web_fetch/005_arxiv_recent_ai-2

{
  "tool_name": "web_fetch",
  "tool_input": {
    "brief": "Get 20 most recent cs.AI submissions",
    "max_age_seconds": 3600,
    "url": "https://arxiv.org/list/cs.AI/recent"
  },
  "dispatch_id": "toolu_014aHqKMU8TuHgRCeNvj6nxv",
  "parent_dispatch_id": "",
  "handle": "",
  "panel_kind": "web_fetch"
}

{
  "tool_name": "web_fetch",
  "dispatch_id": "toolu_014aHqKMU8TuHgRCeNvj6nxv",
  "status": "running",
  "result": null,
  "error": "",
  "elapsed": null,
  "fields": {
    "progress": {
      "message": "Get 20 most recent cs.AI submissions",
      "metadata": {
        "browser_chain": false,
        "url": "https://arxiv.org/list/cs.AI/recent"
      }
    },
    "status": "running",
    "updatedAt": 1789059123582
  }
}

{
  "tool_name": "web_fetch",
  "dispatch_id": "toolu_014aHqKMU8TuHgRCeNvj6nxv",
  "status": "completed",
  "result": {
    "content": "Artificial Intelligence\n\n [Skip to main content](#content)\n\n [https://arxiv.org/IgnoreMe](https://arxiv.org/IgnoreMe) [![archive](/static/base/1.0.1/images/arxiv-logo-primary-light.svg)](https://arxiv.org/)\n\n[Search](https://arxiv.org/search) [Submit](https://arxiv.org/user/create) [Donate](https://info.arxiv.org/about/donate.html)  [Log in](https://arxiv.org/login)\n\n\n\n\n\n\n\n  Search arXiv\n\n Press Enter to search · [Advanced search](https://arxiv.org/search/advanced)\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n# Artificial Intelligence\n\n\n\n## Authors and titles for recent submissions\n\n\n -  [Thu, 10 Sep 2026](/list/cs.AI/recent?skip=0&show=50)\n-  [Wed, 9 Sep 2026](/list/cs.AI/recent?skip=150&show=50)\n-  [Mon, 7 Sep 2026](/list/cs.AI/recent?skip=667&show=50)\n-  [Fri, 4 Sep 2026](/list/cs.AI/recent?skip=872&show=50)\n-  [Thu, 3 Sep 2026](/list/cs.AI/recent?skip=1037&show=50)\n\n\n\nSee today's [new](/list/cs.AI/new) changes\n\n\n\nTotal of 1199 entries : 1-50 [51-100](/list/cs.AI/recent?skip=50&show=50) [101-150](/list/cs.AI/recent?skip=100&show=50) [151-200](/list/cs.AI/recent?skip=150&show=50) ... [1151-1199](/list/cs.AI/recent?skip=1150&show=50)\n\n\n\nShowing up to 50 entries per page: [fewer](/list/cs.AI/recent?skip=0&show=25) | [more](/list/cs.AI/recent?skip=0&show=100) | [all](/list/cs.AI/recent?skip=0&show=2000)\n\n\n\n### Thu, 10 Sep 2026 (showing first 50 of 150 entries )\n\n  [1] [arXiv:2609.10451](/abs/2609.10451) [[pdf](/pdf/2609.10451), [html](https://arxiv.org/html/2609.10451v1), [other](/format/2609.10451)]\n\n\n\nTitle: JarvisGUI: Towards Cross-Device GUI Agents with Dynamic Task Composition\n\n\n\n[Zixiang Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yuheng Lu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zihao Cheng](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zeming Liu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jizeng Bai](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ziye Huang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhiyin Lin](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zihan Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yuhang Guo](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yunhong Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Haifeng Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Accepted to EMNLP 2026 (Main Conference)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [2] [arXiv:2609.10441](/abs/2609.10441) [[pdf](/pdf/2609.10441), [html](https://arxiv.org/html/2609.10441v1), [other](/format/2609.10441)]\n\n\n\nTitle: ConvMem: Convolutional Memory for Long-Context Reasoning\n\n\n\n[Hongming Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhaozhen Gu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Fengshuo Bai](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ming Hao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Qingyang Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yuanyuan Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Shiyang Tang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yanna Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Bo Xu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)\n\n\n\n   [3] [arXiv:2609.10413](/abs/2609.10413) [[pdf](/pdf/2609.10413), [html](https://arxiv.org/html/2609.10413v1), [other](/format/2609.10413)]\n\n\n\nTitle: Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs\n\n\n\n[Ansuman Mullick](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Eray Tüzün](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Preprint, under review. 