 Subjects: Systems and Control (eess.SY); Machine Learning (cs.LG); Neural and Evolutionary Computing (cs.NE) Cite as: arXiv:2611.26412 [eess.SY] (or arXiv:2611.26412v1 [eess.SY] for this version) https://doi.org/10.48550/arXiv.2611.26412 Focus to learn more arXiv-issued DOI via DataCite (pending registration) Related DOI: https://doi.org/10.13140/RG.2.2.13155.41763 Focus to learn more DOI(s) linking to related resources

[LG-10] A Complete Geometric Message-Passing Architecture for Point Clouds

链接: https://arxiv.org/abs/2611.26384
作者: Nadav Dym,Tal Amir,Gia Chi Dang,Christopher Morris
类目: Machine Learning (cs.LG)
*备注: 44 pages, 4 figures, 4 tables. Ancillary files: exact-arithmetic verifier, certificate and companion notebook for the Platonic color-atom computation

点击查看摘要

Abstract:Three-dimensional point clouds are a fundamental data modality, and a large family of machine learning architectures has been developed to respect their natural symmetries to rigid motions and permutations. A principled way to compare these architectures, which vary widely in complexity and cost, is through their expressive power. That is, an architecture is complete if it can distinguish any two inputs not related by the symmetries it respects. Simple distance-based message-passing networks are known to be incomplete, and while there exist complete alternatives, they typically rely on computationally intensive mechanisms like high-order representations of the rotation group or higher-order Weisfeiler-Leman-type tests. In this work, we devise a novel, simple, equivariant message-passing architecture for point clouds, and prove that it is complete. We complement our theoretical result with experiments showing that our architecture separates challenging point clouds that distance-based message passing provably cannot distinguish. Our results suggest that high-order equivariant features may not be necessary to enhance the expressivity of point cloud architectures, and can help settle completeness for other equivariant architectures for which such results are currently unknown. We give initial evidence for this by proving completeness of a slightly modified version of SEGNN. Comments: 44 pages, 4 figures, 4 tables. Ancillary files: exact-arithmetic verifier, certificate and companion notebook for the Platonic color-atom computation Subjects: Machine Learning (cs.LG) Cite as: arXiv:2611.26384 [cs.LG] (or arXiv:2611.26384v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2611.26384 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[LG-11] Evidence from a Multi-Regime Causal Discovery Benchmark: UCMT Algorithms Comparative Performance and Controlled Ablation Studies

链接: https://arxiv.org/abs/2611.26383
作者: Mahyar Eslami
类目: Machine Learning (cs.LG)
*备注: 8 pages, 4 figures, 6 tables. Internal benchmark and controlled ablation study; not peer-reviewed

点击查看摘要

Abstract:We present an internal, pre-registered evaluation of the UCMT causal discovery algorithm family alongside established alternatives across heterogeneous benchmark regimes, complemented by a separately pre-registered, controlled study of UCMT v2 Global components. The contribution is both empirical and methodological: a family-stratified comparison of three actively developed algorithm configurations and targeted evidence about which internal stages affect their outcomes. In the main benchmark, UCMT v2 Global has higher directed F1 than NOTEARS and DAGMA in selected real, semi-synthetic and Gaussian-network comparisons at the paired-run level, while trailing leading permutation-search methods in other settings. The BOSS-based UCMT v4 produces a CSuite-specific directed-F1 improvement over BOSS (mean paired difference +0.279; pair-level Holm p 0.001), but the dataset-level adjusted result is not significant (p = 0.188). In a separate fixed-skeleton ablation, UCMT v2 Global exceeds an adapted R^2-SortnRegress orientation baseline on standardized iSCM data (directed F1 0.537 vs 0.468, dataset-level Holm p = 0.031); absolute orientation accuracy is 0.563 and there is no established broad advantage over RESIT, ANM or DirectLiNGAM. Controlled ablations show that sparsity and BIC-based refinement materially affect some outcomes, whereas the examined arbitration and automatic regime-selection mechanisms do not yield confirmed general improvements. We report performance shortcomings, multiple-testing correction, runtime costs, evaluation limitations and reproducibility checks. The results support UCMT as an implemented and empirically investigated algorithm family, not a claim of overall superiority, independent validation or demonstrated commercial impact.

[LG-12] From Benchmarks to the Bench: Preference Correspondence and Selectivity in Bidirectional Drug-Target Screening

链接: https://arxiv.org/abs/2611.26369
作者: Jiangjie Qiu,Chengrui Xu,Xiaonan Wang
类目: Machine Learning (cs.LG)
*备注: 18 pages, 5 figures and 1 table

点击查看摘要

Abstract:Identifying active molecules for known targets and potential targets for existing drugs is central to drug discovery and repurposing. Drug-target interaction (DTI) models can rank vast numbers of drug-target pairs, but limited experimental capacity restricts testing to a small number of candidates. We compared the bidirectional recommendations of 13 released systems across 711 molecular structures of FDA-approved drugs and 614 human proteins. For the same target or drug, pairs of models produced completely disjoint Top-10 molecule and protein lists in 77.8% and 70.1% of comparisons, respectively. Even where recommendations overlapped, agreement between some systems largely reflected recurring preferences for the same candidates. Published inhibition measurements for 79 molecules and 65 kinases revealed that some systems enriched active candidates, yet their hit yields changed little when recommendations were matched to different drugs or targets, whereas other systems lost substantially more hits. Some of the former systems also rarely ranked both molecules and kinases correctly when two drugs favoured different kinases. We further retrained the same architecture on the same data, with the training objective fixed and random seeds varied. Despite similar test performance, mean pairwise Top-10 overlap on a separate screening catalogue was about two candidates. Our study moves DTI model evaluation beyond benchmark performance to assess whether model predictions can guide meaningful experimental decisions in drug discovery and target identification.

[LG-13] Learning from Context: Leave-One-Out Prediction in Jump Diffusions

链接: https://arxiv.org/abs/2611.26359
作者: Christian Wald
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Continuous diffusion models generate discrete data through stochastic dynamics in an embedding space. We consider jump diffusions that additionally reset coordinates to embeddings of categorical values. A coordinate’s own corrupted state can make prediction easy, potentially reducing the pressure to learn dependencies between coordinates. We study leave-one-out (LOO) prediction, in which each coordinate is predicted from the corrupted states of the other coordinates. We show that, under stated assumptions, jump diffusions with complete coordinate resets and state-independent clocks preserve the prescribed marginals when reset destinations use exact LOO probabilities and the drift uses the full endpoint posterior. We further show that this drift can be constructed from LOO probabilities by analytic likelihood completion without additional network evaluations. We evaluate unrestricted, LOO, and completed predictions on Text8 and LM1B text generation and ZINC250k molecular graph generation. LOO-based variants improve key generation metrics over matched unrestricted controls on all three datasets. Controlled interventions assess reliance on local inputs and context, and we analyze the effect of analytic likelihood completion.

[LG-14] Mamba Flow Matching Neural Processes: Linear-Time Inference for Irregularly Observed Spatial Fields NEURIPS2026

链接: https://arxiv.org/abs/2611.26355
作者: Cosmo Santoni,Giovanni Charles,Timothy James Hitge,Oliver Watson,Elizaveta Semenova
类目: Machine Learning (cs.LG)
*备注: Accepted at NeurIPS 2026. XX pages, X figures. Code: this https URL

点击查看摘要

Abstract:Interpolating a spatial field, such as daily maximum temperature, from sparse, irregularly located observations recurs across the environmental and health sciences. Standard Gaussian-process kriging refits hyperparameters and solves an O(N^3) system for every new field. When many related fields are observed, such as daily maps from one sensor network, neural processes (NPs) instead learn the interpolation map from past fields and predict a new field in one forward pass. Joint-predictive variants, which sample whole fields rather than independent per-location values, typically use self-attention over context and query points, costing O(K^2) in the combined point count K and limiting practical use to a few thousand points. State-space models offer a linear-time alternative, but a direct state-space adaptation compresses the context into one hidden state before reading queries, reducing what the decoder sees. We propose MambaFlowNP, a linear-compute joint-predictive NP trained with conditional flow matching. A multi-directional Mamba-2 scan reads context and query tokens in one shared sequence, so each query’s velocity is read from its own scan position rather than a compressed context summary, and because queries only read the scan state, predictions are consistent across query sets. MambaFlowNP is competitive with the strongest attention-based joint-predictive baselines on synthetic Matérn Gaussian-process benchmarks, and on regional to continental-scale NOAA temperature interpolation it attains the lowest RMSE and CRPS and near-nominal coverage on every tier, while running 114–784 \times faster than the strongest joint-predictive baseline. A single forward pass scales to 10^6 points on one NVIDIA L40S GPU in under a second. This makes joint prediction feasible for continental-scale, irregularly sampled sensor networks; predictive quality is evaluated up to K=5,580 .

[LG-15] RCF: Temporal-Reference Complementary Forecasting for Irregular Multivariate Time Series

链接: https://arxiv.org/abs/2611.26351
作者: Xudong Zuo,Shusen Ma,Qirui Chen,Bing Fan,Yun-Bo Zhao
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Irregular multivariate time series forecasting supports healthcare monitoring, environmental analysis, and activity recognition. Channel-independent (CI) approaches capture each variable’s temporal patterns but do not explicitly use observations from other variables. Channel-dependent (CD) modeling exploits relationships among variables, yet asynchronous measurements make these relationships difficult to learn. Incorporating cross-variable information while preserving each variable’s irregular history therefore remains a key challenge. We propose Temporal-Reference Complementary Forecasting (TRCF), a framework that augments a compact CI forecaster with a complementary CD branch. ANCHOR aggregates observed values and timestamps at shared temporal locations, retaining observation support to form compact CI features. BRIDGE constructs complementary features through history completion. It first estimates each variable’s historical values from its own observations, then uses differences between other variables’ observed and estimated values at nearby times to correct missing entries. The completed history preserves observed values and is encoded to supply context alongside the direct CI representation. Finally, the two representations are fused, combined with the query-time embedding, and passed to a shallow multilayer perceptron for forecasting. Experiments on four real-world datasets show that TRCF achieves the lowest mean squared error and mean absolute error in the main benchmark comparison, while reducing measured peak inference memory by 17.28% to 87.98% relative to APN.

[LG-16] OceanCoder: A Verification-Guided Large Language Model for Automated Simulation Modelling in Ocean Engineering

链接: https://arxiv.org/abs/2611.26333
作者: Can Xue,Changhong He,Ziyi Sun,Anliang Ge,Shimin Yu,Bingchen Liang
类目: Machine Learning (cs.LG); Software Engineering (cs.SE)
*备注: 22 pages, 11 figures, including appendices

点击查看摘要

Abstract:Numerical simulation is important in ocean engineering, but model construction requires expert knowledge. Large language models support automated modelling, but code syntax and successful execution alone do not establish whether generated models and results meet engineering requirements. We present OceanCoder, a Qwen3-8B-based large language model for automated simulation modelling in ocean engineering. To address specialised programming requirements and scarce training data, we develop an extensible data construction pipeline producing 12,000 verified supervised fine-tuning (SFT) examples across 12 engineering task categories. A three-layer verification framework checks model inputs against engineering parameters and task constraints, verifies completion of required simulation analyses, and assesses the engineering plausibility of computed outputs. Its outcomes provide graded rewards for group relative policy optimisation (GRPO), guiding the model towards engineering task completion. We also construct the first benchmark for automated modelling in ocean engineering, covering out-of-distribution tasks and real-world engineering cases. Experiments and ablation studies demonstrate the effectiveness of the data pipeline, verification framework and reward design. OceanCoder outperforms both base models and the closed-source comparison, with Pass@1 increasing from 0% for the base model to 46.77% after SFT and 51.43% after GRPO. These results support engineering verification as a training signal for automated simulation modelling.

[LG-17] Predicting Surface Displacement Fields of Laser Powder Bed Fusion Parts Using STL-Derived Graphs and Part-Level Descriptors

链接: https://arxiv.org/abs/2611.26323
作者: Saima Ahmed Suhi,Lang Yuan,Md Habibor Rahman
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Predicting part distortion before fabrication remains challenging in laser powder bed fusion (LPBF), while thermomechanical simulation is computationally expensive for repeated design evaluations. This study presents a two-branch graph attention network (GAT) to predict three-dimensional surface displacement directly from STL-derived graphs. The first branch learns geometric relationships among connected vertices of a densified STL surface mesh, while the second incorporates 12 part-level descriptors of mass distribution, build-plate contact, overhang, size, and shape. Displacement labels from ANSYS Additive Print simulations are mapped onto STL vertices using Gaussian-kernel weighted averaging. Evaluated on 117 parts using five-fold cross-validation split by part, the model achieves an R^2 of 0.946 and a mean absolute error of 0.0178 mm for displacement magnitude. Relative to a node-wise multilayer perceptron, GAT improves R^2 from 0.913 to 0.946; for build-direction displacement, R^2 increases from 0.903 to 0.939. GAT also outperforms GraphSAGE (R^2 = 0.936) and a graph convolutional network (R^2 = 0.934) across all five folds. Ablation experiments show that removing mass-distribution descriptors reduces magnitude R^2 to 0.900, while removing build-plate contact and overhang descriptors reduces it to 0.920. These results demonstrate the value of combining local mesh connectivity with global geometric descriptors to predict LPBF distortion, enabling faster design iteration and earlier identification of distortion-prone geometries.

[LG-18] Scalable fault vulnerability analysis for ReRAM-based process-in-memory CNN accelerators

链接: https://arxiv.org/abs/2611.26307
作者: Aniseh Dorostkar,Hamed Farbeh,Hamid R. Zarandi,Meysam Sharifzadeh
类目: Hardware Architecture (cs.AR); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Resistive random-access memory (ReRAM)-based processing-in-memory (PIM) accelerators offer a highly efficient platform for parallel matrix-vector multiplications (VMM) in neural networks, leveraging analog computation, high density, low leakage, and non-volatility. However, their susceptibility to permanent faults, such as stuck-at-high (\textitSaH) and stuck-at-low (\textitSaL) resistance states, permanently corrupts the convolutional neural networks (CNNs) weights mapped onto ReRAM cell conductance values, thereby degrading inference accuracy and leading to system unreliability in safety-critical applications. Fault injection is a commonly used technique to assess the vulnerability of neural networks by systematically injecting faults into individual parameters and measuring the resulting accuracy degradation. Although this method is accurate, it is prohibitively time-consuming and computationally expensive, especially for large-scale CNNs. In this paper, we propose a fast, scalable, and analytical framework to quantify the fault vulnerability of convolutional neural networks deployed on ReRAM-based PIM accelerators. This method quantifies vulnerability at different granularities, including weights, kernels, and layers, without the need for exhaustive simulation, enabling the design of fault-tolerant accelerators for edge and safety-critical applications. The comparison between the results of the proposed vulnerability score (VS) metrics with performance metric of CNNs captured with fault injection simulations confirms the effectiveness of our proposed method. Unlike exhaustive fault injection, which requires days or weeks, whereas the proposed approach achieves a speedup of about 10^4\times-10^5\times , enabling fast and accurate vulnerability analysis for large-scale CNNs within only a few minutes or hours.

[LG-19] me Series Structure Learning with Local Causal Parent Discovery

链接: https://arxiv.org/abs/2611.26232
作者: Jacqueline Maasch,Shu Wan,Hank Zhu,Huan Liu,Zhiguang Wang
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注:

点击查看摘要

Abstract:Learning and inference on time series can pose significant statistical challenges, including autocorrelated observations, distribution shift over time, and the curse of dimensionality. These challenges complicate the practical application of causal discovery algorithms, which already suffer from high computational and sample complexity in the general case. To mitigate these issues, this work introduces Iterative Temporal Parent Discovery (ITPD): a sound and complete constraint-based procedure for causal structure learning on multivariate time series with causal nonstationarity. ITPD takes a local-to-global learning approach that discovers the global graph by sequentially discovering the parent set for each variable. ITPD exploits temporal structural priors, graph partitioning, and the local Markov condition to swap higher-order conditional independence tests for lower-order alternatives. With an exact oracle, ITPD performed the least number of tests and used the smallest average conditioning set sizes relative to baselines. In benchmarking simulations, ITPD sat on the Pareto frontier with respect to F1 and wall-clock time and was faster than baselines on large graphs ( \geq 1000 nodes). For graphs with 2000 nodes, ITPD ran 2.3\times , 9.1\times , and 40.9\times faster than iterative IAMB, DF-PC, and PCMCI, respectively.

[LG-20] Attention Head Specialization: Phase Transitions in Sample Complexity

链接: https://arxiv.org/abs/2611.26228
作者: Léo Catteau,O. Duranthon,Lenka Zdeborová
类目: Machine Learning (cs.LG); Disordered Systems and Neural Networks (cond-mat.dis-nn); Machine Learning (stat.ML)
*备注:

点击查看摘要

Abstract:Attention heads specialize into distinct functions during training, often emerging in stages together with sharp changes in model behavior. While this phenomenon has mostly been studied as a function of training time, we ask how head specialization depends on the amount of available training data. Empirically, we find that heads specialized to syntactic relations in BERT emerge as the dataset size increases. We then study this phenomenon in a solvable high-dimensional model of multi-head attention and establish that, in this model, head specialization is a genuine phase transition in sample complexity, corresponding to a non-analyticity in the test error and other observables. Depending on data structure and regularization, specialization can follow uninformative head differentiation or occur through a cascade of phase transitions that progressively recover weaker latent features, providing a statistical form of simplicity bias. Finally, we relate these statistical transitions to the stagewise emergence of specialized heads during online learning, disentangling statistical from dynamical origins of head specialization.

[LG-21] hermodynamically Consistent Neural Networks for a Universal Fluid Gibbs Free Energy Equation of State

链接: https://arxiv.org/abs/2611.26183
作者: Xurui Li,Zhiguo Gan,Jiaming Zhang,Hongxi Zeng,Diannan Lu
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Over the past three centuries, many thermodynamics researchers have worked to develop fluid equations of state with both high accuracy and broad applicability, proposing various theoretical frameworks and models. Building on these efforts, this study uses Gibbs free energy as the starting point and introduces thermodynamically consistent neural networks to construct a fluid equation of state based on Gibbs free energy (GEoS). This approach establishes a novel machine learning framework that provides a unified representation of fluid states. GEoS performs well on the vapor- and liquid-phase density test sets, with average absolute relative deviations (AARDs) of 5.25% and 3.96%, respectively. It outperforms conventional volume-translated cubic equations of state and existing machine learning models, demonstrating its effectiveness. Further analysis using the UMAP dimensionality reduction method shows that most molecules in the test set belong to extrapolation systems, confirming the strong generalization ability of the model. Finally, the test results show that the PC-SAFT equation fitted using GEoS maintains high predictive accuracy, providing a reliable basis for its practical application in industry.

[LG-22] NariQ: Sub-3-Bit Quantization for Long-Horizon Agent ic Tasks

链接: https://arxiv.org/abs/2611.26157
作者: Doyeob Kim
类目: Machine Learning (cs.LG)
*备注: 7 pages, 1 figure, 3 tables

点击查看摘要

Abstract:Long-horizon agentic tasks require a model to call tools over many steps at contexts of 100K tokens or more, yet many low-bit quantization methods are calibrated on short web text and judged on perplexity. This mismatch is costly: existing quantizations of Qwen3.8-27B at or below 3 bits per weight match BF16 on shorter tasks like AIME, yet in published results solve at most 72% as many tasks as BF16 on DeepSWE, a long-horizon agentic benchmark. We propose NariQ, a trellis-coded post-training quantization method that calibrates on tokens generated by the BF16 model in agent sessions. On Qwen3.8-27B, NariQ at 3.0 and 2.5 bits per weight stays level with BF16 on two long-horizon agentic benchmarks, DeepSWE and Terminal Bench, as well as on AIME and RealWorldQA. Our checkpoint at 2.0 bits per weight trails this performance. To our knowledge, these are the first quantizations at or below 3 bits per weight shown to match their full-precision version on long-horizon agentic tasks. We release the files as Qwen3.8-27B-NariQ-GGUF (this https URL), with inference available through our this http URL fork at this https URL.

[LG-23] Loss-Guided Tsetlin Machines: Making the Learning Objective Explicit

链接: https://arxiv.org/abs/2611.26114
作者: Mayur Shende,Runar Helin,Ole-Christoffer Granmo,Vladimir I. Zadorozhny
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:The Tsetlin Machine ™ has no explicit notion of a loss function. Its learning signal is fixed to an implicit, margin-based update rule (MBUR), with no way to choose an objective function suited to a specific task. This paper introduces Loss-Guided TM, which replaces MBUR with an explicit and swappable loss function. In our proposed method, clause feedback and weight updates are driven by the loss, while keeping the Tsetlin Automaton and clause logic unchanged. The approach incorporates cross entropy (CE), symmetric cross entropy (SCE), and asymmetric loss (ASL). We evaluate Loss-Guided TM across six datasets that span multiclass, multilabel, and binary classification tasks. With a suitable loss, Loss-Guided TM matches or improves accuracy and F1 over MBUR on five of the six datasets. The gains are greatest for the imbalanced multilabel datasets, where ASL improves macro F1 by up to 9 percentage points. With CE, Loss-Guided TM is also significantly better calibrated than MBUR on all six datasets, reducing calibration error by a factor of 3 to 11. A qualitative inspection shows that the clause-based global interpretability, examined at the per-class level on image and text data, is preserved when using the Loss-Guided mechanism. These results show that our explicit TM learning objective allows direct targeting of uncertainty calibration, data imbalance, and task structure, opening the TM to the broader literature on loss functions.

[LG-24] How Network Topology and Data Heterogeneity Shape Decentralized Learning: A Spectral Analysis

链接: https://arxiv.org/abs/2611.26104
作者: Luigi Palmieri,Lorenzo Valerio,Chiara Boldrini,Andrea Passarella,Marco Conti,János Kertész
类目: Machine Learning (cs.LG); Social and Information Networks (cs.SI)
*备注:

点击查看摘要

Abstract:Network topology influences decentralized federated learning through information mixing and the relative contribution of individual nodes. While existing analyses primarily focus on spectral mixing properties, they often assume doubly stochastic aggregation, overlooking topology-dependent influence under irregular communication networks. We develop a spectral framework that separates two complementary mechanisms: the dominant eigenmode determines the consensus trajectory and node influence, while non-dominant eigenmodes govern the propagation and attenuation of disagreement. Under uniform neighborhood averaging, we show that the stationary influence is proportional to node degree plus one and derive a bound relating model disagreement to spectral mixing and gradient heterogeneity. We then investigate how these mechanisms interact with the data distribution across nodes. Experiments across regular and heterogeneous network topologies reveal nearly identical learning dynamics under IID data, while network size and connectivity affect how unevenly distributed information is learned under non-IID data. For non-IID settings, we characterize how the distribution of data across nodes shapes the relative contribution of local information to consensus dynamics.

