cs.AI（2026-02-08）

📊 共 23 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) (16 🔗2) 支柱二：RL算法与架构 (RL & Architecture) (6) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (16 篇)

#题目一句话要点标签🔗⭐
1 Geo-Code: A Code Framework for Reverse Code Generation from Geometric Images Based on Two-Stage Multi-Agent Evolution Geo-Code：基于多智能体演化的几何图像逆向代码生成框架 multimodal
2 Can Multimodal LLMs See Science Instruction? Benchmarking Pedagogical Reasoning in K-12 Classroom Videos 提出SciIBI，用于评估多模态LLM在K-12科学课堂视频中教学推理能力。 multimodal
3 Can Large Language Models Implement Agent-Based Models? An ODD-based Replication Study 评估大型语言模型在Agent-Based模型实现中的可靠性，探索其在可复现建模中的潜力与局限。 large language model
4 Large language models for spreading dynamics in complex systems 利用大型语言模型分析复杂系统中传播动力学，应用于数字和生物流行病。 large language model
5 Anonymization-Enhanced Privacy Protection for Mobile GUI Agents: Available but Invisible 提出一种基于匿名化的移动GUI代理隐私保护框架，实现可用但不可见的敏感数据访问。 large language model multimodal ✅
6 MemFly: On-the-Fly Memory Optimization via Information Bottleneck MemFly：基于信息瓶颈的LLM即时记忆优化框架，提升长期记忆能力 large language model
7 AGORA: Agentic Green Orchestration Architecture for Beyond 5G Networks 提出AGORA架构，利用LLM智能调控5G网络能耗，实现绿色运维。 large language model
8 Towards Adaptive, Scalable, and Robust Coordination of LLM Agents: A Dynamic Ad-Hoc Networking Perspective 提出RAPS：基于动态Ad-Hoc网络的LLM Agent自适应协同框架 large language model
9 Accelerating Social Science Research via Agentic Hypothesization and Experimentation 提出EXPERIGEN框架，通过智能体假设生成与实验加速社会科学研究。 multimodal
10 LOCA-bench: Benchmarking Language Agents Under Controllable and Extreme Context Growth LOCA-bench：可控极端上下文增长下的语言Agent基准测试 large language model ✅
11 IV Co-Scientist: Multi-Agent LLM Framework for Causal Instrumental Variable Discovery 提出IV Co-Scientist多智能体框架，利用LLM进行因果工具变量发现。 large language model
12 Rethinking the Value of Agent-Generated Tests for LLM-Based Software Engineering Agents 揭示LLM代码Agent中Agent生成测试的价值：作用有限，需重新评估 large language model
13 Rethinking Latency Denial-of-Service: Attacking the LLM Serving Framework, Not the Model 提出针对LLM服务框架的Fill and Squeeze攻击，提升延迟拒绝服务攻击效果。 large language model
14 Emergent Misalignment is Easy, Narrow Misalignment is Hard 揭示大语言模型涌现式不对齐现象，并提出线性表示用于监控和缓解。 large language model
15 Data Darwinism Part I: Unlocking the Value of Scientific Data for Pre-training 提出Data Darwinism框架，利用高质量科学数据预训练提升大模型性能。 foundation model
16 Learning to Continually Learn via Meta-learning Agentic Memory Designs ALMA：通过元学习自动设计Agentic系统的持续学习记忆模块 foundation model

🔬 支柱二：RL算法与架构 (RL & Architecture) (6 篇)

#题目一句话要点标签🔗⭐
17 Graph-Enhanced Deep Reinforcement Learning for Multi-Objective Unrelated Parallel Machine Scheduling 提出基于图增强深度强化学习的多目标不相关并行机调度方法 reinforcement learning deep reinforcement learning PPO
18 Time Series Reasoning via Process-Verifiable Thinking Data Synthesis and Scheduling for Tailored LLM Reasoning VeriTime：通过可验证过程的思维数据合成与调度，为时序推理定制LLM reinforcement learning large language model multimodal
19 Generative Reasoning Re-ranker 提出生成式推理重排序器（GR2），利用强化学习提升LLM在推荐系统中的重排序性能。 reinforcement learning reward design large language model
20 Objective Decoupling in Social Reinforcement Learning: Recovering Ground Truth from Sycophantic Majorities 提出Epistemic Source Alignment解决社交强化学习中因谄媚导致的客观目标解耦问题 reinforcement learning
21 ToolSelf: Unifying Task Execution and Self-Reconfiguration via Tool-Driven Intrinsic Adaptation ToolSelf：通过工具驱动的内在适应统一任务执行和自我重配置 reinforcement learning large language model
22 SAGE: Scalable AI Governance & Evaluation SAGE：可扩展的AI治理与评估框架，提升大规模搜索系统相关性。 teacher-student distillation

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

#题目一句话要点标签🔗⭐
23 Transforming Science Learning Materials in the Era of Artificial Intelligence AI赋能科学教育：革新学习材料设计，实现个性化、真实化与可访问性 affordance multimodal

⬅️ 返回 cs.AI 首页 · 🏠 返回主页