2 figures. Code, benchmark and run logs: [this https URL](https://doi.org/10.5281/zenodo.20067778)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [4] [arXiv:2609.10350](/abs/2609.10350) [[pdf](/pdf/2609.10350), [other](/format/2609.10350)]\n\n\n\nTitle: Cyber-Financial Contagion: Modeling the Propagation of an AI Vendor Compromise Through the Banking System\n\n\n\n[Alex Leytes](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 11 fig and 10 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computers and Society (cs.CY); Machine Learning (cs.LG)\n\n\n\n   [5] [arXiv:2609.10335](/abs/2609.10335) [[pdf](/pdf/2609.10335), [html](https://arxiv.org/html/2609.10335v1), [other](/format/2609.10335)]\n\n\n\nTitle: From Symbolic Perception to Logical Deduction: A Framework for Guiding Language Models in Geometric Reasoning\n\n\n\n[Weichen Dai](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Rafael Medeiros Cabral](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ziyi Shou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yan Cao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xin Shen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Dongcai Lu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yi Zhou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)\n\n\n\n   [6] [arXiv:2609.10315](/abs/2609.10315) [[pdf](/pdf/2609.10315), [html](https://arxiv.org/html/2609.10315v1), [other](/format/2609.10315)]\n\n\n\nTitle: TRACE: Training Reasoning Agents for Causal Exploration with Synthesized Rewards\n\n\n\n[Rui Sun](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhan Shi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Bing He](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)\n\n\n\n   [7] [arXiv:2609.10263](/abs/2609.10263) [[pdf](/pdf/2609.10263), [html](https://arxiv.org/html/2609.10263v1), [other](/format/2609.10263)]\n\n\n\nTitle: What Should an Agent Forget? Separating What Is Stored from What Is Used\n\n\n\n[Yuhang Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yuchen Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 8 pages, 3 figures, 3 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [8] [arXiv:2609.10221](/abs/2609.10221) [[pdf](/pdf/2609.10221), [html](https://arxiv.org/html/2609.10221v1), [other](/format/2609.10221)]\n\n\n\nTitle: Why Sample What You Can Enumerate? Exact Policy Optimization for Genomic Tool Selection\n\n\n\n[Haoyue Liu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xiaoyu Ma](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ye Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhichao Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xiaoying Tang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [9] [arXiv:2609.10177](/abs/2609.10177) [[pdf](/pdf/2609.10177), [html](https://arxiv.org/html/2609.10177v1), [other](/format/2609.10177)]\n\n\n\nTitle: Beyond Surface Imitation: Contrastive Modeling for Reasoning Path Alignment in Multimodal In-Context Learning\n\n\n\n[Mingbo Yang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wenqiang Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhaolu Kang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Peng Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yannan Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sunshang Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yan Xiao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [10] [arXiv:2609.10144](/abs/2609.10144) [[pdf](/pdf/2609.10144), [html](https://arxiv.org/html/2609.10144v1), [other](/format/2609.10144)]\n\n\n\nTitle: Kernel-Managed Shared Memory for System-Wide Personalization\n\n\n\n[Ryan Lum](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yongfeng Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)\n\n\n\n   [11] [arXiv:2609.10135](/abs/2609.10135) [[pdf](/pdf/2609.10135), [html](https://arxiv.org/html/2609.10135v1), [other](/format/2609.10135)]\n\n\n\nTitle: Agent-Based ML-LLM Fusion with Self-Optimizing Prompts for Plateau Weather Alerts\n\n\n\n[Shuai Yan](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yang Xu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Shan He](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Accepted by ISPDS 2025\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)\n\n\n\n   [12] [arXiv:2609.10092](/abs/2609.10092) [[pdf](/pdf/2609.10092), [html](https://arxiv.org/html/2609.10092v1), [other](/format/2609.10092)]\n\n\n\nTitle: RAP: Research Attention Prediction Reveals Target-Conditioned Evidence Acquisition Biases\n\n\n\n[Yingqian Wu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jingcong Liang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Siyuan Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhenfei Yin](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Philip Torr](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Junchi Yu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhongyu Wei](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)\n\n\n\n   [13] [arXiv:2609.10060](/abs/2609.10060) [[pdf](/pdf/2609.10060), [html](https://arxiv.org/html/2609.10060v1), [other](/format/2609.10060)]\n\n\n\nTitle: Reference-Based Bias Detection in