[LG-25] Revealing Pathway Interactions in Network Computations with Complete Path Contribution

链接: https://arxiv.org/abs/2611.26101
作者: Shenghua Liu,Youssef Faragalla,Surya Ganguli,Stephen A. Baccus
类目: Machine Learning (cs.LG); Neurons and Cognition (q-bio.NC)
*备注: 40 pages, 24 figures

点击查看摘要

Abstract:Networks pass signals through many layers of interconnected units, so explaining what they compute means tracing how information travels from input to output and how pathways interact. Most existing approaches analyze only pairwise interactions between components such as units, layers or attention heads. Even when these measurements are combined post hoc into a circuit, they do not capture higher-order interactions along full causal chains (paths) from input to output, and so cannot address how exponentially many paths interact to construct network computation. This leaves a large set of mechanistic questions unanswerable, for example whether inhibitory pathways veto excitatory ones, and how the two combine to yield network decisions. To address this gap, we introduce Complete Path Contribution, a method to decompose the network output into mathematically complete contributions of individual paths to study path-level interactions directly and at scale. Partial summation then induces definitions of contributions for any circuit at any granularity level, with completeness guaranteed at every level. We further present an efficient algorithm that identifies the k strongest paths without examining all of them, with runtime empirically linear in k rather than exponential in depth. By applying Path Contribution to benchmark image classification networks, we quantify excitatory-inhibitory interactions and show how they collectively construct network decisions at a level of detail previously unmeasurable. Furthermore, in an interpretable model of the mouse retina, it reveals how dynamic receptive fields emerge from excitatory-inhibitory interaction motifs among pathways between different cell types in response to time-varying natural stimuli. Overall, Path Contribution reveals how a network’s causal pathways interact at scale to transform input to output.

[LG-26] Approximate Adjoint Models for Memory and Compute-Efficient Online Learning of Hybrid Models

链接: https://arxiv.org/abs/2611.26095
作者: Emilio González-Zamora,Pierre Tandeo,Laurent Debreu,Said Ouala
类目: Machine Learning (cs.LG); Optimization and Control (math.OC); Atmospheric and Oceanic Physics (physics.ao-ph)
*备注:

点击查看摘要

Abstract:Differentiable programming enables online calibration of numerical models by minimizing errors accumulated along simulated trajectories. This includes estimating physical parameters and training neural sub-models in hybrid configurations. Gradients are typically computed using adjoint methods, or backpropagation through time (BPTT), whose computational and memory costs grow with the length of the training window. This limits the duration and resolution of hybrid-model training, a major bottleneck for realistic geophysical applications. Approaches like truncated BPTT reduces memory use by discarding sensitivities across truncation boundaries, introducing gradient bias and neglecting long-range dependencies. We propose a class of gradient approximation methods that preserve the adjoint recurrence over the full trajectory while approximating only the pullback operators within it. Exact pullbacks are computed by automatic differentiation at sparse anchor states only. Between anchors, sensitivities propagate either through the identity (Zero-order) or through a reused, frozen pullback (Pullback Reuse). This approximation makes the parameter pullback constant within each chunk, allowing the gradient sum to factorize so that a single evaluation per chunk replaces one per time step. We evaluate both on three systems of increasing complexity, an incomplete Lorenz-63 model, a large-eddy closure for two-dimensional quasi-geostrophic turbulence, and bias correction in a multilayer quasi-geostrophic ocean model. Peak memory falls by up to an order of magnitude and runtime up to fourfold, with Pullback Reuse staying closest to the exact gradient direction. In all three cases, approximate gradients support hybrid-model training at comparable accuracy, showing that effective online learning does not require exact gradients.

[LG-27] Phantom of the Masquerade: Graph Neural Network-Based Detection of Masquerade Attacks in Controller Area Networks

链接: https://arxiv.org/abs/2611.26094
作者: Nehal Ameen,Pablo Moriano,Irfan Ahmed
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注: 13 pages, 4 figures, 5 tables. Code: this https URL

点击查看摘要

Abstract:The lack of built-in security in controller area networks (CANs) exposes modern vehicles and industrial systems to masquerade attacks that can compromise safety and reliability. These attacks inject malicious messages that mimic legitimate communication patterns, making them difficult to detect using existing intrusion detection systems (IDS). Prior graph-based IDS approaches often rely on shallow embeddings that decouple representation learning from classification, which may limit their ability to fully exploit structural and signal-level information. In this paper, we propose PhantoMasq, a graph neural network (GNN)-based framework for detecting masquerade attacks in CAN systems. PhantoMasq models CAN traffic as message sequence graphs enriched with lightweight statistical features derived from signal behavior, enabling joint learning of structural dependencies and temporal characteristics. The detection task is formulated as a graph-level classification problem and solved using graph convolutional networks (GCNs). We evaluate PhantoMasq on the ROAD dataset under a protocol that prevents overlapping windows from spanning training and test data. Across five masquerade attack types, PhantoMasq achieves mean AUC-ROC between 0.831 and 0.999, with performance depending on the transmission rate of the targeted identifier and the magnitude of the induced signal deviation. Against a state-of-the-art graph-embedding baseline evaluated on the same sliding-window configurations, PhantoMasq achieves higher AUC-ROC on all five attack types, by 0.13 to 0.36, and is less sensitive to window configuration. Window-processing latency ranges from 2.33 to 5.94 ms per window on CPU, corresponding to a throughput of 168 to 429 windows per second. We make our implementation and experimental artifacts available to support reproducibility.

[LG-28] Learning to Compare Neural Representations from Parameterized Computation Graphs

链接: https://arxiv.org/abs/2611.26081
作者: Ya-Wei Eileen Lin,Daniel Herbst,Stefanie Jegelka
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Comparing internal representations across trained models is important for identifying corresponding layers and selecting components for model reuse. Yet, direct comparisons at inference require evaluating candidate models on known reference inputs and collecting their internal activations. In this paper, we study whether internal representation similarity measured across independently trained models can instead be learned to predict in unseen models from their parameterized computations. We introduce a framework that applies a graph metanetwork for mapping each internal site to an embedding derived from its upstream computations while accounting for hidden-coordinate permutation and internal attention-basis transformations. We analyze the structural properties of our estimators and derive conditional guarantees for layer retrieval and stitching candidate selection. Empirically, we find that the learned predictors support relation estimation and low-regret layer retrieval. They also select useful stitching connections while not being trained on stitching outcomes. This opens up a new way to study representation similarity directly from trained neural networks: representational relations learned across a collection of trained models can generalize to new computations, and that the predicted relations can guide candidate selection before direct measurement or adapter fitting for functional evaluations.

[LG-29] Iterative Scarcity-Guided Exploration: Bootstrapping Generative Auto-bidding from Narrow Support NEURIPS2026

链接: https://arxiv.org/abs/2611.26060
作者: Qingmao Yao,Guangzheng Hu,Yusen Huo,Zhilin Zhang,Chuan Yu,Jian Xu,Xu Chu,Bo Zheng,Xiaotie Deng
类目: Machine Learning (cs.LG)
*备注: Accepted at NeurIPS 2026 (poster)

点击查看摘要

Abstract:Auto-bidding is a core algorithmic component in online advertising auctions. In practical cold-start settings, iterative training is commonly used to compensate for low-quality, narrowly supported historical data. Unfortunately, in iterative training, representative diffusion-based AI-Generated Bidding (AIGB) methods fail to sustain extrapolation beyond the narrow data support and stagnate at a suboptimal level. In this paper, we theoretically attribute this stagnation to Signal-to-Noise Ratio (SNR) collapse: the weak radial return signal is overwhelmed by a curvature-induced penalty. To break this stagnation, we propose \textbfIterative Scarcity-Guided Exploration (ISGE), which introduces a guidance handoff: as the return signal collapses, scarcity subsequently takes over as an exploratory signal to elevate the SNR above a critical threshold. Specifically, ISGE iteratively alternates between a \emphJudger that assigns scarcity scores to trajectories and an \emphExplorer that performs guided generation of high-scarcity and high-return trajectories, thereby bootstrapping from the narrow support with such self-generated trajectories. Extensive experiments on the industrial AuctionNet benchmark demonstrate that ISGE, starting from cold-start datasets, effectively surpasses the performance of the Full-Dataset baseline within three iterations. Code is available at this https URL

[LG-30] When Do Rankings Beat Signs? Gradient Direction Recovery under Even Nuisance

链接: https://arxiv.org/abs/2611.26055
作者: Zhuangzhuang Pan,Yan Xia,Chee Seng Chan
类目: Machine Learning (cs.LG)
*备注: 51 pages, 7 figures, 14 tables, including appendices

点击查看摘要

Abstract:Can ranking a batch of antipodal query pairs recover a gradient direction with fewer evaluations than using within-pair signs alone? We study fixed-center recovery in d dimensions for a linear signal plus an unknown differentiable even nuisance, counting function evaluations and feedback rounds separately. Shared pair offsets create an information hierarchy: Mirror retains within-pair signs, Reverse couples reversed cross-pair orders to cancel offsets, and Full enforces all ranking inequalities. Relative shell variation is the nuisance range over the query sphere normalized by the linear-signal scale. When it is at most c\varepsilon/\sqrtd , Reverse localizes every consistent direction to angular accuracy \varepsilon using \widetildeO(d+\sqrtd/\varepsilon) evaluations in one round at fixed confidence, without a variation bound supplied to the decoder. This matches a lower bound up to logarithmic factors, versus the \Theta(d/\varepsilon) one-round sign-only minimax cost on the same class. Over well-conditioned positive-definite quadratics with no common relative-curvature bound, the R -round minimax cost is instead \Theta(dR(\varepsilon^-1/R-1)) , even with cumulative rankings. We extend validity beyond exact evenness under supplied error bounds. Matched-endpoint experiments certify information beyond signs and short cycles. On two ridge tasks under joint evaluation and round caps, short-cycle recovery attains lower terminal mean gaps than the tested batch-capable controls. An additional optimization benefit from fitting the complete ranking remains unestablished.

[LG-31] Reference-Free Humanoid Skateboarding via Termination-Shaped Reinforcement Learning

链接: https://arxiv.org/abs/2611.26054
作者: Ganeshvar Soenessardien,Michael Przystupa,Kevin Sebastian Luck
类目: Robotics (cs.RO); Machine Learning (cs.LG)
*备注: 8 pages, 15 figures, Accepted at the 2026 IEEE-RAS 25th International Conference on Humanoid Robots (Humanoids 2026)

点击查看摘要

Abstract:Humanoid locomotion research has advanced rapidly on walking and running, but \emph object-mediated locomotion, in which a robot must coordinate its body with a passive external object, remains difficult. One challenging task is skateboarding, which requires simultaneous interaction with the ground and a passive but movable platform to achieve forward motion. Current methods for tackling skateboarding with humanoid robots either require human demonstrations or utilise analytic steering models that require system identification. In this paper, we introduce a deep reinforcement learning framework that does not rely on prior data such as demonstrations and enables robots such as the K1 humanoid to learn to skateboard using model-free learning. The method combines task rewards with exploit-blocking terminations. We introduce a number of constraints restricting behaviours such as losing the planted board foot, standing with both feet on the board, excessive twisting, separation from the board, and unstable board motion. Their changing prevalence during training produces an implicit progression from stance control towards sustained robot–board locomotion.

[LG-32] he Delayed Societal Cost of Unfairness

链接: https://arxiv.org/abs/2611.26030
作者: Giovani Valdrighi,Isabel Valera,Marcos Medeiros Raimundo
类目: Machine Learning (cs.LG)
*备注: Preprint, under review

点击查看摘要

Abstract:Fair machine learning typically evaluates the societal costs of a decision policy – the loss incurred by the decision-maker (DM) and the unfair distribution of costs among demographic groups of decision-subjects (DS) – on static data, where reducing one often comes at the expense of the other. Yet decisions change the populations on which they act, and prior work has shown that fairness interventions designed on static data may fail to benefit, or even harm, the disadvantaged group over time. We investigate a complementary question: When can a policy that is Pareto-optimal on pre-deployment data be outperformed by a fairer policy once each is evaluated on the population it induces, reducing costs for one stakeholder without increasing them for the other? In a setting where qualification scores evolve in response to binary decisions and their outcomes, we show that under Rawlsian fairness, which never lowers any group’s acceptance rate, placing sufficiently high weight on fairness reduces societal costs whenever accepting disadvantaged DS at the DM’s break-even point leaves them in a state less costly to the DM than rejecting them. Under egalitarian fairness, which may lower the advantaged group’s acceptance rate, this additionally requires the reduction in DM loss on the disadvantaged group to outweigh the increase on the advantaged group. Under both fairness notions, however, when DM loss is accuracy-based, improvements in the post-deployment population may arise from declining DS qualification, revealing that gains for the DM do not necessarily translate into better outcomes for DS. Our results show that policy-induced population dynamics can overturn apparent performance-fairness trade-offs, allowing fairer policies to reduce the delayed societal costs of decision-making.

[LG-33] FrontierReplay: Split-Free Exact Prefix Caching for Hybrid Language Models

链接: https://arxiv.org/abs/2611.26020
作者: Zekun Wang,Yizhuo Wang
类目: Machine Learning (cs.LG)
*备注: 15 pages (12 pages main text + references), 12 figures, 4 tables

点击查看摘要

Abstract:Hybrid language models combine self-attention with recurrent layers whose in-place state updates restrict prefix reuse to exact saved positions. Saving states on a coarse grid wastes reusable prefixes, whereas existing fine-grained exact caches must cut the prefill forward pass at each mid-prompt checkpoint, adding scheduler steps that outweigh the reuse gains under load. We observe that across multi-turn chat, agent loops, document question answering, and tree search, prefix reuse almost always ends at structural boundaries predictable in advance, such as turn ends and known branch points. We present FrontierReplay, a prefix caching system that materializes exact checkpoints at these boundaries without cutting the prefill. While the forward pass runs uncut, FrontierReplay logs a short window of recurrent-layer inputs and invokes a cross-layer, state-only replay kernel to build the target state 6-37x cheaper per token than full recomputation without host synchronization. A frontier-aware retention policy prioritizes active anchors and proven branches while lending dead checkpoint slots to the attention KV cache. Across diverse hybrid architectures, model scales, and workloads, FrontierReplay delivers clear gains over existing exact caches, lowering median time to first token (TTFT) by 16-89% (up to 8.9x), raising 80% service-level objective (SLO) capacity by up to 1.73x, and improving goodput by up to 4.26x (3.67x at larger scales). Compared with approximate replay methods, FrontierReplay delivers up to 1.54x higher goodput while preserving exact task accuracy, avoiding the up to 17% accuracy drop suffered by approximation on long contexts. Comments: 15 pages (12 pages main text + references), 12 figures, 4 tables Subjects: Machine Learning (cs.LG) Cite as: arXiv:2611.26020 [cs.LG] (or arXiv:2611.26020v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2611.26020 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[LG-34] Reinforcing Few-Step Diffusion Models via Shadow Posterior Matching

链接: https://arxiv.org/abs/2611.25998
作者: Jingqiao Xiu,Yicong Li
类目: Machine Learning (cs.LG)
*备注: Code available at this https URL

点击查看摘要

Abstract:Few-step diffusion models are widely deployed for their efficiency. However, reinforcing them with black-box rewards, as is routine for multi-step models, faces two structural challenges. First, a few-step generator is a sampler rather than a score model, so its velocity is not the forward velocity of the generated distribution. Second, it must be anchored to the undistilled teacher, which shares no transitions with it. Existing few-step RL methods handle the latter with a marginal KL to the teacher, but each pays for the former in its reward credit. We propose Shadow Posterior Matching (SPM), which renoises generated samples into shadow inputs and obtains the credit from their shadow posterior. From the KL-regularized reward objective, we derive a regression loss that turns reward values into a reward direction while anchoring the generator’s outputs to the teacher and aligning the shadow inputs with the deployed inputs. Applied to 4-step generators, SPM achieves higher rewards with less reward hacking than prior few-step RL methods.

[LG-35] Seeing Beyond the Obvious: Complementary Semantic Guidance for Salient Single-Positive Multi-Label Learning

链接: https://arxiv.org/abs/2611.25986
作者: Tong Jin,Meng-Long Wang,Xingyu Zhao,Sheng-Jun Huang
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Salient Single-Positive Multi-Label Learning (SalSPML) assigns each image its most salient positive label, leaving other classes unlabeled. Existing methods construct class-level Salient Prototypes from observed positives to guide the recognition of unobserved classes. However, since salient annotations favor visually prominent instances, these prototypes may not fully capture the visual diversity of hidden positives, resulting in insufficient positive supervision. To address this limitation, we propose CoST, a Coverage-guided Complementary Soft-Target Refinement framework that supplements Salient Prototypes with complementary class-semantic information. Specifically, CoST identifies semantically relevant candidates from unknown classes by leveraging task-learned category information and image-text semantic correspondence. The selected candidates are then aggregated into class-wise Coverage Prototypes to complement the semantic representations provided by Salient Prototypes. Furthermore, we introduce a ranking-based selective refinement strategy that identifies classes receiving stronger semantic support from Coverage Prototypes and selectively enhances their soft targets while preserving the base soft targets for the rest. Experiments on multiple benchmarks demonstrate the effectiveness of CoST, with strong performance gains over existing baselines.

[LG-36] Power Setpoint Tracking for Large-Scale EV Charging: Transformer-Constrained Charging Policy Learning with Self-Tuning Constraint Prices

链接: https://arxiv.org/abs/2611.25965
作者: Daniyal Sajedi-Hosseini,Yoshihiko Fujisawa,Shubham Singh,Lei Zhou,Katsuki Fujisawa,Sebastien Gros,Mazaher Karimi,Xun Shen
类目: Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 13 pages, 7 figures, 6 tables. Submitted to IEEE Transactions on Transportation Electrification

点击查看摘要

Abstract:In power setpoint tracking (PST), an operator follows a purchased charging profile without overloading shared transformers or undercharging vehicles. PST is cast as a constrained Markov decision process that minimizes tracking error subject to overload and mean user-satisfaction constraints. Transformer-Constrained Charging Policy Learning (TC-CPL) extends soft actor-critic with proportional-integral Lagrange-multiplier updates from measured violations, a residual on the incumbent dispatcher that is warm-started from its demonstrations, and a transformer-aware encoder shared by all ports. It observes no departure times or states of charge. Simulations (five seeds, 100 paired days) compare it with the incumbent, Round-Robin. On a 600-port, 122-substation district, raw overload falls from 3.59 to 0.0017 kWh/day (overloaded seed-days from 77% to 0.4%), and absolute tracking error falls by 27%. Delivered energy rises by 16%, and satisfaction rises from 0.959 to 0.989. On a 42-kW transformer with household load, overload falls from 1.86 to 0.0014 kWh/day and satisfaction rises from 0.974 to 0.985, but absolute tracking error is 22% higher. Against hand-set fixed multipliers, self-tuning is statistically indistinguishable on the district. On the single transformer, it raises satisfaction by 0.011 at similar overload, but a short sweep found a better-serving fixed pair. The overload multipliers stay below their cap.

[LG-37] Gradient Boosting with Kolmogorov-Arnold Units for Interpretable Regression beyond Pairwise Interactions

链接: https://arxiv.org/abs/2611.25942
作者: Minjong Cheon,Hanseon Joo,Eunji Lee
类目: Machine Learning (cs.LG)
*备注: 21 pages, 5 figures

点击查看摘要

Abstract:Interpretable models for tabular regression, such as the explainable boosting machine (EBM), restrict a prediction to main effects and pairwise interactions so that every part of it can be plotted. When the target depends on three or more inputs jointly, this restriction costs accuracy, and a black-box model with post hoc explanations becomes the usual fallback. Here we present MJBoost, a gradient boosting method whose weak learner is a single Kolmogorov-Arnold unit g(\sum_p \phi_p(x_p)) with penalized spline inner and outer functions. Each unit is fitted in closed form by alternating penalized least squares, starting from principal Hessian directions of the residual, and categorical and missing inputs enter the units directly. Because a unit depends on each input only through one inner function, the fitted model has an exact functional ANOVA decomposition into main effects, pair effects and a higher-order remainder, which one-dimensional convolutions compute. On synthetic targets of interaction order three to six, the test error of MJBoost is up to 13 times lower than that of EBM, and its decomposition recovers the true variance share of every order to within 0.04. On the 35 regression tasks of the OpenML-CTR23 benchmark, it has a lower test RMSE than EBM on 21 tasks and a higher one on 2 (geometric-mean RMSE -29.1%). On these tasks the main and pair effects of the fitted model explain a median 0.992 of its variance, so the model reads as plots where the data allow, and the decomposition measures the higher-order part where they do not. With Newton steps the same algorithm classifies: on 32 binary tasks of OpenML-CC18 its test log-loss is on par with that of EBM, and on 11 multiclass tasks its mean log-loss is lower than that of EBM on 8 and than that of tuned LightGBM on 7.

[LG-38] DTS: Delta-Based Tile Scoring for Efficient Long-Context Sparse Attention DATE2027

链接: https://arxiv.org/abs/2611.25938
作者: Emre Uğur,Jiawen Qi,Qinyu Chen
类目: Machine Learning (cs.LG)
*备注: 7 pages, 4 figures. Accepted at DATE 2027. Emre Uğur and Jiawen Qi contributed equally

点击查看摘要

Abstract:Large language models (LLMs) increasingly need to process long documents and codebases, making long-context attention a major prefill bottleneck. Block sparse attention reduces computation by skipping unimportant blocks of the attention matrix, but scoring blocks accurately can itself be expensive. We propose Delta-Based Tile Scoring (DTS), a training-free, content-adaptive tile scoring method for sparse block attention. DTS selects query and key anchors according to local vector changes, using fewer representatives in redundant regions while retaining more detail in heterogeneous regions. To accelerate anchor selection, parallel delta packing partitions the query and key sequences into tile-aligned chunks and performs bounded local scans over these chunks concurrently. The selected tiles are then computed using the original query, key, and value vectors. On LLaMA-3.1-8B-Instruct, DTS keeps average RULER and LongBench scores within 0.10 percentage points of full attention, with RULER evaluated up to 128K tokens. At 128K tokens, DTS reduces measured scoring-pipeline latency by 38% compared with XAttention. On four NVIDIA H100 GPUs, it achieves a 3.3 \times prefill speedup over a PyTorch full-attention baseline using FlashAttention-2 at 512K tokens.

