ArXiv Domain 2026-09-07

数据来源：ArXiv Domain

LLM Domain Papers

1. Where Does Harness-Optimization Value Live? Localized Gains and the Budget-Splitting Trap in Self-Evolving LLM Agents

Abstract:A growing body of work improves frozen large language models (LLMs) as agents by evolving their harness: the textual scaffolding around the model, including persona, strategy, format rules, and control heuristics. Existing reflective prompt-evolution methods usually optimize this harness as one flat string. We instead ask where the optimization value actually resides. We introduce HARNESSEVO, which decomposes the harness into four separately evolvable slots: role, task-strategy, tool/format-rules, and reflection/control. Using the same reflective optimizer under an iso-budget setting, we pair this decomposition with leave-one-in and leave-one-out attribution to measure the contribution of each slot. On ALFWorld with a frozen 7B backbone, HARNESSEVO does not significantly improve the overall binary success rate over either the stock harness or flat-string evolution: 0.657 versus 0.642 and 0.642, respectively. However, the slot-level analysis reveals that nearly all useful optimization value is localized in the reflection/control slot, which achieves a leave-one-in gain of +0.119. The other slots are individually null. We further show that uniform budget splitting is harmful: allocating 64 rollouts across four slots leaves only 16 per slot, below the optimizer’s effective search floor, causing every slot to freeze at its empty seed. Concentrating the budget on the high-credit control slot recovers the lost gain, reaching 0.761 with half the split budget. The effect is task-contingent. On WebShop, all slots freeze empty and all methods tie, indicating a genuine absence of recurrent, verbalizable control failures rather than budget starvation. Overall, our results suggest that harness value is localized, uniform budget splitting can be actively harmful, and credit assignment should precede structured agent-evolution.

中文摘要

摘要：越来越多的研究通过改进冻结的大型语言模型（LLMs）作为智能体的方式，提升了它们的表现：这种改进表现为模型周围的文本支架，包括角色设定、策略、格式规则以及控制启发式方法。现有的反思型提示进化方法通常将这一支架作为一个单一的字符串来优化。而我们则探讨优化价值究竟存在于何处。我们提出了HARNESSEVO，将支架拆分为四个可以单独进化的部分：角色、任务策略、工具/格式规则以及反思/控制。在相同的预算条件下，使用同样的反思优化器，我们结合“保留一个”和“剔除一个”的归因方法来衡量每个部分的贡献。在使用冻结的7B骨干网络的ALFWorld实验中，HARNESSEVO在整体二元成功率上并未显著优于原始支架或单字符串进化：分别为0.657对0.642和0.642。然而，部分层级分析显示，几乎所有有用的优化价值都集中在反思/控制部分，该部分在“保留一个”实验中带来了+0.119的增益。其他部分单独来看则没有效果。我们进一步证明，均匀分配预算是有害的：将64次尝试均分给四个部分，每个部分只有16次，低于优化器的有效搜索下限，导致每个部分停留在其空初始状态。将预算集中在高价值的控制部分可以恢复损失的增益，在预算减半的情况下仍可达到0.761。该效果依赖于具体任务。在WebShop中，所有部分都停留为空状态，所有方法表现相同，这表明并非预算不足，而是确实不存在可复现、可语言化的控制失败。总体而言，我们的结果表明：支架价值具有局部性，均匀分配预算可能会带来负面影响，且在结构化智能体进化之前应先进行信用分配。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02889 (Request failed with status code 502)

Authors: Michael Nguyen, Wei Chen Tan, Nurul Aisyah Hassan, Arvind Raman, Li Hua Lim, Ahmad Faiz Razak

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02889.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02889

Published: 2026-09-07T01:18:07.384Z

2. Bounded Personas Match Retrieval on Classification but Not Regression for a Frozen Agent

Abstract:A personalized language agent must convert a user’s interaction history into behavior on each new request at inference time. Two strategies dominate. Retrieval pulls a few of the user’s most relevant past items into the prompt, which is accurate but pays a per-query selection and context cost that grows with the history. Distillation instead compresses the history once into a compact natural-language persona, which is bounded, query-independent, and interpretable, but is widely assumed to sacrifice accuracy. Whether, and on which tasks, a distilled persona can match retrieval has not been characterized cleanly. We introduce PersonaLink, a training-free method that distills a user’s history into a bounded three-field persona and recursively refines it: each pass self-evaluates the frozen agent on a held-out slice of the user’s own labeled history, rewrites the persona from its errors, and keeps the result only when it does not regress on that slice. Because every comparison shares one frozen 7B backbone and differs only in what is placed in context, the design isolates the effect of representation from that of the model. The result is a clear task-type asymmetry. On 200 users of LaMP-2 (15-way news categorization), PersonaLink reaches 0.745-0.755 accuracy, statistically indistinguishable from BM25 retrieval (0.760-0.765).

中文摘要

摘要：个性化语言代理必须在推理时将用户的交互历史转换为每个新请求的行为。有两种主导策略。检索策略会将用户最相关的一些历史项目拉入提示中，这很准确，但对每次查询都要进行选择，并且上下文成本会随着历史长度增长。蒸馏策略则将历史一次性压缩为紧凑的自然语言角色，这是有限的、与查询无关且可解释的，但普遍认为会牺牲准确性。目前尚未清楚地表征蒸馏角色在何种任务上能够匹配检索。我们提出PersonaLink，一种无需训练的方法，将用户历史蒸馏为有限的三字段角色并递归优化：每一轮都对冻结的代理在用户自己标注的保留历史片段上进行自我评估，根据错误重写角色，并仅在该片段上不退化时保留结果。由于每次比较共享同一个冻结的7B骨干模型，并且上下文中仅放置不同内容，该设计将表示效果与模型效果分离开来。结果显示出明显的任务类型不对称性。在LaMP-2的200名用户（15类新闻分类）上，PersonaLink达到0.745-0.755的准确率，在统计上与BM25检索（0.760-0.765）无显著差异。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02890 (Request failed with status code 502)

Authors: JaeHa Yoon, Minjun Park, Seoyeon Kim, Jiwoo Lee, Hyunwoo Choi, Dohyun Kang

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02890.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02890

Published: 2026-09-07T01:18:07.384Z

3. Counterexamples as Feedback for Agent Self-Correction

Abstract:Single-turn code-generation metrics understate a central property of deployed agents: whether they can repair a wrong artifact after receiving concrete feedback. This paper presents A-CEGIS, a lightweight framework that uses counterexamples as feedback for evaluating multi-turn refinement in natural-language-to-regex synthesis. An agent proposes a regex, a deterministic oracle checks it under full-match semantics, and compact false-positive or false-negative witnesses guide the next turn. On 30 NL-RX-Turk tasks, diagnostic counterexample feedback solves 90\% of tasks within a four-turn ablation budget, compared with 17% for zero-shot generation, 27% for generic self-correction, and 23% for error-only feedback. In a full diagnostic run with hardening, all tasks are solved on the hidden set by the final turn, with mean time-to-success of 2.7 turns and robust success of 77% after targeted probing. These results show that A-CEGIS measures how efficiently an agent improves across turns while adding a practical robustness check beyond the original held-out cases.

中文摘要

摘要：单轮代码生成指标低估了部署代理的一个核心特性：在收到具体反馈后是否能够修复错误的产物。本文提出了 A-CEGIS，这是一种轻量级框架，使用反例作为反馈来评估自然语言到正则表达式合成中的多轮优化。代理提出一个正则表达式，确定性判定器在全匹配语义下进行检查，并由简明的假阳性或假阴性证据引导下一轮。在 30 个 NL-RX-Turk 任务中，诊断反例反馈在四轮消融预算内解决了 90% 的任务，而零样本生成、通用自我修正和仅错误反馈分别解决了 17%、27% 和 23% 的任务。在完整的诊断运行和加固过程中，最终轮在隐藏集上解决了所有任务，平均成功所需轮次为 2.7 轮，在针对性探测后实现了 77% 的稳健成功率。这些结果表明，A-CEGIS 能够衡量代理在多轮改进中的效率，同时为原始保留案例之外提供了实用的鲁棒性检验。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02892 (Request failed with status code 502)

Authors: Sidhesh Badrinarayan, Adithya Parthasarathy

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02892.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02892

Published: 2026-09-07T01:18:07.384Z

4. Probe Generalization as Subspace Selection for OOD Deception Detection

Abstract:Linear probes can be used to detect behaviors and concepts inside language model activations, but may fail to transfer to out-of-distribution examples. When studying the generalization performance of Llama-3.1-8B-Instruct probes over 3 held-out deception detection datasets, we find that projecting inputs onto a small subset of principal components (PCs) from the training distribution of activations enables cross-domain transfer that nearly matches the performance of probes trained directly on the test distribution. Furthermore, we find that PC interpretations can be used to find a subset of those transferable PCs. By using an LLM judge to score each PC on whether its most/ least activating examples imply a transferable deception direction, then probing on the highest-scoring PCs, we close the baseline-to-oracle gap by 78% on Insider Trading Report and by 25% on Sandbagging. The directions a source probe weights heavily appear to encode source-specific surface features, while the directions that actually transfer appear to encode the same contrast more abstractly, in a way natural language descriptions can capture. Broadly, our results suggest that the OOD robustness of probes is largely determined by subspace selection.

中文摘要

摘要：线性探针可以用于检测语言模型激活中的行为和概念，但可能无法很好地迁移到分布外（OOD）的示例。当研究 Llama-3.1-8B-Instruct 探针在 3 个未使用的欺骗检测数据集上的泛化性能时，我们发现，将输入投射到训练激活分布的一小部分主成分（PCs）上，可以实现跨领域迁移，其性能几乎与在测试分布上直接训练的探针相当。此外，我们发现 PC 解释可以用于找到这部分可迁移 PCs 的子集。通过使用大型语言模型（LLM）评估每个 PC，其依据是 PC 上最/最不激活的示例是否暗示可迁移的欺骗方向，然后在评分最高的 PCs 上进行探针，我们在 Insider Trading Report 上将基线与理想差距缩小了 78%，在 Sandbagging 上缩小了 25%。源探针权重较重的方向似乎编码了源特定的表面特征，而实际上可以迁移的方向似乎以更抽象的方式编码了相同的对比，这种方式可以被自然语言描述捕捉。总体而言，我们的结果表明，探针的 OOD 鲁棒性在很大程度上取决于子空间选择。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02893 (Request failed with status code 502)

Authors: Daniel Yoo, Adrians Skapars

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02893.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02893

Published: 2026-09-07T01:18:07.384Z

5. R$^{2}$Adapter: A Routing and Rewriting Adapter for Efficient Hybrid RAG

Abstract:Retrieval-Augmented Generation (RAG) has become a prevailing paradigm for enhancing Large Language Models (LLMs) with non-parametric knowledge. Vanilla RAG efficiently handles simple queries but struggles with relational or multi-hop reasoning. Graph-based RAG alleviates this issue but incurs higher inference complexity and latency. In practice, user queries can differ significantly in their complexity, rendering a fixed RAG strategy suboptimal. However, existing hybrid text-graph RAG methods typically rely on heuristic and LLM-based routing, resulting in unnecessary overhead and strong dependence on the underlying LLM. To address these challenges, we propose R$^{2}$Adapter, a lightweight plug-in Routing and Rewriting Adapter designed to allocate queries between vanilla and graph-based RAG dynamically. By routing only the queries that genuinely benefit from graph-based reasoning, R$^{2}$Adapter reduces unnecessary graph retrieval overhead. Additionally, uncertain graph-routed queries are rewritten to better expose their multi-hop reasoning requirements, improving retrieval quality without additional supervision. Extensive experiments on three multi-hop QA benchmarks demonstrate that R$^{2}$Adapter reduces graph-based RAG usage by up to 59% while maintaining comparable answer accuracy. This adapter is model-agnostic and can be seamlessly integrated into diverse vanilla and graph-based RAG pipelines, providing an efficient and adaptive solution for hybrid RAG systems.

