cs.AI（2026-02-21）

📊 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) (8 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (3) 支柱一：机器人控制 (Robot Control) (1 🔗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) (8 篇)

#题目一句话要点标签🔗⭐
1 Spilled Energy in Large Language Models 提出基于能量模型的LLM推理方法，无需训练即可检测幻觉。 large language model
2 Beyond Description: A Multimodal Agent Framework for Insightful Chart Summarization 提出Chart Insight Agent Flow框架，提升多模态大语言模型在图表总结中洞察力提取能力 large language model multimodal
3 Adaptive Collaboration of Arena-Based Argumentative LLMs for Explainable and Contestable Legal Reasoning 提出ACAL框架，结合论辩LLM与人机交互，提升法律推理的可解释性和可辩论性 large language model chain-of-thought ✅
4 Early Evidence of Vibe-Proving with Consumer LLMs: A Case Study on Spectral Region Characterization with ChatGPT-5.2 (Thinking) 利用消费级LLM验证数学猜想：ChatGPT-5.2在谱区域刻画中的案例研究 large language model
5 Operational Robustness of LLMs on Code Generation 提出场景域分析方法以评估LLMs在代码生成中的鲁棒性 large language model
6 Many AI Analysts, One Dataset: Navigating the Agentic Data Science Multiverse 利用AI分析师群体解决数据分析结果依赖分析决策的问题 large language model
7 Give Users the Wheel: Towards Promptable Recommendation Paradigm 提出解耦可提示序列推荐（DPR）框架，利用自然语言提示动态引导推荐过程。 large language model
8 Orchestrating LLM Agents for Scientific Research: A Pilot Study of Multiple Choice Question (MCQ) Generation and Evaluation 探索LLM智能体在科学研究中的应用：以多项选择题生成与评估为例 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (3 篇)

#题目一句话要点标签🔗⭐
9 Carbon-aware decentralized dynamic task offloading in MIMO-MEC networks via multi-agent reinforcement learning 提出CADDTO-PPO框架，解决MIMO-MEC网络中碳感知去中心化动态任务卸载问题 reinforcement learning PPO spatiotemporal
10 Federated Reasoning Distillation Framework with Model Learnability-Aware Data Allocation 提出LaDa框架，通过模型可学习性感知的数据分配实现联邦推理蒸馏。 distillation large language model
11 GenPlanner: From Noise to Plans -- Emergent Reasoning in Flow Matching and Diffusion Models 提出GenPlanner以解决复杂环境中的路径规划问题 flow matching

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
12 TPRU: Advancing Temporal and Procedural Understanding in Large Multimodal Models TPRU：提升多模态大模型在时序和程序理解能力 manipulation reinforcement learning embodied AI ✅

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

#题目一句话要点标签🔗⭐
13 How Far Can We Go with Pixels Alone? A Pilot Study on Screen-Only Navigation in Commercial 3D ARPGs 提出基于视觉线索的ARPG游戏自动导航方法，探索纯视觉导航的局限性 affordance

⬅️ 返回 cs.AI 首页 · 🏠 返回主页