cs.AI（2026-10-15）

📊 共 69 篇论文 | 🔗 8 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (46 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (22 🔗3) 支柱一：机器人控制 (Robot Control) (1 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (46 篇)

#题目一句话要点标签🔗⭐
1 When Reasoning Is Not Enough: Proactive Inquiry for Cross-Modal Safety in Multimodal Large Language Models 提出SAFE-PRO以解决多模态大语言模型的安全性问题 large language model multimodal
2 MLExpert: Towards a Generalist Agentic Model for Autonomous Machine Learning Engineering 提出MLExpert以解决自主机器学习工程优化问题 generalist agent large language model
3 CART: Compatible Routing for Multimodal Continual Instruction Tuning 提出CART以解决多模态持续指令调优中的适应性问题 large language model multimodal
4 Following Delegation Traces to Study Human Oversight in Interaction with Multimodal AI 提出观察委托轨迹以研究人类与多模态AI的互动监督问题 multimodal
5 Discovering state-of-the-art quantum circuits with large language models 提出进化框架利用大语言模型发现高效量子电路 large language model
6 Agentic-Priors: Using Large Language Models to Address the Cold Start Problem in Preference Elicitation 提出Agentic-Priors以解决偏好引导中的冷启动问题 large language model
7 Functional Subnetwork Localization for Efficient Continual Learning in Large Language Models 提出功能子网络定位方法以解决大语言模型的持续学习问题 large language model
8 Tool-Augmented Large Language Models for Reliable Medical Calculation from Patient Notes 提出工具增强的大语言模型以解决医疗计算的可靠性问题 large language model
9 Full-Duplex Interaction with Speech, Expressions and Gestures 提出ELLSA 2以解决全双工对话中的多模态交互问题 large language model multimodal
10 AutoOnboarder: Autonomously Onboarding LLM Agents 提出AutoOnboarder以解决LLM代理的自动化入职问题 large language model foundation model
11 The Delegation Dilemma: Economies of Scale, Accountability, and Polycentric Governance 提出一种模型以解决代理与问责之间的权衡问题 large language model
12 The Self-Improvement Problem 提出自我改进问题的解决方案以确保安全性 large language model
13 When is a better LLM worth paying for? Stochastic Optimization with Switching Mixed-Fidelity Oracles 提出混合保真度随机优化方法以降低LLM成本 large language model
14 Evaluation Awareness Beyond English: Do LLMs Act Differently When Tested in Indic Languages? 研究大型语言模型在印地语中的评估意识差异 large language model
15 Private Retrieval Augmented Generation via Random Projection 提出随机投影方法以解决RAG系统隐私泄露问题 large language model
16 Multi-sourced AI Tutor for Non-Native English Speaker Students 提出多源AI辅导系统以解决非母语英语学生的学习障碍 large language model
17 Realtime-Bench: Can Frontier Models Solve Real-time Interactive Problems? 提出Realtime-Bench以解决实时交互问题 multimodal
18 RAG-RST: Retrieved Exemplars as In-Context Discourse Schema for RST Parsing 提出RAG-RST框架以提升RST解析性能 foundation model
19 DropKV: Decoupling Residual-Output Perturbation for Near-Optimal KV-Cache Eviction 提出DropKV以解决KV缓存管理中的输出扰动问题 large language model
20 GraphACM: Graph-based Action Consistency Measurement for Uncertainty Quantification in LLM Agents 提出GraphACM以解决LLM代理行为一致性测量问题 large language model
21 CodeHealth Predicts the Token Cost of Agentic Code Review: A Pre-Registered Study of 465 Agent-Authored Pull Requests CodeHealth预测代理代码审查的令牌成本 large language model
22 Evaluating Inquiry Learning in LLM-Based Agents: Hypothesizing, Reasoning, and Exploration 提出统一框架评估LLM在探究学习中的能力 large language model
23 BiShield-TEE: On the (In-)Security of Unilateral Weight Obfuscation in On-Device TEE-Shielded LLM Partition 提出BiShield-TEE以解决单边权重混淆的安全性问题 large language model ✅