中文摘要

摘要：检索增强生成（RAG）已成为增强大型语言模型（LLMs）非参数知识的主要范式。普通 RAG 能够高效处理简单查询，但在关系型或多跳推理方面表现不足。基于图的 RAG 可以缓解此问题，但会增加推理复杂性和延迟。在实际应用中，用户查询的复杂度可能差异很大，使得固定的 RAG 策略并不理想。然而，现有的混合文本-图 RAG 方法通常依赖于启发式和基于 LLM 的路由，导致不必要的开销，并且高度依赖底层 LLM。为了解决这些挑战，我们提出了 R$^{2}$Adapter，一种轻量级的插件式路由和改写适配器，旨在动态地在普通 RAG 和基于图的 RAG 之间分配查询。通过仅将真正受益于基于图推理的查询进行路由，R$^{2}$Adapter 减少了不必要的图检索开销。此外，对于不确定的图路由查询，R$^{2}$Adapter 会进行改写以更好地揭示其多跳推理需求，从而在无需额外监督的情况下提升检索质量。在三个多跳问答基准测试中的大量实验表明，R$^{2}$Adapter 可以在保持相似答案准确率的同时，将基于图的 RAG 使用量最多减少 59%。该适配器与模型无关，可无缝集成到各种普通和基于图的 RAG 流水线中，为混合 RAG 系统提供高效且自适应的解决方案。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02894 (Request failed with status code 502)

Authors: Yucan Guo, Miao Su, Saiping Guan, Long Bai, Zhongni Hou, Zixuan Li, Xiaolong Jin, Jiafeng Guo, Xueqi Cheng

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02894.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02894

Published: 2026-09-07T01:18:07.384Z

6. BharatGather: A Culturally-Informed Benchmark Dataset for Misinformation and Fake News Detection in Indian Public Events

Abstract:Large-scale public events, such as religious festivals, political rallies, and cultural gatherings, are increasingly vulnerable to the rapid dissemination of misinformation, posing substantial risks to public safety and social cohesion. While automated fake news detection has seen significant methodological progress, existing benchmarks frequently fail to capture the socio-cultural nuances and event-specific dynamics characteristic of the Indian context. This paper introduces BharatGather, a curated, multi-source dataset specifically engineered for binary misinformation classification within the ecosystem of Indian mass gatherings. The corpus comprises 14,646 records constructed through a hybrid pipeline involving systematic web scraping of prominent fact-checking platforms, multimedia transcript extraction, and Large Language Model (LLM)-mediated synthetic augmentation to ensure narrative diversity. By providing a resource tailored to the unique complexities of event-aware misinformation in India, this work facilitates the development of culturally informed detection systems and establishes a rigorous benchmark for evaluating their performance in high-stakes public environments.

中文摘要

摘要：大规模公共活动，如宗教节日、政治集会和文化聚会，越来越容易受到错误信息快速传播的影响，对公共安全和社会凝聚力构成重大风险。尽管自动化假新闻检测在方法上取得了显著进展，但现有基准往往未能捕捉印度背景下的社会文化细微差别和事件特定动态。本文提出了BharatGather，这是一个精心策划的、多来源的数据集，专门用于印度大型集会生态系统中的二元错误信息分类。该语料库包含14,646条记录，通过结合系统性网页抓取主要事实核查平台、多媒体文本提取以及大语言模型（LLM）媒介的合成增强的混合流程构建，以确保叙事多样性。通过提供一个针对印度事件感知错误信息独特复杂性的资源，本研究有助于开发文化知情的检测系统，并为在高风险公共环境中评估其性能建立了严格的基准。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02895 (Request failed with status code 502)

Authors: Parth Bramhecha, Smit Deshmukh, Sairaj Bodhale, Adwait Borate, Raviraj Joshi

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02895.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02895

Published: 2026-09-07T01:18:07.384Z

7. PiPMRE: A Pipeline Based on Language Model for Medical Relation Extraction

Abstract:Medical relation extraction (MRE) is commonly known for extracting entities and their relations jointly from a medical text, which has attracted considerable attention in recent years. Previous studies treat MRE as a sequence tagging task, which results in either a challenging design of the tagging schema or a failed extraction of multiple relations, due to intricate relationships among medical entities. In this work, we review the task from the linguistic perspective and propose a novel pipeline framework, PiPMRE, developed on language models to enhance MRE performance. Specifically, PiPMRE consists of a relation generator and a relation filter. Given a text, the generator first yields multiple relational triplets, and then the filter scores each triplet and retains only those that pass the borderline as the final results. Implementing PiPMRE requires no tagging schema; instead, we use a simple template to reformulate the input text, ensuring that entities and relations are generated in a contextual order. Extensive experimental results on two public datasets demonstrate the advancement of PiPMRE. It surpasses the previous state-of-the-art by an average of 5.6 recall points and 4.4 accuracy points. PiPMRE’s superiority is also demonstrated in few-shot settings.

中文摘要

摘要：医疗关系提取（MRE）通常指从医疗文本中联合提取实体及其关系，近年来引起了广泛关注。以往的研究将MRE视为序列标注任务，由于医疗实体之间关系复杂，这导致标注方案设计困难或多重关系提取失败。在本工作中，我们从语言学角度回顾该任务，并提出了一种基于语言模型的新型管道框架PiPMRE，以提升MRE性能。具体而言，PiPMRE由关系生成器和关系过滤器组成。对于给定文本，生成器首先生成多个关系三元组，然后过滤器对每个三元组进行评分，仅保留通过边界线的三元组作为最终结果。实现PiPMRE无需标注方案；相反，我们使用简单模板重新表述输入文本，以确保实体和关系按上下文顺序生成。在两个公开数据集上的大量实验结果表明了PiPMRE的先进性。其召回率平均提升5.6点，准确率平均提升4.4点，超越了之前的最新技术。PiPMRE在少样本设置下的优越性同样得到验证。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02896 (Request failed with status code 502)

Authors: Jiaxin Duan, Fengyu Lu, Junfei Liu

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02896.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02896

Published: 2026-09-07T01:18:07.384Z

8. Margins, Not Windows: Training-Free Per-Step Lossy Speculative Decoding

Abstract:Speculative decoding accelerates LLM inference by drafting candidate tokens and verifying them in parallel. Tree-attention drafters such as EAGLE-3 are widely adopted, yet typically hold two decisions fixed: (1) a strict token-match verification rule and (2) a static draft-tree shape. Prior work relaxes each in isolation under limiting assumptions: long draft chains for training-free lossy verification, and adaptive tree shaping under a fixed token budget. We introduce AdaptiveSpec, a training-free per-step speculative decoding method that adapts both decisions from internal signals already produced during decoding. A per-step margin rule promotes a mismatched draft-proposed token when the ratio of the target’s probability on the drafted token to its top-1 probability exceeds a threshold with no dependence on draft length or underlying drafter architecture. A per-step tree policy adjusts the draft tree’s depth, width, and node count directly from a fused signal of draft top-1 confidence and a rolling acceptance history capturing recent draft-target agreement, allowing the total draft count to vary rather than only be redistributed. The two adaptations operate on orthogonal axes and compound in effect. Implemented on the SGLang production-grade serving engine, AdaptiveSpec improves throughput over the state-of-the-art autoregressive speculative decoding method EAGLE-3 by up to 56%, recovering 93% to fully lossless task accuracy across GSM8K, MATH-500, and HumanEval on three target models (DeepSeek-R1-Distill-Llama-8B, Llama-3.1-8B-Instruct, Qwen3-8B).

中文摘要

摘要：投机性解码通过并行生成候选标记并验证它们来加速大型语言模型（LLM）的推理流程。诸如 EAGLE-3 的树形注意力草稿器被广泛采用，但通常保持两个决策不变：（1）严格的标记匹配验证规则，以及（2）静态草稿树形状。以往的工作在有限假设下分别放宽了每个决策：训练无关的有损验证使用较长的草稿链，而固定标记预算下的自适应树形调整则是另一种方法。我们提出了 AdaptiveSpec，一种无需训练的逐步投机性解码方法，其可以根据解码过程中已生成的内部信号自适应调整这两个决策。逐步边际规则会在草稿标记到目标标记的概率比其最高概率超过阈值时，促进不匹配的草稿提出标记，而无需依赖草稿长度或底层草稿器架构。逐步树策略根据草稿顶1置信度与滚动接受历史融合生成的信号，直接调整草稿树的深度、宽度和节点数，从而允许总草稿数量变化，而不仅仅是重新分配。这两种适应在正交轴上操作，并产生叠加效果。在 SGLang 生产级服务引擎上实现时，AdaptiveSpec 比最先进的自回归投机性解码方法 EAGLE-3 提高了最多 56% 的吞吐量，并在三种目标模型（DeepSeek-R1-Distill-Llama-8B、Llama-3.1-8B-Instruct、Qwen3-8B）上恢复了 GSM8K、MATH-500 和 HumanEval 上任务精度的 93% 的完全无损水平。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02897 (Request failed with status code 502)

Authors: Oszkár Urbán, Young D. Kwon, Stylianos I. Venieris, Cecilia Mascolo

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02897.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02897

Published: 2026-09-07T01:18:07.384Z

9. Distilled Rapid Embedding Transfer (DRET): Parameter-Efficient Biomedical Domain Adaptation via Priority-Based Embedding Transfer

Abstract:Large domain-specific language models such as BioBERT and ClinicalBERT achieve strong performance on biomedical NLP tasks, but their computational demands make them impractical for many real-world deployments. General-purpose, parameter-efficient models such as DistilBERT are lightweight yet lack the domain knowledge required for specialized tasks such as PICO (Population, Intervention, Comparison, Outcome) classification. We introduce Distilled Rapid Embedding Transfer (DRET), a knowledge-transfer paradigm that injects biomedical domain knowledge from large specialized models into a smaller general-purpose model without retraining on the original specialized corpora. DRET is developed as an iterative family of strategies: a unified tokenizer-merge strategy (DRET 1.x), hybrid embedding averaging (DRET 2.0), and a priority-based embedding-transfer mechanism (DRET 3.x) that hierarchically selects embeddings from the most authoritative source models, further combined with embedding-layer freezing, differential learning rates, label propagation, and imbalance-aware loss functions (DRET 4.x). We evaluate DRET on token-level PICO classification using the EBM-NLP corpus under severe class imbalance, across a twelve-metric battery. DRET-enhanced DistilBERT (66M parameters) attains balanced accuracy, recall, and ROC-AUC competitive with, and on several class-wise metrics exceeding, models an order of magnitude larger, while retaining DistilBERT’s efficiency. We further show that transfer occurs at the embedding level through cosine-similarity, semantic-shift, and t-SNE analyses. DRET offers a scalable, resource-efficient route to near-domain-expert performance for biomedical text mining, with direct application to automated systematic literature reviews and clinical decision support.