LLMs via Relative Representations of Hidden States\n\n\n\n[Marek Jeliński](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jan Dubiński](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Maciej Chrabaszcz](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sebastian Cygert](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [14] [arXiv:2609.10055](/abs/2609.10055) [[pdf](/pdf/2609.10055), [other](/format/2609.10055)]\n\n\n\nTitle: OntologyAligner: Ontology-Aligned Retrieval and Hierarchy-Guided Large Language Model Reranking for Biomedical Ontology Normalization\n\n\n\n[Jie Song](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhichuan Xu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ziyu Lu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Meng Xiao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Cheng Bi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yuxin Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xin Zheng](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xiaoran Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Qiongfang Cao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Hao Yang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Bairong Shen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 4 figures\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)\n\n\n\n   [15] [arXiv:2609.10036](/abs/2609.10036) [[pdf](/pdf/2609.10036), [html](https://arxiv.org/html/2609.10036v1), [other](/format/2609.10036)]\n\n\n\nTitle: Belief-State Engine: Augmenting LLMs for Principled Planning Under Partial Observability\n\n\n\n[Arnab Chattopadhayay](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Debdipta Halder](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Total number of pages: 19, total number of figures: 5\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG); Robotics (cs.RO)\n\n\n\n   [16] [arXiv:2609.09928](/abs/2609.09928) [[pdf](/pdf/2609.09928), [html](https://arxiv.org/html/2609.09928v1), [other](/format/2609.09928)]\n\n\n\nTitle: Structural Process Supervision for Latent Chain-of-Thought Reasoning\n\n\n\n[Yiqi Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xu Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Chen Ju](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jiangchao Yao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhaoyang Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jinsong Lan](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xiaoyong Zhu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Bo Zheng](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yu Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [17] [arXiv:2609.09925](/abs/2609.09925) [[pdf](/pdf/2609.09925), [html](https://arxiv.org/html/2609.09925v1), [other](/format/2609.09925)]\n\n\n\nTitle: Time-Frequency Geometric Cross-Attention for Chunked Vision-Language-Action Models\n\n\n\n[Shengye Dong](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Haochen Niu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Hao Liu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Peiwen Lin](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Chuang Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Shanmin Pang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Robotics (cs.RO)\n\n\n\n   [18] [arXiv:2609.09898](/abs/2609.09898) [[pdf](/pdf/2609.09898), [html](https://arxiv.org/html/2609.09898v1), [other](/format/2609.09898)]\n\n\n\nTitle: Grounded Evaluation and Repair for NL-to-PDDL Problem Generation\n\n\n\n[Joana Rosa](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Pedro Santos](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Valdemar Oliveira](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Romão Silva](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [L. Miguel Silveira](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Bruno Martins](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [19] [arXiv:2609.09885](/abs/2609.09885) [[pdf](/pdf/2609.09885), [html](https://arxiv.org/html/2609.09885v1), [other](/format/2609.09885)]\n\n\n\nTitle: Decision Transformer for UAV-Mounted RIS-Assisted Dynamic D2D Communications\n\n\n\n[Yaxuan Liu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Robotics (cs.RO)\n\n\n\n   [20] [arXiv:2609.09882](/abs/2609.09882) [[pdf](/pdf/2609.09882), [html](https://arxiv.org/html/2609.09882v1), [other](/format/2609.09882)]\n\n\n\nTitle: Scored vs. Generated Readouts in Behavioral Language Models: An Empirical Study of Elicitation Format\n\n\n\n[Touchapon Kraisingkorn](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Krittin Pachtrachai](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wachiravit Modecrua](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 12 pages, 1 figure, 2 