cs.AI（2025-12-20）

📊 共 15 篇论文

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (8) 支柱二：RL算法与架构 (RL & Architecture) (4) 支柱一：机器人控制 (Robot Control) (2) 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 External Hippocampus: Topological Cognitive Maps for Guiding Large Language Model Reasoning 提出外部海马体框架，通过拓扑认知地图引导大语言模型推理 large language model
2 Asynchronous Pipeline Parallelism for Real-Time Multilingual Lip Synchronization in Video Communication Systems 提出异步流水线并行Transformer框架，用于实时多语种唇动同步 multimodal
3 Emergent Persuasion: Will LLMs Persuade Without Being Prompted? 研究表明：无需显式提示，大型语言模型可能涌现说服能力，需警惕潜在风险。 large language model
4 Bidirectional RAG: Safe Self-Improving Retrieval-Augmented Generation Through Multi-Stage Validation 提出Bidirectional RAG，通过多阶段验证实现安全自提升的检索增强生成。 large language model
5 VeruSAGE: A Study of Agent-Based Verification for Rust Systems VeruSAGE：基于Agent的Rust系统验证研究，提升LLM代码正确性证明能力 large language model
6 Intelligent Human-Machine Partnership for Manufacturing: Enhancing Warehouse Planning through Simulation-Driven Knowledge Graphs and LLM Collaboration 提出基于仿真驱动知识图谱和LLM协作的智能人机伙伴关系，优化仓库规划。 large language model
7 MSC-180: A Benchmark for Automated Formal Theorem Proving from Mathematical Subject Classification 提出MSC-180基准，用于评估LLM在数学分支上的自动定理证明能力。 large language model
8 Propose, Solve, Verify: Self-Play Through Formal Verification 提出PSV框架，通过形式化验证的自博弈提升代码生成能力 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (4 篇)

#题目一句话要点标签🔗⭐
9 Large Language Models as Discounted Bayesian Filters 提出基于折扣贝叶斯滤波的大语言模型在线推理评估框架 world model large language model
10 CosmoCore-Evo: Evolutionary Dream-Replay Reinforcement Learning for Adaptive Code Generation CosmoCore-Evo：进化梦境回放强化学习，用于自适应代码生成 reinforcement learning PPO
11 Monitoring Monitorability 提出监控可监控性以提升AI系统的安全性 reinforcement learning chain-of-thought
12 NL2CA: Auto-formalizing Cognitive Decision-Making from Natural Language Using an Unsupervised CriticNL2LTL Framework 提出NL2CA，实现从自然语言到认知决策模型的自动形式化转换。 reinforcement learning large language model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
13 Breaking Minds, Breaking Systems: Jailbreaking Large Language Models via Human-like Psychological Manipulation 提出心理学越狱方法HPM，利用类人心理操纵攻破大语言模型安全防线 manipulation penetration large language model
14 SoK: Understanding (New) Security Issues Across AI4Code Use Cases 综述AI4Code应用中的新型安全问题，并提出未来发展方向 manipulation

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

#题目一句话要点标签🔗⭐
15 Neural Brain Fields: A NeRF-Inspired Approach for Generating Nonexistent EEG Electrodes 提出Neural Brain Fields以解决EEG电极不足问题 NeRF neural radiance field

⬅️ 返回 cs.AI 首页 · 🏠 返回主页