中文摘要

摘要：大型特定领域语言模型如 BioBERT 和 ClinicalBERT 在生物医学自然语言处理任务中表现出色，但其计算需求使其在许多实际部署中不切实际。通用、高参数效率的模型如 DistilBERT 虽然轻量化，但缺乏完成 PICO（人群、干预、比较、结果）分类等专门任务所需的领域知识。我们提出了蒸馏快速嵌入迁移（Distilled Rapid Embedding Transfer, DRET），一种知识迁移范式，可将大型专用模型中的生物医学领域知识注入较小的通用模型中，而无需在原始专用语料上重新训练。DRET 被开发为一系列迭代策略：统一分词器合并策略（DRET 1.x）、混合嵌入平均（DRET 2.0）、以及基于优先级的嵌入迁移机制（DRET 3.x），该机制从最权威的源模型中分层选择嵌入，并进一步结合嵌入层冻结、差异化学习率、标签传播和不平衡感知损失函数（DRET 4.x）。我们在 EBM-NLP 语料库上，在严重类别不平衡情况下，对 token 级 PICO 分类进行了十二项指标评估。经过 DRET 增强的 DistilBERT（66M 参数）在平衡准确率、召回率和 ROC-AUC 上与大十倍规模的模型竞争，并在若干类别指标上超过这些模型，同时保持了 DistilBERT 的高效性。我们进一步通过余弦相似度、语义漂移和 t-SNE 分析展示了迁移发生在嵌入层。DRET 提供了一种可扩展、资源高效的方式，使生物医学文本挖掘达到近领域专家水平，直接应用于自动系统文献综述和临床决策支持。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02898 (Request failed with status code 502)

Authors: Girish Sundaram, Daniel Berleant

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02898.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02898

Published: 2026-09-07T01:18:07.384Z

10. Contamination Inflates Scores but Rarely Reorders Large Language Model Leaderboards

Abstract:Benchmark contamination, the leakage of test items into training data, is widely described as a threat to the reliability of large language model (LLM) leaderboards. We argue that this concern conflates two distinct questions: whether contamination inflates absolute scores, and whether it reorders the ranking of models. We recast contamination as a violation of anchor-item invariance and measure it through the differential functioning of original versus semantically equivalent paraphrased items, a within-item contrast that holds the measured skill fixed and isolates memorization from capability. Using per-instance responses from 47 publicly released models and 74 models finetuned with a known dose of contamination, across four benchmarks (ARC, GSM8K, HellaSwag, MMLU), we first calibrate the measure against ground truth: it recovers injected contamination dose-responsively (a corrected effect of +0.187 accuracy points for test-set leakage) and never flags a negative-control model trained only on the legitimate training split (-0.012). We then quantify leaderboard impact: the rank correlation between a standard leaderboard and a paraphrase-controlled leaderboard is 0.997, and a sensitivity analysis shows that the observed differential contamination is far below the level needed to move rankings, with only 3 of 188 model-by-benchmark cases showing differential contamination corroborated across two references. Contamination among these public models is therefore largely uniform: it inflates absolute scores without reordering the leaderboard, and ranking distortion requires the rare case of differential contamination. We provide a calibrated invariance audit, released as a reference implementation, and recommend that leaderboards report paraphrase-controlled rankings alongside confidence intervals.

中文摘要

摘要：基准污染，即测试项目泄漏到训练数据中，被广泛认为是大型语言模型（LLM）排行榜可靠性的威胁。我们认为，这一担忧混淆了两个不同的问题：污染是否会抬高绝对分数，以及它是否会重新排序模型排名。我们将污染重新定义为锚点项不变性的违反，并通过原始项目与语义等价的改写项目的差异功能进行测量，这种项内对比固定了测量技能，并将记忆与能力区分开来。利用47个公开发布模型和74个已知受污染微调模型的每实例响应数据，跨四个基准（ARC、GSM8K、HellaSwag、MMLU），我们首先将该测量方法与真实情况进行校准：它以剂量响应的方式恢复注入的污染（测试集泄漏的修正效应为+0.187准确率点），并且从未标记仅在合法训练集上训练的负控制模型（-0.012）。然后我们量化排行榜影响：标准排行榜与改写控制排行榜之间的排名相关性为0.997，并且敏感性分析显示，观察到的差异性污染远低于移动排名所需的水平，188个模型-基准案例中仅有3个显示出在两个参考中得到证实的差异性污染。因此，这些公开模型中的污染大体上是均匀的：它抬高了绝对分数但不改变排行榜顺序，而排名扭曲需要罕见的差异性污染。我们提供了校准的不变性审计，作为参考实现发布，并建议排行榜报告改写控制排名及其置信区间。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02899 (Request failed with status code 502)

Authors: Xingyao Xiao, Yihong Cheng

Categories: cs.CL

PDF URL: https://arxiv.org/pdf/2609.02899.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02899

Published: 2026-09-07T01:18:07.384Z

Agent Domain Papers

1. Structure and Implementation of New Practical English Textbooks Driven by Artificial Intelligence

Abstract:Artificial intelligence is changing the form of applied English materials from fixed paper sequences to adaptive learning systems that can diagnose learners, recommend tasks, and provide formative feedback. This paper studies the structure and application of a new practical English textbook driven by artificial intelligence. A five-layer architecture is proposed: knowledge mapping, learner profiling, task generation, feedback orchestration, and teacher-side governance. A prototype was tested on 186 non-English-major undergraduates for eight weeks of teaching. Compared with a static digital textbook, the proposed system increased the unit completion accuracy from 72.4% to 84.9%, raised the average score for speaking tasks by 10.8 points, and reduced the teacher’s correction time by 31.6%. Therefore, an AI-driven textbook can maintain the stability of the curriculum while providing personalised learning paths, rich practice materials and traceable classroom data.

中文摘要

摘要：人工智能正在将应用型英语教材的形式从固定的纸质序列转变为能够诊断学习者、推荐任务并提供形成性反馈的自适应学习系统。本文研究了由人工智能驱动的新型实用英语教材的结构与应用。提出了五层架构：知识映射、学习者画像、任务生成、反馈协调和教师端管理。对186名非英语专业本科生进行了为期八周的教学原型测试。与静态数字教材相比，该系统使单元完成准确率从72.4%提高至84.9%，口语任务平均分提高10.8分，并将教师批改时间减少了31.6%。因此，人工智能驱动的教材可以在保持课程稳定性的同时，提供个性化学习路径、丰富的练习材料和可追踪的课堂数据。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02981 (Request failed with status code 502)

Authors: Ya Wang, Lei Zhang, Xueguang Yang, Bo Chen

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.02981.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02981

Published: 2026-09-07T01:18:49.920Z

2. MasterControl Seventeen Every Time

Abstract:We study a governed approach to enterprise analytics: a language model interprets the question, while deterministic policy selects and runs a pre-approved analytical program that returns both results and evidence. We show that this restriction can remain expressive within a defined analytical class, using relational operations plus aggregation, comparison, windows, ranking, and similarity. Fixed meaning, policy, data, and execution rules also make results replayable. Across 440 runs, three 8B models generated SQL and selected tools at runtime, while Qwen3-8B interpreted intent only and policy executed the approved program. None of 330 runtime-planning episodes matched the full answer-and-evidence contract across all test datasets; the policy-executed analyzer matched 110 of 110. This is a configuration-specific result, not evidence that runtime agents cannot succeed under other designs.

中文摘要

摘要：我们研究了一种受控的企业分析方法：语言模型解释问题，而确定性策略选择并运行预先批准的分析程序，该程序返回结果和证据。我们展示了这种限制在定义的分析类别内仍然可以保持表达能力，使用关系操作加上聚合、比较、窗口、排序和相似性。固定的意义、策略、数据和执行规则还使结果可重放。在440次运行中，三种8B模型在运行时生成SQL并选择工具，而Qwen3-8B仅解释意图，策略执行批准的程序。在330个运行时规划的实例中，没有一个完全匹配所有测试数据集的完整答案与证据合同；策略执行的分析器匹配了110/110次。这是特定配置下的结果，不是证明运行时代理在其他设计下无法成功的证据。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03209 (Request failed with status code 502)

Authors: MasterControl AI Lab

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03209.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03209

Published: 2026-09-07T01:18:49.920Z

3. Speculative Macro Commit for Faster Tool-Using Agents

Abstract:Tool-using LLM agents spend wall-clock time not only on model inference but also in serial action—observation turns, where each tool call, environment transition, and observation can delay subsequent decisions. We introduce \textbf{Speculative Macro Commit} (SMC), a runtime mechanism for a two-tier agent system: a large authoritative actor model produces the official trajectory, while a faster speculative drafter model continuously predicts and executes future action chains on an isolated environment snapshot. SMC mines recurring multi-action skeletons from training traces and stores them in a macro library used to match against action chains predicted by the drafter at runtime. When the actor’s next tool call matches the first drafted action, SMC commits the remaining pre-executed draft steps, together with their observations, to the official trajectory. Using Qwen3.5-27B INT4 as the authoritative actor model and Qwen3.5-4B as the speculative drafter model, SMC matches the sequential agent’s overall accuracy while reducing latency by 10.23\% over the Speculative Actions (SA) baseline and 18.59\% over sequential execution on the $\tau^2$-Bench Telecom subset. On AppWorld, SMC reduces wall time by 7.7\% over SA baseline and 44.9\% over sequential execution, with a small reduction in task completion. Overall, SMC provides a practical way to reuse multi-step speculative execution and reduce agent latency beyond single-step speculative actions. Our code is publicly available \href{this https URL}{\textcolor{magenta}{here}}.

中文摘要

摘要：使用工具的LLM代理不仅花费大量时间在模型推理上，还在串行操作中——观察回合，每个工具调用、环境切换和观察都可能延迟后续决策。我们引入了\textbf{Speculative Macro Commit}（SMC），这是一种两层代理系统的运行时机制：大型权威演员模型生成官方轨迹，而更快的推测起草模型则持续预测并执行隔离环境快照中的未来动作链。SMC从训练轨迹中挖掘重复的多动作骨架，并将其存储在宏库中，用于与生成器在运行时预测的动作链匹配。当actor的下一个工具调用与第一个草拟动作匹配时，SMC将剩余的预执行草稿步骤及其观察数据提交到官方轨迹中。以Qwen3.5-27B INT4为权威行为者模型，Qwen3.5-4B作为推测性设计模型，SMC在$\tau^2$-Bench电信子集上，匹配顺序行动（SA）基线的延迟10.23\%，在$\tau^2$-Bench电信子集的顺序执行中降低18.59%。在AppWorld上，SMC较SA基线减少7.7%的墙壁时间，较顺序执行减少44.9%，任务完成度略有下降。总体而言，SMC提供了一种实用方式，可重用多步推测执行，并降低代理延迟，超越单步推测操作。我们的代码公开发布 \href{this https URL}{\textcolor{magenta}{here}}。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03236 (Request failed with status code 502)

Authors: Zeyu Liu, Souvik Kundu, Peter A. Beerel

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03236.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03236

Published: 2026-09-07T01:18:49.920Z

4. Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory

Abstract:Distributed LLM-agent teams can read the latest shared facts and still act on an obsolete plan. A planner may derive an action from requirement $r_3$, another agent may commit $r_4$, and an executor may receive $r_4$ without replacing the plan derived from $r_3$. We call this \emph{stale-plan execution}: state freshness does not establish that the plan authorizing an action remains valid. We introduce PlanFence, a dependency-scoped action-validation protocol. Plans cite the exact public records they used, and an executor validates only the records that can affect the pending external action, replanning once or blocking when validation is incomplete. In 30 controlled live workflows with a post-plan revision, a freshness-only executor acts on the obsolete plan in every task, whereas PlanFence completes all tasks without an invalid action. Controlled replay reveals two conditional boundaries: proactive synchronization yields lower coordination stall at low churn, while PlanFence avoids repeated update-path coordination as churn grows and avoids validating unrelated state as the shared keyspace grows. These are controlled safety and systems-cost results, not general task-accuracy gains.