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [21] [arXiv:2609.09875](/abs/2609.09875) [[pdf](/pdf/2609.09875), [other](/format/2609.09875)]\n\n\n\nTitle: AgentAudit: An Open, Extensible Framework for Full-Lifecycle Trust Evaluation of AI Agents\n\n\n\n[Shrey Nag](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sachita](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Abhishek Kumar Singh](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Lipi Goel](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Rajeshwar Singh Janwar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 23 pages, 12 figures\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [22] [arXiv:2609.09864](/abs/2609.09864) [[pdf](/pdf/2609.09864), [html](https://arxiv.org/html/2609.09864v1), [other](/format/2609.09864)]\n\n\n\nTitle: Shifting Relational Paradigms for Affective Computing: Affective Resonance, Vitality Affects, and Vocal Interaction Fields\n\n\n\n[Cy Gorman](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yihang Yao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Accepted at Interspeech 2026 for poster presentation. 5 pages, 2 figures, 2 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [23] [arXiv:2609.09853](/abs/2609.09853) [[pdf](/pdf/2609.09853), [html](https://arxiv.org/html/2609.09853v1), [other](/format/2609.09853)]\n\n\n\nTitle: The Era by Eon Benchmark: A Generated Enterprise Estate with Exact Ground Truth for Benchmarking LLM Agents\n\n\n\n[Benjamin Gruenbaum](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Doron Porat](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Assaf Natanzon](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Roy Zavida](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Chen Dinachi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Or Itzahary](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 12 pages\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [24] [arXiv:2609.09815](/abs/2609.09815) [[pdf](/pdf/2609.09815), [html](https://arxiv.org/html/2609.09815v1), [other](/format/2609.09815)]\n\n\n\nTitle: UnitBoost: Managing Compound LLM Systems with a Merge Operator, Not a Model\n\n\n\n[Xing Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Guanghui Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yanwei Cui](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Mengdie Flora Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Peiyang He](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Multiagent Systems (cs.MA)\n\n\n\n   [25] [arXiv:2609.09776](/abs/2609.09776) [[pdf](/pdf/2609.09776), [html](https://arxiv.org/html/2609.09776v1), [other](/format/2609.09776)]\n\n\n\nTitle: Proof-Carrying Cognition: Closing the Verification Gap with Reality-Settled Reward\n\n\n\n[Eshwar Reddy M](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sourav Karmakar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 21 pages, 13 figures\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Machine Learning (cs.LG)\n\n\n\n   [26] [arXiv:2609.09774](/abs/2609.09774) [[pdf](/pdf/2609.09774), [html](https://arxiv.org/html/2609.09774v1), [other](/format/2609.09774)]\n\n\n\nTitle: Procedural Memory Under Change: Reuse and Interference in Controlled Web Tasks\n\n\n\n[Yanze Cao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 17 pages, 2 figures, 11 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [27] [arXiv:2609.09754](/abs/2609.09754) [[pdf](/pdf/2609.09754), [html](https://arxiv.org/html/2609.09754v1), [other](/format/2609.09754)]\n\n\n\nTitle: LexAgentHallu: A Hierarchical Benchmark for Profiling Hallucinations in Legal Agents\n\n\n\n[Yujin Zhou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Mingxuan Zheng](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Chuxue Cao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Huang Yidan](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jiale Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yike Guo](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sirui Han](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: EMNLP 2026 Main\n\n\n\nJournal-ref: EMNLP 2026 Main\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [28] [arXiv:2609.09735](/abs/2609.09735) [[pdf](/pdf/2609.09735), [html](https://arxiv.org/html/2609.09735v1), [other](/format/2609.09735)]\n\n\n\nTitle: Can Artificial Intelligence Support Healthcare and Mental Health Through Early Cyberbullying Detection ? The Impact of Emotion-Aware AI on Proactive Online Safety\n\n\n\n[Hamed Jelodar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Amir Firouzi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yen-Wu Lo](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Maryam Tanha](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sajjad Dadkhah](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)\n\n\n\n   [29] [arXiv:2609.09707](/abs/2609.09707) [[pdf](/pdf/2609.09707), [html](https://arxiv.org/html/2609.09707v1), [other](/format/2609.09707)]\n\n\n\nTitle: Which Tokens Should SFT Actually Learn? A Token-Trimming Perspective on Mathematical Reasoning\n\n\n\n[Yaning Jia](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Chunhui Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wenxuan Xu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xingjian Diao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xiaoyuan Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Soroush Vosoughi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Accepted to Findings of the Association for Computational Linguistics: EMNLP 2026. 