[LG-39] A Lightweight Hybrid Approach for Milling Force Reconstruction from Internal Machine Signals

链接: https://arxiv.org/abs/2611.25921
作者: Eric Hirsch,Christian Friedrich
类目: Machine Learning (cs.LG)
*备注: Author-prepared preprint. 7 pages, 4 figures, 5 tables

点击查看摘要

Abstract:Data-driven monitoring based on milling force can enable advanced manufacturing services including process optimization and condition monitoring. However, industrial implementation is often hindered by costly and intrusive hardware and large training data requirements. This paper therefore presents a hybrid approach using machine-internal signals and machining metadata to enable force-based monitoring on brownfield machines without additional sensors. Commonly available machine signals have relatively low sampling rates and cannot directly represent high-frequency milling-force variations, especially at higher spindle speeds. Instead of learning the complete force signal, which would increase model complexity, data requirements, and overfitting risk, a supervised machine-learning model predicts only the average milling force from selected machine-signal features and machining parameters. This low-frequency estimation parameterizes a static milling force model that reconstructs the force behavior over one spindle revolution. Although the static model does not reproduce all dynamic effects, it captures the force distribution, load levels, and angular regions of highest tool engagement. The approach is evaluated on one milling machine across varying process parameters, engagement conditions, cutting tools, and two workpiece materials. Dynamometer measurements are required only for development and validation. Ridge regression achieves an R^2 of 0.984 and 0.991 for average-force prediction on the test data ( \barF_x and \barF_y ). The reconstructed force behavior over one spindle revolution achieves an R^2 of 0.962 and 0.982 ( F_x and F_y ) compared with the reconstruction based on measured force features. The method provides a cost and computationally efficient basis for process monitoring, tool condition assessment, and process optimization on existing machines.

[LG-40] LIFT: Latent Flow Matching for Probabilistic Cell Fate Dynamics with Single Cell Snapshots

链接: https://arxiv.org/abs/2611.25917
作者: Bowen Yang,Junda Ying,Yuxuan Wang,Peijie Zhou,Lei Zhang
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Lineage tracing links cell states through clonal ancestry, but does not generally identify unique source-descendant pairings. Learning from such data therefore requires a model of conditional outcomes, rather than only the terminal population distribution. We introduce LIFT, a latent flow framework for probabilistic cell-state generation from lineage-informed couplings. Variational inference encodes alternative outcomes in a latent space, while a conditional prior enables sampling from the initial state alone. A Mean Flow decoder predicts future cell state directly, requiring one decoder evaluation per sampling, which makes it convenient to predict cell fate through large scale Monte-Carlo sampling. This construction combines a representation of one-to-many relationships with efficient sampling and an analysis of conditional prediction error. Experiments on synthetic and real datasets demonstrate LIFT’s ability to recover both population distribution and cell fate probabilities.

[LG-41] Parameter-Efficient Adaptation of Self-supervised Vision Backbones Survives Distribution Shift in Radar Micro-Doppler HAR ICASSP2027

链接: https://arxiv.org/abs/2611.25874
作者: Keryan Chelouche(1 and 2),Petr Dobias(1),Nistor Grozavu(1),Olivier Romain(1) ((1) ETIS UMR 8051, CY Cergy Paris Université, ENSEA, CNRS, Cergy, France, (2) Fondation Léonie Chaptal, Sarcelles, France)
类目: Machine Learning (cs.LG)
*备注: Submitted to ICASSP2027

点击查看摘要

Abstract:Micro-Doppler radar enables contactless, privacy-preserving human activity recognition (HAR), but deployment robustness remains underexplored. We compared full fine-tuning, frozen probing, and parameter-efficient fine-tuning (PEFT) using subject-disjoint in-distribution cross-validation and few-shot cross-age, cross-orientation, and cross-dataset transfer. Full fine-tuning suffered negative cross-dataset transfer, while frozen features under-adapted on cross-orientation. Only PEFT avoided both failures. We paired DINOv3 with all-layer principal-singular-direction adaptation (PiSSA), to our knowledge the first application of a self-supervised image foundation backbone to radar micro-Doppler HAR. DINOv3-PiSSA ranked first among PEFT methods in every evaluated setting. Under our protocol, its Small variant matched the in-distribution accuracy of the prior Base-size PEFT state of the art using 4.5x fewer total parameters, with only 3% trainable. It surpassed a size-matched variant of that baseline in every evaluated setting, with transfer gains increasing with shift severity: 2.1, 6.6, and 8.9 percentage points, respectively. Ablations attributed improvements to both pretraining and adapter recipe, with the pretraining contributing most under compound shift. Code and protocol are available at this https URL.

[LG-42] X Hedging: towards an eXplainable artificial intelligence option Hedging framework

链接: https://arxiv.org/abs/2611.25870
作者: Bjørn André Aaslund,Johannes Berge,Ayoub Haida,Ying Ni,Rita Pimentel
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注:

点击查看摘要

Abstract:We propose X Hedging, an alternative to the Deep Hedging framework that addresses computational and explainability limitations of neural-network-based hedging strategies. Deep Hedging leverages neural networks to learn flexible hedging policies but relies on opaque models and typically requires large training datasets, which can complicate interpretation and limit its effectiveness in data-constrained settings. X Hedging takes a step toward addressing these limitations by replacing neural networks with tree ensemble methods, specifically LightGBM, to estimate hedging strategies. This design provides a computationally efficient alternative while preserving strong hedging performance and moving toward more user-centered explainability. To integrate LightGBM into a reinforcement-learning-style hedging framework, we introduce an optimization approach that sequentially trains tree-ensemble models while introducing stochasticity through random training orders. In controlled experiments, X Hedging achieves performance comparable to Deep Hedging while offering greater user-centered explainability, improved computational efficiency, and substantially higher robustness when training data are limited. The numerical results demonstrate that tree-ensemble methods provide an effective computational alternative to neural networks for data-driven dynamic hedging.

[LG-43] Full Rank and Stable Sampling of tanh Networks

链接: https://arxiv.org/abs/2611.25867
作者: Philipp Grohs,Matěj Trödler
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:We study the recovery of feedforward \tanh networks of fixed architecture with P parameters from finitely many point evaluations. First, we characterize the set of parameters where the differential of the realization map has full rank and show that it coincides with the regular parameters in the sense of Vlačić and Bölcskei (2021). Second, we make use of this characterization to prove that almost every tuple of 2P+1 sampling points yields uniform-norm Lipschitz stability on compact subsets of this parameter set. Third, we prove that for exact observations, projected gradient descent and projected Gauss–Newton converge locally from sufficiently close feasible initial guesses. The same evaluations also determine the realized function for arbitrary parameters and identify parameters in the characterized set up to permutations and sign changes of hidden neurons. Finally, we develop a general framework for real-analytic networks that highlights the role of full rank in stability and local reconstruction.

[LG-44] Beyond Alignment: Whitening Geometry Profile Synthesis and Heterogeneous Ensembles for Cross-Environment WiFi Action Recognition

链接: https://arxiv.org/abs/2611.25831
作者: Lei Wang
类目: Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: 15 pages, 3 figures, 4 tables. Preprint of a manuscript submitted to IEEE Transactions on Mobile Computing

点击查看摘要

Abstract:Cross-environment deployment is the central obstacle to practical WiFi sensing: a pipeline reaching 0.89+ top-1 in its training environments collapses to 0.408 under Leave-One-Domain-Out (LODO) evaluation with only K-shot labels. A ceiling analysis shows the collapse is not an alignment problem: the action signal is strong across seen environments (about 0.89 K10), environment identity is decodable from the encoder (0.965), and the failure concentrates in the encoder’s extrapolation to the novel environment. A pre-registered falsification chain confirms that “align the domains” methods–training-time alignment, readout-level prototype alignment, adversarial confusion, and covariance/subspace alignment (CORAL -3.8pp, subspace -8.3pp)–degrade performance, and that source-domain statistics anti-correlate with cross-environment contribution. The effective levers lie elsewhere; we combine them into Dec-H: (i) closed-form source within-class whitening, provably re-geometrizing the readout’s L2 regularization (+8.6pp, zero cost); (ii) profile-mixing augmentation (+1.0pp); (iii) few-shot adaptation with alternating whitening refit (+1.5pp; fixed preconditioners are 20/20 negative); (iv) heterogeneous decision ensembles (+1.7pp); homogeneous scaling and source-weighted stacking are saturated or harmful. Dec-H reaches 0.560 K10 (0.812 K40), +15.2pp over the baseline, consistent across all four environments and seeds; all thirteen configurations were pre-registered with frozen criteria.

[LG-45] A Sparse Low-Rank Biclique Decomposition for Graphs NEURIPS2026

链接: https://arxiv.org/abs/2611.25826
作者: Antoine Vialle,Sergei Gerasimov,Aref Einizade,Antonio Ortega,Fragkiskos D. Malliaros,Jhony H. Giraldo
类目: Machine Learning (cs.LG)
*备注: Accepted at NeurIPS 2026. 29 pages, including references and appendices

点击查看摘要

Abstract:Low-rank plus sparse decompositions, such as robust PCA, typically model the low-rank component as dense and the sparse component as corruption. This viewpoint is poorly suited to graph adjacency matrices, where sparsity is structural and dense low-rank factors destroy the combinatorial and computational properties of the graph. We propose signed biclique decomposition (SBD), a sparse low-rank decomposition for integer-weighted graphs that represents the structured component as a sum of sparse signed binary rank-one factors, together with a sparse signed residual. The resulting representation remains discrete, interpretable, sparse, and compatible with standard linear algebra. We formulate fixed-rank and greedy SBD objectives and derive a gain-based greedy algorithm with theoretical guarantees. Empirically, SBD recovers block structure in noisy graphs, while its factorized representation enables efficient sparse matrix-matrix multiplication (SpMM). On real graphs with strong shared-neighborhood structure, SBD achieves compression factors above 10 and SpMM speedups up to 6\times over cuSPARSE, suggesting a practical alternative to dense low-rank decompositions for graphs. Code is available at this https URL.

[LG-46] How Reliable Are Explanations of Language Models That Detect Insider Threats?

链接: https://arxiv.org/abs/2611.25796
作者: Suramya R. Angdembay
类目: Machine Learning (cs.LG)
*备注: 34 pages, 9 figures

点击查看摘要

Abstract:Organizations monitor their workers’ activity for insider threats, in which workers misuse their legitimate access, by copying confidential files before leaving. Detectors score how unusual each day of activity is, flagging the most unusual for an analyst to review. Prominent current approaches use opaque neural networks, so explanations of a flag point to internal features: patterns of neuron activity most active on days of misuse. We audit such explanations for language-model detectors trained on CERT’s simulated workplace logs. To test an explanation, we see what happens to the detector’s score when we remove its features inside the model. We compare this to a control: a change that is as similar as possible to the removal, but without actually removing the features. Because defining “similar” requires choosing a metric, our central contribution is the tipping radius: a first-order prediction of how far that metric must move before the difference between removal and control disappears, a measure of how fragile the result is to that choice. In 33 of 34 runs of seven language models, the tipping radius is at most 18% of the distance between two standard choices of metric. Moving that far removes at least 58% of the difference or reverses it, while random moves of the same size leave it largely intact. The result replicates in thirteen of fourteen pre-declared primary runs on openly released models, though on our trained detectors the difference is often too small to measure. We also find that a flag’s score partly reflects who the worker is, not what they did, and that the features named depend on whom they are compared against. Audits should therefore fix the control’s metric in advance, report its tipping radius, and disclose that comparison.

[LG-47] Disentangling Monitor Hardware Artifacts and Smoothing Lag for Shift-Robust Intraoperative EEG and Hemodynamic Risk Calibration

链接: https://arxiv.org/abs/2611.25765
作者: Zeyu Jia
类目: Machine Learning (cs.LG)
*备注: 16 pages, 4 figures, 5 tables

点击查看摘要

Abstract:Commercial intraoperative electroencephalography (EEG) and blood pressure monitors introduce cyclic impedance pulses, smoothing delays, direct-current (DC) drift, and cuff holds that confound risk forecasting and uncertainty calibration. We developed a causal signal-conditioning and shift-robust conformal calibration framework across 5,439 public anesthesia procedures. Streaming filters blank 10.13-min 64-Hz impedance checks, apply unclipped post-notch conditioning, separate 63-s suppression-ratio lag, and distinguish three-channel cuff inflations from dropouts across propofol-remifentanil total intravenous anesthesia (TIVA; N = 3,214), volatile anesthesia (N = 2,046), European endoscopy sedation (N = 171), and DC-coupled surgery (N = 8). De-impedanced cortical tree ensembles were compared against deep sequence models, paired with patient-cluster Subgroup-Covariate-Weighted Conformal Risk Control (CRC) and lag-buffered online Conformal Proportional-Integral-Derivative (PID) tracking. Pre-filter blanking removed 97.90% and 97.60% of false flatline conflicts while preserving 99.23% and 98.96% of deep suppression. On unseen TIVA surgeries (N = 525), cortical features achieved area under the precision-recall curve (AUPRC) 0.3135 (1.94% prevalence) and 0.1627 (0.82% lag-free prevalence), with no detectable difference from deep encoders (p = 0.36) and no cortical gain for hypotension (0.3432 vs. 0.3437, p = 0.784). Zero-retraining volatile transfer (N = 1,048) showed no detectable difference from an end-tidal gas oracle on lag-free suppression (0.0820 vs. 0.0736, p = 0.472). Subgroup-Covariate-Weighted CRC bounded all 70 TIVA shift cells at = 4.38% (alpha = 5.0%) in 66.32 ms per 30-s epoch. Causal artifact disentangling and shift-aware conformal calibration support reliable real-time intraoperative risk bounds within explicit physiological observability limits.

[LG-48] Goal-Guided Reinforcement Learning via Fine-Tuning Vision-Language Models

链接: https://arxiv.org/abs/2611.25742
作者: Yajie Qi,Wei Wei,Lin Li,Lijun Zhang,Jianye Hao
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Exploration remains a major challenge for reinforcement learning (RL) agents in Craftax-style environments due to sparse rewards and long horizons. Large language models (LLMs) can guide exploration by generating high-level goals, but methods using models without explicit task adaptation often rely on carefully designed prompts and may produce inconsistent or contextually inappropriate goals. We introduce CCGP-2M, a multimodal goal-preference dataset constructed from interaction trajectories in Craftax-Classic, and propose a vision-language model (VLM)-guided RL framework that learns environment-specific goal guidance from this dataset. The framework consists of two stages: offline VLM fine-tuning and online goal-guided RL. In the offline stage, we align the VLM with goal preferences using Kullback-Leibler Token-level Direct Preference Optimization (KLTDPO), which combines token-level KL regularization with distributionally robust reweighting. Experiments on Craftax-Classic show that our method collects diamonds earlier than state-of-the-art baselines, achieves the highest score and reward, and attains the highest success rates on most achievements.

[LG-49] Measurable Does Not Imply Transportable: A Prospective Audit of Model Explanations in Multi-Year Field Trials NEURIPS2026

链接: https://arxiv.org/abs/2611.25692
作者: Rayhaneh Shabani Nia,Ali Karkehabadi
类目: Machine Learning (cs.LG)
*备注: Accepted at AISciK @ NeurIPS 2026

点击查看摘要

Abstract:Model explanations are increasingly used as scientific evidence, yet it remains unclear what they license scientists to conclude beyond the data on which they were constructed. When a fine-grained explanation fails to replicate, the failure may be attributed to noisy measurement of its target, raising a fundamental question: does reliable measurement imply a reproducible explanation? We test this directly. We formalize prospective explanation transportability: the extent to which an explanation learned from past data is recovered in a genuinely future context. This property is distinct from feature importance and measurement reliability. Using multi-year Genomes-to-Fields maize trials, where each new year is a real future context rather than a random resample, we evaluate 16 agronomically interpretable environmental concepts across five forward splits and one prospectively locked test year. Coarse environment-level explanations transport reliably (median cosine 0.83); their locked 2024 value, 0.811, fell within a prespecified 80% bootstrap percentile interval. For grain yield, genotype-conditioned explanations have a historical median cosine of 0.30 and range from 0.006 to 0.822 across years. Reliability does not order these outcomes: the year with the highest split-half reliability had the lowest explanation transport. Within-context resampling showed that GxE explanations were reproducible within a fixed context (0.77-0.95) while exhibiting lower prospective alignment across the 2020-2023 transitions (0.01-0.36). A second phenotype transports strongly at both resolutions, and non-temporal validation inflates apparent interaction-level transport by +0.22. These results show that target reliability and within-context explanation reproducibility do not by themselves establish when a model explanation can support a reproducible scientific claim.

[LG-50] Explainability of genotype embedding models through observed cohort paths

链接: https://arxiv.org/abs/2611.25681
作者: Amashi Niwarthana,Aloysius Tan,Jagath C. Rajapakse
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Deep learning models that combine genotypes with neuroimaging predict neurological diseases more accurately than either modality alone. Their clinical use, however, requires identifying which genetic variants a prediction is based on. Path-based attribution methods like Integrated Gradients address this by integrating the model’s gradient from a baseline to the input. These methods assume the model accepts any point along that path as a valid input. Genotype models violate this assumption. They map the three discrete allele dosages \0,1,2\ to vectors through a learned embedding defined only at those values, so a straight line passes through fractional genotypes at which no embedding, and therefore no gradient, exists. Attribution must therefore proceed in embedding space where the choice of path can affect the resulting attribution. We show that the commonly used per-variant estimator does not generally satisfy completeness. Additive separability guarantees completeness for this estimator, but interactions between variants generally break that property. Instead, we integrate along piecewise-linear embedding-space paths connecting genotype waypoints observed in the training cohort. Each path satisfies completeness in the continuum although the interiors of its segments are not necessarily observed. Our method improves faithfulness over Integrated Gradients and discretized Integrated Gradients under both perturbation protocols, achieves the strongest results for Parkinson’s disease classification under support-preserving perturbations, and is competitive with a support-background Shapley baseline under per-variant perturbations. It also shows higher seed stability and better recovery of planted non-additive effects in a synthetic benchmark.

[LG-51] GPU-Accelerated Secure Inference for Drug–Target Interaction Prediction

链接: https://arxiv.org/abs/2611.25664
作者: Sikha Pentyala,Steven Golob,Daniil Filienko,Ming-Chun Kao,Martine De Cock
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Deep learning models that predict drug–target interactions (DTI) can speed up drug discovery, but the candidate drugs they score are valuable intellectual property that their owners cannot reveal. Secure multiparty computation (MPC) lets two servers compute a prediction on a secret-shared drug, without either server learning the drug. However, secure inference of deep models is slow. Non-linear operations require communication between the servers, and MPC arithmetic does not map naturally to GPUs. In this paper, we present our solution to Track~III of the iDASH 2026 privacy and security competition, which asked for fast two-party secure DTI inference on GPUs. Our solution is built on the co-design of MPC-friendly models and MPC protocols. On the model side, we design SlimDTA, a DTI model that reduces the operations on the secret drug, and in particular the non-linear operations that require communication. On the protocol side, we build a GPU framework that evaluates the linear layers with additive secret sharing and the non-linear layers with function secret sharing, including a custom CUDA kernel for the comparisons. Our solution won first place in the competition, at under 10,ms per prediction on average, and its secure predictions match the plaintext model to within 0.02%.

[LG-52] NIMIS: A Structure-Agnostic Chemically Informed Slot-Attention Framework for Demixing 1D 1H-NMR Mixture Spectra

链接: https://arxiv.org/abs/2611.25629
作者: Anna Marija Sumrova,Jonathan M. Goodman
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:The products of chemical reactions are usually complex mixtures that must be separated into their components before interpretable ^1 H-NMR spectra can be obtained. Direct analysis of the spectra of crude mixtures would accelerate reaction analysis and facilitate the exploration of new chemical space. We present NIMIS, a structure-agnostic, attention-based deep learning framework that groups the peaks of a 1D ^1 H-NMR mixture spectrum according to the component from which they originate. J-coupling information is introduced directly into the attention mechanism of a multi-head Transformer encoder, and the resulting contextualised peak embeddings are separated into components by slot attention. NIMIS was trained on 158,005 synthetic mixtures constructed from experimental spectra in the NMRexp database, covering molecules of up to 20 heavy atoms composed of C, H, N, O, S, F, Cl, Br and I. On held-out molecules it reaches a peak-to-component assignment accuracy of 82.28% for noise-free input and 77.34% with 10% integral noise, with a mean NIMIS score of 0.83 for correct assignments. NIMIS outperforms a conventional Transformer and simple baselines, and its performance carries over to a scaffold-disjoint evaluation. By grouping peaks into component-level sets, NIMIS provides a foundation for more robust mixture analysis methods, including methods capable of handling overlapping signals.

[LG-53] QUBO-based feature selection for MRI undersampling pattern design

链接: https://arxiv.org/abs/2611.25625
作者: Xiaoxin Huang,Masayuki Ohzeki
类目: Machine Learning (cs.LG); Signal Processing (eess.SP)
*备注: 20 pages, 3 figures

点击查看摘要

Abstract:We formulate undersampling mask design for Cartesian magnetic resonance imaging (MRI) as a feature subset selection problem. Each sampling line is represented by a binary feature indicating whether it is included in each candidate mask. Using candidate masks and their training-set evaluations, we estimate mutual information (MI) and conditional mutual information (CMI). These estimates and a sampling-count penalty define a quadratic unconstrained binary optimization (QUBO) model. The model is solved using simulated annealing and quantum annealing to determine a feature subset and its corresponding sampling mask. We assess the selected subsets on the test set through zero-filled reconstruction in Fourier undersampling simulations using fastMRI knee magnitude reference images, with patient-disjoint training and test sets. At all six tested target sampling rates from 15% to 40%, the resulting reconstructions have lower mean root mean square error than those obtained with four conventional sampling methods. They also have higher mean structural similarity than those obtained using univariate mutual-information ranking at all tested rates. These results support information-theoretic feature selection as an approach to MRI undersampling mask design under the tested zero-filled simulation settings.