24 CriticaLLM: Using LLMs as Autonomous Game Master for Tabletop Role-Playing Games 提出CriticaLLM以实现桌面角色扮演游戏的自动化游戏主持 large language model
25 SCRIBE: Scalable Contextual Representations of Interpretable Brain-Grounded Semantics 提出SCRIBE框架以研究人脑对语言语义的理解 large language model
26 Decomposing Brain-LLM Alignment with Interpretable Semantic Representations 利用可解释语义表示分解大脑与LLM的对齐 large language model
27 FlashAttention with Statistics: Low-Overhead Attention Features for Compact Hallucination Detectors 提出低开销注意力特征以提升紧凑型幻觉检测器的性能 large language model
28 Ontology Refinement with Embeddings and LLMs: A Case Study on CASTON 提出基于嵌入和大语言模型的本体精炼方法以解决CASTON问题 large language model ✅
29 JevAgentHarness: Benchmarking, Deploying and Training Decision Models That Share Control with LLMs 提出JevAgentHarness以解决LLM决策效率低下问题 large language model ✅
30 MemCut: Every Secret Leaves a Trail in the Long-Term Memory of LLM Agents 提出MemCut以解决大语言模型的用户数据遗忘问题 large language model
31 PriorityFEC: Deadline-Aware Streaming FEC for KV-Cache Handoff in Edge Multi-Agent Systems 提出PriorityFEC以解决边缘多智能体系统中的KV缓存交接问题 large language model
32 FedSkillGit: Federated Skill Evolution via Version Control 提出FedSkillGit以解决联邦技能演化中的数据隐私问题 large language model
33 Bias Control and Injection for Multi-Agent Aggregation 提出多代理聚合框架以解决决策偏见问题 large language model
34 Brain-Aligned Multimodal LLMs for Social and Emotional Understanding of Cartoon Movies 提出脑对齐的多模态大语言模型以理解卡通电影中的社会与情感 large language model multimodal
35 Recoverability-Controlled Safety Miscalibration in Large Language Models 提出RiskCueBench以解决语言模型安全行为评估问题 large language model
36 Does Latent Chain-of-Thought Actually Reason? A Single, Unifying Taxonomy for Latent Reasoning 提出统一分类法以解析潜在推理的计算机制 chain-of-thought
37 Graph Serializations for Reasoning LLMs: A Study Across Knowledge Extraction and Structural Reasoning 提出动态图序列化方法以提升LLM推理性能 instruction following
38 LEMON: Equivariant Learning over Attention Maps for Zero-Shot Misbehavior Monitoring on New LLMs 提出LEMON以解决新LLM的零-shot失误监控问题 large language model
39 Are Large Reasoning Models Good Everyday Reasoners? A Comparison with Non-Reasoning Models on Logical Reasoning Tasks 评估大型推理模型在日常逻辑推理中的表现 chain-of-thought
40 Testing the Planning Claims of LLM-Based Agents against Formal Semantics 通过形式语义验证LLM代理的规划能力 chain-of-thought
41 Political Narratives in the 2024 Moldovan Presidential Election: Stance, Amplification, and Influence 通过多语言立场检测与网络分析研究摩尔多瓦2024年总统选举中的政治叙事 large language model
42 From Natural Language Intent to Executable Control: Hierarchical Feedback-Guided LLM Planning for Multi-component Cyber-Physical Systems 提出分层反馈引导的LLM规划以解决CPS执行问题 large language model
43 BiShield-TEE: On the (In-)Security of Unilateral Weight Obfuscation in On-Device TEE-Shielded LLM Partition 提出BiShield-TEE以解决单向权重混淆在TEE保护下的安全性问题 large language model ✅
44 OnBodyAI - Why the Next AI Needs a Body 提出OnBodyAI以实现更个性化的人工智能交互 large language model
45 SSR: Structural Signal Retriever for Scalable Agent Discovery in LLM-Orchestrated Multi-Agent Systems 提出SSR以解决大规模多代理系统中的代理发现问题 large language model
46 Sci$Π$vot: Benchmarking Principal-Investigator Judgment at Pivotal Moments of Scientific AutoResearch Agents 构建Sci$Π$vot基准以评估科学研究中的判断能力 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (22 篇)