14 pages. Code available at [this https URL](https://github.com/karpning/TrimSFT)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [30] [arXiv:2609.09702](/abs/2609.09702) [[pdf](/pdf/2609.09702), [html](https://arxiv.org/html/2609.09702v1), [other](/format/2609.09702)]\n\n\n\nTitle: Decision Shifts, Lost Label Functionality, and an Inconclusive Grounding Audit in Correctness-Gated Multi-Teacher Distillation\n\n\n\n[Xiaofei Feng](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [31] [arXiv:2609.09678](/abs/2609.09678) [[pdf](/pdf/2609.09678), [html](https://arxiv.org/html/2609.09678v1), [other](/format/2609.09678)]\n\n\n\nTitle: Safe to Stop? Risk-Constrained Stopping for Sequential Clinical Diagnosis Agents\n\n\n\n[Yuexin Wu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Vasile Rus](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [32] [arXiv:2609.09664](/abs/2609.09664) [[pdf](/pdf/2609.09664), [html](https://arxiv.org/html/2609.09664v1), [other](/format/2609.09664)]\n\n\n\nTitle: PRAGMA: Evaluating Personalized Guidance with Memory Alignment in Lifelong Conversations\n\n\n\n[Hyojeong Yu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Hyukhun Koh](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Minsung Kim](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yunah Jang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Kyomin Jung](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Accepted to EMNLP 2026\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [33] [arXiv:2609.09657](/abs/2609.09657) [[pdf](/pdf/2609.09657), [html](https://arxiv.org/html/2609.09657v1), [other](/format/2609.09657)]\n\n\n\nTitle: RESCUE-BENCH: Towards Relation-Aware Multi-Party Emotional Support Conversation Systems\n\n\n\n[Haichuan Hu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yang Xiao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Mingni Tang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jiawen Duan](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Quanjun Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Congqing He](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Hao Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jiashuo Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Johan F. Hoorn](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wenjie Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: accepted as AACL findings\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [34] [arXiv:2609.09647](/abs/2609.09647) [[pdf](/pdf/2609.09647), [html](https://arxiv.org/html/2609.09647v1), [other](/format/2609.09647)]\n\n\n\nTitle: Black-Box Red Teaming of Agentic AI: A Taxonomy-Driven Framework for Automated Risk Discovery\n\n\n\n[Divyanshu Kumar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Nitin Aravind Birur](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Tanay Baswa](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sahil Agarwal](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Prashanth Harshangi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [35] [arXiv:2609.09646](/abs/2609.09646) [[pdf](/pdf/2609.09646), [html](https://arxiv.org/html/2609.09646v1), [other](/format/2609.09646)]\n\n\n\nTitle: RobustSGPO: Search-Space Control for Agent Harness Evolution\n\n\n\n[Zibo Zhao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jijun Shi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Mo Zhou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhongyuan Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Shifu Bie](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yunfei Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xuanting Zhou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xiangyu Wu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Bin Liu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ruiming Tang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wenwu Ou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Kun Gai](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 7 pages, 7 figures, 3 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [36] [arXiv:2609.09627](/abs/2609.09627) [[pdf](/pdf/2609.09627), [html](https://arxiv.org/html/2609.09627v1), [other](/format/2609.09627)]\n\n\n\nTitle: