cs.AI（2026-10-27）

📊 共 39 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (25 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (11 🔗4) 支柱一：机器人控制 (Robot Control) (3 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (25 篇)

#题目一句话要点标签🔗⭐
1 CogArena: Benchmarking Multimodal Agents on Interactive Behavioral Experiments 提出CogArena以评估多模态智能体在互动行为实验中的表现 large language model multimodal
2 Correctness Conditioning as Selection in Chain-of-Thought Faithfulness Evaluation 提出链式思维忠实度评估中的选择性修正方法 chain-of-thought
3 Natural Hard-Negative Mining from Positives for Instruction-Following Retrieval 提出自然困难负样本挖掘以提升指令跟随检索性能 instruction following
4 Who Knows What - and What Does It Support? Perspective-Bounded Evidence Calibration for Multimodal Theory of Mind 提出PIP-ToM以解决多模态心智理论中的证据可接受性问题 multimodal
5 ADTGuard: Adaptive Depth-Time Safeguard for Streaming Inference in Large Language Models 提出ADTGuard以解决大语言模型流媒体推理中的安全问题 large language model
6 LM-DSB: A Lightweight Multi-agent System for Automated Daily Safety Bulletin Generation in Construction 提出LM-DSB系统以自动生成建筑工地日常安全公报 large language model multimodal
7 Learning Where to Trust: Reliability-Guided Expert Routing and Velocity Gating for 3D Deformable Registration 提出ReliMorph以解决3D变形配准中的可靠性问题 multimodal
8 AnalogScout: Simulation-Efficient PVT-Aware Analog Sizing with LLM Agents 提出AnalogScout以解决模拟器件PVT敏感性优化问题 large language model
9 Blindfold: Keeping the Audit Coin Secret from LLM Agents 提出Blindfold以解决LLM代理审计隐私问题 TAMP
10 God Does Not Play Dice: Enforcing Batch Invariance in Agentic Kernel Generation 提出 ool以解决GPU内核生成中的批次不变性问题 large language model
11 Timestral 提出Timestral以解决时间序列预测问题 foundation model
12 Concordat: Typed Sessions for LLM Agents 提出Concordat以解决大型语言模型代理的协议执行问题 large language model
13 FeatureArc: Benchmarking Coding Agents on Evolving Features 提出FeatureArc以评估编码代理在特征演变中的表现 large language model
14 Beyond the Readout: A Latent Numerical Frame Shared Across Modality and Operation 提出跨模态共享潜在数字框架以解决数字理解问题 multimodal
15 Clearing the FoG: Training KG Agents to Retrieve the Evidence Behind Their Answers 提出FoG框架以解决KG代理缺乏证据支持的问题 large language model
16 Uncovering Why LLMs Fail: From Internal Mechanisms to Adversarial Reasoning Trajectories 提出CauTrace框架以揭示大型语言模型失败的根本原因 large language model
17 ReForest: Interpretability Meets Reasoning in Isolation Forest 提出ReForest以解决孤立森林模型的语义理解问题 large language model
18 Grounding LLM Agents in Business Processes: Process-Aware Knowledge Graphs for Reliable Multi-Step Service Invocation 提出基于过程感知知识图的LLM代理以解决API调用可靠性问题 large language model
19 Rènyi-Regularized Flow-Based Variational Inference for Bayesian Calibration 提出Rényi正则化流基变分推断以解决贝叶斯校准中的多模态问题 multimodal
20 AutoVend: Benchmarking LLM Agents on Long-Horizon Autonomous Vending Operations 提出AutoVend基准以评估LLM代理在长期自动售货操作中的表现 large language model
21 When the Critic Falls Behind: Co-Evolving Rubrics and Harnesses for Continual Agent Self-Improvement 提出RuHarness框架以解决自主体自我改进中的评估滞后问题 large language model
22 Dependency Graphs of Mathematical Arguments: Representation, Structural Measurement, and Proof Review 提出基于依赖图的数学论证表示与测量框架 large language model
23 Can LLMs Reliably Grade Olympiad Proofs? A Study of Mathematical Verification with LLMs 提出MABGrader以提升奥林匹克数学证明的验证准确性 large language model ✅
24 Learning to Formalize: Bridging Open-Ended Workloads and Fast Decision Models 提出小候选形式化器以解决开放式工作负载与快速决策模型之间的接口问题 large language model
25 Benchmarking LLM Agents for Behaviorally Adequate Test Generation from Software Requirements 提出Req2Test-Bench以解决软件需求测试生成的行为适应性问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (11 篇)

#题目一句话要点标签🔗⭐
26 AC-JEPA: Anchored Cross-Modal Joint Embedding Predictive Architecture for Visual-to-Symbolic Translation 提出AC-JEPA以解决跨模态翻译中的数据不完美问题 JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture
27 JEPA-Inspired Latent Field Prediction for Sparse-Pilot MIMO-OFDM Channel Reconstruction 提出JEPA启发的潜在场预测以解决稀疏导频MIMO-OFDM信道重建问题 world model world models JEPA
28 GUI-JEPA: Candidate-Conditioned JEPA Latents as a Fast Action Policy for Mobile GUI Agents 提出GUI-JEPA以解决移动GUI代理的速度和内存问题 JEPA Joint-Embedding Predictive Architecture joint-embedding predictive architecture ✅
29 PailitaoMMSearch2.0 Technical Report: Scaling Generative Retrieval to Unified End-to-End Multimodal Search 提出PailitaoMMSearch以解决多模态检索统一性问题 reinforcement learning multimodal
30 RouteDRAG: Learning to Route and Control Agentic Execution for Multimodal Document RAG 提出RouteDRAG以优化多模态文档的检索与执行问题 policy learning multimodal ✅
31 Strategyproof Deep Reinforcement Learning for Online Resource Allocation 提出策略无关的深度强化学习以解决在线资源分配问题 reinforcement learning deep reinforcement learning
32 Guided-RLVR: Guided Exploration in Reinforcement Learning with Verifiable Rewards 提出Guided-RLVR以解决RLVR探索空间狭窄问题 reinforcement learning large language model
33 RPSD: Retrieval-Privileged On-Policy Self-Distillation with Optimal Transport 提出RPSD以解决序列推荐中的信用盲优化与协作盲监督问题 reinforcement learning distillation
34 From Solver Trajectories to Teaching Trajectories: Cognition-Aligned Reasoning Distillation 提出Cognition-Aligned Reasoning Distillation以解决知识蒸馏中的认知不匹配问题 RLHF distillation ✅
35 Vision-Language Agents Can Learn in Their Own Latent Worlds 提出Dwell框架以解决视觉语言模型在线学习的效率问题 reinforcement learning world model world models ✅
36 Noise Steering: Exploiting Wake Steering and Acoustic Directivity for Noise-Constrained Wind Farm Control via Reinforcement Learning 提出噪声引导策略以解决风电场噪声控制问题 reinforcement learning

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
37 Correction or Manipulation? Measuring Epistemic Vigilance in LLM Agents Under User Pressure 提出测量LLM代理在用户压力下的认知警觉性的方法 manipulation
38 Polarity--Time Interaction and Multi-Level Reliability for Event-Based Spacecraft Pose Estimation 提出一种可靠性感知框架以解决航天器姿态估计问题 trajectory optimization spatiotemporal
39 Evolving Agent Teams 提出进化代理团队以解决GPU内核优化中的局部最优问题 manipulation ✅

⬅️ 返回 cs.AI 首页 · 🏠 返回主页