cs.AI（2026-02-07）

📊 共 28 篇论文 | 🔗 6 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (17 🔗4) 支柱二：RL算法与架构 (RL & Architecture) (10 🔗2) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (17 篇)

#题目一句话要点标签🔗⭐
1 Steer2Adapt: Dynamically Composing Steering Vectors Elicits Efficient Adaptation of LLMs Steer2Adapt：通过动态组合steering vectors实现LLM的高效适应 large language model
2 Efficient Table Retrieval and Understanding with Multimodal Large Language Models 提出TabRAG框架，解决多模态大语言模型在海量表格图像中检索与理解的难题。 large language model foundation model multimodal
3 VGAS: Value-Guided Action-Chunk Selection for Few-Shot Vision-Language-Action Adaptation VGAS：面向少样本视觉-语言-动作自适应的价值引导动作块选择 vision-language-action VLA multimodal ✅
4 How does longer temporal context enhance multimodal narrative video processing in the brain? 研究视频时序上下文长度如何影响大脑对多模态叙事视频的处理，并与模型对齐 large language model multimodal
5 M2A: Multimodal Memory Agent with Dual-Layer Hybrid Memory for Long-Term Personalized Interactions 提出M2A：双层混合记忆多模态Agent，用于长期个性化交互 multimodal ✅
6 Evaluating Large Language Models for Detecting Architectural Decision Violations 利用大型语言模型检测软件架构决策违规 large language model
7 MSP-LLM: A Unified Large Language Model Framework for Complete Material Synthesis Planning 提出MSP-LLM，用于完整材料合成规划的统一大语言模型框架 large language model
8 Are Reasoning LLMs Robust to Interventions on Their Chain-of-Thought? 研究推理LLM在思维链中受干扰时的鲁棒性，揭示其恢复机制与效率权衡。 chain-of-thought
9 SupChain-Bench: Benchmarking Large Language Models for Real-World Supply Chain Management SupChain-Bench：用于供应链管理的大语言模型基准测试 large language model
10 Linguistic properties and model scale in brain encoding: from small to compressed language models 研究表明：3B规模语言模型在脑编码预测中可媲美更大模型，且对压缩具有鲁棒性 large language model
11 SoK: DARPA's AI Cyber Challenge (AIxCC): Competition Design, Architectures, and Lessons Learned AIxCC挑战赛分析：利用AI自主发现并修复开源软件漏洞 large language model
12 MemPot: Defending Against Memory Extraction Attack with Optimized Honeypots MemPot：通过优化蜜罐防御大语言模型Agent的记忆提取攻击 large language model
13 Reverse-Engineering Model Editing on Language Models 揭示模型编辑漏洞：提出KSTER攻击以逆向工程语言模型编辑数据 large language model ✅
14 Can LLMs Truly Embody Human Personality? Analyzing AI and Human Behavior Alignment in Dispute Resolution 评估LLM在冲突解决中模拟人类个性的能力，揭示其与人类行为的差异 large language model
15 AgentSys: Secure and Dynamic LLM Agents Through Explicit Hierarchical Memory Management AgentSys：通过显式分层内存管理实现安全动态的LLM Agent foundation model ✅
16 AgentTrace: A Structured Logging Framework for Agent System Observability AgentTrace：用于提升Agent系统可观测性的结构化日志框架 large language model
17 Progressive Searching for Retrieval in RAG 提出渐进式搜索算法，提升RAG系统中检索效率与准确性，适用于大规模数据库。 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
18 SleepMaMi: A Universal Sleep Foundation Model for Integrating Macro- and Micro-structures 提出SleepMaMi睡眠基础模型，整合宏观睡眠结构与微观信号特征，提升睡眠分析通用性。 masked autoencoder MAE contrastive learning
19 Joint Reward Modeling: Internalizing Chain-of-Thought for Efficient Visual Reward Models 提出联合奖励建模(JRM)，提升视觉奖励模型在图像编辑等复杂任务中的效率和语义理解能力。 reinforcement learning preference learning chain-of-thought
20 Debugging code world models 研究代码世界模型的错误根源，提出改进监督和状态表示的建议。 world model chain-of-thought
21 High Fidelity Textual User Representation over Heterogeneous Sources via Reinforcement Learning 提出基于强化学习的文本用户表示方法，解决异构数据源融合与LLM兼容问题 reinforcement learning large language model
22 Secure Code Generation via Online Reinforcement Learning with Vulnerability Reward Model SecCoderX：基于在线强化学习和漏洞奖励模型的安全代码生成框架 reinforcement learning large language model ✅
23 RAPiD: Real-time Deterministic Trajectory Planning via Diffusion Behavior Priors for Safe and Efficient Autonomous Driving RAPiD：基于扩散行为先验的实时确定性轨迹规划，保障自动驾驶安全高效 policy learning imitation learning multimodal ✅
24 VERIFY-RL: Verifiable Recursive Decomposition for Reinforcement Learning in Mathematical Reasoning VERIFY-RL：基于可验证递归分解的强化学习方法，提升数学推理能力 reinforcement learning curriculum learning
25 Semantic Search At LinkedIn LinkedIn提出基于LLM的语义搜索框架，显著提升AI职位和人才搜索效率。 distillation large language model
26 EventCast: Hybrid Demand Forecasting in E-Commerce with LLM-Based Event Knowledge EventCast：利用LLM事件知识增强电商混合需求预测 MAE large language model
27 Adaptive Scaffolding for Cognitive Engagement in an Intelligent Tutoring System 提出自适应脚手架，通过动态选择教学示例提升智能辅导系统中学生的认知参与度。 reinforcement learning deep reinforcement learning DRL

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
28 Agent-Fence: Mapping Security Vulnerabilities Across Deep Research Agents 提出AgentFence以评估深度代理的安全漏洞 manipulation large language model

⬅️ 返回 cs.AI 首页 · 🏠 返回主页