中文摘要

摘要：分布式大型语言模型代理团队可以读取最新的共享事实，但仍可能在一个过时的计划上采取行动。一个计划者可能会从需求 $r_3$ 推导出一个动作，另一个代理可能会提交 $r_4$，而执行者可能会接收 $r_4$，而不替换由 $r_3$ 推导出的计划。我们将这种情况称为\emph{过时计划执行}：状态的新鲜度并不能确保授权某个动作的计划仍然有效。我们引入了 PlanFence，一种依赖范围内的动作验证协议。计划会引用其使用的确切公共记录，而执行者只验证可能影响待执行外部动作的记录，在验证不完整时要么重新规划一次，要么阻塞。在 30 个经过控制的现场工作流中进行计划后修订时，仅依赖新鲜度的执行者在每个任务中都会基于过时计划采取行动，而 PlanFence 则完成所有任务而不会产生无效动作。受控重放揭示了两个条件边界：在低变动情况下，主动同步可降低协调停滞，而随着变动增加，PlanFence 避免了重复的更新路径协调，并避免在共享键空间增大时验证无关状态。这些是受控的安全性和系统成本结果，而非一般任务准确性提升。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03340 (Request failed with status code 502)

Authors: Evan Chen, Shiqiang Wang, Christopher G. Brinton

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03340.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03340

Published: 2026-09-07T01:18:49.920Z

5. A Prompt-Engineering Approach to Develop Scalable, Flexible, and Real-Time Hybrid Micro-Level Personalization in a General Purpose AI Teaching Assistant

Abstract:Artificial intelligence (AI) teaching assistants powered by large language models (LLMs) offer scalable educational support but often provide limited personalization. This study presents a prompt-engineering-based framework for personalizing general-purpose LLM/RAG-based AI teaching assistants such as Jill Watson across academic disciplines and courses. The framework adapts responses using six learner-specific dimensions: self-assessment, abstraction preference, verbosity preference, perceptual orientation, information processing style, and level of understanding, yielding 96 distinct learner profiles. Student queries are additionally analyzed using Bloom’s Taxonomy to estimate cognitive complexity at the interaction level. Learner attributes and cognitive assessments are encoded in structured prompts that condition the LLM without requiring model retraining. The framework is evaluated through experiments using NLP metrics and a human study with five participants. Results show perceived differences in response style and structure across personalization conditions, with statistical analyses identifying learner attributes associated with measurable response changes. These findings provide preliminary evidence that prompt-based personalization can support adaptive behavior in LLM-powered educational agents.

中文摘要

摘要：由大型语言模型（LLM）驱动的人工智能（AI）教学助手能够提供可扩展的教育支持，但通常个性化程度有限。本研究提出了一种基于提示工程的框架，用于在各学科和课程中个性化通用的LLM/RAG AI教学助手，如Jill Watson。该框架通过六个学习者特定维度来调整响应：自我评估、抽象偏好、啰嗦程度偏好、感知取向、信息处理风格和理解水平，从而生成96个不同的学习者档案。此外，学生的问题还使用布鲁姆认知分类法进行分析，以估算互动层面的认知复杂度。学习者属性和认知评估被编码为结构化提示，以条件化LLM，而无需重新训练模型。该框架通过使用NLP指标的实验和与五名参与者的人类研究进行评估。结果显示，在不同个性化条件下，响应风格和结构存在感知差异，统计分析识别出与可测量响应变化相关的学习者属性。这些发现提供了初步证据，表明基于提示的个性化可以支持由LLM驱动的教育代理的自适应行为。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03402 (Request failed with status code 502)

Authors: Saptarshi Basu, Sandeep Kakar, Ashok Goel

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03402.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03402

Published: 2026-09-07T01:18:49.920Z

6. Caught in the Story: Narrative Captivity in Multi-turn LLMs Conversation

Abstract:People increasingly turn to large language models (LLMs) for everyday advice, making ethically charged interpersonal problems a practical moral-advisory context. Most prior work has studied this context through single-turn judgments or pressure-laden rebuttals, assumptions that poorly match how guidance is sought in real-world contexts. These assumptions leave unclear whether narration alone, without an explicit opposing position, can shift model judgments during multi-turn moral consultation. Yet real-world moral-conflict conversation often elicits one party’s self-justifying account, which can unfold over multiple turns and create information asymmetry. We introduce \textbf{narrative captivity}, a failure mode in which a model treats an unopposed one-sided account as complete and aligns with the narrator’s interpretation without seeking missing perspectives. To measure this phenomenon, we build a benchmark of $5{,}078$ interpersonal-conflict scenarios spanning six moral dimensions. Across 17 LLMs, narrative captivity is widespread: end-state judgments under multi-turn narration shift by 25 percentage points on average beyond the matched single-turn baseline. Stage-level analysis identifies preference optimization as a major contributor, while four inference-time strategies provide only partial mitigation. We hope our project fosters LLM advisors that preserve independent judgment in real-world consultation.

中文摘要

摘要：人们越来越多地依赖大型语言模型（LLM）来获取日常建议，这使得涉及伦理问题的人际关系问题成为一个实际的道德咨询场景。以往的大多数研究通过单轮判断或压力驱动的反驳来研究这一场景，这些假设与现实世界中获取指导的方式不匹配。这些假设使得尚不清楚单纯叙述——即没有明确对立立场的情况下——是否能够在多轮道德咨询中改变模型的判断。然而，现实世界中的道德冲突对话通常会引发表达自我辩护的一方，该叙述可能跨越多轮并造成信息不对称。我们提出了\textbf{叙事俘获}概念，这是模型将无对抗的单方面叙述视为完整，并在未经寻求缺失视角的情况下与叙述者的解释保持一致的一种失效模式。为了测量这一现象，我们构建了一个包含 $5{,}078$ 个跨越六个道德维度的人际冲突场景的基准数据集。在17个LLM中，叙事俘获现象普遍存在：在多轮叙述下的最终判断比匹配的单轮基线平均偏移25个百分点。阶段层面的分析表明偏好优化是一个主要因素，而四种推理时策略仅能提供部分缓解。我们希望我们的项目能够促进LLM顾问在现实咨询中保持独立判断。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03407 (Request failed with status code 502)

Authors: Yuhe Wu, Guangyu Wang, Yujie Chen, Jiatong Zhang, Yuran Chen, Yutong Zhang, Xiyin Cheng, Wenpeng Cao, Zhuang Liu, Guang Zhang

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03407.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03407

Published: 2026-09-07T01:18:49.920Z

7. Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection

Abstract:LLM-empowered paper-code discrepancy detection has received growing concern since the scaling of research submissions exceeds the manual review capability. However, the limited context capacity and one-sided discrepancy detection of existing single-agent LLM paradigms lead to an inferior recall performance in detecting discrepancies. In this paper, we propose Dude, the first Dual-Detection Multi-Agent System for paper-code discrepancy detection. We discover that the granularity asymmetry of the paper-language and code-language introduces over-interpretation and over-reporting challenges in a multi-agent system design for discrepancy detection, resulting in increasing false positives. To address this, we propose a granularity-aligned negotiation and a two-stage salience-filtering mechanism in Dude, which effectively prevents agents from falsely reporting discrepancies. Experimental results in real-world paper-code discrepancy datasets showcase Dude’s significant recall and precision improvement by up to 22.8%, increasing F1 score by up to 18.7% compared to baseline methods.

中文摘要

摘要：随着研究提交量超过人工审查能力，利用大型语言模型（LLM）进行论文-代码不一致检测越来越受到关注。然而，现有单代理 LLM 范式的有限上下文容量和片面不一致检测导致在检测不一致时召回率表现较差。在本文中，我们提出了 Dude，这是第一个用于论文-代码不一致检测的双重检测多代理系统。我们发现，论文语言与代码语言的粒度不对称在多代理系统设计中引入了过度解释和过度报告的挑战，导致假阳性增加。为了解决这一问题，我们在 Dude 中提出了粒度对齐协商和两阶段显著性过滤机制，有效防止代理错误报告不一致。在真实世界论文-代码不一致数据集上的实验结果显示，Dude 在召回率和精确率上相较于基线方法显著提升，提升幅度最高可达 22.8%，F1 分数最高提升 18.7%.

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03416 (Request failed with status code 502)

Authors: Weijie Liu, Running Zhao, Wenhao Yuan, Jinfeng Xu, Zhanfeng Xu, Xiaoxi Zhang, Edith Cheuk-Han Ngai

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03416.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03416

Published: 2026-09-07T01:18:49.920Z

8. DuplexSpeechBench-IFEval: Evaluating Implicit Instruction Following in Full-Duplex Voice Agents

Abstract:Full-duplex voice agents must continuously decide when to listen, backchannel, interrupt, handle speech overlaps, take the floor, and yield. Existing benchmarks largely test these behaviors through explicit turn-management instructions, while deployed agents are often configured through roles or personas from which the appropriate conversational behavior must be inferred. We introduce DuplexSpeechBench-IFEval (DSB-IFEval) for evaluating implicit instruction-following in real-time spoken interaction. (DSB-IFEval) comprises 1,038 test cases spanning eight diverse assistant roles and evaluates five conditioning protocols for instruction-following: default behavior, explicit behavioral instructions, persona-implied behavior, combined persona—rule conditioning, and instruction conflict. We measure real-time floor management using a deterministic Instruction Adherence Score (IAS) and persona-consistent content using LLM-judged Persona Adherence Score (PAS). Across six real-time speech systems, we find architecture-dependent trade-offs. Full duplex models like F-Actor and PersonaPlex are more sensitive to whether conversational behavior is stated explicitly or must be inferred from a persona, with adherence dropping by 9.7% and 4.5%, respectively, under persona-only conditioning. In contrast, GPT-Realtime, MiniCPM-o, and Fun-Audio-Chat strongly adhere to persona-consistent content, but their floor behavior does not adapt across explicit and persona-only instructions and remains constrained on several proactive actions. We further find that even if systems reliably follow conflicting directives to their prescribed persona, they still struggle to override them under safety conflict. These results show that inferring the behavior implied by a role, executing it at the appropriate conversational moment, and resolving competing instructions remain distinct challenges for full-duplex voice agents.