[LG-54] Deviations from the NTK Regime in CNNs under Joint Depth-Width Scaling

链接: https://arxiv.org/abs/2611.25624
作者: Maria Matveev,Pascal Esser,Gitta Kutyniok,Mariia Seleznova
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:The concentration and training-time constancy of the neural tangent kernel (NTK) depend on the interplay between network width, depth, and initialization. While joint depth-width scaling has been studied in fully-connected, residual, and recurrent networks, its effects in convolutional neural networks (CNNs) remain less understood. We investigate these effects numerically in CNNs, examining NTK dispersion at initialization and kernel rotation during training. Let depth L be the number of convolutional layers, and width C the number of channels per layer. We find that NTK dispersion increases with L/C in the critical and chaotic initialization regimes, but remains small in the ordered regime. At critical initialization, dispersion curves approximately collapse as a function of L/C , suggesting a nontrivial joint scaling regime. During training, kernel rotation generally increases with L/C , even in the ordered regime where initial dispersion is small. Together, these results suggest that L/C may govern NTK deviations from the infinite-width limit in certain regimes, and that NTK concentration at initialization and stability during training exhibit different scaling behaviors.

[LG-55] Hierarchical Spatiotemporal World Modeling with Content-Adaptive Spatial Addressing

链接: https://arxiv.org/abs/2611.25610
作者: Tiantian Li,Junfeng Lu,Yifei Wang,Zhaoyang Zhang,Lei Wang,Tianyi Qian
类目: Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Long-horizon world modeling in partially observable environments requires compact memories that preserve spatial structure despite accumulated localization drift. Inspired by hippocampal episodic and spatial memory and the multi-periodic coding of entorhinal grid cells, we propose HiSTFormer, a hierarchical spatiotemporal world model, which adopts a layer-wise organization of local temporal RoPE and multi-periodic spatial RoPE, together with content-conditioned address offsets, to capture recent temporal context and long-range spatial structure. A fixed-capacity memory keeps KV storage and per-step attention cost bounded as the sequence grows. On MemoryMaze and HM3D, HiSTFormer outperforms strong world-model baselines within the training sequence length and generalizes substantially better to prediction horizons far beyond those seen during training, while using a compact model. The hierarchical spatiotemporal design remains effective across deeper Transformer stacks, with consistent gains as depth increases. These results show that hierarchical spatiotemporal organization and content-adaptive spatial addressing provide effective inductive biases for efficient long-horizon world modeling.

[LG-56] PSTO: A PDE-structured transport operator with shared implicit updates for multiscale CDR systems

链接: https://arxiv.org/abs/2611.25607
作者: Kaiwen Li,Dezhi Zhou,Kun Wu,Xingcai Lu
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Multiscale convection-diffusion-reaction (CDR) systems couple spatial transport with local stiff reaction kinetics, making their numerical simulation computationally demanding. Conventional neural surrogates require costly coupled CDR training data, and their parameter counts scale with the number of transported variables, limiting scalability to high-dimensional CDR systems. Here we propose a PDE-Structured Transport Operator (PSTO) that learns convection–diffusion dynamics solely from reaction-free data. By embedding the PDE structure into the neural architecture, PSTO generalizes to full CDR systems at inference without retraining. A single transport operator block is shared across all transported variables and refined through relaxed fixed-point iterations, performing an implicit-like correction in latent space. This improves stability during recursive multi-step prediction without adding trainable parameters. We evaluate PSTO on three benchmarks: a canonical 2D CDR system, a multiscale Robertson counterflow flame, and a multiscale hydrogen planar jet flame. In every case, PSTO reduces the mean relative L_2 error after 100 coupled steps by at least 50% compared to the strongest neural baseline. Ablation studies on the fixed-point iterations and detailed error decomposition further confirm that PSTO achieves both high one-step accuracy and long-term stability in coupled rollouts. Time-step sensitivity analysis shows that PSTO maintains accuracy even with a tenfold larger time step, whereas baseline accuracies decay rapidly as the time step increases. This framework enables efficient, modular modeling of multiscale CDR dynamics by separating transport learning from the treatment of local kinetics.

[LG-57] From Data Alignment to Parameter Generation: Few-Shot Cross-City Traffic Prediction with Conditional Diffusion

链接: https://arxiv.org/abs/2611.25592
作者: Shengchun Wang,Yuhang Lin,Tong Li,Pengyuan Zhou,Andy Chi Lok Tai,Wenjie Chen,Zhu Xiao,Zhu Han
类目: Machine Learning (cs.LG)
*备注: 25 pages

点击查看摘要

Abstract:Accurate mobile traffic forecasting is crucial for network service management, supporting proactive capacity provisioning, congestion mitigation, and energy-efficient operation. However, many cities and regions are data-scarce (e.g., newly deployed or poorly instrumented areas), thus making traffic prediction a few-shot learning problem and limiting the practicality of standard data-driven models. Existing cross-city solutions primarily rely on data-space alignment to match traffic distributions between data-rich and data-scarce cities, but such alignment is often brittle under real-world distribution shifts, measurement noise, and city-specific bias. To address these limitations, we propose a Conditional Diffusion-based Parameter Generation (CDPG) framework. Instead of aligning raw data, CDPG learns the conditional distribution of high-performing predictor parameters for source cities and directly generates node-specific model parameters for the target city. To support reliable adaptation for network traffic prediction, CDPG conditions parameter generation on multi-modal spatio-temporal context. Specifically, CDPG captures spatial environment cues by combining satellite imagery with large language model (LLM)-derived semantic descriptions, and summarizes complex temporal behaviors via unsupervised pattern clustering. These contextual components guide CDPG in producing parameters tailored to each base station’s operational context, thereby improving robustness under cross-city and cross-region shifts. Extensive experiments on three large-scale datasets show that CDPG consistently outperforms state-of-the-art cross-city few-shot baselines by up to 9.72% on standard forecasting metrics and remains robust for few-shot multi-step prediction. The code is available at this https URL.

[LG-58] Reinforcement-Aligned Probabilistic Weather Forecasting with Discrete Tokens

链接: https://arxiv.org/abs/2611.25570
作者: Jinhan Li,Zhisheng Chen,Yuxuan Li,Yuan Gao,Fan Xu,Hao Wu,Jinlong Du,Kun Wang
类目: Machine Learning (cs.LG)
*备注: 28 pages, 11 figures, 15 tables. Code: this https URL

点击查看摘要

Abstract:Global medium-range weather forecasting must not only generate ensembles efficiently to characterize uncertainty in future states, but also serve downstream applications such as extreme-event warnings and cross-sphere coupled simulations. We introduce WeatherCraft, a framework that unifies probabilistic generation and reinforcement learning alignment through discrete weather tokens, bringing the GPT-style pretraining–post-training paradigm to global medium-range weather forecasting. A multiscale tokenizer represents atmospheric states as a coarse-to-fine token hierarchy, while a Transformer models causal dependencies across time and scale and predicts tokens within each scale in parallel. Direct sampling generates diverse ensembles, and exact token log-probabilities enable Group Relative Policy Optimization (GRPO) to maximize downstream task rewards without differentiating through evaluators or coupled models. On the ERA5 benchmark, WeatherCraft’s eight-member ensemble mean reduces day-15 normalized RMSE by 17.6% relative to the strongest evaluated deterministic baseline and, with KV caching, achieves an approximately 18\times inference speedup over GenCast. Direct fair-CRPS adaptation improves ensemble probabilistic skill, while reward-based post-training improves forecasts of persistent land temperature extremes and coupled ocean heat extremes. This work establishes discrete autoregressive modeling as a general foundation for scalable, task-alignable Earth system foundation models.

[LG-59] Is a Description Worth a Thousand Demonstrations? Specificity and Reliability of Task Descriptions in In-Context Learning

链接: https://arxiv.org/abs/2611.25560
作者: Bruce Nguyen,Bei Jiang,Linglong Kong
类目: Machine Learning (cs.LG)
*备注: 39 pages, 6 figures, 17 tables

点击查看摘要

Abstract:A prompt for in-context learning (ICL) can describe a task in addition to, or instead of, demonstrating it through examples. Yet how task descriptions work remains poorly understood theoretically. Through the Bayesian view of ICL, we model descriptions as informative priors characterized by their specificity and reliability. We adapt the effective sample size (ESS) from Bayesian statistics, quantifying a description’s predictive value as an equivalent number of demonstrations. For in-context linear regression, we show that a reliable description’s ESS grows with its specificity and with the task’s dimension. A description that may be irrelevant, however, gains little from its specificity unless demonstrations are given. In practice, do in-context learners use descriptions as a Bayes-optimal predictor would? Our experiments show that meta-trained Transformers come close to the Bayes-optimal ESS when tested on their training distribution. Pretrained language models, however, make too little use of highly specific descriptions when answering without reasoning; most of them closely match the Bayes-optimal predictor only when reasoning is enabled. These results extend the Bayesian analysis of ICL from demonstrations to descriptions.

[LG-60] When and why can integral constraints improve the accuracy of physics-informed neural networks?

链接: https://arxiv.org/abs/2611.25515
作者: Myeong-Su Lee
类目: Numerical Analysis (math.NA); Machine Learning (cs.LG); Analysis of PDEs (math.AP)
*备注:

点击查看摘要

Abstract:Adding constraints obtained by integrating governing equations to the loss has been reported to improve the accuracy of physics-informed neural networks (PINNs) in several applications, but mathematical understanding of these benefits remains limited. For several partial differential equations (PDEs), we derive stability estimates for losses with and without integral constraints. The comparison reveals that in certain parameter regimes, the standard loss has very weak control over specific error components, and that the integral constraints restore uniform control of these components. This explains when and why the constraints improve accuracy. Finally, numerical experiments support this interpretation. In parameter regimes where the standard loss is predicted to have weak control over those components, the two losses indeed show clear differences in accuracy. Conversely, in regimes where the standard loss retains sufficient control, both losses yield similar accuracy.

[LG-61] Sequential Probability Assignment against Smoothed Adversaries with Unknown Base Measure NEURIPS2026

链接: https://arxiv.org/abs/2611.25514
作者: Ziyi Liu,Daniel M. Roy
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: NeurIPS 2026

点击查看摘要

Abstract:Smoothed online learning has recently been studied as a way to bypass hardness results for the fully adversarial setting, which can be overly pessimistic. In this framework, the adversary is constrained to generate contexts from distributions having a bounded density with respect to some fixed base measure \mu . Prior work makes the strong assumption that \mu is known to the learner, with notable exceptions including Block et al. (2024) and Blanchard (2025). In this paper, we study sequential probability assignment (a.k.a., online learning with log loss) with smooth, well-specified data in the more general setting where \mu is \emphunknown, without the Lipschitz condition imposed in the work of Block et al. (2024) and Blanchard (2025). We prove regret upper bounds against both adaptive and oblivious smoothed adversaries: the adaptive bound is controlled by empirical \ell_\infty Hellinger entropy, while the oblivious bound improves this dependence to empirical \ell_2 Hellinger entropy, the notion that has been shown to characterize the complexity of learning with i.i.d. data (Bilodeau et al., 2023). Both bounds are obtained through algorithms based on truncated empirical Hellinger covers. We complement these results with a general lower bound, showing that our upper bounds are essentially tight for some natural classes, which implies a separation in the difficulty of smoothed online learning between regimes where \mu is known and where it is unknown.

[LG-62] Dimension-Free Buffer Bounds for Compressed-Momentum Error Feedback

链接: https://arxiv.org/abs/2611.25463
作者: Alireza Moayedikia
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: Submitted to a journal and details will be confirmed once paper accepted

点击查看摘要

Abstract:Analyses of error feedback bound the memory buffer through a compressor contraction constant – r/p for the best rank- r approximation of a p -dimensional spectrum – which survives into the rate. We show it need not appear at all. When the compressed momentum is re-injected into the pre-compression accumulator, the previous compressed momentum \beta_1 M_t-1 is an admissible competitor in the projection, and the buffer obeys |E_t|_F \le (1-\beta_1)|G_t|_F + \gamma|E_t-1|_F , free of the dimensions m,n and of the rank r . The argument uses only that the compressor is a metric projection onto a closed cone, so it covers top- k selection as well as truncated SVD, holds for every selection from the projection set with no condition on singular-value crossings, and in its trajectory-tracking form assumes nothing about the gradients. In the two-stage template of the standard analyses it replaces a Stage-2 constant of order p^2/r^2 by one of order 1 , against \Theta(p/r) for the sharpest contraction-based bounds. Transferring the full rate costs a leakage term, which we bound by the step-weighted gradient mass \sum_u\le t\eta_u|G_u|_F and show tight up to a factor \sqrt2 ; the price is an additive O(\log^2 T) floor, not a slower rate. On BERT fine-tuning at rank ratio 32/768 the feedback decay costs under 0.008 validation F_1 .

[LG-63] FamilySpace: From Emergent Chemical Semantics to Controllable Crystal Generation

链接: https://arxiv.org/abs/2611.25452
作者: Haowei Hua,Irwin King,Wanyu Lin
类目: Machine Learning (cs.LG); Materials Science (cond-mat.mtrl-sci)
*备注:

点击查看摘要

Abstract:Latent crystal generative models typically combine a variational autoencoder (VAE) for learning atom-wise representations with a flow model that transports a simple source distribution to the learned latent distribution. However, previous studies have mainly focused on the quality of generated crystals, leaving the structure and generative utility of the atom-wise latent space largely unexplored. In this work, we show that even a vanilla crystal VAE implicitly organizes atom-wise latent representations according to elemental families, with different families occupying largely separated regions of the latent space. We term this phenomenon emergent chemical semantics. This structure suggests that sampling from a family-specific latent region can bias the decoder toward atoms from the corresponding elemental family. Building on this insight, we propose FamilySpace, a framework for controlling the elemental-family composition of generated crystals. FamilySpace assigns each elemental family a distinct source component, forming a family-indexed multi-source distribution, and trains a flow model using source-target pairs matched by family. In inference, a user-specified family template selects the source component for each atom-wise representation, enabling direct control over the family composition of generated crystals. Experiments on MP-20 and MPTS-52 show that FamilySpace achieves competitive crystal-generation performance while matching prescribed elemental-family compositions. Moreover, selected family templates increase the fraction of generated crystals predicted to be stable or metastable, demonstrating that latent-space composition control can support targeted exploration of promising chemical regions without additional optimization techniques such as reinforcement learning. Code is available at this https URL

[LG-64] Inverse ECG Reconstruction via a Preconditioned Variational Quantum Linear Solver

链接: https://arxiv.org/abs/2611.25450
作者: Xizhuo Zhang,Chao Lu,Bing Yao
类目: Machine Learning (cs.LG)
*备注: 51 pages, 8 figures, 3 tables

点击查看摘要

Abstract:Inverse electrocardiography (ECG) reconstructs heart-surface potentials from body-surface measurements, enabling noninvasive mapping of cardiac electrical activity. However, the reconstruction requires solving dense and ill-conditioned linear systems, whose computational cost grows rapidly with the refined heart mesh. Quantum computing offers new opportunities for solving such systems. This paper presents a hybrid quantum-classical framework for inverse ECG reconstruction, which, to the best of our knowledge, is the first to apply a quantum linear solver to this problem. We first propose a Laplacian-regularized pre-encoding procedure to reduce the dimension of the system while improving its conditioning. Second, the resulting reduced system is encoded and solved using a coherent variational quantum linear solver (VQLS). Finally, we provide a rigorous theoretical proof establishing that the proposed framework preserves the exact solution of the original reconstruction problem throughout all stages of the computational process. Experimental results on a realistic heart-torso geometry demonstrate that our approach holds significant ability for accurate inverse ECG reconstruction. These findings highlight the potential of quantum solvers for noninvasive cardiac imaging, providing a foundation for further research into quantum-enhanced cardiac assessment and other healthcare applications.

[LG-65] Interference Bounds and Merge-Aware Training Dynamics for Composable Fine-Tuned Models

链接: https://arxiv.org/abs/2611.25430
作者: Alireza Moayedikia
类目: Machine Learning (cs.LG); Optimization and Control (math.OC)
*备注: Submitted to a journal. The journal info will be confirmed once the paper is accepted

点击查看摘要

Abstract:Combining fine-tuned models by composing their parameter updates has become a standard alternative to multi-task training. We formalize composability as a property of a collection of task vectors and prove that the excess loss of the merged model is controlled by an interference functional of the cross-task curvature overlap and the signed alignment of the updates. In the quadratic regime this functional equals the excess loss, giving a criterion that is necessary as well as sufficient and that measures the angle between task vectors in the curvature metric rather than the Euclidean one. Fine-tuning under a curvature-guided regularizer – constructed from curvature statistics of the kind memory-efficient optimizers maintain – keeps displacement in the base curvature metric below a ceiling set by the penalty coefficient at every iterate, at the standard O(1/\sqrtT) non-convex rate. Under a single Loewner comparison between the base statistic and the task curvatures, this control converts into a ceiling on the interference functional, inversely proportional to the penalty coefficient, so that the curvature term of the bound is controlled during training. A first-order term, present because regularized solutions are not stationary for the task loss, is not reduced in the same way. Experiments on an exact non-quadratic model confirm the budget and the ceiling, show the first-order term dominating for strong penalties, and find the curvature-guided penalty slightly better than an isotropic one. In a three-seed transformer study the functional is rank-correlated with merging degradation, but no more strongly than the Euclidean displacement norm.

[LG-66] Coordinate Basis Linear Bandits with Sublinear Regret over Low-Rank Clusters

链接: https://arxiv.org/abs/2611.25424
作者: Hantao Yang,Hong Xie,Defu Lian
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:We study the problem of computational efficiency for linear bandits in high-dimensional settings with a finite arm set. In linear bandits, the increase in the dimension d of the feature vectors leads to growing computational costs of O(d^2) at each round of update. Traditional low-rank approximation methods such as SOFUL reduce computation via fixed-size matrix sketching, yet run the risk of incurring vacuous linear regret when the spectral tail of the data is heavy and the approximation rank is inadequately selected. To guarantee regret convergence and effectively reduce computational costs, we introduce a clustering mechanism and propose the Coordinate Basis Linear Bandit (CBL) algorithm. Our method preserves the full covariance information in each cluster by maintaining an exact orthonormal basis of the cluster’s column space together with a low-dimensional coordinate representation, which guarantees robust sublinear regret without spectral-tail vulnerabilities. It performs cluster switching by assigning a sentinel for each cluster, and reduces per-round update computation to O(\bar r^2) in the cluster’s effective dimension \bar r=\min(r,l) , where r is the intrinsic rank of the data and ld is a tunable rank bound.

[LG-67] From Sequence Uncertainty to Phenotype Hypotheses: Uncertainty-Conditioned Inference over Biomedical Knowledge Graphs

链接: https://arxiv.org/abs/2611.25422
作者: Huixin Zhan,Jason H. Moore
类目: Machine Learning (cs.LG); Genomics (q-bio.GN)
*备注:

点击查看摘要

Abstract:Genomic foundation models provide variant-level evidence, but connecting that evidence to phenotype hypotheses requires accounting for uncertainty throughout structured reasoning. We introduce an uncertainty-conditioned path-energy framework that converts sequence predictions and calibrated predictive uncertainty into explicit distributions over graph-supported phenotype hypotheses, without patient-level phenotype supervision. An interpretable learned energy scores admissible gene–disease–phenotype routes, query uncertainty controls the target temperature, and path probabilities are aggregated into phenotype hypotheses. This formulation makes evidence quality an explicit determinant of hypothesis breadth while retaining traceable multi-path support. Exact normalization over enumerable candidate sets provides a reference for evaluating a Hastings-corrected learned revision operator. On AlzKB and Hetionet, phenotype-level entropy tracks upstream uncertainty, with Spearman correlations of 0.951 and 0.878, respectively; ablations identify the contribution of uncertainty conditioning. In masked-edge recovery, the model exceeds chance and degree-matched probability-permutation controls. Training the energy under the uncertainty-dependent temperature yields higher held-out recall@10 than training under a constant temperature of comparable magnitude. Under gene hold-out, recall@10 exceeds that of a trained DistMult embedding on 24 of 30 graph masks (0.183 versus 0.123); this difference is attributable mainly to three genes with many variants and is not established for all genes. These results establish an uncertainty-aware interface from genomic predictions to graph-supported phenotype hypotheses that combines interpretable probability allocation with held-out association recovery of explicitly stated scope.

[LG-68] Antom 3-in-1 Transformer: A Payment Foundation Model

链接: https://arxiv.org/abs/2611.25403
作者: Weiyu Han,Haoyu Jiang,Jie Sheng,Youxi Pan,Xianglong Liu,Zeqi Ren,Dongquan Yi,Yinhao Liu,Ruobing Zuo,Ze Zhao,Huilin He,Nairong Zhang,Xinyi He,Xiao Tan,Jun Wang,Pan Fang,Wei Zeng,Jianan Su,Kaixin Shen,Jingtong Wu,Hao Li,Shuai Chen,Jiangming Yang,Tianyi Zhang
类目: Machine Learning (cs.LG)
*备注: Technical report, 18 pages

点击查看摘要

Abstract:We report the development of Antom 3-in-1 Transformer, a payment foundation model that integrates sequence, tabular, and graph representations for payment risk detection. The model contains approximately 7B parameters, with approximately 1.7B activated per task route. Antom 3-in-1 Transformer is pretrained to learn reusable behavioral and relational representations, primarily on Ant International’s internal payment data. In a cross-business transfer experiment, pretraining improves the target business metric by 34.5% over the same architecture trained from scratch. With only 25% of the training labels, the pretrained model outperforms the full-label scratch baseline in Average Precision (AP) and Recall@0.5% on different merchant domains. In a four-month online experiment on Ant International’s billion-scale payment platform, the model reduced the overall loss rate by 37.88% relative to XGB-Online.