Seven Sources of Physical AI Capability Formation\n\n\n\n[Gang Chen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [37] [arXiv:2609.09625](/abs/2609.09625) [[pdf](/pdf/2609.09625), [html](https://arxiv.org/html/2609.09625v1), [other](/format/2609.09625)]\n\n\n\nTitle: From State Synchronization to Cognitive Self-Evolution: An Operational Architecture for Cognitive Digital Twins\n\n\n\n[Haoran Gao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [An Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhen Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jun Cai](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [38] [arXiv:2609.09589](/abs/2609.09589) [[pdf](/pdf/2609.09589), [html](https://arxiv.org/html/2609.09589v1), [other](/format/2609.09589)]\n\n\n\nTitle: A Function-Space Approach to the Statistical Mechanics of Learning Dynamics\n\n\n\n[Yizhou Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Weichen Wu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Lun Du](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zhengjie Miao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [39] [arXiv:2609.09578](/abs/2609.09578) [[pdf](/pdf/2609.09578), [html](https://arxiv.org/html/2609.09578v1), [other](/format/2609.09578)]\n\n\n\nTitle: CityPlanner: A Sandbox Agent for Executable Urban Planning\n\n\n\n[Wentao Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jingyuan Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Zetong Zhou](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Yifan Yang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wenrui Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: EMNLP Under Review\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL)\n\n\n\n   [40] [arXiv:2609.09565](/abs/2609.09565) [[pdf](/pdf/2609.09565), [html](https://arxiv.org/html/2609.09565v1), [other](/format/2609.09565)]\n\n\n\nTitle: Multi-Agent Agentic Graph Learning via Structural Signatures\n\n\n\n[Liang Qu](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jianxin Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Hua Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Under review\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [41] [arXiv:2609.09458](/abs/2609.09458) [[pdf](/pdf/2609.09458), [html](https://arxiv.org/html/2609.09458v1), [other](/format/2609.09458)]\n\n\n\nTitle: ContractEval: Query-Conditioned Execution Matching for Procedural Instruction Conformance\n\n\n\n[Praphul Singh](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Shanu Kumar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Akshat Agarwal](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Ganesh Kumar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [42] [arXiv:2609.09448](/abs/2609.09448) [[pdf](/pdf/2609.09448), [html](https://arxiv.org/html/2609.09448v1), [other](/format/2609.09448)]\n\n\n\nTitle: Do Agents Know When They Succeed? Calibrating Agent Confidence from Internal Representations\n\n\n\n[Priyanka Mary Mammen](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Emil Joswin](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Srujananjali Medicherla](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [43] [arXiv:2609.09428](/abs/2609.09428) [[pdf](/pdf/2609.09428), [other](/format/2609.09428)]\n\n\n\nTitle: XAI-Arena: Can LLMs Assess the Quality of XAI Explanations?\n\n\n\n[Yanfei Hu Fleischhauer](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Alona Zharova](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Nadja Klein](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Stefan Feuerriegel](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [44] [arXiv:2609.09418](/abs/2609.09418) [[pdf](/pdf/2609.09418), [html](https://arxiv.org/html/2609.09418v1), [other](/format/2609.09418)]\n\n\n\nTitle: Valerant: An Automatic Navigable Game Map Generator via Action-Conditioned World Model Exploration\n\n\n\n[Yiran Qiao](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Feng Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jing Ma](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [45] [arXiv:2609.09413](/abs/2609.09413) [[pdf](/pdf/2609.09413), [html](https://arxiv.org/html/2609.09413v1), [other](/format/2609.09413)]\n\n\n\nTitle: Decision-Focused Active Learning for Scale-Aware Critical-Materials Recovery\n\n\n\n[Niranjan Srinivas](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Debajyoti Ray](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Elias Nakouzi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computational Engineering, Finance, and Science (cs.CE); Robotics (cs.RO)\n\n\n\n   [46] [arXiv:2609.09395](/abs/2609.09395) [[pdf](/pdf/2609.09395), [html](https://arxiv.org/html/2609.09395v1), [other](/format/2609.09395)]\n\n\n\nTitle: