cs.AI（2025-09-26）

📊 共 30 篇论文 | 🔗 3 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (18 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗2) 支柱一：机器人控制 (Robot Control) (2) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱四：生成式动作 (Generative Motion) (1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (18 篇)

#题目一句话要点标签🔗⭐
1 Can Large Language Models Develop Gambling Addiction? 研究发现大语言模型可能表现出类似人类赌博成瘾的行为模式 large language model
2 UniMIC: Token-Based Multimodal Interactive Coding for Human-AI Collaboration UniMIC：面向人机协作的Token化多模态交互编码框架 multimodal
3 The Emergence of Altruism in Large-Language-Model Agents Society 提出基于Schelling模型的LLM智能体社会模拟框架，揭示利他主义涌现机制与模型异质性。 large language model
4 Large Language Models as Nondeterministic Causal Models 提出基于非确定性因果模型的大语言模型反事实生成方法 large language model
5 Patient-specific Biomolecular Instruction Tuning 提出KRONOS图-LLM框架，结合CPTAC-PROTSTRUCT数据集，提升肿瘤精准医疗中患者个体化蛋白质组学理解。 large language model multimodal
6 Guiding Evolution of Artificial Life Using Vision-Language Models ASAL++：利用视觉-语言模型引导人工生命演化，实现开放式探索 foundation model multimodal
7 You Can't Steal Nothing: Mitigating Prompt Leakages in LLMs via System Vectors 提出SysVec，通过系统向量编码缓解大语言模型中的提示词泄露问题 large language model instruction following
8 Efficient Fine-Grained GPU Performance Modeling for Distributed Deep Learning of LLM 提出一种高效细粒度的GPU性能建模方法，用于预测LLM分布式训练性能。 large language model
9 Hilbert: Recursively Building Formal Proofs with Informal Reasoning Hilbert：结合非形式推理与形式验证，递归构建数学证明 large language model
10 Dynamic Experts Search: Enhancing Reasoning in Mixture-of-Experts LLMs at Test Time 提出动态专家搜索(DES)，提升MoE LLM在推理时的性能和稳定性。 large language model
11 TrueGradeAI: Retrieval-Augmented and Bias-Resistant AI for Transparent and Explainable Digital Assessments TrueGradeAI：一种检索增强且抗偏置的透明可解释AI数字评估框架 large language model
12 Estimating the Empowerment of Language Model Agents 提出EELMA算法，通过信息论中的Empowerment评估语言模型Agent的能力。 chain-of-thought
13 AutoPK: Leveraging LLMs and a Hybrid Similarity Metric for Advanced Retrieval of Pharmacokinetic Data from Complex Tables and Documents AutoPK：利用LLM和混合相似度量从复杂表格和文档中高效检索药代动力学数据 large language model ✅
14 Bridging Language Models and Formal Methods for Intent-Driven Optical Network Design 提出结合LLM与形式化方法的意图驱动光网络设计框架 large language model
15 Toward a Theory of Generalizability in LLM Mechanistic Interpretability Research 提出LLM可解释性研究中的泛化性理论框架，并验证1-back注意力头的泛化能力 large language model
16 InfiAgent: Self-Evolving Pyramid Agent Framework for Infinite Scenarios InfiAgent：面向无限场景的自进化金字塔型智能体框架 large language model
17 SecureAgentBench: Benchmarking Secure Code Generation under Realistic Vulnerability Scenarios SecureAgentBench：在真实漏洞场景下评估代码Agent的安全代码生成能力 large language model
18 The Thinking Spectrum: An Empirical Study of Tunable Reasoning in LLMs through Model Merging 通过模型融合实现LLM可调推理能力：大规模实证研究 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
19 WaveMind: Towards a Conversational EEG Foundation Model Aligned to Textual and Visual Modalities WaveMind：面向文本和视觉模态对齐的会话式脑电图基础模型 representation learning large language model foundation model
20 InfiMed-Foundation: Pioneering Advanced Multimodal Medical Models with Compute-Efficient Pre-Training and Multi-Stage Fine-Tuning InfiMed-Foundation：通过高效预训练和多阶段微调，构建先进的多模态医学模型 distillation large language model multimodal ✅
21 Benefits and Pitfalls of Reinforcement Learning for Language Model Planning: A Theoretical Perspective 理论分析强化学习在语言模型规划中的优劣，揭示探索与多样性的重要性 reinforcement learning policy learning reward design
22 Towards Efficient Online Exploration for Reinforcement Learning with Human Feedback 提出在线RLHF高效探索算法，解决奖励模型不确定性问题 reinforcement learning RLHF large language model
23 StepORLM: A Self-Evolving Framework With Generative Process Supervision For Operations Research Language Models StepORLM：一个自进化框架，通过生成过程监督提升运筹学语言模型性能。 reinforcement learning DPO direct preference optimization
24 Towards Strategic Persuasion with Language Models 提出基于贝叶斯劝说的LLM战略劝说框架，并用强化学习提升劝说能力 reinforcement learning large language model
25 Structured Sparse Transition Matrices to Enable State Tracking in State-Space Models 提出PD-SSM：一种结构化稀疏状态空间模型，提升有限状态自动机模拟能力。 SSM state space model ✅

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
26 GeoSketch: A Neural-Symbolic Approach to Geometric Multimodal Reasoning with Auxiliary Line Construction and Affine Transformation GeoSketch：提出一种神经-符号几何多模态推理框架，支持辅助线构造和仿射变换。 manipulation reinforcement learning large language model
27 EMMA: Generalizing Real-World Robot Manipulation via Generative Visual Transfer EMMA：通过生成式视觉迁移实现真实世界机器人操作的泛化 manipulation vision-language-action VLA

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
28 REMA: A Unified Reasoning Manifold Framework for Interpreting Large Language Model REMA：统一的推理流形框架，用于解释大型语言模型 spatial relationship large language model multimodal

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
29 Red Teaming Quantum-Resistant Cryptographic Standards: A Penetration Testing Framework Integrating AI and Quantum Security 提出AI驱动的量子密码协议红队评估框架，提升量子网络安全性 penetration

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

#题目一句话要点标签🔗⭐
30 Generative Modeling and Decision Fusion for Unknown Event Detection and Classification Using Synchrophasor Data 提出基于生成模型和决策融合的电力系统未知事件检测与分类框架 spatiotemporal

⬅️ 返回 cs.AI 首页 · 🏠 返回主页