cs.AI（2026-10-26）

📊 共 43 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) (29) 支柱二：RL算法与架构 (RL & Architecture) (10 🔗2) 支柱一：机器人控制 (Robot Control) (2) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (29 篇)

#题目一句话要点标签🔗⭐
1 DocTalk: Enhancing Medical Knowledge Retrieval via Fine-tuned Large Language Models with Chain-of-Thought Reasoning for Robust Clinical Decision Support 提出DocTalk以解决医疗知识检索中的安全性与有效性问题 large language model chain-of-thought
2 Faithful Explanations under Missing Modalities in Multimodal Dementia Prognosis 提出多模态缺失情况下的可信解释方法以改善阿尔茨海默病预后 large language model multimodal
3 Patient-specific rubric-based evaluation of large language models across longitudinal cancer care 提出NPC-RubricBench以评估癌症护理中的语言模型表现 large language model
4 Reva: Real-Time Vulnerability Detection through Contextual Probing of Large Language Models 提出REVA以解决实时漏洞检测问题 large language model
5 EEGaze: Multimodal Attention Prediction in Visual Search via EEG-Based Saliency Refinement 提出EEGaze以解决视觉搜索中的多模态注意力预测问题 multimodal
6 Explainable Multimodal AI Approach to Reduce Unnecessary Breast Biopsies 提出多模态AI方法以减少不必要的乳腺活检 multimodal
7 CANS: Capability-Preserving Neuron Steering for Large Language Models 提出CANS以解决大语言模型行为控制与能力保持问题 large language model
8 AgenticCity-Bench: Longitudinal Multimodal Observations of Urban Dynamics with Agentic Reasoning 提出AgenticCity-Bench以解决城市动态建模不足问题 multimodal
9 LongSearchEval: Evaluating Long-Horizon Web Search Capabilities of Large Language Models via a Trajectory-Aware Agent Framework with Failure-Node Localization 提出LONGSEARCHEVAL框架以评估长时间网页搜索能力 large language model
10 Co-Design Bench: Multimodal Collaborative Coding Evaluation for Agentic Models 提出Co-Design Bench以评估人机协作编码能力 multimodal
11 Zero-Shot Multi-Class Wafer Map Classification with Rule-Ordered Prompts for Large Multimodal Models 提出无监督六类晶圆图缺陷分类框架以应对生产条件变化 multimodal
12 Beyond Behaviour: A Stimulus-Response Safety Evaluation of Vision-Language-Action Models 提出视觉-语言-行动模型的刺激-反应安全评估方法 vision-language-action
13 Omni-InteractScope: Benchmarking Proactive and Agentic Interaction in Continuous Omnimodal Streams 提出Omni-InteractScope以解决Omni-LLMs在连续多模态流中的交互问题 large language model
14 Self-Supervised Pretraining for PHM: What Transfers, What Does Not, and What Federating It Costs 提出自监督预训练方法以提升PHM任务性能 foundation model
15 Toward Fast Human Reaction Modeling from High-Frequency Digital Interaction Data 提出Ego-Knowledge以解决快速人类反应建模问题 foundation model
16 A Unified Arena for LLM Agents across Diverse Eurogame 提出DECE以评估大语言模型在多样化Eurogame中的表现 large language model
17 Governing Enterprise AI Traffic at Scale: Design and Operation of a Policy-Driven, Multi-Vendor LLM Gateway on Envoy 提出Cosmos LLM网关以解决企业AI流量治理问题 large language model
18 MarkovSim: an LLM Agent Harness with Bounded Context for Long-Horizon Computational Simulation 提出MarkovSim以解决长时间计算仿真中的上下文管理问题 large language model
19 Satellite Intelligence: Distributed Training, Collaborative Inference,and LLM Adaptation in LEO Constellation 提出卫星智能框架以解决低轨卫星网络中的AI训练与推理问题 large language model
20 Context Is Not Free: Carbon-Aware LLM Routing for Multi-Turn LLM Web Services across Geo-Distributed Data Centers 提出CarbonRouter以解决多轮LLM服务的碳排放问题 large language model
21 V$^3$AST-Bench: Verifiable, Valid, and Versatile Safety Evaluation for Coding Agents 提出V$^3$AST-Bench以解决编码代理的安全评估问题 large language model
22 AutoEnergyDesign: An LLM-Driven Agent for Automatic Energy System Design Optimization 提出AutoEnergyDesign以优化建筑能源系统设计 large language model
