cs.AI（2026-06-26）

📊 共 16 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (8) 支柱九：具身大模型 (Embodied Foundation Models) (6) 支柱六：视频提取与匹配 (Video Extraction) (1 🔗1) 支柱一：机器人控制 (Robot Control) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
1 Tandem Reinforcement Learning with Verifiable Rewards 提出串联强化学习以解决可验证奖励问题 reinforcement learning large language model chain-of-thought
2 MER-R1: Multimodal Emotion Reasoning via Slow-Fast Thinking Synergy 提出MER-R1框架以优化多模态情感识别 reinforcement learning multimodal
3 Optimizing Teacher-Student Partitioning for Scalable Knowledge Distillation on HPC Systems 提出HPC感知的知识蒸馏优化方法以提升训练效率 teacher-student distillation
4 Grounded Iterative Language Planning: How Parameterized World Models Reduce Hallucination Propagation in LLM Agents 提出GILP以减少LLM代理中的幻觉传播问题 world model world models
5 Understanding Rollout Error in Graph World Models 提出错误感知图世界模型以解决长时间预测误差问题 world model world models
6 ATOD: Annealed Turn-aware On-policy Distillation for Multi-turn Autonomous Agents 提出ATOD以解决多轮自主智能体训练效率问题 reinforcement learning distillation
7 Verifiable Geometry Problem Solving: Solver-Driven Autoformalization and Theorem Proposing 提出SD-GPS框架以解决几何问题求解中的瓶颈 reinforcement learning multimodal
8 Bifocal Diffusion Language Models: Asymmetric Bidirectional Context for Parallel Generation 提出Bifocal dLLMs以解决生成质量与推理效率的矛盾问题 Mamba SSM

🔬 支柱九：具身大模型 (Embodied Foundation Models) (6 篇)

#题目一句话要点标签🔗⭐
9 Robust Harmful Features Under Jailbreak Attacks: Mechanistic Evidence from Attention Head Specialization in Large Language Models 提出机制性证据以揭示大语言模型中的监狱突破攻击特征 large language model
10 Single and Multi Truth Data Fusion using Large Language Models 利用大型语言模型解决单一与多重真值数据融合问题 large language model
11 End-to-End Dynamic Sparsity for Resource-Adaptive LLM Inference 提出L2A框架以解决动态资源适应的LLM推理问题 large language model
12 ToolPrivacyBench: Benchmarking Purpose-Bound Privacy in Tool-Using LLM Agents 提出ToolPrivacyBench以解决工具使用LLM代理的隐私评估问题 large language model
13 Agentic AI-Powered Re-Identification: An Emerging, Scalable Threat to Mobility Microdata Privacy 提出基于代理AI的再识别方法以应对移动微数据隐私威胁 large language model
14 Towards Reliable and Robust LLM Planning: Symbolic Feedback-Driven Iterative Self-Refinement Framework 提出符号反馈驱动的自我精炼框架以增强LLM规划的可靠性 large language model

🔬 支柱六：视频提取与匹配 (Video Extraction) (1 篇)

#题目一句话要点标签🔗⭐
15 NormAct: A Benchmark for Hidden Social Norm Compliance in Embodied Planning 提出NormAct基准以解决隐性社会规范遵循问题 egocentric large language model multimodal ✅

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
16 Agent-Native Immune System: Architecture, Taxonomy, and Engineering 提出Agent-Native免疫系统以解决自主智能体的安全问题 manipulation

⬅️ 返回 cs.AI 首页 · 🏠 返回主页