[LG-69] Protein Functional Semantic Alignment via Representation Rectification

链接: https://arxiv.org/abs/2611.25396
作者: Chuyang Zhou,Bingxin Zhou,Andi Han,Chang Xu
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Protein language models (PLMs) are commonly pretrained with likelihood-based objectives such as masked language modeling or denoising. These objectives capture sequence-level regularities, but can produce representations in which functional signals are entangled with non-functional variation (e.g., sequence identity, compositional bias, and frequent local motifs), creating a readout bottleneck for downstream function prediction. We introduce Functional Semantic Alignment via Representation Rectification (SAVR), a contrastive post-training method using Pfam family co-membership as a homology-guided discrete semantic signal to reshape PLM representations. SAVR reduces directional dominance and redistributes variance across embedding dimensions, decreasing variance concentration, and increasing effective dimensionality of the embedding covariance spectrum. Without observing any task-specific labels during post-training, SAVR achieves state-of-the-art results on Enzyme Commission (EC) benchmarks and competitive results on Gene Ontology (GO) benchmarks under frozen-backbone evaluation. These results suggest that homology-guided contrastive post-training can rectify PLM representation geometry and improve sequence-based protein function prediction.

[LG-70] uringSpace: Grounding Spatial Intelligence in Action with Interactive CAPTCHAs

链接: https://arxiv.org/abs/2611.25387
作者: Xuwei Ding,Skylar Zhai,Linxin Song,Jiarui Feng,Dongqi Fu,Huajun Xi,Junnan Li,Jingcheng Liang,Jieyu Zhao,Yinglong Xia,Hong Li,Jian Kang
类目: Machine Learning (cs.LG)
*备注: 45 pages. Project page: this https URL

点击查看摘要

Abstract:Humans act in the visual world by inferring the 3D layout behind a 2D view, predicting how an action changes it, and correcting themselves after seeing the result. Can multimodal large language models (MLLMs) do the same? CAPTCHAs offer a natural test, since humans solve them through this very loop, clicking and dragging on the screen and adjusting after each result. Yet existing CAPTCHA benchmarks are largely solved and rarely spatial, while spatial benchmarks rarely require action. We therefore present TuringSpace, a synthetic web environment of 125 interactive CAPTCHAs that span five spatial abilities and must be solved through clicks and drags on the screen. Every task is a program that generates new instances with a checker and an oracle, so the tasks also form a scalable training gym. Humans solve 94.1% of the tasks, whereas the best model, GPT-6 Astra, still fails 25.9% of them, and the best open-weight model, Kimi K3, fails 68.0%. Our further analysis shows that most failures come from misreading the screenshot, and that model reasoning rarely corrects a wrong belief but hallucinates evidence to defend it. To test whether such spatial abilities can be learned, we fine-tune Qwen3.5-9B on this gym. Its success rate rises from 1.6% to 33.1%, above Kimi K3, and it improves on all seven out-of-distribution benchmarks, including a 9.9-point gain on the real-world indoor videos of VSI-Bench. These results show that grounding spatial intelligence in screen actions exposes what MLLMs lack and offers a scalable way to teach it, with gains that transfer to real-world scenes. We will release TuringSpace to support further study.

[LG-71] Drifting Discriminating and Directioning for Action Coordination in Higher-dimensional Action Space

链接: https://arxiv.org/abs/2611.25385
作者: Taisuke Kobayashi
类目: Machine Learning (cs.LG)
*备注: 15 pages, 6 figures

点击查看摘要

Abstract:This paper proposes a lightweight policy model to mitigate the difficulty of handling higher-dimensional action spaces in reinforcement learning (RL), alongside an algorithm to effectively train it. To address this, we focus on drifting models, which can implicitly represent complex distributions and infer actions in a single forward pass. However, while drifting models can be straightforwardly trained via backpropagation of the Q-gradient, this alone cannot be expected to appropriately capture the desired action coordination. Therefore, this paper constructs a three-step learning algorithm to encourage the action coordination. First, the source distribution of the drifting model is co-optimized as a diagonal policy model. This conditions the drifting model to attach the coordination, which cannot be expressed by a diagonal distribution, as residuals. Second, a discriminator model before and after drifting is introduced with adversarial training for a regularization term. This actively promotes the extraction of the coordination that fundamentally separates the two distributions. Third, best-of-N sampling is applied for action selection during rollouts in order to give directionality toward/beyond the regions where the coordination is insufficiently extracted/sufficiently discovered. The proposed method consistently improves the sample efficiency and the final control performance over the conventional diagonal distribution modeling across three locomotion tasks with high-dimensional action spaces. Furthermore, the analysis indicates that this improvement is achieved by reducing variance in certain action dimensions, effectively coupling them with other actions as action coordination.

[LG-72] Fundamentals for Binary Classification Under Label Noise

链接: https://arxiv.org/abs/2611.25372
作者: Dimitrios I. Diochnos
类目: Computational Complexity (cs.CC); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 59 pages, 6 figures, 3 tables. Preprint of a chapter published in the book Mathematics and Artificial Intelligence (Springer)

点击查看摘要

Abstract:This chapter surveys fundamental results for binary classification under label noise. Binary classification under label noise involves training models on datasets where class labels are partially incorrect, leading to degraded performance and misclassification. We distinguish poisoning attacks from noise processes and formalize random classification noise, Massart noise, and malicious classification noise within the probably approximately correct (PAC) framework. These models capture different adversarial strengths and data-generation realities. We then highlight some core ideas and representative results. In particular, we explain why random classification noise preserves the expected ordering of hypotheses and thus supports learning using the empirical risk minimization (ERM) principle, how sample complexity depends on the noise rate, and why learning algorithms often require an upper bound on the corruption level, as well as how such a bound may be estimated from data in the case of random classification noise. We also discuss computational aspects of ERM under noise and present a canonical negative result technique – the method of induced distributions – within the malicious noise model where the adversary has more power and may affect not only the label of an example but also the instance. We conclude with pointers to related work and a taxonomy that organizes classical label-noise models using the NCAR/NAR/NNAR lens.

[LG-73] Class-Conditional Conformal Triage for ICAO Face-Image Compliance under Class Imbalance

链接: https://arxiv.org/abs/2611.25342
作者: Mustafa Vedat Yildirim,Inna Skarga-Bandurova
类目: Machine Learning (cs.LG)
*备注: 12 pages, 8 figures, 4 tables

点击查看摘要

Abstract:Automated ICAO/ISO face-image compliance analysers provide criterion-level pass/fail decisions, yet these verdicts do not quantify uncertainty when deciding whether to accept, reject, or retake a photograph. We investigate conformal prediction as a triage layer for BioGaze, evaluating 14 compliance criteria on the synthetic TONO benchmark. Without retraining the analyser, we fit per-criterion logistic classifiers and compare marginal and label-conditional (Mondrian) conformal prediction with conventional abstention rules using subject-disjoint splits. Marginal conformal prediction achieved approximately 0.90 overall coverage but substantially under-covered violations, with a ten-partition mean of 0.37 for one criterion. The corresponding Mondrian means ranged from 0.89 to 0.92 across criteria. However, most of the reduction in criterion-level false positives came from the logistic layer (0.145 to 0.048), while tuned abstention rules achieved lower false-positive rates than Mondrian at comparable retake rates. Bonferroni correction for joint coverage across 14 criteria required retaking nearly all images. These findings show that class-conditional conformal prediction can achieve near-target empirical coverage for minority violation classes and reveal trade-offs between statistical coverage, classification performance, and retake requirements in multi-criterion face-image compliance assessment.

[LG-74] Now You See Me Now You Dont: A Framework To Detect and Reduce Proxy Capacity

链接: https://arxiv.org/abs/2611.25327
作者: Nirek Sharma
类目: Machine Learning (cs.LG)
*备注: 22 pages, 8 figures, 4 tables

点击查看摘要

Abstract:In regulated domains such as lending, the two most salient fairness concepts are disparate impact and disparate treatment: disparate impact is generally defined by differential rates of adverse outcomes between demographic groups, while disparate treatment is generally defined by the use of demographic information or proxies to issue decisions. There already exist established methods to quantitatively detect and mitigate disparate impact; however, there has been less research on doing the same for disparate treatment. From a practitioner’s perspective, the manner in which a conditional independence test is formulated and operationalized to identify proxies is nuanced and underexplored. As Machine Learning (ML) systems increase in scope and application in regulated domains, practitioners need a defensible framework to identify proxies at scale. Current disparate treatment frameworks typically rely upon some combination of quantitative and qualitative review of models’ inputs to identify proxies. Features that are identified as proxies are then typically removed from the model in practice, and predictive performance degrades in favor of fairness through unawareness. Here, we propose a two-part framework that reduces the disparate treatment risk of a feature and keeps the predictive performance that removal of the feature loses. First, we propose a quantitative test for proxy capacity that identifies whether some of a feature’s values reveal membership in a particular demographic group. Second, we propose a pre-processing methodology that adjusts those proxy features to reduce their proxy capacity. We show empirically on synthetic and public datasets that this framework identifies features that proxy for demographic information and reduces their proxy capacity while the adjusted feature retains task-relevant signal (i.e. improves predictive performance versus removal).

[LG-75] RE-CAM: Readout-Enhanced Content-Addressable Memory for Quantitative Similarity Search

链接: https://arxiv.org/abs/2611.25326
作者: Yi-Chun Liao,Dongyoung Lee,Jiazheng Chen,Tergel Molom-Ochir,Tomohisa Kawakami,Tania Roy,Ik-Joon Chang,X. Sharon Hu,Hai “Helen” Li,Yiran Chen
类目: Hardware Architecture (cs.AR); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Similarity search is essential to many applications but incurs massive data movement when executed on conventional architectures. Content-addressable memory (CAM) is a promising hardware substrate, as it allows similarity search directly in the CAM array, minimizing data movement and offering both low latency and high energy efficiency. However, conventional CAM array readout produces only binary match results, discarding quantitative similarity information that could be further exploited. Extracting this information requires interpreting the CAM array output voltage that evolves during a CAM search, making the sensing timestamp a crucial design variable alongside the reference voltage and decoding parameter. In this work, we propose RE-CAM, a readout enhancement framework for CAM that is agnostic to memory technology and application, enabling CAM array readout to approximate linear and nonlinear functions with one CAM search. By reusing the comparators to sense the time-varying output voltage multiple times, jointly optimizing the three design variables under system constraints and hardware nonidealities, and co-designing the optimization algorithm and reconfigurable hardware, RE-CAM achieves accurate approximation with low energy and area overhead without introducing a readout throughput bottleneck. Across linear, piecewise-linear, sigmoid, and exponential functions, RE-CAM reduces normalized root mean square error by 10.89% to 24.34% for general approximation and 29.43% to 46.73% when exploiting dataset distribution, compared to a shared uniform-reference ADC baseline with the same comparison budget. Relative to the same baseline, RE-CAM also achieves 8\times higher search throughput while reducing energy per search and area by 47% and 39%, respectively.

[LG-76] Improving Imitation Policies for Multi-Phase Tasks through Residual Learning Allocation

链接: https://arxiv.org/abs/2611.25320
作者: Clinton Enwerem,Saleh Kemal,John S. Baras,Calin Belta
类目: Machine Learning (cs.LG); Robotics (cs.RO); Systems and Control (eess.SY)
*备注: 21 pages, 5 figures, 6 tables

点击查看摘要

Abstract:Residual reinforcement learning corrects an imitation-trained base policy using a limited budget of environment interaction. In a task with several phases, this budget must be divided among the phases, and the phase in which the base policy fails is not necessarily the phase in which a correction most improves task completion. We formalize this distinction through the completion gain: the change in full-task completion probability when, during a single phase, a state-feedback controller replaces selected components of the base policy’s action. We decompose the completion gain into the probability of reaching the phase and the conditional improvement in completion from the start of the phase, and we estimate it from paired episodes. The positive estimates determine which action components the residual policy corrects and how training episodes are distributed across phases. In stochastic point-mass docking experiments, most failures occurred in the final docking phase, whereas the largest completion gain occurred in the preceding channel phase; gain-based allocation improved completion over allocation proportional to failure counts and, in a variant with a longer approach, over uniform allocation at half of the budget. In simulated bimanual robotic assembly, the completion-gain estimates were positive only for insertion and release, although few failures occurred during release; in a second assembly configuration, a residual policy trained in the release phase improved completion from held-out states at the start of release by 16.9%.

[LG-77] Prescribed Bayesian Efficiency under Lifetime Adjustment Costs

链接: https://arxiv.org/abs/2611.25311
作者: Viet Bac Nguyen
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 47 pages

点击查看摘要

Abstract:A Bayesian learner may retain an issued action while updating its posterior, trading predictive loss against adjustment charges. We assume population curvature, boundary control, an order- 1/n lower bound on expected posterior uncertainty about the optimal action, and an observable unbiased gradient witness whose squared norms are uniformly integrable. Sample losses may be unbounded or nonconvex. Cumulative expected excess measures the loss increase relative to current predictive Bayes actions. For bounded nonnegative nondecreasing charges c with c(0)=0 and nondecreasing profiles G: \mathbbN \to [1,\infty) with G(J)=o(J) , we characterize joint attainability of cumulative expected excess O(G(J)) at horizons T=2^J and finite expected lifetime charge. Here J indexes logarithmic time. The criterion is finiteness of a prefix-constrained allocation of tracking-error allowances across action-distance scales, equivalently its decreasing-weight dual. Adaptive timing and private randomization do not enlarge the attainable region. For each feasible pair, a horizon-independent deterministic checkpoint policy attains the profile with lifetime-cost moments of every positive order. For excess O((\log T)^r) , 0r1 , every policy needs \Omega((\log T)^2-r) revisions in expectation, and a charge-independent schedule achieves a matching pathwise upper bound for every feasible charge. We verify the assumptions for logistic decision models with nondegenerate covariates, finite second moments and a locally positive smooth prior density, including fixed-ridge models with unbounded parameter priors.

[LG-78] SplitReAlign: Partition-Aware Representation Realignment for Split Federated Unlearning

链接: https://arxiv.org/abs/2611.25301
作者: Jia Chen,Yecheng Lv,Xinrong Hu,Jun Bai
类目: Machine Learning (cs.LG); Cryptography and Security (cs.CR); Distributed, Parallel, and Cluster Computing (cs.DC)
*备注: 5 pages, 2 figures, 2 tables

点击查看摘要

Abstract:Split federated learning (SFL) distributes neural networks across clients and a server, coupling class-specific knowledge across their representations. This coupling complicates split federated unlearning (SFU), where class-wise forgetting must remove target knowledge while preserving retained utility under heterogeneous representation spaces and asymmetric update constraints. We propose SplitReAlign, a partition-aware representation realignment framework for class-wise SFU. It constructs endpoint-specific discriminative subspaces coordinated through a shared semantic anchor, selectively realigning forgotten representations while penalizing changes outside the active subspaces. To accommodate asymmetric update constraints, SplitReAlign leverages gradients through frozen servers for Client-only updates and fixed client-feature statistics for Server-only guidance, while Joint updates optimize both partitions. Experiments on CIFAR-10 and CIFAR-100 under heterogeneous client distributions demonstrate that SplitReAlign achieves the lowest average gap to retraining among all compared methods across six settings. Notably, Server-only achieves the closest agreement with retraining on both datasets, highlighting the effectiveness of partition-aware guidance under restricted updates.

[LG-79] Federated Learning of Car Aerodynamics Surrogates with Spectral Optimizers

链接: https://arxiv.org/abs/2611.25300
作者: Arman Bolatov,Nikolay Spitsyn,Daniil Medyakov,Vladimir Aletov,Dmitry Bylinkin,Martin Takáč,Aleksandr Beznosikov
类目: Machine Learning (cs.LG)
*备注: 24 pages, 9 figures, 12 tables

点击查看摘要

Abstract:Neural surrogates predict the aerodynamic forces on a car in seconds, compared with hours for a CFD simulation, but training them requires thousands of CFD simulations of different car designs, which car manufacturers keep proprietary. Federated learning allows several manufacturers to train a shared surrogate without exchanging data. We study which federated optimizer brings such a surrogate closest to one trained on pooled data. On DrivAerNet++, we train a Transolver surrogate across six clients that each hold cars of a single body type, and we compare three spectral optimizers, which orthogonalize the update of each weight matrix (FedMuon and two methods we propose, FedMARS-M and FedOLion), with three non-spectral ones (FedAvg, SCAFFOLD-Adam and FedLion), all tuned with the same protocol. FedMuon and FedMARS-M reach a drag error within 2% of centralized training with Muon, using two to four local epochs per round. The non-spectral methods perform best with sixteen local epochs, the largest number we tested, and still have a 17 to 37% higher drag error than FedMARS-M. We relate this gap to the low stable rank of the weight gradients, for which an orthogonalized step yields about twice the first-order decrease of a gradient step of the same spectral norm. Code: this https URL

[LG-80] Cross-Question Reliable Reinforcement Learning

链接: https://arxiv.org/abs/2611.25292
作者: Hector G. Rodriguez,Marcus Rohrbach
类目: Machine Learning (cs.LG)
*备注: 27 pages. Code: this https URL

点击查看摘要

Abstract:AI systems operating near the limits of their capabilities can maximize utility while minimizing the risk of error by abstaining or deferring questions they are not confident about to a human or a stronger model. Recently, large language models (LLMs) have been post-trained to estimate confidence in their own answers. To achieve this, methods reinforce confidence scores based on the correctness signal to each question-answer pair individually, with rewards that encourage assigning confidence 1 to correct answers and 0 to incorrect answers. However, estimating confidence scores is useful in that it enables selecting only the most confident answers, to maximize the share of solved questions while accommodating the user’s risk tolerance. Assigning confidence 1 to correct answers and 0 to incorrect ones is theoretically optimal, but unattainable in practice and detrimental for selective prediction with a non-zero risk tolerance: confidences pushed to 0 and 1 carry no information about how answers compare to one another. We therefore introduce CARE (Cross-question Abstention Reward Estimation), which computes rewards over a set of question-answer pairs: At each of several risk tolerances, the confidences across the set induce an abstention threshold, and each confidence is rewarded for lying on the correct side of it. Since the threshold itself comes from the set, every verbalized confidence is reinforced relative to other questions’ confidence, not against its own binary correctness label. On math, CARE increases the coverage at 5% risk (C@5) across MATH-500, GSM8K, and Big-Math-Digits from 0.8% to 42.4%, compared to existing pointwise correctness-matching rewards. When trained on multi-hop Wikipedia QA from HotpotQA, CARE improves C@5 in out-of-distribution general-knowledge, reasoning, and math benchmarks (including CommonsenseQA, GPQA, SimpleQA, and TriviaQA) from 12.3% to 23.6%.

[LG-81] DS3M: Deep structured state space models for nonlinear system identification on modern hardware

链接: https://arxiv.org/abs/2611.25274
作者: Marco Roschkowski,Hannes Gernandt
类目: Machine Learning (cs.LG); Systems and Control (eess.SY); Optimization and Control (math.OC)
*备注:

点击查看摘要

Abstract:Neural state-space identification learns dynamical models from measured inputs and outputs. SUBNET combines this task with an encoder that estimates initial states from short measurement histories. Its continuous-time version requires many successive neural-network evaluations during numerical integration, resulting in slow training speed which is a major limitation for practical use cases. We propose deep structured state space models (DS3M), which combine compact nonlinear dynamical components in parallel and in series. Custom GPU kernels implemented with Triton reduce the launch and memory overhead of their sequential evaluation. We explain this construction, describe a port-Hamiltonian variant, and present various benchmark comparisons of accuracy and training time.

[LG-82] Scone: Symmetry-aware Neural Combinatorial Optimization for Graph Coloring

链接: https://arxiv.org/abs/2611.25273
作者: Paolo Pellizzoni,Melanie Weber
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Graph coloring, which entails assigning colors to nodes such that no adjacent nodes have the same color, is a fundamental combinatorial optimization problem. In general, however, finding a node coloring with no monochromatic edges given a fixed palette is NP-hard. This motivates learning-based methods that can learn to exploit recurring structure across instances to efficiently produce high-quality colorings. Crucially, the validity of a coloring is invariant to permutations of the colors, and we show that exploiting this symmetry reduces the size of the effective state space over which the model has to learn its policy. At the same time, the coloring algorithm must be able to break some symmetries, for example to assign different colors to two adjacent nodes belonging to the same automorphism orbit. In this paper, we introduce Scone, a model that uses a color-equivariant graph neural network to construct colorings sequentially, allowing information from previous assignments to propagate and break node symmetries. Scone achieves perfect solution quality with low inference times across several graph coloring tasks, including on sudoku_extreme and register allocations, and showcases generalization to graph families not encountered during training.

[LG-83] Learning to Verify: Signature-Guided Training for Adaptive Adversarial Detection

链接: https://arxiv.org/abs/2611.25271
作者: Arash Vashagh,Roozbeh Razavi-Far
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Adversarial example detectors often fail when attackers explicitly attempt to bypass them. We propose a dual-model approach that keeps the original classifier fixed and trains a separate verifier to assess its predictions. During training, we identify patterns of internal neural network activations that distinguish genuine examples of each class from adversarial examples targeting that class. These patterns, called activation signatures, guide the generation of challenging adversarial examples that mislead the classifier, challenge the verifier, and internally resemble the target class. The verifier learns to recognize these examples while preserving its performance on clean inputs through knowledge distillation. At inference, the verifier evaluates the classifier’s prediction without requiring activation signatures. We evaluate the approach on publicly available datasets against several adaptive attacks, including attacks unseen during training. The proposed verifier substantially outperforms directly attacked confidence-based and feature-based detectors under the evaluated L-infinity attacks. Ablation studies show that adaptive verifier training is the main source of this improvement, while signature guidance primarily benefits clean accuracy. Additional L2 evaluations reveal limitations in robustness across perturbation types.