#题目一句话要点标签🔗⭐
47 CoWorld Reasoner: Agentic Multimodal Reasoning with the Evolving Video World Model-Driven Harness 提出CoWorld Reasoner以解决多模态推理中的灵活性问题 world model world models spatiotemporal
48 Progress Reward Model for Reinforcement Learning via Large Language Models 提出进展奖励模型以解决稀疏奖励的强化学习问题 reinforcement learning reward shaping large language model ✅
49 The Density of Thought: Predicting and Steering Reasoning in Large Language Models 提出信息密度预测与引导方法以提高大语言模型推理准确性 reinforcement learning large language model chain-of-thought
50 CIM-Embed: Learning Instruction-Conditioned Multimodal Embeddings for Clustering 提出CIM-Embed以解决多模态聚类中的指令依赖问题 contrastive learning multimodal
51 Learning Intrinsic Dynamics in Latent World Models 提出Concat IDM以解决内在动态学习问题 world model world models
52 Learning from Interaction Experience via Localized Context Distillation 提出局部上下文蒸馏以解决交互经验学习中的偏差问题 distillation privileged information large language model
53 Constrained Multi-Agent Reinforcement Learning in Imagination 提出CA-MAMBA以提升多智能体强化学习的效率 reinforcement learning world model world models
54 Incentivizing Tool-Use Reinforcement Learning for Post-Arrival Emergency Care Prediction 提出TIR-EC框架以提升急救预测的工具使用强化学习 reinforcement learning large language model
55 MemComposer: Learning to Assemble Memories for LLM Agents via Reinforcement Learning 提出MemComposer以优化LLM代理的记忆组装问题 reinforcement learning large language model
56 DeciCAD: Harness-Grounded Learning of Action Policies for CAD Agents 提出DeciCAD以优化CAD领域的智能体决策与学习 reinforcement learning policy learning large language model
57 X-RIC: An Agentic RAN-Intelligent Controller Benchmark for Enhancing Reinforcement Learning 提出X-RIC以增强强化学习在O-RAN中的应用 reinforcement learning
58 Learning from Unverified Peers via Shared Residual Distillation 提出共享残差蒸馏方法以从未验证的同伴学习 distillation
59 Hierarchical Interpretable Multitask Audio Representation Learning: An Analysis of Event-Scene Relationships through the Information Bottleneck 提出层次化可解释的多任务音频表示学习以解决声学场景分类问题 representation learning
60 Rubric Judgments Lack Uncertainty: Conformal Weighting for LLM Evaluation 提出不确定性感知的评分加权方法以改进LLM评估 reinforcement learning large language model
61 Condition-Contrastive Extrapolation for On-Policy Distillation of Language Models 提出条件对比外推以解决语言模型蒸馏中的错误推理问题 distillation
62 Mallea: Building, Steering, and Memorizing Worlds with Coding Agent 提出Mallea以构建、操控和记忆文本驱动的世界模型 world model world models ✅
63 PROBE: Learning to Audit Policy Compliance in Tool-Using LLM Agents 提出PROBE以解决工具使用LLM代理的政策合规审计问题 DPO reward design
64 FlexUMM: Flexible Representation Learning for Unified Multimodal Models 提出FlexUMM以解决统一多模态模型的表示冲突问题 representation learning multimodal ✅
65 GlobeBench: Evaluating Text-Based World Model Capabilities for Scalable Environment Simulations 提出GlobeBench以评估文本基础世界模型在环境模拟中的能力 world model world models
66 All-in-One Memory: Process-Rewarded Reinforcement Learning for Memory Agents 提出All-in-One Memory以解决长时记忆优化问题 reinforcement learning DPO large language model
67 Diverse Reasoning Paths Matter: Symbolic Equivalence Augmentation for Enhancing RL Exploration 提出EquivAug以增强RLVR探索能力 reinforcement learning chain-of-thought
68 CleanRM: Reward Modeling from Noisy Feedback with a Few Clean Feedback for RLHF 提出CleanRM以解决反馈噪声对奖励建模的影响 RLHF

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
69 Reinforcement Learning-driven Diffusion Models: Language, Vision, and Beyond 提出强化学习驱动的扩散模型以解决生成路径探索不足问题 trajectory optimization reinforcement learning RLHF ✅

⬅️ 返回 cs.AI 首页 · 🏠 返回主页