The Menu Is an Execution Prior: State-Path Tool Menus for Online Agents\n\n\n\n[Bo Yan](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Weikai Lin](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Song Wang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Accepted to EMNLP 2026 Main\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [47] [arXiv:2609.09374](/abs/2609.09374) [[pdf](/pdf/2609.09374), [html](https://arxiv.org/html/2609.09374v1), [other](/format/2609.09374)]\n\n\n\nTitle: An Autonomous GeoAI Agent for Arctic Eco-Navigation\n\n\n\n[Samira Alkaee Taleghan](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Younghyun Koo](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Farnoush Banaei-Kashani](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI)\n\n\n\n   [48] [arXiv:2609.09306](/abs/2609.09306) [[pdf](/pdf/2609.09306), [html](https://arxiv.org/html/2609.09306v1), [other](/format/2609.09306)]\n\n\n\nTitle: Gradland: On Phenomenal Experience, Differentiated Across Many Dimensions\n\n\n\n[David Balduzzi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: Code: [this https URL](https://github.com/dbalduzzi/gradland/)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Neural and Evolutionary Computing (cs.NE)\n\n\n\n   [49] [arXiv:2609.09233](/abs/2609.09233) [[pdf](/pdf/2609.09233), [html](https://arxiv.org/html/2609.09233v1), [other](/format/2609.09233)]\n\n\n\nTitle: Subagents vs Agent Skills: Executing Reusable Knowledge for Long-Horizon Agentic Tasks\n\n\n\n[Wasu Top Piriyakulkij](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Rachel Lawrence](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Alicia Curth](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Sushrut Karmalkar](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Niranjani Prasad](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Computation and Language (cs.CL); Machine Learning (cs.LG)\n\n\n\n   [50] [arXiv:2609.09226](/abs/2609.09226) [[pdf](/pdf/2609.09226), [other](/format/2609.09226)]\n\n\n\nTitle: Adaptive Entangled Game Modules in Artificial General Intelligence\n\n\n\n[Haochen Li](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Xinshuai Guo](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Jingdong Ouyang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Wei Zhang](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D), [Leilei Shi](https://arxiv.org/search/cs?searchtype=%5BREDACTED%5D&query=%5BREDACTED%5D)\n\n\n\nComments: 22 pages, 13 figures, and 3 tables\n\n\n\nSubjects: Artificial Intelligence (cs.AI); Physics and Society (physics.soc-ph); Neurons and Cognition (q-bio.NC); General Finance (q-fin.GN); Quantum Physics (quant-ph)\n\n\n\n\n\nTotal of 1199 entries : 1-50 [51-100](/list/cs.AI/recent?skip=50&show=50) [101-150](/list/cs.AI/recent?skip=100&show=50) [151-200](/list/cs.AI/recent?skip=150&show=50) ... [1151-1199](/list/cs.AI/recent?skip=1150&show=50)\n\n\n\nShowing up to 50 entries per page: [fewer](/list/cs.AI/recent?skip=0&show=25) | [more](/list/cs.AI/recent?skip=0&show=100) | [all](/list/cs.AI/recent?skip=0&show=2000)\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n We gratefully acknowledge support from our **major funders**, [**member institutions**](https://info.arxiv.org/about/ourmembers.html), ****, and all contributors.\n\n\n\n [About](https://info.arxiv.org/about) · [Help](https://info.arxiv.org/help) · [Contact](https://info.arxiv.org/help/contact.html) · [Subscribe](https://info.arxiv.org/help/subscribe) · [Copyright](https://info.arxiv.org/help/license/index.html) · [Privacy](https://info.arxiv.org/help/policies/privacy_policy.html) · [Accessibility](https://info.arxiv.org/help/web_accessibility.html) · [Operational Status (opens in new tab)](https://status.arxiv.org)\n\n\n\n\n\n\n\nMajor funding support from\n\n\n\n [![Simons Foundation](/static/base/1.0.1/images/funders/simons-foundation.png)](https://www.simonsfoundation.org/) [![Simons Foundation International](/static/base/1.0.1/images/funders/simons-foundation-international.png)](https://www.sfi.org.bm/) [![Schmidt Sciences](/static/base/1.0.1/images/funders/schmidt-sciences.png)](https://www.schmidtsciences.org/)",
    "content_length": 36500,
    "description": "",
    "status_code": 200,
    "success": true,
    "title": "Artificial Intelligence",
    "url": "https://arxiv.org/list/cs.AI/recent"
  },
  "error": "",
  "elapsed": 0.039146732
}

{
  "tool_name": "send_message",
  "tool_input": {
    "final": true,
    "in_response_to": "14f702dc",