[LG-84] Process-Knowledge-Guided Latent Diffusion for Generative Design of Porous Architectures: A Case Study in Aerosol Jet Printing

链接: https://arxiv.org/abs/2611.25264
作者: Ngoc Dang Khoa Nguyen,Hyunwoong Ko
类目: Machine Learning (cs.LG); Computational Engineering, Finance, and Science (cs.CE)
*备注: 12 pages, 6 figures, 3 tables (under review for NAMRC55 - 2027)

点击查看摘要

Abstract:Printed porous layouts interleave functional material with open channels, and the layout that uses the material best depends on how the process settings shape the deposit. Generative models propose layouts quickly, but a generator that does not receive the process setting samples from the same distribution at every setting, even when the process-dependent score ranks candidate layouts differently. This paper examines whether process knowledge can steer a latent diffusion generator toward layouts suited to a given process setting in a prescribed design domain. We formulate a process-dependent utilization score in which a process model sets the length over which the utility of printed material decays with distance from a channel. The generator combines a vector-quantized variational autoencoder (VQ-VAE) latent representation, a topology U-Net, a ControlNet domain branch, and a separately trained regressor that predicts the score from noised latents; the domain branch conditions each denoising step and the regressor gradient steers it. The method is applied to aerosol jet printing, where the atomizer and sheath gas flows enter the score through the flow-focusing relation. Across seven modeled flow settings, the lowest-scoring candidate layout of a domain differed between settings in 191 of 500 domains. In a sampling study, the mean directly computed score on three validation domains and one test domain was 0.642 with domain conditioning alone, 0.636 at a weight of 300, and 0.450 at 12,000 (112 scores per weight; lower is better). The results support process-knowledge-guided generation as an approach to process-specific porous design under the stated assumptions.

[LG-85] Moving Generality from Model to Modeler: LLM Agents for PHM on a Deployment-Style Benchmark

链接: https://arxiv.org/abs/2611.25252
作者: Zhenling Chen,Zhiguo Zeng,Anne Barros
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Coding agents built on large language models (LLMs) now help develop many models published as general solutions. Should generality instead reside in the agent that builds a model for each new task? We examine this question in prognostics and health management (PHM), where every model must work on unseen equipment. Our deployment-style benchmark, NextUnit-Bench, covers 24 public datasets split into 58 folds that hold out units or operating conditions; each agent has two hours to deliver an executable predictor, then run on the hidden units. First, frontier agents outperform AutoML, which reaches less than half the best agent’s score, and faithful implementations of published methods on four datasets, of which only 6 of 30 beat the strongest agent’s own design; published accuracies, where reproducible, rest on the papers’ own conventions, and a single convention, capping the remaining life of test engines, can halve the reported error. Second, the agents’ advantage is a design made on each machine’s data; once made, a plain-text plan of a few hundred words carries most of the gap between two frontier agents, and given a weaker agent’s plan, the strongest agent falls to that agent’s level, often to identical predictions. Before touching the data, however, it matters little whose plan an agent carries out, and domain toolkits appear to help neither: what sets the strongest apart is how much it learns from the data. Third, such designs reach only implementers that can execute them: given the strongest agent’s designs, a locally deployable open-weight model delivers half as often, even when they are lightened, and scores no higher than without them. Generality, we argue, is not disappearing but moving, from the model to the modeler.

[LG-86] Stability Bands for Feature Attribution: The Core the Nucleolus and the Bargaining Set of Explained-Variance Games

链接: https://arxiv.org/abs/2611.25238
作者: Fernando Delbianco,Fernando Tohmé
类目: Computer Science and Game Theory (cs.GT); Machine Learning (cs.LG); Econometrics (econ.EM)
*备注: Presented at the XXV Jornadas Latinoamericanas de Teoría Económica (JOLATE 2026), San Luis, Argentina, October 29, 2026. 24 pages, 5 figures

点击查看摘要

Abstract:Shapley-value attributions are the default way of explaining machine-learning predictions in economics, but they report a single number per feature and say nothing about whether a coalition of features that jointly accounts for more than it receives could contest that allocation. We study feature attribution through the core, the nucleolus and the bargaining set of an explained-variance game, in which the worth of a set of features is the variance of the model’s predictions that those features generate. For every feature we report a stability band (its range of shares over the core, or over the least core when the core is empty), with the nucleolus as distinguished point and a certificate that the allocation admits no justified objection. Three results organise the approach. First, for linear models the game is exactly 2-additive, with pairwise Harsanyi dividends 2\beta_i\beta_j\sigma_ij , so multicollinearity is what creates coalitional structure. Second, in every 2-additive game the prenucleolus coincides with the Shapley value, the (least) core is point-symmetric around it, and the core is non-empty if and only if every cut of the interaction graph has non-negative weight. The Shapley value thus always lies in the prebargaining set, and stability adds information only through band widths and interactions of order three or more. Third, for convex games the bands have a closed form and any allocation can be certified with one minimum cut. In a hedonic price model, the usual mean-SHAP explanation admits a justified objection by the coalition Ratings, ROM, Battery. In a gradient-boosting forecaster of retail orders the exact game has an empty core and strong higher-order interactions; the nucleolus departs from the Shapley value by up to 0.19 variance shares, and sparse third-order surrogates recover it where pairwise ones cannot.

[LG-87] ϕTD: Distributional Reinforcement Learning using Characteristic Functions NEURIPS2026

链接: https://arxiv.org/abs/2611.25208
作者: Thomas Mousseau,Tyler Kastner,Davide Baldelli,Amir-massoud Farahmand
类目: Machine Learning (cs.LG)
*备注: Accepted at NeurIPS 2026. First two authors contributed equally

点击查看摘要

Abstract:Distributional Reinforcement Learning methods aim to learn the entire distribution of returns, yet current algorithmic approaches are limited to a narrow class of distributions. We propose \varphi TD, a method to overcome this limitation by computing the loss in the frequency domain. We demonstrate that this further unifies quantile and categorical approaches to distributional RL, and enables the learning of distributions that were previously impossible to capture, such as mixtures of continuous distributions. We empirically study \varphi TD, across synthetic benchmarks and large-scale deep RL environments and demonstrate that the flexibility obtained from wider parametric families can lead to improved performance.

[LG-88] From Self-Play RL to Constrained Control: Distilling DiffMPC Students with DAgger

链接: https://arxiv.org/abs/2611.25197
作者: Luca M. Hartmann,Parinaz Naghizadeh
类目: Robotics (cs.RO); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 18 pages, 10 main content pages, 3 figures, L4DC submission

点击查看摘要

Abstract:Self-play reinforcement learning (RL) produces strong driving policies, but these policies are hard to verify or interpret. Differentiable model predictive control (DiffMPC) can enforce explicit constraints, but it is usually trained by imitating logged data or with costly RL. To add constraints to an already trained policy, we keep a PPO teacher’s backbone and replace its action head with a DiffMPC that enforces road boundary half-planes and agent ellipses. We then distill the student with DAgger in the teacher’s self-play simulator. Over 9 seeds, the constrained student of a smaller teacher matches its teacher on nuPlan and beats it on CARLA and on out-of-distribution Waymo scenes. For a stronger teacher, the student has fewer failures on nuPlan and Waymo, while staying level on CARLA. Its distillation takes no more than 30% of the teacher’s training time.

[LG-89] PDE-State Conditioning for Learned Optimization Using Physics-Informed Neural Networks

链接: https://arxiv.org/abs/2611.25185
作者: Farzana Yasmin,Ricardo Vilalta
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Physics-informed neural network (PINN) optimization depends on global properties of the underlying PDE and how the physics residual evolves across the domain. Standard optimizers such as Adam use fixed gradient-based statistics, while learned PINN optimizers replace the fixed rule with a learned update built from local parameters and optimization-history features. However, none explicitly use PDE-level information. We study two coupled design questions: what PDE-level information should a learned optimizer observe, and how should that information influence its update? We introduce an 11-dimensional PDE state-capturing operator structure, residual geometry, and short-horizon update response, and use it to condition a structured meta-learned optimizer through bounded feature-wise modulation. Repeated design studies identify the compact residual-geometry state G2 and bounded FiLM as the strongest overall choices among those tested. The resulting PDE-State-conditioned optimizer improves all evaluation metrics (relative-L2, L1, solution MSE, and physics MSE) over Adam and the unconditioned learned optimizer on Linear Advection, KdV, and Burgers.

[LG-90] Separating Interpretability and Reconstruction in Sparse Autoencoders

链接: https://arxiv.org/abs/2611.25165
作者: Anh Tien Nguyen,Dimitris Samaras,Mahdi S. Hosseini
类目: Machine Learning (cs.LG)
*备注: Code: this https URL

点击查看摘要

Abstract:Sparse Autoencoders (SAEs) are effective tools for mechanistic interpretability by decomposing neural representations into sparse and interpretable features. However, a single sparse representation in traditional SAEs is responsible for both reconstruction fidelity and interpretability, which are not guaranteed to align. To separate these objectives, we propose DualSAE, a modular framework that augments standard SAEs with a plug-in Residual Branch. A standard SAE serves as the Sparse Branch, learning a sparse code that contains the dominant semantics of the input. The added Residual Branch models the non-semantic information omitted by the Sparse Branch. We ensure that the contributions of the two branches towards the overall reconstruction have minimal overlap through a decorrelation loss. We also adopt a two-stage training strategy that first stabilizes sparse representation learning before coupling with the Residual Branch. Across diverse encoder-SAE configurations on two vision datasets and two language corpora, DualSAE achieves mean relative improvements of approximately 29% in reconstruction fidelity and 34% in interpretability (measured by monosemanticity score). Moreover, with more interpretable representations, DualSAE yields consistent performance gains in the downstream applications of topic modeling, hypothesis generation, and LLM steering. Qualitative analysis further suggests that the Sparse Branch indeed preserves dominant semantic content, whereas the Residual Branch captures non-semantic information to improve reconstruction fidelity.

[LG-91] Bayesian κ-Explorer for Minimax Model Estimation in MDPs

链接: https://arxiv.org/abs/2611.25160
作者: Xihe Gu,Urbashi Mitra,Tara Javidi
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:We study active exploration for minimax transition estimation in unknown Markov decision processes (MDPs). Frequentist \kappa -Explorer~\citepgu2026kappa addresses this problem through a parameterized family of exploration objectives, but requires confidence-bound constructions and joint optimization over plausible transition kernels and occupancy measures, introducing a computational burden that limits its practical applicability. We develop Bayesian \kappa -Explorer, which simplifies this procedure by placing Dirichlet priors on the transition distributions. Under the posterior mean estimator, the posterior expected squared \ell_2 error equals the total posterior variance, providing a closed-form exploration signal. The algorithm plans under the posterior mean model, eliminating confidence-set optimization. Under the stated assumptions, this simpler planning procedure retains the \widetildeO(T^-1/3) surrogate convergence rate of frequentist \kappa -Explorer. We further bound the maximum posterior variance relative to the optimal asymptotic minimax allocation benchmark. Experiments on Pendulum, Mountain Car, and Taxi demonstrate lower worst-case transition-estimation errors than the tested baselines.

[LG-92] Rate-distortion Assessment for Hierarchical Content Summarization

链接: https://arxiv.org/abs/2611.25151
作者: Yunye Gong,Indranil Sur,Jared Claypoole,Pritish Sahu,Ajay Divakaran
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Powerful generative models including large language and vision models have demonstrated success across a wide range of content generation, editing and manipulation tasks via end-to-end prompt-based generation. However, current assessment of the quality of generative content manipulation remains subjective, limiting scientific understanding of model capabilities. Inspired by rate-distortion theory from Shannon information compression theory, in this work we formalize hierarchical content summarization as an information encoding problem. We design a principled framework to extract the semantic hierarchy of the source content to support quantitative characterization of the rate-distortion trade-off between compression efficiency (rate) against information loss (distortion) during generative content manipulation, which allows us to benchmark various models with respect to performance conditional on rate budget of interest. We demonstrate the use of the framework with empirical experiments on content summarization for long-context question-answering with state-of-the-art large language models, offering new insights into their efficiency and comprehension capabilities. We further demonstrate the use of the semantic hierarchy as guidance for performing effective retrieval augmented generation, showing improved rate-distortion characteristics against current baselines.

[LG-93] CAIMS: Scalable Cross-Array In-Memory Search

链接: https://arxiv.org/abs/2611.25141
作者: Yi-Chun Liao,Liu Liu,Ruiyang Qin,Yiran Chen,X. Sharon Hu
类目: Hardware Architecture (cs.AR); Emerging Technologies (cs.ET); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Search, especially top-k similarity search, plays a vital role in many applications, but its performance can be severely limited by memory access overhead. In-memory search (IMS), an alternative to CPU/GPU-based approaches, exploits massive parallelism within a content-addressable memory (CAM) array to perform search in situ and provides both low latency and high energy efficiency. However, the capacity of a single CAM array is often insufficient for modern workloads, requiring data to be distributed across multiple arrays. Deploying scalable top-k similarity search across multiple CAM arrays poses a fundamental problem that existing designs fail to address: how to accurately and efficiently reduce local binary match results to global top-k results. This work introduces CAIMS, a cross-array IMS framework that is agnostic to memory technologies and applications, enabling scalable and efficient top-k similarity search while preserving accuracy. To enable accurate top-k result reduction, CAIMS leverages the existing threshold match functionality with a combined offline and online parameter selection approach to transform local binary match results that lack globally comparable similarity information into global coarse-grained similarity scores. To further improve efficiency, CAIMS conducts hierarchical reduction and transfers only the selection results and necessary control signals with lightweight local peripherals. Implemented on a 3D NAND Flash model and demonstrated using the retrieval phase of retrieval-augmented generation, CAIMS improves throughput by up to 22.3\times/10.1\times and energy efficiency by 426.6\times/23.7\times compared to the server-class CPU baseline and prior 3D NAND Flash accelerators, respectively, while maintaining comparable accuracy.

[LG-94] mporal Extrapolation of Elastic Seismic Wave Propagation in Layered Media Using Neural Operators

链接: https://arxiv.org/abs/2611.25130
作者: Dibyajyoti Nayak,Yuze Pu,Kami Mohammadi,Somdatta Goswami
类目: Machine Learning (cs.LG); Geophysics (physics.geo-ph)
*备注: 31 pages including appendix, 14 figures, D. Nayak and Y. Pu contributed equally to this work

点击查看摘要

Abstract:Temporal prediction of seismic velocity fields beyond the time interval represented in training remains challenging for neural operator models. This study develops a coupled neural operator framework for stress-conditioned temporal prediction in layered elastic media. A time-integrator-embedded Deep Operator Network (TI-DeepONet) advances the velocity field through an embedded fourth-order Runge-Kutta (RK4) scheme, while a Fourier Neural Operator (FNO) learns a residual correction to the predicted spatial field. The two components are trained jointly in a predictor-corrector configuration. Prediction is initialized with reference velocity fields at the first two time levels and proceeds autoregressively, with reference stress fields and prescribed excitation supplied throughout rollout, including beyond the training interval. Four cases consider horizontally polarized shear (SH) waves in four-layer soil columns with fixed total depths of 50 and 280 m, excited by third-order Gaussian-derivative (DOG3) wavelets with central frequencies between 0.5 and 6 Hz. Layer thicknesses and shear-wave velocities vary across profiles. Evaluations on test profiles extend beyond the training interval by up to twice its duration. The predictions reproduce principal velocity arrivals and depth-dependent response patterns, with accuracy varying across profiles. Across the four cases, mean peak-normalized root-mean-square errors (NRMSEs) range from 0.53% to 2.31% over the full evaluation windows and from 0.61% to 2.82% over the extrapolation intervals. Both metrics use each profile’s peak absolute reference velocity over the full evaluation window as the normalizer. These results demonstrate temporal velocity prediction under reference-stress conditioning for the tested configurations; autonomous continuation of the coupled velocity-stress wavefield remains to be established.

[LG-95] Auditing Learned Dynamics in Weather Models

链接: https://arxiv.org/abs/2611.25119
作者: Skylar Gu,Christina Last
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Learned weather models are experimentally accessible computational systems whose internal representations can be inspected and manipulated. Using sparse autoencoder features from GraphCast, DeepMind’s autoregressive weather forecasting model, we test whether causal relevance requires physical semantics, a question that matters for auditing whether AI forecasters used in high-assurance settings rely on reliable mechanisms. We also look at whether physically associated latent directions can be steered to improve extreme-event forecasts, where interventions are applied to the model’s internal activations during the rollout. We find that ablating a feature locked to GraphCast’s computational mesh increases global geopotential height error despite lacking conventional physical semantics. Second, suppressing convection- and low-level-vorticity-associated features weakens predicted tropical-cyclone deepening (the fall in central pressure during intensification), while amplifying them reduces intensity error, with amplification of convection features through feature scaling achieving a median of 51% error reduction over seven storms. Together, these results separate interpretable, physical association from causal utility and establish sparse features as experimental coordinates for auditing and intervening on learned weather dynamics.

[LG-96] Fast and Memory-Light Evaluation of Density-Based Clusterings with the DISCO Index in R

链接: https://arxiv.org/abs/2611.25114
作者: Amin Entezari
类目: Machine Learning (cs.LG); Computation (stat.CO)
*备注: 12 pages, 4 figures. R package discoCVI available on CRAN

点击查看摘要

Abstract:DISCO is a recent internal cluster validity index for density-based clustering that, unlike earlier indices, also evaluates the quality of noise labels. The discoCVI package makes DISCO available in R. This article describes the design of the package and a typical workflow, in which DISCO is used to compare clusterings and to choose the parameters of HDBSCAN without ground truth, followed by the implementation introduced in version 0.2.0. DISCO is defined through density-connectivity distances between all pairs of points, so direct implementations store several n-by-n matrices. We show that every quantity DISCO needs can be read off the single-linkage merge tree of the mutual-reachability minimum spanning tree, so that the index can be computed exactly with memory that grows linearly with the number of points, while the running time remains quadratic as in the original definition. The implementation is tested against the previous version, agrees with the original Python implementation to within 3 \times 10^-13 , and corrects the score of singleton clusters to match the published definition. In benchmarks, discoCVI is 23 to 138 times faster than its previous version for 500 to 4,000 points and 3 to 9 times faster than the Python reference for 500 to 8,000 points, using a fraction of the memory of both, and it scores 100,000 points in about 50 seconds with less than 200 MB of memory.

[LG-97] HapBO: Harnessed α-π Bayesian Optimization via Expert Natural Language Priors

链接: https://arxiv.org/abs/2611.25100
作者: Jean-Sébastien Giroux,Rose Guay Hottin,Marco Bonizzato
类目: Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 41 pages, 22 figures, 8 tables. Code: this https URL

点击查看摘要

Abstract:Bayesian Optimization (BO) can be substantially more sample-efficient when informed by prior knowledge about the problem. In practice, however, such knowledge is often available only in natural language, e.g., expert intuitions or documentation, rather than as the explicit probability distribution landscape that prior-guided BO methods may require. Leveraging it effectively demands both an expressive mapping from language to priors and robustness to information that may be misleading. We propose HapBO (Harnessed \alpha - \pi Bayesian Optimization), which elicits beliefs once from a human expert, a large language model (LLM), or both, and deterministically compiles a structured specification into a reusable belief function ( \pi: \mathcalX \to[0,1] ), with a rich representation not restricted to a Gaussian family. Building on \alpha - \pi BO, HapBO applies this belief function through a bounded positive multiplier on any nonnegative acquisition, preserving positive acquisition support and requiring no LLM calls during optimization. We prove that, for the Expected Improvement acquisition function (EI), HapBO preserves the standard simple-regret rate up to a bounded factor, even under adversarial beliefs. We evaluate HapBO on 40 HPOBench tasks, a synthetic suite in 6-15 dimensions and two real-world applications: cortical neurostimulation and constrained solar design. HapBO achieves the lowest aggregate HPOBench regret and strong real-world task performance, with zero in-loop token cost. In real-world applications it improves the optimization outcome even at budgets of 12-20 queries, and at a small fraction of the wall-clock and token costs of other LLM-based methods.

[LG-98] PriTabICL: Differentially Private Tabular In-Context Learning with Task-Aligned Cluster-Wise Synthesis

链接: https://arxiv.org/abs/2611.25073
作者: Syed Ahzaz Shah,Evan Bramblett,Héber H. Arcolezi,Chen Gong,Qianqian Tong,Xiaochen Li
类目: Cryptography and Security (cs.CR); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Tabular in-context learning (TabICL) enables large language models (LLMs) to predict target records using a handful of labeled demonstrations in their prompts. However, these demonstrations often contain sensitive personal information. Differential privacy (DP) offers principled protection, but noise-based methods introduce substantial noise, distorting the feature-label relationships essential for accurate prediction, especially as data dimensionality grows. In this work, we study DP data synthesis for generating privacy-preserving TabICL demonstrations. We identify two limitations of directly applying general-purpose DP synthesis to entire datasets: modeling a global distribution can obscure local structure and minority patterns in heterogeneous data, while measuring broad attribute dependencies under a limited privacy budget can reduce the accuracy available for relationships most relevant to prediction. More broadly, global statistical fidelity may fail to preserve heterogeneous structures needed by downstream predictive and generative models. To address these challenges, we propose PriTabICL, a general framework compatible with existing DP synthesis algorithms. It partitions records into DP clusters and synthesizes each independently, using a fixed feature-target workload to prioritize prediction-relevant statistical relationships. At inference, each query is routed to a DP cluster, and demonstrations are sampled from its synthetic dataset without additional privacy loss for the original records. Experiments on four real-world datasets demonstrate the effectiveness of our methods across heterogeneous datasets. In particular, one of our instantiations, PriTabICL-PrivSyn, improves over global PrivSyn in 8 out of 12 settings and outperforms existing private TabICL approaches in nearly all evaluated settings. Open-source code is available. Subjects: Cryptography and Security (cs.CR); Machine Learning (cs.LG) Cite as: arXiv:2611.25073 [cs.CR] (or arXiv:2611.25073v1 [cs.CR] for this version) https://doi.org/10.48550/arXiv.2611.25073 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[LG-99] Watch Where You Linearize: Predicting and Correcting Model-Merging Interference

链接: https://arxiv.org/abs/2611.25000
作者: Ido Nitzan Hidekel,Dan Raviv
类目: Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Model merging combines separately fine-tuned models by adding their task vectors to a shared backbone. Interference is typically assessed only after constructing and evaluating a candidate merge, making coefficient search expensive. We show that a first-order directional derivative of task A’s outputs along task B’s vector can predict interference, but only when evaluated at the appropriate model. At the receiver model, this derivative forecasts nearby merges per task and per example, halves the accuracy error of a no-change baseline, and identifies most examples whose predictions flip. For full multi-task merges, a single evaluation is insufficient; two evaluations along the merge path are needed. At the shared pretrained backbone, however, the derivative is nearly orthogonal to realized changes in our vision experiments and only partly predictive in LoRA language-model pilots. Controls link this failure to displacement along the receiver’s own task vector, not simply parameter distance; the mismatch appears within the first few dozen fine-tuning steps. Freezing receiver-local responses produces a surrogate affine in the merge coefficients. We use it to construct LR-Merge, a label-free correction that optimizes coefficients through convex teacher matching on unlabeled probes, without backpropagation through merged networks. LR-Merge improves all tested closed-form, isotropic, and sparsified base mergers, with gains generally increasing as more tasks are merged. Using pretrained rather than receiver-local responses instead degrades the base merge. LR-Merge matches exact teacher matching through a merged network, which takes roughly three times longer to converge and performs similarly at an equal compute budget. Once built, the surrogate answers new coefficient and task-subset queries in seconds. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2611.25000 [cs.LG] (or arXiv:2611.25000v1 [cs.LG] for this version) https://doi.org/10.48550/arXiv.2611.25000 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[LG-100] PsyWorld: Profile-Conditioned World Modeling for Human Behavior Prediction KDD2027

链接: https://arxiv.org/abs/2611.24999
作者: Ziyuan Qin,Yang Qiao,Summer B. Frandsen,Michael T. Treadway,Liang Zhao
类目: Machine Learning (cs.LG)
*备注: Accepted to the KDD 2027 AI for Sciences Track

点击查看摘要

Abstract:Understanding how stable psychometric traits and time-varying states jointly inform behavior is an important problem in computational behavioral science. Existing personalized agents commonly condition actions directly on a static profile, leaving realized state transitions unused as an explicit source of supervision. We introduce a profile-conditioned state world model that uses observed environment–behavior state transitions to learn transition-informed representations for behavioral prediction. During training, a posterior encoder quantizes realized transitions, while a prior predicts the resulting codes from the psychometric profile and pre-transition context. The prior code and predicted next state then condition macro-level behavior and simulator-grounded execution. Thus, the world model provides structured transition supervision instead of treating personality as a direct action descriptor. We evaluate the framework in VEET, a controlled behavioral simulator with human trajectories, and in AgentSociety as a structurally different synthetic environment. A participant-disjoint split of 240 training, 25 validation, and 48 test participants evaluates generalization to unseen people. On VEET, the model reaches 0.904 Top-1 accuracy (95% CI: [0.895, 0.913]) and 0.0033 Macro KL. Ablations associate continuous next-state modeling with most of the Top-1 gain, while discrete codes and prior alignment mainly improve Top-3 and distributional alignment. These results establish transition-informed world-model supervision as useful for subject-generalized behavior prediction, without implying causal psychological mechanisms or calibrated long-horizon state dynamics.

