cs.AI（2026-10-16）

📊 共 36 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (21 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (13 🔗1) 支柱一：机器人控制 (Robot Control) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (21 篇)

#题目一句话要点标签🔗⭐
1 Self-Reflective Chain-of-Thought (SRCoT): Evidence-Based Text Generation via a Closed-Loop Retrieve--Reflect--Revise Pipeline on Edge LLMs 提出自反思链式思维(SRCoT)以解决LLMs生成文本的支持不足问题 large language model chain-of-thought
2 OwlSight: A Panoramic Monitor for Chain-of-Thought Reasoning 提出OwlSight以解决多任务链式思维监控问题 chain-of-thought
3 PresenceBench: An Exploratory Evaluation of Large Language Model Adaptation to User State 提出PresenceBench以评估大语言模型对用户状态的适应性 large language model
4 Trapped in Your Story: Evaluating Narrative-Induced Sycophancy in Large Language Models 提出叙事诱导的谄媚性评估方法以研究大型语言模型的偏见 large language model
5 S-GAP: Stealthy Graph Attack Primitives for Graph Large Language Models 提出S-GAP以解决图大语言模型的攻击脆弱性问题 large language model
6 CoTA++: Understanding the Repeat Curse in dMLLMs from an Information Flow Perspective 提出CoTA++以解决dMLLMs中的重复诅咒问题 large language model multimodal
7 Adaptation of Computer Use Agents to New Software with User Manuals 提出MAGE框架以提升计算机使用代理对新软件的适应性 foundation model multimodal ✅
8 QD-Rec: Open-Ended Evolution of Recommendations via Quality-Diversity 提出QD-Rec以解决开放式推荐中的多样性与质量问题 foundation model
9 Reasoning Models Can Control Their Chains of Thought When They Tell Themselves To 提出自我指令以提升推理模型的思维链可控性 chain-of-thought
10 Beyond SAEs: Consistent Concept Discovery with Sparse Coding 提出一致稀疏字典学习方法以解决SAE不一致性问题 foundation model
11 On the possibilities and limitations of LLM-based multi-agent systems for multi-criteria decision-making in healthcare operations management 提出框架评估LLM生成偏好在医疗决策中的可靠性 large language model
12 From Trajectories to Differentiated Skills: Cognitive Differentiation for Self-Improving Agents 提出认知差异化技能进化方法以提升自我改进代理的能力 large language model
13 Darwin-Claw: Environment-Adaptive Experience Consolidation for Recursively Self-Evolving Agents 提出Darwin-Claw以解决自我进化代理的环境适应性问题 large language model
14 SkillBrowser: Selective Skill Exposure for Scalable Skill-Using Agents 提出SkillBrowser以解决可扩展技能使用中的选择性技能暴露问题 large language model
15 LocalIntent: Outcome-Specific Structured Generation for Backend-Compatible Smart Building Automation with Local LLMs 提出LocalIntent以提升智能建筑自动化的结果特定生成能力 large language model
16 Designing for Imperfect Prompts: A Simulation-Based Evaluation of an LLM-Based Enterprise Data Assistant Using Knowledge Graphs 提出知识图谱介导以解决企业数据助手中的不完美提示问题 large language model
17 Not All Test-Time Compute Is Evidence: Model-Specific Bias in LLM Reasoning 提出DIA以解决LLM推理中的模型特定偏差问题 chain-of-thought
18 Validating LLM-generated political networks with survey-reported partisan exposure 基于LLM的政治网络生成与调查报告的验证方法 large language model
19 LexPilot: Benchmarking the Proactive Interaction Ability of Legal LLMs with Laypeople in Chinese Civil Case Consultation 提出LexPilot以解决法律LLM在民事咨询中的主动交互能力不足问题 large language model
20 NL2SQL with Iterative Self-Refinement 提出多代理驱动的NL2SQL框架以解决高精度数据访问问题 large language model
21 Multiple Turns, Short Horizons: Measuring Effective-Horizon Compression in Jailbreaks 提出有效视野压缩方法以优化多轮攻击查询 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (13 篇)

#题目一句话要点标签🔗⭐
22 Beyond Large Language Models: A Mental World Model Framework for Mental Modeling, Simulation, Evolution, and Interpretability 提出心理世界模型框架以解决主观认知建模问题 world model world models large language model
23 Beyond Reward Scoring: Online Multi-Agent Preference Learning for Multimodal Response Generation 提出PrefMind以解决多模态响应生成中的偏好理解问题 reinforcement learning preference learning RLHF
24 World Action Critic (WAC) Model: Joint Foresight and Criticism for Policy Learning 提出世界动作评论家模型解决反事实动作评估问题 policy learning world model world models
25 CIIR - Class-Indexed Interface Representation Learning 提出CIIR以解决多模态数据集无对应问题 representation learning multimodal
26 Sep-Latent: Separator Tokens as Natural Compression Anchors for Implicit Reasoning 提出Sep-Latent以解决隐式推理中的表示崩溃问题 distillation large language model chain-of-thought
27 SimOPD: Relational Knowledge Transfer for On-Policy Distillation 提出SimOPD以解决在政策蒸馏中的知识提取问题 contrastive learning distillation
28 OfficeTown: Generative Organizations for Evolving Agent Intelligence 提出OfficeTown框架以提升智能体在专业环境中的协作能力 distillation generalist agent
29 WebCoder: Scaling General Frontend Development with Artifact-Grounded Agentic Reinforcement Learning 提出WebCoder以解决前端开发中的多任务学习问题 reinforcement learning
30 Distill Until Divergence: Error-Truncated Hindsight Distillation for Agentic RL 提出错误截断的后见蒸馏方法以解决长时间决策问题 distillation
31 Where Research Ideation Begins: Forward-Citation Supervision for Actionable Limitation Generation 提出基于前向引用监督的行动限制生成方法以提升研究构思 reinforcement learning large language model
32 Causal Value Attribution of Marketing Labels in LLM-Generated Ads: A Deconfounder Approach 提出D-VAE框架以解决生成广告中的因果价值归因问题 representation learning multimodal
33 DiffusionRVGE: Reinforcement Learning via Regression-Based Value Gradient Estimation 提出DiffusionRVGE以提高扩散模型的强化学习效率 reinforcement learning ✅
34 Agentic In-Context Learning: What Algorithms Can Transformers Learn in Closed-Loop Interaction? 提出Agentic ICL以解决固定参数Transformer学习局限性 reinforcement learning distillation

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
35 MobileInject: Evaluating Reasoning-to-Action Security of MLLM-Based Mobile Agents under Cross-Modal Consistency Injection 提出MobileInject以评估MLLM基础移动代理的安全性问题 manipulation large language model multimodal

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
36 ASE: Attribution-Guided Semantic Evolution for Time Series Forecasting 提出ASE框架以提升时间序列预测中的文本信息利用 ASE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页