cs.AI（2026-02-04）

📊 共 13 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) (7 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (6 🔗1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)

#题目一句话要点标签🔗⭐
1 BrainVista: Modeling Naturalistic Brain Dynamics as Multimodal Next-Token Prediction 提出BrainVista以解决自然主义脑动态建模问题 multimodal
2 ReThinker: Scientific Reasoning by Rethinking with Guided Reflection and Confidence Control ReThinker：通过引导式反思和置信度控制实现科学推理 large language model foundation model
3 OMG-Agent: Toward Robust Missing Modality Generation with Decoupled Coarse-to-Fine Agentic Workflows OMG-Agent：通过解耦的粗到细智能体工作流实现鲁棒的缺失模态生成 multimodal
4 LLM-Empowered Cooperative Content Caching in Vehicular Fog Caching-Assisted Platoon Networks 提出基于LLM的车载雾计算辅助车队网络协同内容缓存架构 large language model
5 From Assumptions to Actions: Turning LLM Reasoning into Uncertainty-Aware Planning for Embodied Agents 提出PCE框架，将LLM推理转化为不确定性感知规划，提升具身智能体多智能体协作效率。 large language model
6 ProxyWar: Dynamic Assessment of LLM Code Generation in Game Arenas 提出ProxyWar框架以动态评估LLM代码生成质量 large language model ✅
7 Empirical-MCTS: Continuous Agent Evolution via Dual-Experience Monte Carlo Tree Search 提出 Empirical-MCTS，通过双重经验蒙特卡洛树搜索实现连续Agent进化 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
8 Learning the Value Systems of Agents with Preference-based and Inverse Reinforcement Learning 提出基于偏好和逆强化学习的智能体价值系统学习方法，解决人机协作中的价值对齐问题。 reinforcement learning inverse reinforcement learning
9 WideSeek-R1: Exploring Width Scaling for Broad Information Seeking via Multi-Agent Reinforcement Learning 提出WideSeek-R1，通过多智能体强化学习扩展LLM宽度，解决广域信息检索问题。 reinforcement learning large language model
10 Dual Mind World Model Inspired Network Digital Twin for Access Scheduling 提出基于双心智世界模型的数字孪生网络接入调度框架 reinforcement learning world model
11 Agent-Omit: Training Efficient LLM Agents for Adaptive Thought and Observation Omission via Agentic Reinforcement Learning Agent-Omit：通过Agentic强化学习训练高效LLM Agent，自适应省略思考和观察 reinforcement learning ✅
12 Steering LLMs via Scalable Interactive Oversight 提出可扩展交互监督框架，解决大语言模型复杂任务中人工指导难题 reinforcement learning large language model
13 InterPReT: Interactive Policy Restructuring and Training Enable Effective Imitation Learning from Laypersons InterPReT：交互式策略重构与训练，助力非专业人士进行有效的模仿学习 imitation learning

⬅️ 返回 cs.AI 首页 · 🏠 返回主页