cs.AI（2026-10-23）

📊 共 39 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) (28 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (8) 支柱一：机器人控制 (Robot Control) (2) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (28 篇)

#题目一句话要点标签🔗⭐
1 CASE: Confusion-Aware Skill Evolution for Frozen Multimodal Emotion Models 提出CASE框架以解决多模态情感识别中的情感混淆问题 large language model multimodal
2 Beyond Post-Hoc Rewards: Privileged Reasoning Exploration for Multimodal Embeddings 提出PILOT框架以优化多模态嵌入模型的推理探索 large language model multimodal
3 Adaptive Control of Scientific Instruments with Nonstationary Responses through Multimodal Agents 提出多模态智能体以解决科学仪器非平稳响应问题 large language model multimodal
4 MedCaseAgent: From Multimodal Clinical Evidence to Rare Disease Case Reports 提出MedCaseAgent以解决稀有疾病案例报告生成问题 large language model multimodal
5 A Foundation Model for Policy In-Context Longitudinal Causal Effect Estimation 提出PICLE-PFN以解决动态治疗政策下的因果效应估计问题 foundation model
6 LLaCO: Continuous Diffusion for Large Language Models with Competition-Aware Semantic Scheduling 提出LLaCO以解决连续扩散模型的时间压缩问题 large language model
7 Beyond Channel Data: Configuration-Enriched Radio Foundation Models 提出配置增强的无线基础模型以解决标签稀缺问题 foundation model
8 Behaviour analysis for large language model-driven agents 提出一种框架以分析大语言模型驱动代理的行为 large language model
9 CEMIR vs. FISER: a comparative study of reasoning-based and enrichment-based approaches to pragmatic intent recognition 比较CEMIR与FISER以解决人机交互中的意图识别问题 large language model multimodal
10 DriveMind: Mind-Evolving Belief Tracking for Closed-Loop Autonomous Driving 提出DriveMind以解决闭环自主驾驶中的信念追踪问题 large language model multimodal
11 Closing the Modality Gap via Simplex-Constrained Representations 通过简单约束表示法解决多模态对齐问题 multimodal
12 Do Language Models Trust Their Own Justifications? A Study on Functional Consistency 提出功能自一致性评估框架以检验语言模型的自我解释可靠性 large language model
13 Not All Information Is Worth Asking For: Decision-Informed Information Seeking with LLMs 提出DIVA以解决LLM信息获取中的决策效率问题 large language model
14 MPQ: A Message-Passing View of Post-Training Quantization 提出MPQ以解决后训练量化中的信息损失问题 large language model ✅
15 Can Test-Time Memory Evolution Outperform Test-Time Training for Scientific Discovery? 提出测试时间记忆演化以解决测试时间训练的效率问题 large language model
16 Green Code or Costly Code? Evaluating the Energy Costs of Iterative LLM-Based Code Optimizations 评估LLM代码优化的能耗与性能之间的关系 large language model
17 AI Alignment for Digital Democracy: Fairer Voters Representation against LLM Biases 提出AI对齐方法以解决LLM偏见导致的选民代表性不公问题 large language model
18 LGMT: A Tool for Testing the Logical Reasoning Reliability of LLMs 提出LGMT工具以解决LLMs推理一致性检测问题 large language model ✅
19 When Agents Fail: Model or Harness--Who's to Blame? 提出模型与执行环境交互评估框架以优化LLM代理性能 multimodal
20 Async-RBench: Benchmarking Asynchronous Task Execution in LLM Agents 提出Async-RBench以解决LLM代理异步任务执行评估问题 large language model
21 LeafSkill: Efficient Leaf-Level Action Grounding for Recursive LLM Agents 提出LeafSkill以解决递归LLM代理的叶级动作定位问题 large language model
22 DealsFlow: An Evidence-Gated Agentic Harness for Venture Capital Deal Triage 提出DealsFlow以解决风险投资评分不准确问题 large language model
23 TGR-EvoDAG: A DAG-based Multi-Agent Code Evolution Framework for Generative Recommendation on Tencent 提出TGR-EvoDAG以解决生成推荐中的代码演化问题 large language model
24 Solvent: Long-Horizon Reasoning for Multi-Step Retrosynthesis 提出Solvent以解决多步逆合成中的长远推理问题 large language model
25 Training an MoE LLM in MXFP4: How We Closed the Gap to BF16 提出ALTO以解决FP4格式训练大语言模型的优化问题 large language model
26 PEAR: Adaptive Advantage Reshaping for Perception and Exploration in Large Vision-Language Models 提出PEAR以解决多模态推理中的探索不足问题 multimodal
27 AgentTrace: Execution-Level Fingerprinting of Black-Box LLM Agents 提出AgentTrace以解决黑箱LLM代理的来源验证问题 foundation model
28 Do AI-for-Science Agents Preserve Model Safety? 评估AI科学代理中的模型安全性问题 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
29 When to Stop Penalizing Actions? Martingale GRPO for VLA Reinforcement Learning 提出Martingale GRPO以优化视觉-语言-动作强化学习中的反馈机制 reinforcement learning vision-language-action VLA
30 Financial Context-Rich Strategy for Actor-Critic based Agents: A Deep Reinforcement Learning Approach for Stock Trading 提出基于深度强化学习的金融情境丰富策略以优化股票交易 reinforcement learning deep reinforcement learning DRL
31 From Visual Affinity to Learning Credit in Multimodal Reinforcement Learning 提出Outcome-Conditioned Visual Credit以解决多模态强化学习中的视觉关联问题 reinforcement learning multimodal
32 $Ψ$-0.6: Adaptive Streams of Thought in a Unified Vision-Language World Model 提出$Ψ$-0.6以解决多模态推理的灵活性问题 reinforcement learning world model world models
33 FOCUS: A Lightweight Mamba-2 Model for EEG-Based Cognitive Workload Classification 提出FOCUS模型以解决EEG认知负荷分类问题 Mamba SSM
34 NgramRL: Accelerating RL Post-Training for Structured Generation via Rollout-Aware N-gram Speculative Decoding 提出NgramRL以加速结构化生成中的强化学习后训练 reinforcement learning large language model
35 Spatial Heuristic Strategy based Diffusion Reinforcement Learning for Online 3D Bin Packing 提出基于空间启发式策略的扩散强化学习以解决在线3D装箱问题 reinforcement learning
36 StarRTT: An End-to-End Real-Time Framework for Blind Zone-Free RTT Prediction in Starlink LEO Satellite Networks 提出StarRTT框架以解决Starlink LEO卫星网络中的RTT预测问题 MAE spatiotemporal

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
37 PPO vs. SAC: A Comparative Study for Cooperative Motion Control of Dual-Arm Space Robots 提出协调强化学习框架以解决双臂空间机器人运动控制问题 bi-manual dual-arm motion planning
38 Retiming Robot Demonstrations with a Learned Speed Prior 提出学习速度先验以解决机器人演示时序不一致问题 teleoperation vision-language-action

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

#题目一句话要点标签🔗⭐
39 Learning Nonlinear Waveform Fusion Across Scales for Integrated Sensing and Communication 提出非线性波形融合方法以解决集成感知与通信问题 PULSE

⬅️ 返回 cs.AI 首页 · 🏠 返回主页