    "message": "Here are the 20 newest [cs.AI submissions](https://arxiv.org/list/cs.AI/recent) (Thu, 10 Sep 2026), grouped:\n\n| Theme | Papers |\n|---|---|\n| Agent memory, forgetting, personalization (4) | [ConvMem: Convolutional Memory for Long-Context Reasoning](https://arxiv.org/abs/2609.10441); [Fortunate Recall: Ontology-Driven Memory Lifecycle Management for Persistent Coherence in LLMs](https://arxiv.org/abs/2609.10413); [What Should an Agent Forget? Separating What Is Stored from What Is Used](https://arxiv.org/abs/2609.10263); [Kernel-Managed Shared Memory for System-Wide Personalization](https://arxiv.org/abs/2609.10144) |\n| Agents that act: tools, GUIs, planning (6) | [JarvisGUI: Cross-Device GUI Agents with Dynamic Task Composition](https://arxiv.org/abs/2609.10451); [TRACE: Training Reasoning Agents for Causal Exploration with Synthesized Rewards](https://arxiv.org/abs/2609.10315); [Why Sample What You Can Enumerate? Exact Policy Optimization for Genomic Tool Selection](https://arxiv.org/abs/2609.10221); [Belief-State Engine: Planning Under Partial Observability](https://arxiv.org/abs/2609.10036); [Grounded Evaluation and Repair for NL-to-PDDL Problem Generation](https://arxiv.org/abs/2609.09898); [Agent-Based ML-LLM Fusion with Self-Optimizing Prompts for Plateau Weather Alerts](https://arxiv.org/abs/2609.10135) |\n| Reasoning supervision and multimodal reasoning (3) | [From Symbolic Perception to Logical Deduction: Guiding Language Models in Geometric Reasoning](https://arxiv.org/abs/2609.10335); [Beyond Surface Imitation: Contrastive Modeling for Reasoning Path Alignment in Multimodal In-Context Learning](https://arxiv.org/abs/2609.10177); [Structural Process Supervision for Latent Chain-of-Thought Reasoning](https://arxiv.org/abs/2609.09928) |\n| Evaluation, bias, and model diagnostics (3) | [RAP: Research Attention Prediction Reveals Target-Conditioned Evidence Acquisition Biases](https://arxiv.org/abs/2609.10092); [Reference-Based Bias Detection in LLMs via Relative Representations of Hidden States](https://arxiv.org/abs/2609.10060); [Scored vs. Generated Readouts in Behavioral Language Models](https://arxiv.org/abs/2609.09882) |\n| Domain and embodied applications (4) | [Cyber-Financial Contagion: Propagation of an AI Vendor Compromise Through the Banking System](https://arxiv.org/abs/2609.10350); [OntologyAligner: Biomedical Ontology Normalization](https://arxiv.org/abs/2609.10055); [Time-Frequency Geometric Cross-Attention for Chunked Vision-Language-Action Models](https://arxiv.org/abs/2609.09925); [Decision Transformer for UAV-Mounted RIS-Assisted Dynamic D2D Communications](https://arxiv.org/abs/2609.09885) |\n\nMost common direction: making LLM-based agents reliable over long horizons. Half the batch is either managing what an agent remembers and forgets, or structuring how it plans, selects tools, and gets rewarded, rather than scaling the base model."
  },
  "dispatch_id": "toolu_01YN2rSkBcnuxD5NQaTcHAF2",
  "parent_dispatch_id": "",
  "handle": "",
  "panel_kind": ""
}

{
  "tool_name": "send_message",
  "dispatch_id": "toolu_01YN2rSkBcnuxD5NQaTcHAF2",
  "status": "completed",
  "result": "Delivered (confirmed) as 3 messages. Your draft was polished into texting register, so the wording differs from what you wrote; this is exactly what the user sees:\n1) 20 newest cs.AI papers grouped by theme:\n2) ### cs.AI submissions (Sep 10, 2026) · Theme and papers\n\n| Theme | Papers |\n|---|---|\n| Agent memory, forgetting, personalization (4) | https://arxiv.org/abs/2609.10441; https://arxiv.org/abs/2609.10413; https://arxiv.org/abs/2609.10263; https://arxiv.org/abs/2609.10144 |\n| Agents that act: tools, GUIs, planning (6) | https://arxiv.org/abs/2609.10451; https://arxiv.org/abs/2609.10315; https://arxiv.org/abs/2609.10221; https://arxiv.org/abs/2609.10036; https://arxiv.org/abs/2609.09898; https://arxiv.org/abs/2609.10135 |\n| Reasoning supervision and multimodal reasoning (3) | https://arxiv.org/abs/2609.10335; https://arxiv.org/abs/2609.10177; https://arxiv.org/abs/2609.09928 |\n| Evaluation, bias, and model diagnostics (3) | https://arxiv.org/abs/2609.10092; https://arxiv.org/abs/2609.10060; https://arxiv.org/abs/2609.09882 |\n| Domain and embodied applications (4) | https://arxiv.org/abs/2609.10350; https://arxiv.org/abs/2609.10055; https://arxiv.org/abs/2609.09925; https://arxiv.org/abs/2609.09885 |\n3) Most common direction: making LLM-based agents reliable over long horizons. Half the batch is either managing what an agent remembers and forgets, or structuring how it plans, selects tools, and gets rewarded.\nThese are your words now, and this is the whole reply: the delivery layer chose its density and shape by design. Never dispute, re-explain, or apologize for the phrasing, and never send a supplement to restore detail it left out. Only a final send settles its request. Never chase wording with another send. If a fact in it is wrong and would change what the user does, correct it when their next message arrives.\n\nYour turn continues.\nanswers: 14f702dc-7365-4062-8f89-2298a9aa42c1",
  "error": "",
  "elapsed": 8.105325831
}