中文摘要

摘要：全双工语音代理必须持续决定何时倾听、反向通道、中断、处理语音重叠、接听和让路。现有基准测试主要通过显式的回合管理指令测试这些行为，而部署中的代理通常通过角色或角色设定，需从中推断出适当的对话行为。我们介绍了DuplexSpeechBench-IFEval（DSB-IFEval），用于实时语音交互中的隐式指令跟随评估。（DSB-IFEval）包含1038个测试用例，涵盖八个不同助手角色，评估五种指令遵循条件协议：默认行为、显式行为指令、人物隐含行为、组合角色-规则条件反射和指令冲突。我们使用确定性指令依从评分（IAS）和使用LLM判定的人格依从评分（PAS）测量实时地板管理和角色一致性内容。在六个实时语音系统中，我们发现了依赖架构的权衡。全双工模型如F-Actor和PersonaPlex对对话行为是明确陈述还是必须从角色推断更为敏感，在仅角色条件下，依从率分别下降了9.7%和4.5%。相比之下，GPT-Realtime、MiniCPM-o和Fun-Audio-Chat严格遵守角色一致性内容，但其地板行为无法适应明确和仅限persona指令，且在多项主动行动中仍受限。我们进一步发现，即使系统可靠地遵循其指定persona的冲突指令，在安全冲突下仍难以覆盖这些指令。这些结果表明，推断角色所隐含的行为、在适当的对话时刻执行以及解决竞争指令，仍然是全双工语音代理面临的独特挑战。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03423 (Request failed with status code 502)

Authors: Puneet Mathur, Dinesh Manocha

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03423.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03423

Published: 2026-09-07T01:18:49.920Z

9. Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents

Abstract:Graphical user interface (GUI) agents are increasingly used to execute natural-language instructions on user interfaces, yet real users may issue infeasible instructions due to benign mistakes. A reliable agent should not only know how to act, but also when not to act. In this work, we introduce CONFLICTGUI, a benchmark covering instruction-internal conflicts and instruction-GUI context conflicts to study conflict-aware termination. Our evaluation reveals severe execution-biased overcompliance: agents that perform well on feasible tasks often continue to execute blindly under conflicting instructions. To mitigate this behavior, we propose CONFLICTGUARD, an inference-time framework that aligns an agent’s feasibility awareness with its action generation. CONFLICTGUARD contains two coupled components: a feasibility verification protocol that guides the agent to assess instruction logic and GUI-side evidence before acting, and a conditional action modulation mechanism that steers agents from over-compliant execution into termination-oriented behavior. Experiments across five widely-used agents demonstrate that CONFLICTGUARD improves average conflict task success rate significantly, while preserving normal GUI-task performance. These results validate that a lightweight inference-time intervention can substantially boost GUI Agent’s competence to identify inappropriate execution scenarios and refrain from unnecessary actions.

中文摘要

摘要：图形用户界面（GUI）代理越来越多地用于在用户界面上执行自然语言指令，但真实用户可能由于无意的错误而发出不可行的指令。一个可靠的代理不仅应该知道如何行动，还应知道何时不行动。在本工作中，我们引入了CONFLICTGUI，一个涵盖指令内部冲突和指令-GUI上下文冲突的基准，用于研究冲突感知的终止。我们的评估揭示了严重的执行偏向过度遵从问题：在可行任务中表现良好的代理在面对冲突指令时往往仍盲目执行。为减轻这种行为，我们提出了CONFLICTGUARD，这是一个在推理时对齐代理的可行性意识与动作生成的框架。CONFLICTGUARD包含两个耦合组件：一个可行性验证协议，引导代理在行动前评估指令逻辑和GUI端证据；以及一个条件动作调节机制，将代理从过度遵从的执行引导到以终止为导向的行为。在五种广泛使用的代理上的实验表明，CONFLICTGUARD显著提高了冲突任务的平均成功率，同时保持了正常的GUI任务性能。这些结果验证了轻量级的推理时干预可以显著增强GUI代理识别不适当执行情境并避免不必要操作的能力。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03438 (Request failed with status code 502)

Authors: Zhaoyuan Huang, Tianjie Ju, Pengzhou Cheng, Zheng Wu, Yansi Li, Chuanbiao Song, Jun Lan, Huijia Zhu, Weiqiang Wang, Zhuosheng Zhang

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03438.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03438

Published: 2026-09-07T01:18:49.920Z

10. Beyond “Made with AI”: Visualizing Provenance Density to Mitigate the Transparency Penalty

Abstract:As generative AI makes polished prose cheap to produce, users can no longer rely on fluency as a proxy for truth. We call this failure mode the Fluency Trap: users trust fluent hallucinations while also discounting accurate content once it is disclosed as AI-generated. Binary ``Made with AI’’ labels respond with authorship disclosure, but they do not show what supports a claim. We propose Provenance Density, an evidence-visualization interface that shows the density of verified claims in a text. In a user study with 81 participants, an idealized Provenance Density interface produced a large discernment gap between truth and fabrication ($+4.15$ points, $d=1.82$), whereas participants given no signal showed no detectable discrimination. A technical audit with 200 samples shows that retrieval density alone is insufficient; unexpectedly, the Consistency Veto carries most of the discriminative signal on dynamic queries. As AI-generated content becomes indistinguishable from human writing, effective transparency must move from authorship disclosure toward evidence visualization.

中文摘要

摘要：随着生成式人工智能让精炼的文字变得廉价，用户不再能依赖流畅性来判断真实性。我们将这种失效模式称为“流畅性陷阱”：用户会信任流畅的幻觉内容，同时在内容被披露为 AI 生成后会贬低其准确性。“AI 创作”二元标签虽然提供了作者身份披露，但并未显示支持某一论断的依据。我们提出了来源密度（Provenance Density），一种证据可视化界面，用于显示文本中经过验证的论断密度。在一项涉及 81 名参与者的用户研究中，理想化的来源密度界面在真实与虚构内容之间产生了较大的鉴别差距（+4.15 分，d=1.82），而没有信号提示的参与者则未显示出可察觉的辨别能力。对 200 个样本的技术审计显示，仅凭检索密度并不足以区分；出乎意料的是，一致性否决（Consistency Veto）在动态查询中承载了大部分判别信号。随着 AI 生成内容越来越难以与人类写作区分，有效的透明度必须从作者身份披露转向证据可视化。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03460 (Request failed with status code 502)

Authors: Qing Zhang, Yifei Huang, Juyoung Lee, Thad Starner, Jun Rekimoto

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03460.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03460

Published: 2026-09-07T01:18:49.920Z

Evaluation Domain Papers

1. Structure and Implementation of New Practical English Textbooks Driven by Artificial Intelligence

中文摘要

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.02981 (Request failed with status code 502)

Authors: Ya Wang, Lei Zhang, Xueguang Yang, Bo Chen

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.02981.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.02981

Published: 2026-09-07T01:19:32.874Z

2. MasterControl Seventeen Every Time

中文摘要

摘要：我们研究了一种受控的企业分析方法：语言模型解释问题，而确定性策略选择并运行预先批准的分析程序，该程序返回结果和证据。我们展示了这种限制在定义的分析类别内仍然可以保持表达能力，使用关系操作加上聚合、比较、窗口、排序和相似性。固定的意义、策略、数据和执行规则还使结果可重放。在440次运行中，三种8B模型在运行时生成SQL并选择工具，而Qwen3-8B仅解释意图，策略执行批准的程序。在330个运行时规划实验中，没有任何一次完全满足所有测试数据集的完整答案与证据契约；策略执行的分析器则在110次实验中全部匹配。这是特定配置的结果，并不能证明运行时代理在其他设计下无法成功。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03209 (Request failed with status code 502)

Authors: MasterControl AI Lab

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03209.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03209

Published: 2026-09-07T01:19:32.874Z

3. Speculative Macro Commit for Faster Tool-Using Agents

中文摘要

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03236 (Request failed with status code 502)

Authors: Zeyu Liu, Souvik Kundu, Peter A. Beerel

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03236.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03236

Published: 2026-09-07T01:19:32.874Z

4. Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory

中文摘要

摘要：分布式大型语言模型代理团队可以读取最新的共享事实，但仍可能在一个过时的计划上采取行动。一个计划者可能会从需求 $r_3$ 推导出一个动作，另一个代理可能执行 $r_4$，而执行者可能收到 $r_4$ 却没有替换从 $r_3$ 推导出的计划。我们称之为\emph{过时计划执行}：状态的新鲜度并不能保证授权某个动作的计划仍然有效。我们提出了 PlanFence，一种依赖范围限定的动作验证协议。计划引用它们使用的确切公共记录，执行者仅验证可能影响待处理外部动作的记录，验证不完整时会重新规划一次或阻塞。在 30 个经过后期计划修订的受控实时工作流中，仅依赖新鲜度的执行者在每个任务中都在过时计划上执行，而 PlanFence 完成了所有任务而没有执行无效动作。受控回放揭示了两个条件边界：在低变动环境下，主动同步可降低协调停滞，而随着变动增加，PlanFence 避免了重复的更新路径协调，并且随着共享键空间增大，避免了验证无关状态。这些都是受控的安全性和系统成本结果，而非一般任务准确度的提升。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03340 (Request failed with status code 502)

Authors: Evan Chen, Shiqiang Wang, Christopher G. Brinton

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03340.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03340

Published: 2026-09-07T01:19:32.874Z

5. A Prompt-Engineering Approach to Develop Scalable, Flexible, and Real-Time Hybrid Micro-Level Personalization in a General Purpose AI Teaching Assistant

中文摘要

摘要：由大型语言模型（LLM）驱动的人工智能（AI）教学助手能够提供可扩展的教育支持，但通常个性化程度有限。本研究提出了一种基于提示工程的框架，用于在各类学科和课程中对通用 LLM/RAG 驱动的 AI 教学助手（如 Jill Watson）进行个性化。该框架通过六个学习者特定维度调整响应：自我评估、抽象偏好、冗长偏好、感知取向、信息处理风格和理解水平，从而生成 96 种不同的学习者画像。学生的查询还会使用布卢姆认知分类法进行分析，以在交互层面估算认知复杂性。学习者属性和认知评估被编码在结构化提示中，以条件化 LLM，而无需重新训练模型。通过使用自然语言处理指标的实验和五名参与者的人类研究评估了该框架。结果显示在个性化条件下，响应风格和结构存在感知上的差异，统计分析识别出与可测量的响应变化相关的学习者属性。这些发现提供了初步证据，表明基于提示的个性化可以支持 LLM 驱动教育代理的适应性行为。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03402 (Request failed with status code 502)

Authors: Saptarshi Basu, Sandeep Kakar, Ashok Goel

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03402.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03402

Published: 2026-09-07T01:19:32.874Z

6. Caught in the Story: Narrative Captivity in Multi-turn LLMs Conversation

中文摘要

摘要：人们越来越多地依赖大型语言模型（LLM）来获取日常建议，使得带有伦理色彩的人际问题成为实际的道德咨询场景。以往的大多数研究都是通过单轮判断或压力导向的反驳来研究这一场景，这些假设与现实世界中获取指导的方式不匹配。这些假设使得尚不清楚单纯叙述——即没有明确对立立场的情况下——是否能够在多轮道德咨询中改变模型的判断。然而，现实世界中的道德冲突对话通常会引发一方的自我辩解，这种过程可能持续多轮并产生信息不对称。我们提出了\textbf{叙事困境}的概念，这是一个失败模式，其中模型将未受反对的一面之词视为完整，并与叙述者的解释保持一致，而不去寻求缺失的视角。为了衡量这一现象，我们建立了一个涵盖六个道德维度的$5{,}078$个人际冲突场景基准。在17个LLM中，叙事困境普遍存在：在多轮叙述下的最终判断平均相较于匹配的单轮基线偏移了25个百分点。阶段级分析表明，偏好优化是一个主要因素，而四种推理时策略仅能部分缓解这一问题。我们希望我们的研究能够推动LLM顾问在现实咨询中保持独立判断。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03407 (Request failed with status code 502)

Authors: Yuhe Wu, Guangyu Wang, Yujie Chen, Jiatong Zhang, Yuran Chen, Yutong Zhang, Xiyin Cheng, Wenpeng Cao, Zhuang Liu, Guang Zhang

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03407.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03407

