cs.AI（2025-12-24）

📊 共 13 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (10) 支柱二：RL算法与架构 (RL & Architecture) (2) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (10 篇)

#题目一句话要点标签🔗⭐
1 Beyond Context: Large Language Models Failure to Grasp Users Intent 大型语言模型未能理解用户意图，易被恶意利用绕过安全机制 large language model
2 Logic Sketch Prompting (LSP): A Deterministic and Interpretable Prompting Method 提出逻辑草图提示（LSP）框架，提升LLM在规则遵循任务中的确定性和可解释性。 large language model chain-of-thought
3 Agentic Explainable Artificial Intelligence (Agentic XAI) Approach To Explore Better Explanation 提出Agentic XAI框架，通过迭代优化解释提升农业推荐系统性能。 large language model multimodal
4 RoboSafe: Safeguarding Embodied Agents via Executable Safety Logic RoboSafe：通过可执行安全逻辑保障具身智能体的安全性 multimodal
5 Shape of Thought: When Distribution Matters More than Correctness in Reasoning Tasks 利用错误CoT轨迹训练提升语言模型推理能力，关注数据分布而非绝对正确性 chain-of-thought
6 Decomposing LLM Self-Correction: The Accuracy-Correction Paradox and Error Depth Hypothesis 解构LLM自纠错能力：揭示准确率-纠错悖论与误差深度假设 large language model
7 Scaling Laws for Economic Productivity: Experimental Evidence in LLM-Assisted Consulting, Data Analyst, and Management Tasks 量化LLM算力与经济生产力关系，揭示AI模型进步对专业任务效率的提升 large language model
8 Casting a SPELL: Sentence Pairing Exploration for LLM Limitation-breaking SPELL：通过句子配对探索LLM恶意代码生成限制突破 large language model
9 Mesh-Attention: A New Communication-Efficient Distributed Attention with Improved Data Locality 提出Mesh-Attention，通过优化数据局部性提升分布式Attention的通信效率，加速LLM训练。 large language model
10 Tree of Preferences for Diversified Recommendation 提出Tree of Preferences (ToP)框架，利用LLM挖掘用户未探索偏好，提升推荐多样性。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)

#题目一句话要点标签🔗⭐
11 Agentic Software Issue Resolution with Large Language Models: A Survey 综述：基于大语言模型的Agentic软件问题解决 reinforcement learning large language model
12 Policy-Conditioned Policies for Multi-Agent Task Solving 提出基于策略条件策略的程序化迭代最佳响应算法，解决多智能体任务中的策略动态适应问题。 reinforcement learning deep reinforcement learning large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
13 One Tool Is Enough: Reinforcement Learning for Repository-Level LLM Agents RepoNavigator：利用强化学习和单一代码跳转工具解决大型代码库问题定位 manipulation reinforcement learning distillation

⬅️ 返回 cs.AI 首页 · 🏠 返回主页