[LG-101] When Is Program Structure Required? Certificate-Grounded Calibration of Program-Induction Difficulty

链接: https://arxiv.org/abs/2611.24975
作者: Wenzhang Du
类目: Programming Languages (cs.PL); Machine Learning (cs.LG)
*备注: 26 pages, 1 figure, 24 tables, including appendices. Code and reproduction artifacts at this https URL

点击查看摘要

Abstract:A program-induction benchmark is informative when its examples require the structure it aims to test. Yet a two-rule generator can produce demonstrations that a single rule also explains. We introduce certificate-grounded calibration, which constructs matched tasks with minimum solution lengths of one and two in a bounded string-replacement language. Complete one-rule enumeration and an executable two-rule witness certify the change before model inference. We evaluate 120 reference families with three to six shared examples, using 1,440 new model trajectories and 480 saved trajectories. Across Qwen3-8B and Olmo-3-7B-Think, displayed-task completion is 23.53-45.83 percentage points lower on certified separation tasks, while success on a wholly unshown witness pair is 8.40-13.45 points higher. These witness gains accompany lower agreement on uniformly sampled strings and 1.93-5.78% agreement on inputs that activate both reference rules. Recertification also reveals that a single unchanged example can exclude a shortcut and determine whether two rules are required. Together, these results show how the evidence supplied by examples shapes both the required solution and the programs models produce. The construction provides an executable testbed for comparing program search, final-answer delivery, and behavior on new inputs under verified task requirements.

[LG-102] Simple Private Noisy SGD is Minimax-Optimal under Population Strong Convexity

链接: https://arxiv.org/abs/2611.26515
作者: Valentin Roth,Marco Avella Medina
类目: atistics Theory (math.ST); Machine Learning (cs.LG); Machine Learning (stat.ML)
*备注: 34 Pages

点击查看摘要

Abstract:Stochastic gradient descent (SGD) is at the center of modern machine learning and is known to be statistically minimax-optimal in many regression settings, e.g., for convex Generalized Linear Models (GLMs). Yet, it is unknown whether its simplest differentially private Noisy SGD counterpart fulfills similar favorable guarantees even on strongly convex losses. We resolve this question affirmatively. We upper bound the parameter estimation error for a single pass of Noisy SGD with a fixed stepsize schedule and constant batch sizes on smooth and strongly convex Lipschitz losses over a compact convex domain under Renyi DP. Specialized to strongly convex GLMs, our error bounds match minimax lower bounds we derive based on the Score Attack framework. In particular, this provides the first minimax-optimal rate for Noisy SGD in private logistic regression, avoiding poly-logarithmic terms present in existing results. Our analysis is versatile as it relies on standard convex analysis and establishes the same rates in expectation and with high probability.

[LG-103] Bayesian optimization on amortized search space

链接: https://arxiv.org/abs/2611.26459
作者: Tristan Fauvel,Federico Pavone,Titouan Guillou,Pau Redon Muñoz,Chloé Geoffroy
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:In real-world black-box optimization problems, a mechanistic simulator is often available that approximately predicts the behavior of the system as a function of the parameters being optimized. Here, we show how an amortized inverse map of such a simulator can be used in Bayesian optimization to restrict the search space while using a Gaussian process kernel that is aware of the original problem geometry. This allows for efficiently leveraging the low intrinsic dimensionality of the optimization problem. We characterize its approximation bias and information gain, and derive an intrinsic-dimension-dependent regret rate under additional geometric and Gaussian Process assumptions. We illustrate the method with an application to drug dosing personalization in Parkinson’s disease. We use, as a simulator, a pharmacokinetic/pharmacodynamic (PK/PD) model of a drug, a mechanistic model calibrated on population data with quantified parameter uncertainty. A dosing neural network is trained offline against the model, amortizing the map from patient PK/PD parameters to the regimen predicted to be optimal, and online Bayesian optimization then searches over this learned, low-dimensional representation rather than over raw dosages. Numerical simulations demonstrate the improved performance of our method in this application, as well as under simulator misspecification.

[LG-104] Scaling classical hardware for accurate real-time quantum error correction

链接: https://arxiv.org/abs/2611.26447
作者: Chenyang Jin,Susanne F. Yelin,Andi Gu
类目: Quantum Physics (quant-ph); Hardware Architecture (cs.AR); Machine Learning (cs.LG)
*备注: 13 pages, 5 figures, 4 Extended Data figures, 1 Extended Data table

点击查看摘要

Abstract:Fault-tolerant quantum computation requires a classical decoder fast enough to process syndrome measurements at the rate at which they are generated, and to return its decisions with low latency. This is most demanding for superconducting processors, whose fast syndrome cycles force a trade-off between decoding speed and accuracy. Decoders fast enough to keep up sacrifice accuracy, which must be compensated for with larger codes and thus additional physical qubits. Here we introduce Rapids, a streaming neural decoder that is both fast and accurate. We achieve this by co-designing the decoder with a field-programmable gate array (FPGA): the network is built from operations that map cheaply onto the device and trained in its integer arithmetic. For the distance-19 surface code, Rapids decodes each round of syndromes in under a microsecond on a single FPGA, fast enough for superconducting processors, and returns its decisions with the low latency that fault-tolerant algorithms require. Its logical error rate is 240 times lower than that of existing real-time hardware decoders. The same design extends to the color code and the [[144,12,12]] bivariate-bicycle code, on which our decoder is 16 times more accurate than the real-time Relay decoder. We further show that the decoder scales: spread over several FPGAs or hardened into custom silicon, it becomes faster, more accurate, or able to decode larger codes. Classical hardware thus becomes a further resource, alongside qubits, that can be scaled to improve fault-tolerant quantum computers.

[LG-105] Exact Bayesian layers via characteristic functions

链接: https://arxiv.org/abs/2611.26426
作者: William H. W. Thompson,Guillaume St-Onge,Laurent Hébert-Dufresne,Jean-Gabriel Young
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Computation (stat.CO)
*备注:

点击查看摘要

Abstract:Bayesian neural networks provide principled uncertainty estimates, but training them with Monte Carlo sampling can be expensive. Sampling-free variational inference algorithms make training Bayesian neural networks feasible, but existing approaches assume a Gaussian output distribution, which can be inaccurate in applied problems. Here, we introduce CF-Prop, an uncertainty propagation method that uses characteristic functions to compute the exact output distribution of a Bayesian ReLU layer, without any sampling, even in non-Gaussian settings. Because CF-Prop propagates characteristic functions, the evidence lower bound—the objective of variational training—can be computed in closed form for a broad class of likelihoods, including the Gaussian, Poisson, and Gamma. This allows to implement an efficient resulting variational inference algorithm, CF-VI. On CIFAR-100, CF-VI achieves the lowest calibration error of any tested method, improving on the next best method by about 35%. On a heavy-tailed insurance claim dataset, CF-VI accurately captures tail uncertainty that is severely underestimated by both other sampling-free methods and Monte Carlo methods.

[LG-106] Statistical Inference with Ill-Conditioned Random Weight Neural Network Sieves

链接: https://arxiv.org/abs/2611.26410
作者: Giovanni Ballarin,Lyudmila Grigoryeva,Lukas Gonon
类目: Methodology (stat.ME); Machine Learning (cs.LG); Econometrics (econ.EM); Machine Learning (stat.ML)
*备注:

点击查看摘要

Abstract:Random weight neural networks (RWNNs) are shallow networks whose hidden-layer parameters are randomly generated and then kept fixed, while only the output coefficients are estimated. Conditional on the sampled parameters, RWNN regression is a linear sieve problem with random approximation space. We show that such random sieves cannot remain uniformly well-conditioned as their dimension grows. For parametrized feature families satisfying polynomial covering and Lipschitz conditions, the eigenvalues of the population covariance matrix are proven to have polynomial upper bounds determined by the covering geometry of the parameter space, leading to a vanishing smallest eigenvalue. We develop approximation, estimation, and inference theory that explicitly treats the ill-conditioning. Under a \beta -Hölder source condition, we obtain both faster-than-Monte-Carlo L^2 approximation rates and new L^\infty approximation bounds based on stratified sampling. We then derive improved consistency guarantees for ridge-regularized least squares driven by ridge effective dimension, spectral, and leverage-based stability conditions. For inference, we establish asymptotic linearity and Gaussian limits for linear functional directions while allowing the smallest eigenvalue of the population covariance matrix to vanish. Nondegeneracy is directional: we characterize it with a ridge-normalized Rayleigh quotient and extend it to smooth functionals via Riesz representation. Consistency of ridge-adjusted variance estimators is proved, yielding feasible studentized inference. Numerical experiments illustrate the deterioration of random-feature conditioning and the finite-sample behavior of the resulting estimation and inference procedures.

[LG-107] Simulation-pretrained autoencoders versus PCA for compressing multimode fibre transmission matrices

链接: https://arxiv.org/abs/2611.26391
作者: Callum J. Wilson,Terry Wright,Qing Yang,Zhong Wen,Michael P. Pound,George S.D. Gordon
类目: Optics (physics.optics); Machine Learning (cs.LG)
*备注: 31 pages, 14 figures, including supplementary material. Code, trained weights and data are linked in the Data and code availability section

点击查看摘要

Abstract:Multimode optical fibres can image through a thin, flexible probe once their transmission matrix ™ has been measured, but a TM is large and must be re-measured whenever the fibre moves. A compact model of feasible TMs would make storing and re-estimating them cheaper, so we ask how compactly TMs can be represented, and from how few measured TMs. Here we benchmark a permutation-equivariant autoencoder, pretrained on simulated fibres and adapted to measured TMs, against principal component analysis (PCA). At 62x compression, adapting the pretrained autoencoder with just two measured TMs reconstructs the TM of a 50um fibre well enough to focus light at 98% of the ideal enhancement, and it stays accurate when the TMs are shifted by 25 pixels or zoomed by 10%, which quickly degrades a linear basis. It is not, however, a better compressor: PCA is about ten times more accurate when fitted on all 163 real TMs (1.1% against 11.9% error on held-out TMs) and is already more accurate from ten. The autoencoder’s value is as a prior when few or no real TMs exist: with none it reaches 29% error, against 91% for PCA fitted on the same simulation. Transfer to a different fibre is only partial. The most important extension, not yet tested, is to recover pixel-basis TMs from fewer measurements than a full TM, which would turn the pretrained prior from a compressor into a measurement-saving tool.

[LG-108] Perfumer: A Large Language Model-based Agent ic Framework Accelerating Structural Relaxation and Electronic Band Prediction for First-Principle Calculation

链接: https://arxiv.org/abs/2611.26349
作者: Zhiwei Li,Yuhao Zhou,Yikai Yang,Haonan Wang,Qing Ye,Jiancheng Lv,Kazunari Matsuda
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
*备注: 33 pages

点击查看摘要

Abstract:First-principles structural relaxation and band calculations require substantial computation and expertise. Here we present Perfumer, a large language model-based agentic framework that couples Quantum ESPRESSO with two predictive models. Across tested structures, pre-relaxation with Perfumer-relax model followed by QE refinement yielded a 2.75-fold overall speedup over QE alone. Perfumer-band model combines a periodic graph neural network with a latent Hermitian spectral head returns 12 band energies at any requested wavevector, preserving reciprocal lattice periodicity and time-reversal symmetry without constructing Hamiltonian. Trained on 125,153 self-curated band structures, Perfumer-band model reaches a median per-material mean absolute error of 0.157 eV on 2,607 test materials, including lanthanide, actinide and magnetic compounds, from held-out composition families spanning 85 elements, taking about 97 ms per material on two consumer CPU cores. The agent turns natural-language requests into relaxation-to-band workflows, from a single uploaded structure to automatically constructed new candidates and returns predicted properties.

[LG-109] Conformal Selection for Hierarchical Streaming Data

链接: https://arxiv.org/abs/2611.26328
作者: Zhenyuan Sun,Chengyao Yu,Ruixing Ming,Bingyi Jing
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 17 pages, 6 figures

点击查看摘要

Abstract:Many real-world tasks involve making sequential decisions about whether to select candidates before their responses of interest are revealed. In this work, we address the online selection problem in the context of hierarchical data, where candidates arrive sequentially from different populations. This extends existing online conformal selection frameworks beyond the standard i.i.d./exchangeable data assumption. The primary challenge lies in constructing valid statistics for each candidate sequentially under the hierarchical data structure. To this end, we first develop Online Hierarchical Selection (OHS), a distribution-free method that enables principled candidate selection for hierarchical data streams. Specifically, we introduce methods to construct rank-based e-values for both between-group and within-group samples, and apply averaged e-values to the online e-BH procedure. We then develop OHS with feedback (OHSF), which boosts the power of OHS by leveraging potential response feedback. Theoretically, the proposed methods achieve false discovery rate control under mild hierarchical exchangeability and data arrival assumptions at any time. We also generalize methods to broader settings, such as data with covariate shift and selection under effect-related responses. Extensive experiments on both simulated and real-world datasets demonstrate the effectiveness of the proposed methods.

[LG-110] Who Moved My Z Boson? Revealing Resonance Features Inside a Neural Network

链接: https://arxiv.org/abs/2611.26291
作者: Kevin Alberto Urquía Calderón,Midori Kato,Inar Timiryasov,Oleg Ruchayskiy
类目: High Energy Physics - Phenomenology (hep-ph); Machine Learning (cs.LG); High Energy Physics - Experiment (hep-ex); Data Analysis, Statistics and Probability (physics.data-an)
*备注: 10 pages (+4 pages of references, +14 pages of supplementary material, 29 total pages), 7 figures in the main text (+14 figures in the supplementary material, total of 21 pages)

点击查看摘要

Abstract:Methods used in mechanistic interpretability have revealed human-aligned concepts inside language and image models. In this paper, we search for physics concepts inside a neural a generative collider physics model trained on real LHC data (ShellFlow). We discover features inside ShellFlow that represent resonances, such as the Z , J/\psi , and \Upsilon . The network was never told which muons come from a resonance, it was only trained to generate the kinematics of the final-state particles. We use these features to alter aspects of the Z boson: we change its mass between \sim80 and 105 GeV, and removing such a feature from the neural network produces a broader Z peak. In both cases the rest of the invariant mass spectrum remains mostly unchanged. The feature was built from the gradients of the neural network, and its selectivity was built using pattern recognition methods. A generative particle physics model trained on unlabeled data can be modified internally to change a physical quantity.

[LG-111] A Statistical Framework for Training Data Auditing of Large Language Models

链接: https://arxiv.org/abs/2611.26264
作者: Shengjie Niu,Yeheng Ge,Jian Huang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Applications (stat.AP)
*备注:

点击查看摘要

Abstract:Training-data auditing asks whether data of interest were used to train a large language model (LLM), a question central to privacy auditing, copyright protection, and benchmark-contamination detection. These concerns are particularly relevant to the use of copyrighted books and scientific papers in pre-training, as well as synthetic text and medical question-answer records in fine-tuning. In practice, training membership signals may be concentrated in only a few passages of a long document, making it important to aggregate evidence within each document without diluting these localized signals. Beyond individual documents, auditing a target collection requires an overall assessment based on the document-level evidence. In this paper, we propose a unified statistical framework for document-level membership inference, target-collection global auditing, and suspicious-subset identification. We integrate complementary chunk-level membership evidence with Cauchy combination to construct document scores for document-level membership inference. The document scores and their calibrated p-values further enable an overall assessment of training-data collection use and the identification of suspicious documents through multiple testing. Under the stated exchangeability conditions, we establish finite-sample validity of document-level p-values, type-I error control for global auditing, and false discovery rate control for subset identification. Case studies on BookTection, Pile-arXiv, Cosmopedia, and HealthCareMagic demonstrate the effectiveness and reliability of the proposed framework in auditing pre-training and fine-tuning data.

[LG-112] Reward misalignment as an opportunity for materials discovery

链接: https://arxiv.org/abs/2611.26186
作者: Ryotaro Okabe,Shoki Kishida,Abhijatmedhi Chotrattanapituk,Dominic Bollettieri,Takuya Isogawa,Yongqiang Cheng,Mingda Li,Masaki Adachi
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG)
*备注: 33 pages (15 pages main text and 18 pages Supplementary Information), 4 figures

点击查看摘要

Abstract:Reward misalignment is usually regarded as a failure of artificial intelligence. In materials discovery with generative AI models, however, disagreement between predefined rewards and expert judgment may reveal materials that the rewards overlook. We investigate this possibility using electronic flat bands as a prime example. Experts mostly agree with a numerical bandwidth score, but their disagreements are systematic and cannot be explained by an estimated 5% rate of mistaken labels. We learn the experts’ preferences from pairwise comparisons and use them to guide a crystal diffusion model. The learned preferences and numerical score agree on the flattest bands but diverge elsewhere. For example, the preferences favor KCo2As2 and CoGa, reported compounds with partially flat bands that the score would discard. The framework also extends to synthesizability and phonon properties. Our findings suggest that reward misalignment can expose opportunities for materials discovery beyond what fixed rewards are designed to find.

[LG-113] Physics-Informed Deep Learning for Shear Wave Speed Estimation in MR Elastography

链接: https://arxiv.org/abs/2611.26111
作者: Stefan Martin,Jakob Schattenfroh,Patrick Schuenke,Felix Frederik Zimmermann,Ingolf Sack,Christoph Kolbitsch,Andreas Kofler
类目: Image and Video Processing (eess.IV); Machine Learning (cs.LG); Medical Physics (physics.med-ph)
*备注: 13 pages (incl. cover page), 6 figures, 2 tables. Accepted version of the article published in IEEE Transactions on Biomedical Engineering. Please cite the published version (DOI: https://doi.org/10.1109/TBME.2026.3666306 ). Code: this https URL

点击查看摘要

Abstract:Objective: Magnetic Resonance Elastography (MRE) is a non-invasive imaging technique for mapping biomechanical properties of in vivo tissue, including shear wave speed (SWS), but involves intrinsically slow data acquisition and an ill-posed wave inversion. Instead of relying on handcrafted image priors, we propose a data-driven approach jointly combining image reconstruction and MRE inversion for robust SWS estimation from undersampled k-space data. Methods: Our physics-informed reconstruction framework comprises two blocks: a model-based neural network (NN)-regularized reconstruction module and a phase-gradient inversion (k-MDEV) calculating SWS from the reconstructed images. Concatenating both blocks yields an end-to-end trainable method to estimate SWS directly from measured k-space data. We evaluated the method on retrospectively highly undersampled brain MRE data and compared it to a total variation (TV) minimization-based approach. We assessed the impact of end-to-end training (qualitative images and SWS maps as targets) versus pre-training (qualitative images as targets) and applied the method also to in vivo data. Results: Our approach significantly reduces NRMSE by 30% compared to TV. End-to-end training improves SWS estimation over separate image reconstruction and SWS calculation. Conclusion: Accurate SWS quantification is possible at acceleration factors up to 19. Our method significantly outperforms TV, highlighting the need for data-driven regularization in this challenging MR problem. Further, our approach successfully generalizes to in vivo data. Significance: We present the first end-to-end trainable MRE reconstruction method for estimating SWS maps directly from k-space. NN-based reconstruction can enable rapid stiffness mapping for dynamic studies, functional imaging, and real-time clinical feedback.