Published: 2026-09-07T01:19:32.874Z

7. Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection

中文摘要

摘要：随着研究提交量超过人工审查能力，利用大型语言模型（LLM）进行论文-代码不一致检测越来越受到关注。然而，现有单代理 LLM 范式的有限上下文容量和片面不一致检测导致在检测不一致时召回率表现较差。在本文中，我们提出了 Dude，这是第一个用于论文-代码不一致检测的双重检测多代理系统。我们发现，论文语言与代码语言的粒度不对称在多代理系统设计中引入了过度解释和过度报告的挑战，导致假阳性增加。为此，我们在 Dude 中提出了粒度对齐协商和两阶段显著性过滤机制，有效防止代理错误报告不一致。现实世界的论文-代码不一致数据集实验结果显示，与基线方法相比，Dude 在召回率和精确度上显著提升，最高提高 22.8%，F1 分数提高最高达 18.7%。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03416 (Request failed with status code 502)

Authors: Weijie Liu, Running Zhao, Wenhao Yuan, Jinfeng Xu, Zhanfeng Xu, Xiaoxi Zhang, Edith Cheuk-Han Ngai

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03416.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03416

Published: 2026-09-07T01:19:32.874Z

8. DuplexSpeechBench-IFEval: Evaluating Implicit Instruction Following in Full-Duplex Voice Agents

中文摘要

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03423 (Request failed with status code 502)

Authors: Puneet Mathur, Dinesh Manocha

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03423.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03423

Published: 2026-09-07T01:19:32.874Z

9. Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents

中文摘要

摘要：图形用户界面（GUI）代理越来越多地用于在用户界面上执行自然语言指令，但真实用户可能会由于无意的错误而发出不可行的指令。一个可靠的代理不仅应该知道如何采取行动，还应该知道何时不采取行动。在本工作中，我们提出了CONFLICTGUI，这是一个涵盖指令内部冲突和指令与GUI上下文冲突的基准，用于研究冲突感知的终止行为。我们的评估显示了严重的执行偏向过度遵从：在可行任务上表现良好的代理在面对冲突指令时仍常盲目执行。为了缓解这种行为，我们提出了CONFLICTGUARD，这是一个推理时框架，用于将代理的可行性意识与其动作生成对齐。CONFLICTGUARD包含两个耦合组成部分：一个可行性验证协议，引导代理在行动前评估指令逻辑和GUI端证据，以及一个条件动作调节机制，将代理从过度遵从的执行引导向以终止为导向的行为。在五种广泛使用的代理上的实验表明，CONFLICTGUARD显著提高了冲突任务的平均成功率，同时保持了正常GUI任务的性能。这些结果验证了轻量级推理时干预可以显著增强GUI代理识别不适当执行场景并避免不必要动作的能力。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03438 (Request failed with status code 502)

Authors: Zhaoyuan Huang, Tianjie Ju, Pengzhou Cheng, Zheng Wu, Yansi Li, Chuanbiao Song, Jun Lan, Huijia Zhu, Weiqiang Wang, Zhuosheng Zhang

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03438.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03438

Published: 2026-09-07T01:19:32.874Z

10. Beyond “Made with AI”: Visualizing Provenance Density to Mitigate the Transparency Penalty

中文摘要

摘要：随着生成式人工智能让精炼文章的制作变得廉价，用户无法再依赖流畅性来判断真实性。我们将这种失效模式称为“流畅性陷阱”：用户会信任流畅的幻觉内容，同时在内容被披露为 AI 生成后会贬低其准确性。“AI 创作”二元标签虽然提供了作者身份披露，但并未显示支持某一论断的依据。我们提出了来源密度（Provenance Density），一种证据可视化界面，用于显示文本中经过验证的论断密度。在一项涉及 81 名参与者的用户研究中，理想化的来源密度界面在真实与虚构内容之间产生了较大的鉴别差距（+4.15 分，d=1.82），而没有信号提示的参与者则未显示出可察觉的辨别能力。对 200 个样本的技术审计显示，仅凭检索密度并不足以区分；出乎意料的是，一致性否决（Consistency Veto）在动态查询中承载了大部分判别信号。随着 AI 生成内容越来越难以与人类写作区分，有效的透明度必须从作者身份披露转向证据可视化。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03460 (Request failed with status code 502)

Authors: Qing Zhang, Yifei Huang, Juyoung Lee, Thad Starner, Jun Rekimoto

Categories: cs.AI

PDF URL: https://arxiv.org/pdf/2609.03460.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03460

Published: 2026-09-07T01:19:32.874Z

VLM Domain Papers

1. IDSPACE: A Novel Document Generator for Reliable Evaluation of Digital Identity Verification Systems [Extended Technical Report]

Abstract:As services move online, trust institutions such as banks, lenders, and governments must verify the identity of remote users. Fraud detection tools are widely available, but evaluating and fine-tuning them remains difficult because identity documents are sensitive and therefore scarce. Synthetic data generation offers a path forward, and demand is clear: our prior work in this area has been downloaded over $11{,}000$ times (aggregated from eight parts). We introduce IDSpace, extending this line of research in three directions. First, we propose model-guided Bayesian optimization, which tunes generation parameters to maximize both visual similarity and prediction consistency with target-domain models given only a few samples from a target domain. Second, we decouple user-specified metadata (demographics, fraud patterns, capture device) from automatically tuned control parameters (font styles, noise levels, image quality), allowing users to configure evaluations without low-level expertise. Third, we expand beyond template images to support scanned and mobile-captured documents. Experiments show IDSpace improves evaluation consistency by $15-45\%$ over baselines including CycleGAN, diffusion inpainting, and non-guided optimization, using only a few real samples, while improving training accuracy by up to $9\%$ and SSIM similarity with the target domain by $10\%$. We also released a new dataset consisting of $359{,}240$ high-quality synthetic documents across ten European ID types.

中文摘要

摘要：随着服务迁移到线上，银行、贷款机构和政府等信任机构必须验证远程用户的身份。欺诈检测工具广泛可用，但评估和微调这些工具仍然困难，因为身份文件具有敏感性，因此数量有限。合成数据生成提供了一条前进的道路，而且需求明确：我们在这一领域的先前工作已被下载超过11,000次（汇总自八个部分）。我们引入IDSpace，在三个方向上扩展了该研究路线。首先，我们提出了模型引导的贝叶斯优化，该方法通过仅使用目标域的少量样本来调整生成参数，以最大化与目标域模型的视觉相似性和预测一致性。其次，我们将用户指定的元数据（人口统计、欺诈模式、捕获设备）与自动调优的控制参数（字体样式、噪声水平、图像质量）分离，使用户无需低级专业知识即可配置评估。第三，我们超越模板图像，支持扫描和移动设备拍摄的文档。实验表明，IDSpace在使用少量真实样本的情况下，相比包括CycleGAN、扩散修复和非引导优化在内的基线方法，提高了15-45%的评估一致性，同时将训练准确率提高最多9%，目标域的SSIM相似性提高10%。我们还发布了一个新数据集，包含10种欧洲身份证类型的359,240份高质量合成文档。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03052 (Request failed with status code 502)

Authors: Lulu Xie, Yancheng Wang, Kanchan Chowdhury, Rolando Garcia, Yingzhen Yang, Jia Zou

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03052.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03052

Published: 2026-09-07T01:20:17.006Z

2. Position: Unlabeled IS NOT Equal to No Human Supervision in Visual Learning

Abstract:This position paper argues that the absence of labels does not imply the absence of human supervision in visual learning, and urges the research community to identify sources of supervision more explicitly. Many recent methods in computer vision build upon representations learned from large-scale unlabeled data, and are therefore grouped under the same umbrella term unsupervised.'' However, different data curation schemes and training objectives embed substantially different human priors on which models rely, and we argue that oneunsupervised’’ umbrella term is no longer capturing these distinctions. This ambiguity makes it harder to compare unsupervised learning research conducted under different assumptions, coinciding with a sharp decline in papers titled with ``unsupervised’’ in flagship computer vision conferences since 2021, despite continued growth of the field. While we fully embrace pre-training as a strong foundation for modern computer vision, we advocate for a community-level effort toward greater conceptual clarity: authors are encouraged to disclose priors in data selection and learning objectives, and to specify which components of a learning pipeline depend on which assumptions. Standardized disclosure practices can improve academic communication, ensure fairer comparisons, and preserve methodological diversity in unsupervised learning.

中文摘要

摘要：本文立场性论文认为，标签的缺失并不意味着视觉学习中缺乏人为监督，并敦促研究社区更明确地识别监督来源。计算机视觉中的许多最新方法都建立在从大规模无标签数据中学到的表示之上，因此被归入同一个“无监督”总称。然而，不同的数据整理方案和训练目标会嵌入模型依赖的显著不同的人类先验，我们认为单一的“无监督”总称已无法反映这些差异。这种模糊性使得在不同假设下进行的无监督学习研究难以进行比较，同时，自2021年以来，尽管该领域持续增长，旗舰计算机视觉会议上标题中带有“无监督”的论文数量却急剧下降。虽然我们完全认同预训练作为现代计算机视觉的坚实基础，但我们倡导社区层面的努力以实现更清晰的概念：鼓励作者披露数据选择和学习目标中的先验，并明确学习管道的哪些组件依赖于哪些假设。标准化的披露实践可以改善学术交流、确保更公平的比较，并保持无监督学习的方法多样性。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03077 (Request failed with status code 502)

Authors: Dong Lao

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03077.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03077

Published: 2026-09-07T01:20:17.006Z

3. Exemplar: Classical Priors Complement Frozen Features for Few-Shot Microscopy Segmentation at Native Resolution

Abstract:Segmenting a new biomedical dataset usually means a domain-specific model trained on substantial annotation, or a foundation model steered at inference time. We present Exemplar, a few-shot segmenter that fuses a frozen DINOv3 backbone with a fixed bank of classical native-resolution filter responses in one lightweight head, fitted from the support masks alone. In the few-mask, native-resolution regime, classical priors and frozen self-supervised features are complementary: fused in one head, a single fixed configuration spans eleven biomedical imaging datasets. Under the same head, the classical bank alone reaches 0.693 on the eleven-dataset panel, scored by foreground intersection-over-union or centreline Dice, and the frozen features alone 0.672; the bank leads on seven of the eleven and the features on the rest, and fused they reach 0.782. Against five forward-pass few-shot methods, Exemplar leads in 54 of 55 method-dataset comparisons, 52 of them significant after Holm correction. From a single annotated mask it reaches 0.703 on the same panel, against 0.682 for a from-scratch nnU-Net trained on that same mask. At eight masks nnU-Net overtakes it on the panel mean, chiefly on centreline agreement, but takes 16-77x longer to fit.