23 Beyond Single-Attribute Auditing: A Configurable Multi-Attribute Benchmark for Intersectional Fairness in LLMs 提出ComFair基准以解决多属性公平性审计问题 large language model
24 PatentAgentBench: Benchmarking LLM Agents with Professional Patent Task Suites 提出PatentAgentBench以评估LLM在专利任务中的表现 large language model
25 Critic's Eye: Review-Grounded Graph Reasoning for Long-Form Motif Discovery 提出Critic's Eye以解决长篇文学中的动机发现问题 foundation model
26 C-SafeGen: Certified Safe LLM Generation with Claim-Based Streaming Guardrails 提出C-SafeGen以解决大型语言模型生成内容安全性问题 large language model
27 PhysEvolve: Verifier-grounded Self-evolution for Physical Reasoning with Executable Worlds 提出PhysEvolve以解决物理推理中的数据稀缺问题 zero-shot transfer
28 Collective Emergent Alignment: Measuring and Amplifying Interaction-Induced Reasoning Alignment in Multi-Agent LLMs 提出REMAP框架以量化多智能体LLM的推理对齐 large language model
29 Silent and Costly: Characterizing and Detecting Prompt-Caching Mis-Integration in LLM Agents 提出CACHELINT以解决LLM代理中的提示缓存误集成问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
30 UniGFM-RAG: Unify Graph Foundation Models and Large Language Models for End-to-End Retrieval and Reasoning 提出UniGFM-RAG以解决复杂推理与知识整合问题 distillation large language model foundation model
31 When Fusion Order Matters: Insights on Progressive Multimodal Representation Learning for Cross-modal Retrieval 提出UniPRO以优化多模态视频检索中的融合顺序问题 representation learning multimodal
32 EvoRouter: Cost-Efficient Multi-Agent Routing via Insight-Guided On-Policy Distillation 提出EvoRouter以解决多智能体路由的成本效率问题 reinforcement learning distillation large language model ✅
33 Movable-Element Multi-Functional RIS-Assisted Hybrid RSMA-NOMA: A Constrained Hybrid Reinforcement Learning Approach 提出可移动元素多功能RIS架构以提升通信灵活性和能效 reinforcement learning deep reinforcement learning
34 Evolve on Graph: Self-Evolving Knowledge Graph Reasoning without Question Annotations 提出Evolve on Graph框架以解决无QA注释的知识图推理问题 reinforcement learning large language model ✅
35 Test-Time Confidence Adaptation under RL-Induced Trajectory Shift 提出StaR-TTA以解决RL引起的轨迹偏移问题 reinforcement learning large language model
36 An Interactive Evolutionary Multiobjective Optimization Algorithm Based on Interpretable Active Preference Learning 提出可解释的交互式进化多目标优化算法以解决决策透明性问题 preference learning
37 MuSID-KD: Multigranularity Knowledge Distillation for Generative Recommendation with Semantic IDs 提出MuSID-KD以解决生成推荐中的知识蒸馏问题 distillation
38 Centralized Critics for N-Agent Ad Hoc Teamwork 提出集中式评论者以解决N-agent临时团队合作问题 reinforcement learning privileged information
39 Uncovering the Turing Test's Latent Signal: Indistinguishability as a Reward 提出基于图灵测试的自我归属概率作为强化学习奖励机制 reinforcement learning teacher-student

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
40 SoK: Timestamps Are Not a Time Model: Rethinking Temporal Reasoning in Host-Based Intrusion Detection 提出时间模型以解决主机入侵检测中的时间推理问题 manipulation TAMP
41 Provably Efficient Q-learning for Robust Reinforcement Learning 提出高效的Q学习算法以解决稳健强化学习问题 sim-to-real reinforcement learning

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
42 Geometric Consistency Based Label-Free Validation for Checkpoint Selection in Audio Active Learning 提出几何一致性无标签验证以优化音频主动学习中的检查点选择 geometric consistency

🔬 支柱三：空间感知与语义 (Perception & Semantics) (1 篇)

#题目一句话要点标签🔗⭐
43 Reasoning Patterns for Efficient AI Coding: An Empirical Study 提出推理模式以提升AI编码效率 semantic mapping semantic map

⬅️ 返回 cs.AI 首页 · 🏠 返回主页