[LG-114] SlotFlow-GB: A Neural Framework for Trans-Dimensional Inference of Overlapping Galactic Binaries

链接: https://arxiv.org/abs/2611.26057
作者: Niklas Houba,Giovanni Giarda,Davide Piras
类目: Instrumentation and Methods for Astrophysics (astro-ph.IM); Machine Learning (cs.LG); General Relativity and Quantum Cosmology (gr-qc); Data Analysis, Statistics and Probability (physics.data-an); Instrumentation and Detectors (physics.ins-det)
*备注: 24 pages, 14 figures

点击查看摘要

Abstract:The Laser Interferometer Space Antenna (LISA) will observe the millihertz gravitational-wave sky, including millions of long-lived Galactic binaries (GBs), tens of thousands of which are expected to be individually resolvable throughout the mission’s lifetime. Extracting these sources requires jointly determining their number and physical parameters from strongly overlapping signals, posing a computationally demanding trans-dimensional inference problem. We introduce SlotFlow-GB, a neural framework that constructs a learned approximation to the trans-dimensional posterior, producing a probability distribution over source multiplicity and parameter densities for individual binaries in a single evaluation, without requiring their number to be specified in advance. We demonstrate this framework using year-long LISA simulations containing up to ten binaries within a 2 \mu Hz frequency interval, focusing on a high-SNR regime in which source overlap, rather than individual-source detectability, is the primary challenge. The inferred source count differs from the true count by at most one in 94% of cases, while the model identifies individual binaries even when their signals overlap strongly in frequency. A complete evaluation, including parameter sampling, takes just 41 ms per band on one NVIDIA GH200 GPU, or 530 ms on eight laptop CPU threads. We further demonstrate that these distributions can be used to initialize subsequent likelihood-based inference, connecting fast amortized predictions with joint parameter estimation.

[LG-115] Hellinger Decision Trees for Density Ratio Estimation: Finite-Sample Guarantees

链接: https://arxiv.org/abs/2611.25936
作者: Ren-Rui Liu,Zheng-Chu Guo
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 45 pages, 5 figures, 2 tables

点击查看摘要

Abstract:Density ratios play a central role in two-sample comparison, distribution-shift adaptation, and generative-model evaluation. However, estimating a density ratio through separate estimation of the underlying densities can be unnecessarily challenging. In this paper, we develop a direct approach to the log density ratio estimation by minimizing a balanced Hellinger risk over additive ensembles of axis-aligned decision trees, trained using Newton boosting. We establish finite-sample guarantees for the estimator produced by the algorithm and derive an explicit convergence rate. Under suitable tail and Hölder smoothness conditions, together with an additional assumption on the progress made at each boosting iteration, our rate matches, up to logarithmic factors, the classical minimax-optimal \mathscrL^2 rate for function estimation. A key technical contribution is an optimization analysis that exploits the special exponential structure of the Hellinger risk. In particular, we show that the decrease in the second-order Taylor approximation (which is computable) of the empirical risk controls the actual decrease in that risk, yielding a direct optimization guarantee for the boosting algorithm. This allows us to establish statistical convergence guarantees for the algorithm’s output without relying on an idealized empirical-risk minimizer. Experiments on synthetic data corroborate the predicted convergence behavior, while experiments on public datasets demonstrate the practical effectiveness of the proposed estimator.

[LG-116] Beyond Error Mitigation: Using Boosting for Noisy Quantum Machine Learning

链接: https://arxiv.org/abs/2611.25752
作者: Anastasiia Lukianenko,Yash J. Patel,Nurbolat Kenbayev,Vedran Dunjko
类目: Quantum Physics (quant-ph); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Quantum error mitigation (QEM) can recover information about ideal quantum computations from noisy circuit executions, but may require a large sampling overhead. Fundamental lower bounds show that recovering noiseless observables without problem-specific information can require superpolynomially many samples, with exponential overhead for broad classes of noisy circuits. However, problem-specific information can reduce the number of samples needed to obtain the required answer. We investigate this possibility in supervised quantum machine learning, where labeled examples provide information about the function to be learned. Binary classification provides such a setting, since correct prediction depends on the sign and margin of the classifier output rather than on accurate reconstruction of every noiseless expectation value. We develop a theoretical and empirical framework for comparing boosting of noisy parameterized quantum circuits with probabilistic error cancellation (PEC), requiring both methods to reach the same learning target. For the specified circuit architectures under Pauli noise, we establish a common hypothesis-class complexity bound: matching empirical margin loss gives the same upper bound on expected classification risk. We derive resource criteria and evaluate AdaBoostR and Real AdaBoost in finite-shot teacher-student simulations across circuit depth and physical noise strength. Under the same classification-accuracy target, boosting requires fewer circuit executions as depth and noise increase. Boosting also achieves higher test accuracy than the corresponding reference. For the supervised-learning problems studied, these results demonstrate a task-dependent resource advantage: exploiting information retained in noisy outputs can avoid the cost of ideal-output reconstruction without sacrificing observed classification performance.

[LG-117] Safeguarded Learning-Assisted Nonlinear Conjugate Gradient Methods

链接: https://arxiv.org/abs/2611.25707
作者: Morteza Kimiaei,Saman Babaie–Kafaki
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注: 56 pages, 7 figures

点击查看摘要

Abstract:A learning-assisted framework for Nonlinear Conjugate Gradient (NCG) methods is developed in which lightweight online prediction is used to support, rather than replace, the classical optimization process. The proposed approach retains the standard two-term NCG structure and introduces two independent online extreme learning machine models. The first model predicts an NCG coefficient during restart recovery or when persistent stagnation is detected, whereas the second model is used only after a line-search failure in finite precision arithmetic to propose a bounded correction to the deterministic fallback step size. Neural predictions are accepted only when they satisfy deterministic numerical safeguards, including descent, scale, and out-of-distribution requirements; otherwise, the corresponding deterministic recovery or fallback mechanism is retained. Both learning models are updated online using quantities already generated along the optimization trajectory and therefore require no additional function or gradient evaluations for training. These learning frameworks are applied to six well-known classical NCG methods, producing corresponding low-storage learning-assisted variants with deterministic safeguards. Numerical experiments on 480 CUTEst test problems show improved robustness for all six variants at broadly comparable evaluation costs. A PDE-constrained conductivity-identification example further illustrates the potential of the proposed approach to reduce evaluation costs and computational time for problems with expensive function and gradient evaluations. These findings suggest that restricted online learning can improve the robustness and computational performance of classical NCG methods while preserving their low-storage structure and deterministic optimization safeguards, and is particularly promising when function or gradient evaluations are expensive.

[LG-118] Lightweight Low-Frequency Non-Intrusive Load Monitoring via Karhunen-Loève Feature Distillation and Temporal Convolutional Networks

链接: https://arxiv.org/abs/2611.25678
作者: Sanka Peeris,Pandula Thennakoon,Praveen Sumanasekara,Roshan Godaliyadda,Parakrama Ekanayake,Janaka Ekanayake
类目: ignal Processing (eess.SP); Machine Learning (cs.LG); Systems and Control (eess.SY)
*备注: 10 pages, 6 figures. Submitted to IEEE Transactions on Smart Grid. S. Peeris, P. Thennakoon and P. Sumanasekara contributed equally

点击查看摘要

Abstract:Non-Intrusive Load Monitoring (NILM) is a critical enabler for demand-side management which allows for smart grid optimization. A deployment-ready NILM solution necessitates three key features: accurate disaggregation, low-frequency sampling, and low computational cost. This paper proposes a lightweight, edge-oriented NILM framework that combines activity-boundary-aware synthetic data augmentation with Karhunen-Loève Transform (KLT) feature extraction and a Temporal Convolutional Network (TCN). The augmentation stage extracts and recombines complete or near-complete appliance activation segments so that the model can be trained without new fully labeled household aggregates. The KLT filter bank extracts decorrelated, appliance-informed temporal features which are then fed into a compact TCN along with the original aggregate for energy disaggregation. Evaluations on two standard NILM datasets, UK-DALE and REDD, at 10-s sampling and with leave-one-house-out testing, demonstrate that the proposed architecture achieves comparable or superior disaggregation accuracy relative to NILMFormer, BERT4NILM, BiLSTM, and BiGRU. The computational footprint of the framework is 31,009 parameters, the fewest of the five models compared, and on a single CPU thread it needs 2.78 ms per 1000 samples, 84-95% less than the baselines. These results demonstrate a favorable accuracy-efficiency tradeoff for low-frequency smart-meter NILM.

[LG-119] AdsNet: A Graph Neural Network for Direct Adsorption Energy Prediction DSN

链接: https://arxiv.org/abs/2611.25646
作者: Changzhi Ai,Frank Abild-Pedersen,Johannes Voss
类目: Materials Science (cond-mat.mtrl-sci); Machine Learning (cs.LG); Computational Physics (physics.comp-ph)
*备注: 24 pages, 6 figures, 1 table; includes Supplementary Information (4 pages, 2 tables). AdsNet is MIT-licensed. Code: this https URL - Documentation: this https URL - Dataset (AdsTraj): this https URL - Model weights: this https URL

点击查看摘要

Abstract:Fast and accurate prediction of adsorption energies is critical for high-throughput computational catalyst screening, existing methods compromise either on evaluation speed (full-graph machine learning potentials) or geometric sensitivity (coordination-based models like the alpha-scheme). Here, we present AdsNet, an equivariant graph neural network that directly predicts adsorption energies in a single forward pass, which operates on a local graph restricted only to adsorbate atoms and its immediate surface neighbors within a cutoff radius. This local graph formulation decouples computational complexity from total system size while encoding continuous geometric features (interatomic distances and directional vectors) enabling sensitivity to subtle variations in local bonding environments. AdsNet adopts a message-passing architecture with scalar and vector node features over multiple interaction layers. Trained on AdsTraj, a dataset of 33,920 density functional theory (DFT) calculated structures spanning 26 transition metals and 6 adsorbates (H*, C*, O*, CO*, OH*, HOCO*), the model achieves an adsorption energy root-mean-square error (RMSE) of 0.032 eV and a force RMSE of 0.047 eV/A. Analytic forces obtained via automatic differentiation enable structure relaxation through a projected gradient descent algorithm with adaptive line search, eliminating entire geometry optimizations. Additionally, AdsNet uniquely decomposes predicted adsorption energies into contributions from individual atoms within the local adsorbate graph - an interpretability feature unattainable with conventional machine-learning potentials. AdsNet thus establishes an accurate, highly efficient, and interpretable framework for large-scale adsorption energy prediction and local interface relaxation across various material surfaces.

[LG-120] Dynamics of Protected State Evolution

链接: https://arxiv.org/abs/2611.25541
作者: Qinyou Wang
类目: Optimization and Control (math.OC); Machine Learning (cs.LG)
*备注: 48 pages, 10 figures, 7 tables

点击查看摘要

Abstract:States with identical protected observations can admit different future interventions. Each transition specifies its successor, response increment, required initial reserve and net consumption. Composing these records in execution order distinguishes protocols with equal net consumption and different initial reserve requirements. At a fixed state, a weighted Gram operator determines the linear response ellipsoid and a quotient identifies directions attainable only through cooperation between blocks. Nonlinear systems with identical complete affine data can have incomparable finite-budget response sets. In regular charts, the first correction to quadratic minimum cost is a covariant cubic form involving the third derivative of cost and the Hessians of protection and response. At singular gates, sequential feasibility depends on the peak reserve along the path. A history-weighted operator detects responses outside the old channel’s range closure. Controlled congruence characterizes exact reductions of deterministic labelled dynamics, and a two-cut value identity separates the effect of paid verification from that of subsequent action selection. On Qwen2.5-7B Base, two projected policies preserved 2,048 reference top tokens under a per-position KL ceiling of 2\times10^-5 . Over 200 arrivals they recovered 391 of 396 initial rejections. Each yielded over 1,200 times the held-out cross-entropy reduction of historical-channel-only AdamW. Under an expanded 2,560-position contract, fixed continuation and neutral renewal began on the same protected logit fiber in distinct structural strata. Both admitted all 200 subsequent writes. Fixed continuation yielded 1.74 times the held-out loss reduction per unit squared write cost. Renewal had a smaller KL excess on the first candidate tested at every arrival and used fewer full verification scans. Comments: 48 pages, 10 figures, 7 tables Subjects: Optimization and Control (math.OC); Machine Learning (cs.LG) MSC classes: 93B03 (Primary) 93C55, 49K40, 68T07 (Secondary) Cite as: arXiv:2611.25541 [math.OC] (or arXiv:2611.25541v1 [math.OC] for this version) https://doi.org/10.48550/arXiv.2611.25541 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[LG-121] Differentially Private Score Estimation with Applications to Score-Based Generative Models

链接: https://arxiv.org/abs/2611.25536
作者: Mengxiao Gao,Elly K. H. Hung,Yuchen Wu,Yi Yu
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Statistics Theory (math.ST); Methodology (stat.ME)
*备注: 57 pages, 18 figures

点击查看摘要

Abstract:Estimating the score function of a distribution plays a central role in statistics and machine learning, with applications in distribution estimation, sampling, and score-based generative modeling. However, learning scores from sensitive data raises privacy concerns. These concerns are particularly pronounced in score-based generative modeling, where trained models may memorize individual training examples and expose sensitive information through membership inference and data extraction attacks. Although non-private score estimation has been extensively studied, the statistical limits of its private counterpart remain less understood. In this work, we study score estimation under (\alpha,\beta) -differential privacy for d -dimensional sub-Gaussian distributions with locally \gamma -Hölder-continuous scores. We develop a sampling-based private score estimator using quadratic B-splines and the exponential mechanism, and we implement our proposal through projected Langevin Monte Carlo with runtime polynomial in the sample size n . The proposed estimator satisfies (\alpha,\beta) -differential privacy, and achieves expected density-weighted squared error \mboxpolylog(n) \max(n^-2\gamma/(d+2\gamma+2),,(n\alpha)^-2\gamma/(d+\gamma+1)) . We also establish a nearly matching private minimax lower bound, thereby characterizing the privacy-accuracy trade-off for score estimation. As an application, we construct a differentially private diffusion model whose privacy guarantee is independent of the number of denoising steps. We show that, with an appropriate discretization scheme, the resulting diffusion model attains nearly minimax optimal expected total variation error among all private distribution estimators. We validate our theoretical findings through experiments on both synthetic and real-world datasets. Comments: 57 pages, 18 figures Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG); Statistics Theory (math.ST); Methodology (stat.ME) Cite as: arXiv:2611.25536 [stat.ML] (or arXiv:2611.25536v1 [stat.ML] for this version) https://doi.org/10.48550/arXiv.2611.25536 Focus to learn more arXiv-issued DOI via DataCite (pending registration)

[LG-122] Instance-Optimal Moment Estimation in the Add-Remove Model of Differential Privacy

链接: https://arxiv.org/abs/2611.25425
作者: Shokichi Takakura
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:In this paper, we study the problem of estimating empirical means, variances, and covariances under pure differential privacy in the add-remove model. While existing mechanisms achieve optimal worst-case error, they may not be optimal for individual datasets. To measure the performance of mechanisms on individual datasets, we introduce instance-dependent error metrics and prove sharp lower bounds on them. Then, we design mechanisms that achieve these lower bounds and demonstrate that they achieve asymptotically better utility compared to existing methods.

[LG-123] Causal Discovery with False Positive Error Control NEURIPS2026

链接: https://arxiv.org/abs/2611.25376
作者: Erik Jahn,Venkat Chandrasekaran,Frederick Eberhardt,Leonard Schulman
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: Accepted at NeurIPS 2026. 26 pages, 8 figures, 1 table, including appendices

点击查看摘要

Abstract:Causal discovery methods produce data-driven hypotheses about causal relationships, which may become new scientific findings or inform decisions about downstream interventions. In such settings, false positive causal claims can be more consequential than false negatives. Yet few methods target false positive error control in causal discovery, and existing error guarantees are often asymptotic and require strong versions of the faithfulness assumption. We introduce the m -conservative PC algorithm and prove that it controls false discovery rate in the linear Gaussian setting under an extremely weak version of the faithfulness assumption. In our framework, false positive error for causal graphs is defined based on the conditional-independence information they entail, yielding a metric that penalizes both false adjacencies and incorrect edge orientations. The parameter m quantifies a trade-off between algorithmic complexity and the strength of the required faithfulness assumption. The m -conservative PC algorithm combines undirected graph learning, a Benjamini-Yekutieli-style procedure for conditional independence testing, and a conservative orientation step that marks unresolved or contradictory edge directions as ambiguous. Our algorithm produces valid outputs and retains error guarantees even in the finite-sample setting, despite possible random errors in the conditional independence tests.

[LG-124] Statistical Machine Learning and Deep Learning Methods for Missing Covariates: An Empirical Evaluation with Applications to Alzheimers Disease Research

链接: https://arxiv.org/abs/2611.25219
作者: Yiqing Pan,Sharon X. Xie,Shujie Ma,Angela L. Jefferson,Panpan Zhang
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:Missing data are common in biomedical research, particularly when biomarkers are costly or require invasive procedures to obtain, such as cerebrospinal fluid measures in Alzheimer’s disease. Although a wide range of missing data methods exists, from traditional statistical approaches to modern machine learning (ML) and deep learning (DL) algorithms, their comparative performance under different missingness mechanisms remains insufficiently understood. In this study, we provide an empirical evaluation of eight methods for handling missing covariates, including complete-case analysis, multiple imputation, ML-based techniques, and DL frameworks such as multilayer perceptrons, variational autoencoders, and generative adversarial networks. We summarize their conceptual foundations, practical implementation considerations, strengths, and limitations. Through extensive simulations, we assess estimation bias, efficiency, standard error accuracy, confidence interval coverage, and computational burden across different missingness mechanisms and proportions. An application to the Vanderbilt Memory and Aging Project further illustrates the implications of different methods for regression analysis in a real-world biomedical setting. Finally, we provide practical recommendations for method selection and highlight remaining challenges that warrant future investigation.

[LG-125] Empirical Variational Bayes Factorization for Longitudinal Matrix Data

链接: https://arxiv.org/abs/2611.25065
作者: Jiakun Li,Eric F. Lock
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注: 46 pages, including 18 pages of supplementary materials

点击查看摘要

Abstract:Collecting high-dimensional data over time for multiple individuals is increasingly common in biomedical research, particularly in studies of disease progression or molecular time-course experiments. Often data are missing or the timepoints are irregularly spaced across individuals. There is a lack of flexible methodology to explore latent patterns of change over time and account for missing data. We propose an empirical variational Bayes (EVB) approach to factorize high-dimensional longitudinal data, in which the underlying factors have temporal correlation over time. The EVB approach is computationally efficient and accommodates flexible covariance structures over time with automatic determination of all hyper-parameters (including the underlying rank, i.e., number of factors). Simulations demonstrate that the approach captures the underlying signal more precisely than other strategies, and performs well with respect to missing data imputation under different scenarios. An application to a gene time-course data for multiple sclerosis treatment provides a factorization with interpretable components over time and improved imputation accuracy for missing data.

[LG-126] Variance-Adaptive Upper Confidence Bounds for Multi-Armed Bandits with Kaplan–Meier Survival Estimation

链接: https://arxiv.org/abs/2611.24988
作者: Yangfan Cui,Andrew Papanicolaou
类目: Machine Learning (stat.ML); Machine Learning (cs.LG); Probability (math.PR)
*备注:

点击查看摘要

Abstract:We study the stochastic multi-armed bandit problem in which the reward associated with each arm is the survival probability at a pre-specified time horizon, and the observed outcome is a right-censored time-to-event variable. Standard bandit algorithms face delayed and partial observations in this setting. Prior adaptations for the Kaplan–Meier (KM) estimator rely on Hoeffding-type bounds that ignore the inherent heteroscedasticity of continuous survival processes. To bridge this gap, we propose KM-UCB, a variance-adaptive policy that unites continuous-time survival analysis with discrete sequential decision-making. By leveraging martingale theory and continuous influence functions, we derive a fixed-cohort empirical Bernstein bound parameterized by Greenwood’s variance formula. For realtime allocation, we retain the Greenwood index and use a survival bound from completed records to review the recommended arm. Under catch-up reporting, the policy has logarithmic expected regret with an arm-specific variance term in the exploration bound. We evaluate the original policy in a dual-layer delayed simulation and a retrospective replay of breast cancer trial data, and evaluate the correction in a separate paired simulation.

[LG-127] Risk-optimal active learning with informative abstention

链接: https://arxiv.org/abs/2611.24986
作者: Fariborz Setoudehtazangi,Geoffrey J. McLachlan
类目: Machine Learning (stat.ML); Machine Learning (cs.LG)
*备注:

点击查看摘要

Abstract:We study label acquisition when an oracle may abstain and its response probability shares parameters with the classifier. A query can then provide information through both a returned label and the response outcome. Under a correctly specified parametric model, we derive the classifier information after adjustment for response-model nuisance parameters. Combining this information with the local curvature of multiclass zero-one risk gives an acquisition criterion for minimizing the leading asymptotic excess risk. We characterize the optimal allocation under a fixed query budget and show how query allocation affects response identifiability. The oracle criterion reduces to response-discounted acquisition when response probability does not depend on the classifier parameters. Simulations examine the resulting query allocations and isolate the contribution of the response-outcome term through an ablation. A Consumer Complaints replay retains observed human abstentions and reveals benchmark labels upon response. Response-aware acquisition has lower mean test error than response-discounted acquisition at the two largest budgets, with paired differences significant after Holm correction. Entropy sampling performs better at the smallest budget. The gains depend on the response mechanism and query budget.

[LG-128] Improving On-board Shape Models Using Sun Sensors and Implicit Representations

链接: https://arxiv.org/abs/2611.24976
作者: Zacharia A. Rudge,Giacomo Acciarini,Dario Izzo,Dominik Dold,Moritz Fieback,Said Hamdioui
类目: Instrumentation and Methods for Astrophysics (astro-ph.IM); Machine Learning (cs.LG)
*备注: 3 page, SPAICE 2026 conference (21-23 October 2026)

点击查看摘要

Abstract:We propose an original, end-to-end approach to generate and refine the onboard 3D shape model of irregular small bodies, such as asteroids, using only low-cost Sun sensor eclipse data and coarse a priori knowledge, tailored for resource-constrained missions like CubeSats. Leveraging EclipseNets, a differentiable neural model of eclipse boundaries, our pipeline first refines 2D silhouettes by learning from real-time entry and exit events during spacecraft orbits, avoiding the limitations of traditional imaging or lidar in bandwidth-starved environments. These silhouettes then supervise a compact neural signed-distance field with periodic activations, enabling incremental updates to a high-fidelity implicit geometry representation capable of capturing fine-scale features and concavities. In simulations of the body Eros under realistic orbital dynamics, the method increases silhouette IoU from 0.70 to 0.87 and produces coherent 3D reconstructions, suggesting our approach could find a place in future autonomous architectures for small-body exploration.

附件下载

点击下载今日全部论文列表