中文摘要

摘要：对一个新的生物医学数据集进行分割通常意味着需要一个在大量标注数据上训练的特定领域模型，或一个在推理时进行引导的基础模型。我们提出了 Exemplar，一种少样本分割器，它将冻结的 DINOv3 主干与固定的经典原生分辨率滤波器响应库融合在一个轻量级的头部中，仅通过支持掩码进行拟合。在少掩码、原生分辨率的情况下，经典先验和冻结的自监督特征是互补的：融合在一个头部中，一个固定配置可以覆盖十一种生物医学成像数据集。在相同的头部下，仅使用经典库在十一数据集组合上通过前景交并比或中心线 Dice 得分为 0.693，而仅使用冻结特征为 0.672；经典库在十一数据集中的七个领先，冻结特征在其余数据集领先，融合后得分为 0.782。与五种前向传递少样本方法相比，Exemplar 在 55 个方法-数据集对比中领先 54 个，其中 52 个在 Holm 校正后显著。从单个标注掩码出发，它在同一组合上的得分为 0.703，而使用从零开始训练的 nnU-Net 在相同掩码上为 0.682。在八个掩码时，nnU-Net 在组合平均上超过了它，主要体现在中心线一致性上，但拟合时间是其的 16-77 倍。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03080 (Request failed with status code 502)

Authors: Michal Průšek, Adam Novozámský, Filip Šroubek

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03080.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03080

Published: 2026-09-07T01:20:17.006Z

4. Solving the Needle-in-a-Haystack Problem in Mammography Vision-Language Model with Differentiable Subset Sampling

Abstract:There is growing interest in adopting CLIP-style vision—language model (VLM) pretraining for mammography. However, models that directly employ the standard CLIP architecture and training objective exhibit limited zero-shot performance in clinically important tasks such as cancer, finding-type, and BI-RADS predictions. We argue that this underwhelming performance is due to neglecting two characteristics of mammography data: (1) its high-res nature, and (2) homogeneity of radiology reports, largely driven by a predominance of negative/benign findings on examinations. We propose TopKSigLIP, a VLM designed to address these two limitations through a novel architecture and learning objectives. Instead of downscaling high-res mammography images to satisfy GPU memory constraints, TopKSigLIP introduces TopK-Patch module that learns to sample a sparse set of high-res patches likely to contain lesions, sidestepping the resolution—batch size tradeoff of VLM training. The sampled patch locations additionally serve as a built-in localization tool. To address report homogeneity, we replace the contrastive loss, which falsely repels semantically similar pairs, with a Sup-sigmoid loss. Sup-sigmoid loss extends the sigmoid loss from SigLIP with soft labels derived from structured data. TopKSigLIP outperforms existing open-source mammography and general medical VLMs on both internal and external benchmarks on density assessment, BI-RADS classification, finding subtyping, and cancer prediction under zero-shot evaluation. TopKSigLIP remains competitive under linear probing despite using a significantly smaller vision encoder and smaller training batches than baselines. The TopK-Patch module additionally achieves superior lesion localization over post-hoc Grad-CAM. Code and weights are made public:this https URL.

中文摘要

摘要：对于乳腺X线影像，采用CLIP风格的视觉-语言模型（VLM）预训练的兴趣日益增加。然而，直接使用标准CLIP架构和训练目标的模型，在诸如癌症、发现类型和BI-RADS预测等临床重要任务中的零样本性能有限。我们认为，这种不理想的性能是由于忽视了乳腺X线影像数据的两个特征：（1）其高分辨率特性；（2）放射学报告的同质性，这主要是由于检查中阴性/良性发现占多数。我们提出了TopKSigLIP，一种旨在通过新型架构和学习目标解决这两个限制的VLM。TopKSigLIP并不将高分辨率乳腺X线图像降采样以满足GPU内存限制，而是引入了TopK-Patch模块，学习采样可能包含病变的稀疏高分辨率图像块，从而规避了VLM训练中的分辨率与批量大小之间的权衡。采样的图像块位置还可作为内置定位工具。为了解决报告同质性问题，我们用Sup-sigmoid损失替代了错误地排斥语义相似对的对比损失。Sup-sigmoid损失将SigLIP中的sigmoid损失扩展为使用结构化数据导出的软标签。TopKSigLIP在零样本评估下，在密度评估、BI-RADS分类、发现分型和癌症预测方面，超越了现有的开源乳腺X线与通用医学VLM在内部和外部基准上的表现。尽管使用了远小于基线模型的视觉编码器和训练批量，TopKSigLIP在线性探测下仍保持竞争力。TopK-Patch模块在病变定位上也优于事后Grad-CAM。代码和权重已公开：此https URL。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03085 (Request failed with status code 502)

Authors: Young Seok Jeon, Beatrice Brown-Mulry, Rohan Satya Isaac, Anjana Dissanayaka, Theo Dapamede, Mohammadreza Chavoshi, Judy Gichoya, Hari Trivedi

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03085.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03085

Published: 2026-09-07T01:20:17.006Z

5. WireSeg-32K: A Physics-Grounded Synthetic Dataset for Wire Instance Segmentation

Abstract:Deformable linear objects such as wires and cables are difficult to segment because they are thin, highly deformable, and frequently self-occluded, while large-scale instance-level annotations are expensive to obtain in real scenes. Existing resources either focus on cable tracing or semantic segmentation under constrained settings, or generate visually plausible images without physically grounded wire deformation. We present WireSeg-32k, a synthetic dataset for wire instance segmentation with 32,000 RGB images, instance masks, depth maps, and a complementary real-world test set with annotations. To generate this dataset, we develop DeformX, a co-simulation pipeline that couples Cosserat-rod dynamics with photorealistic Isaac Sim rendering, enabling physically plausible, contact-consistent wire shapes, CAD-based wire assets, and diverse visually grounded scenes. As a simple baseline, LoRA fine-tuning SAM3 on WireSeg-32k alone improves real-world mAP@75 by 10.2% over the off-the-shelf model, showing that physically grounded synthetic data can transfer to real wire perception.

中文摘要

摘要：可变形线性物体，如电线和电缆，由于它们细长、高度可变形且经常自我遮挡，因此很难进行分割，同时在真实场景中获得大规模实例级标注也成本高昂。现有资源要么专注于受限条件下的电缆追踪或语义分割，要么生成在视觉上合理但没有物理基础的电线形变的图像。我们提出了 WireSeg-32k，这是一个用于电线实例分割的合成数据集，包括 32,000 张 RGB 图像、实例遮罩、深度图以及一个带有标注的真实世界测试集。为了生成该数据集，我们开发了 DeformX，这是一种联合仿真管道，它将 Cosserat 杆动力学与逼真 Isaac Sim 渲染相结合，实现物理上合理且接触一致的电线形态、基于 CAD 的电线素材以及多样的、视觉上合理的场景。作为一个简单的基线，对 WireSeg-32k 进行 LoRA 微调 SAM3 模型，相较于现成模型将真实世界 mAP@75 提升了 10.2%，表明具有物理基础的合成数据能够迁移到真实电线感知任务中。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03102 (Request failed with status code 502)

Authors: Zilin Dai, Lehong Wang, Yi Yang, Xiang Fei

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03102.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03102

Published: 2026-09-07T01:20:17.006Z

6. SLIDEFORGE: An LLM Agent for Controllable Editing of Slides as Structured Artifacts

Abstract:Current AI agents compellingly describe slides. However, AI-assisted slide editing requires more than understanding: the output must retain layout, style, component structure, and native editability. Towards, AI-assisted slide editing, existing agents operate on screenshots or weak document representations and often fragment coherent visual units, rasterize editable content, or break layout. In contrast, for controllable slide editing, we introduce an agentic framework, SLIDEFORGE, which builds a Deck State Graph, an executable slide state that links visual decomposition, native pptx object structure, and perceptual organization. By recovering human-referable components while retaining fine-grained editable structure, SLIDEFORGE supports theme-preserving reconstruction through slide-native operations and rendered-state verification. We further introduce an evaluation paradigm for controllable slide transformation that jointly measures component recovery, preservation, restyling consistency, visual quality, and native editability. Experiments show that SLIDEFORGE outperforms direct prompting, screenshot-based agents, and generic code-agent baselines across these dimensions. Code is available at this https URL.

中文摘要

摘要：当前的人工智能代理能够生动地描述幻灯片。然而，AI辅助的幻灯片编辑不仅仅需要理解能力：其输出必须保持布局、风格、组件结构以及本地可编辑性。针对AI辅助的幻灯片编辑，现有的代理通常操作于截图或薄弱的文档表示，经常会破坏连贯的视觉单元，将可编辑内容栅格化，或者破坏布局。相比之下，为实现可控的幻灯片编辑，我们引入了一个代理框架SLIDEFORGE，它构建了一个幻灯片状态图（Deck State Graph），这是一个可执行的幻灯片状态，将视觉分解、本地pptx对象结构和感知组织联系起来。通过在保留可精细编辑结构的同时恢复可参考的人类组件，SLIDEFORGE支持通过幻灯片本地操作和渲染状态验证进行主题保持的重建。我们进一步引入了一个可控幻灯片转换的评估范式，该范式联合衡量组件恢复、保持、重设计一致性、视觉质量和本地可编辑性。实验表明，SLIDEFORGE在这些方面均优于直接提示、基于截图的代理以及通用代码代理基线。代码可在此https链接获得。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03109 (Request failed with status code 502)

Authors: Haozhen Zheng, Fulin Wang, Tianhu Xiong, Yingjie Yu, Shengyi Qian, Hanchao Yu, Alex Schwing, Klara Nahrstedt, Mingyuan Wu

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03109.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03109

Published: 2026-09-07T01:20:17.006Z

7. Beyond Small Patches: Black-Box Detection and Purification of Diverse Backdoor Triggers

Abstract:Deep neural networks (DNNs) are increasingly deployed in real-world vision systems, yet their predictions can be covertly manipulated by backdoor attacks, in which malicious triggers cause targeted misclassification while preserving high clean accuracy. Existing defenses often rely on model internals, training data, or clean validation samples, making them difficult to deploy when only black-box access to a trained model is available. We propose TRIM (Trigger Removal by Identifying Manipulated Regions), a deployment-oriented black-box defense that detects and selectively removes backdoor triggers at inference time without requiring model internals, training data, or clean samples. The key insight behind TRIM is to identify image regions that are responsible for anomalous model behavior and purify only those regions while preserving benign content. TRIM innovates via three key components: (i) region-based segmentation with deep feature representations, (ii) adaptive trigger discovery through inpainting and diffusion-based reconstruction to isolate regions responsible for misclassification—-without assumptions about trigger type, shape, or location, and (iii) selective region purification that cleans poisoned regions while retaining benign content. To support practical deployment, TRIM further caches feature embeddings of previously identified triggers, enabling efficient recognition and avoiding redundant detection and purification. Extensive experiments across diverse datasets and backdoor types, including blended, sparse, varying-size, and multiple triggers, show that TRIM consistently outperforms existing black-box defenses, reducing attack success rates (ASR) to as low as 1.16% while preserving clean accuracy of up to 87.87%. These results demonstrate that effective backdoor mitigation is possible at inference time even when the defender has no access to any auxiliary data.

中文摘要

摘要：深度神经网络（DNNs）正日益部署于现实世界的视觉系统中，但其预测可能会被后门攻击隐秘操纵，在这种攻击中，恶意触发器会导致特定的错误分类，同时保持高的干净准确率。现有防御方法通常依赖于模型内部、训练数据或干净的验证样本，当仅能黑箱访问已训练模型时，这些方法难以部署。我们提出了TRIM（通过识别被操控的区域移除触发器），一种面向部署的黑箱防御方法，可在推理时检测并选择性地移除后门触发器，而无需模型内部信息、训练数据或干净样本。TRIM的核心理念是识别导致异常模型行为的图像区域，仅净化这些区域，同时保留正常内容。TRIM通过三个关键组成部分实现创新：（i）基于区域的分割与深度特征表示，（ii）通过修复和基于扩散的重建进行自适应触发器发现，从而隔离导致错误分类的区域——无需对触发器类型、形状或位置作任何假设，以及（iii）选择性区域净化，清理受污染区域的同时保留正常内容。为了支持实际部署，TRIM进一步缓存先前识别触发器的特征嵌入，实现高效识别，避免重复检测和净化。在多种数据集和后门类型（包括混合型、稀疏型、不同尺寸及多触发器）上的大量实验证明，TRIM始终优于现有黑箱防御方法，将攻击成功率（ASR）降至低至1.16%，同时保持高达87.87%的干净准确率。这些结果表明，即使防御方无法获取任何辅助数据，仍可在推理阶段实现有效的后门缓解。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03139 (Request failed with status code 502)

Authors: Ahmed Abdelnaby, Mohamed Elmahallawy

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03139.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03139

Published: 2026-09-07T01:20:17.006Z

Abstract:Visual fluency in generated video does not imply physical reliability, and a scalar quality score alone is incapable of indicating the obligation a clip violates or the moment it fails. We present VeriPhy, an auditable physical-verification system in which a text-only planner compiles the prompt into typed physical obligations and a statically validated execution plan before any frame is observed. During execution, observations gate and scope only declared calls to frozen low-level experts (e.g., segmentation and tracking, counting, eleven typed physical measurements over the resulting tracks, depth, OCR, and audio-event detection). Each action returns a provenance-carrying evidence record whose payload, when usable, is either a typed measurement or an explicitly tagged learned state. Typed resolvers and fixed composition map usable records to a three-valued state (supported, contradicted, or unknown, surfaced as plausible, implausible, or abstain) with full provenance, so that every verdict is traceable to the evidence that produced it. We anchor evaluation in a 1,500-clip corpus of human-annotated flaw records that localize real generation failures in prompt reference, space, and time. On a 149-clip core carrying 304 such records, VeriPhy accounts for 228, against 164 for a published question-decomposition evaluator given the same clips and the same claims. Recall alone does not separate it from prompting the same backbone monolithically, which reaches 222; what separates them is that each decision retains its evidence record and provenance, making the traces auditable one verdict at a time and usable as the interface through which a critic verdict could be written back into generation.

