cs.AI（2026-06-17）

📊 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 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (7) 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 ThinkDeception: A Progressive Reinforcement Learning Framework for Interpretable Multimodal Deception Detection 提出ThinkDeception以解决多模态欺骗检测的可解释性问题 reinforcement learning large language model multimodal
2 DeFAb: A Verifiable Benchmark for Defeasible Abduction in Foundation Models 提出DeFAb基准以解决可否定推理的评估问题 DPO foundation model chain-of-thought ✅
3 Rethinking Reward Supervision: Rubric-Conditioned Self-Distillation 提出基于评分标准的自蒸馏方法以改善推理语言模型训练 reinforcement learning distillation chain-of-thought
4 Skill-Guided Continuation Distillation for GUI Agents 提出技能引导的延续蒸馏方法以解决GUI代理的监督缺口问题 behavior cloning distillation
5 R2D-RL: A RoboCup 2D Soccer Environment for Multi-Agent Reinforcement Learning 提出R2D-RL以解决RoboCup 2D足球环境的多智能体强化学习问题 reinforcement learning reward shaping
6 Reference-Driven Multi-Speaker Audio Scene Generation from In-the-Wild Priors 提出ScenA以解决多说话者音频场景生成问题 flow matching foundation model
7 ProfiLLM: Utility-Aligned Agentic User Profiling for Industrial Ride-Hailing Dispatch 提出ProfiLLM以解决工业网约车调度中的用户画像问题 DPO large language model
8 MIDS: Detecting Stealthy Masquerade and Tampering Attacks on CAN Bus via Bidirectional Mamba 提出MIDS以解决CAN总线隐蔽伪装和篡改攻击问题 Mamba

🔬 支柱九：具身大模型 (Embodied Foundation Models) (7 篇)

#题目一句话要点标签🔗⭐
9 What Must Generalist Agents Remember? 提出记忆机制以提升通用智能体在多环境中的表现 generalist agent
10 Beyond Safe Data: Pretraining-Stage Alignment with Regular Safety Reflection 提出安全反思预训练以增强大语言模型的安全性 large language model
11 Skill-MAS: Evolving Meta-Skill for Automatic Multi-Agent Systems 提出Skill-MAS以解决多智能体系统中的经验保留与模型能力矛盾问题 large language model
12 X+Slides: Benchmarking Audience-Conditioned Slide Generation 提出X+Slides以解决观众条件下幻灯片生成问题 large language model
13 A Technical Taxonomy of LLM Agent Communication Protocols 提出技术分类法以解决LLM代理通信协议互操作性问题 large language model
14 TRAP: Benchmark for Task-completion and Resistance to Active Privacy-extraction 提出TRAP以解决任务完成与隐私泄露的矛盾问题 instruction following
15 CAPRA: Scaling Feedback on Software Architecture Deliverables with a Multi-Agent LLM System 提出CAPRA以解决软件架构交付物反馈自动化问题 large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
16 Generative-Model Predictive Planning for Navigation in Partially Observable Environments 提出BeliefDiffusion以解决部分可观测环境中的导航问题 MPC model predictive control reinforcement learning

⬅️ 返回 cs.AI 首页 · 🏠 返回主页