中文摘要

摘要：生成视频中的视觉流畅性并不意味着物理可靠性，而且单靠标量质量分数无法指示片段违反的义务或其失败的时刻。我们提出了 VeriPhy，这是一种可审计的物理验证系统，其中纯文本规划器在观察任何帧之前，将提示编译为类型化的物理义务和静态验证的执行计划。在执行过程中，观测仅限制和作用于声明的冻结低级专家调用（例如分割和跟踪、计数、在生成的轨迹上进行的十一种类型化物理测量、深度、OCR 和音频事件检测）。每个动作都会返回一个携带来源信息的证据记录，其有效负载在可用时要么是类型化的测量，要么是明确标记的学习状态。类型解析器和固定组合将可用记录映射为三值状态（支持、矛盾或未知，以合理、不合理或弃权呈现），并附有完整来源信息，因此每个裁决都可以追溯到产生它的证据。我们在一个包含1,500个片段的人类标注缺陷记录语料库中进行评估，该语料库定位了提示参考、空间和时间上的实际生成失败。在一个包含149个核心片段、含304条此类记录的数据集中，VeriPhy 说明了228条记录，而针对相同片段和相同声明，公开的问题分解评估器仅说明了164条。单靠召回率无法将其与单一骨干提示法区分，该方法达到222条；它们的区别在于每个决策都保留了其证据记录和来源信息，使得裁决的追溯可被逐条审计，并且可用作将评论裁决写回生成的接口。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03153 (Request failed with status code 502)

Authors: Wenzhuo Xu, Yuchen Zhu, Chongjian Ge, Xuan Shen, Jing Shi, Jason Kuen, Yongxin Chen, Molei Tao, Christopher McComb, Noelia Grande Gutiérrez, Jiuxiang Gu

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03153.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03153

Published: 2026-09-07T01:20:17.006Z

9. Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization

Abstract:Visual token pruning reduces the inference cost of vision-language models (VLMs), but most methods only ask which tokens to keep. This retained-token view can keep redundant high-scoring tokens while leaving discarded evidence without a close representative. We propose CoverPruner, a training-free pruner that asks the complementary demand-side question: after a token is removed, which surviving original token represents it for the target VLM? CoverPruner formulates pruning as Representational Coverage Maximization (RCM), covering the full projected visual-token set with query-weighted demand. It instantiates RCM with projector-space coverage and a lightweight first-layer attention probe. Across multiple VLM architectures and compression rates, CoverPruner achieves the best average accuracy among all compared methods, with the largest gains usually appearing under aggressive compression.

中文摘要

摘要：视觉 token 剪枝可以降低视觉-语言模型（VLM）的推理成本，但大多数方法仅关注保留哪些 token。仅保留 token 的视角可能会保留冗余的高分 token，同时使被丢弃的证据缺少近似的代表。我们提出了 CoverPruner，一种无需训练的剪枝器，它提出了互补的需求侧问题：在移除某个 token 之后，哪些存活的原始 token 可以代表它以适应目标 VLM？CoverPruner 将剪枝表述为表示覆盖最大化（Representational Coverage Maximization, RCM），通过查询加权的需求覆盖完整的投影视觉 token 集。它通过投影空间覆盖和轻量的第一层注意力探针实现了 RCM。在多种 VLM 架构和压缩率下，CoverPruner 在所有比较方法中取得了最高的平均准确率，并且在高压缩率下通常获得最大增益。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03158 (Request failed with status code 502)

Authors: Qingchan Zhu, Weihang You, Hanqi Jiang, Changdi Yang, Tianming Liu, Geng Yuan

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03158.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03158

Published: 2026-09-07T01:20:17.006Z

10. RoboTok: An Internet-Scale Data Engine for Human Demonstration Retrieval and Dexterous Manipulation Learning

Abstract:Robot learning increasingly depends on broad and diverse demonstrations, yet collecting robot data remains expensive and poorly suited to covering the long tail of real-world tasks. To address this bottleneck, we introduce RoboTok, an internet-scale data engine that, given a query human manipulation video, retrieves manipulation-relevant human demonstrations from web videos for training dexterous robot policies. Specifically, we learn a latent motion space from 3D hand trajectories expressed in estimated actor-centered reference frames. This representation enables manipulation behaviors to be compared across variations in camera viewpoint, scene appearance, and actor occlusions, while remaining compact enough for efficient search and continual indexing over internet-scale video collections. We evaluate RoboTok against existing robot-data retrieval approaches on retrieval benchmarks and downstream robot policy performance. Our results show that RoboTok retrieves more relevant manipulation demonstrations and improves downstream task success, establishing hand-pose trajectory-aware retrieval as a way to make web video a scalable and continuously growing source of supervision for robot learning.

中文摘要

摘要：机器人学习越来越依赖广泛而多样的示范，但收集机器人数据仍然成本高昂，且难以覆盖现实世界任务的长尾部分。为了解决这一瓶颈，我们引入了RoboTok，一个互联网规模的数据引擎。给定一个查询的人类操作视频，它可以从网络视频中检索与操作相关的人类示范，用于训练灵巧的机器人策略。具体来说，我们从以估计的以演员为中心的参考系表示的3D手部轨迹中学习潜在运动空间。这种表示能够在摄像机视角、场景外观和演员遮挡变化下比较操作行为，同时保持足够紧凑，以便在互联网规模的视频集合中进行高效搜索和持续索引。我们在检索基准和下游机器人策略表现上，将RoboTok与现有的机器人数据检索方法进行了评估。结果表明，RoboTok能够检索更多相关的操作示范并提升下游任务成功率，确立了手部姿态轨迹感知的检索方式，使网络视频成为机器人学习可扩展且持续增长的监督来源。

LLM Analysis

LLM Analysis Failed: Error: Kimi内容未就绪或抓取失败: 2609.03199 (Request failed with status code 502)

Authors: Howard Qian, Yiting Chen, Yunfei Xie, Kejia Ren, Podshara Chanrungmaneekul, Gaotian Wang, Bowen Wen, Chen Wei, Kaiyu Hang

Categories: cs.CV

PDF URL: https://arxiv.org/pdf/2609.03199.pdf

CoolPaper URL: https://papers.cool/arxiv/2609.03199

Published: 2026-09-07T01:20:17.006Z

ArXiv Domain 2026-09-07

LLM Domain Papers

1. Where Does Harness-Optimization Value Live? Localized Gains and the Budget-Splitting Trap in Self-Evolving LLM Agents

2. Bounded Personas Match Retrieval on Classification but Not Regression for a Frozen Agent

3. Counterexamples as Feedback for Agent Self-Correction

4. Probe Generalization as Subspace Selection for OOD Deception Detection

5. R$^{2}$Adapter: A Routing and Rewriting Adapter for Efficient Hybrid RAG

6. BharatGather: A Culturally-Informed Benchmark Dataset for Misinformation and Fake News Detection in Indian Public Events

7. PiPMRE: A Pipeline Based on Language Model for Medical Relation Extraction

8. Margins, Not Windows: Training-Free Per-Step Lossy Speculative Decoding

9. Distilled Rapid Embedding Transfer (DRET): Parameter-Efficient Biomedical Domain Adaptation via Priority-Based Embedding Transfer

10. Contamination Inflates Scores but Rarely Reorders Large Language Model Leaderboards

Agent Domain Papers

1. Structure and Implementation of New Practical English Textbooks Driven by Artificial Intelligence

2. MasterControl Seventeen Every Time

3. Speculative Macro Commit for Faster Tool-Using Agents

4. Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory

5. A Prompt-Engineering Approach to Develop Scalable, Flexible, and Real-Time Hybrid Micro-Level Personalization in a General Purpose AI Teaching Assistant

6. Caught in the Story: Narrative Captivity in Multi-turn LLMs Conversation

7. Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection

8. DuplexSpeechBench-IFEval: Evaluating Implicit Instruction Following in Full-Duplex Voice Agents

9. Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents

10. Beyond “Made with AI”: Visualizing Provenance Density to Mitigate the Transparency Penalty

Evaluation Domain Papers

1. Structure and Implementation of New Practical English Textbooks Driven by Artificial Intelligence

2. MasterControl Seventeen Every Time

3. Speculative Macro Commit for Faster Tool-Using Agents

4. Fresh Memory, Stale Plans: Dependency-Scoped Validation for Distributed LLM-Agent Memory

5. A Prompt-Engineering Approach to Develop Scalable, Flexible, and Real-Time Hybrid Micro-Level Personalization in a General Purpose AI Teaching Assistant

6. Caught in the Story: Narrative Captivity in Multi-turn LLMs Conversation

7. Dude: A Dual-Detection Multi-Agent System for Paper-Code Discrepancy Detection

8. DuplexSpeechBench-IFEval: Evaluating Implicit Instruction Following in Full-Duplex Voice Agents

9. Do GUI Agents Know When Not to Act? Enabling Conflict-Aware Termination for Multimodal GUI Agents

10. Beyond “Made with AI”: Visualizing Provenance Density to Mitigate the Transparency Penalty

VLM Domain Papers

1. IDSPACE: A Novel Document Generator for Reliable Evaluation of Digital Identity Verification Systems [Extended Technical Report]

2. Position: Unlabeled IS NOT Equal to No Human Supervision in Visual Learning

3. Exemplar: Classical Priors Complement Frozen Features for Few-Shot Microscopy Segmentation at Native Resolution

4. Solving the Needle-in-a-Haystack Problem in Mammography Vision-Language Model with Differentiable Subset Sampling

5. WireSeg-32K: A Physics-Grounded Synthetic Dataset for Wire Instance Segmentation

6. SLIDEFORGE: An LLM Agent for Controllable Editing of Slides as Structured Artifacts

7. Beyond Small Patches: Black-Box Detection and Purification of Diverse Backdoor Triggers

8. VeriPhy: Agentic Physical Reasoning for World Model Evaluation and Refinement

9. Who Speaks for the Pruned? Visual Token Pruning as Coverage Optimization

10. RoboTok: An Internet-Scale Data Engine for Human Demonstration Retrieval and Dexterous Manipulation Learning