cs.AI（2026-06-29）

📊 共 31 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) (14 🔗1) 支柱二：RL算法与架构 (RL & Architecture) (13 🔗2) 支柱一：机器人控制 (Robot Control) (3) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (14 篇)

#题目一句话要点标签🔗⭐
1 ManimAgent: Self-Evolving Multimodal Agents for Visual Education 提出ManimAgent以解决多轮反思任务间知识孤岛问题 large language model multimodal
2 Making Multimodal LLMs Reliable Chart Data Extractors: A Benchmark and Training Framework 提出多模态大语言模型以解决图表数据提取问题 large language model multimodal
3 Using Large Language Models as Low-Cost Statistical Estimators for Human-Response Data 利用大型语言模型作为低成本统计估计器解决人类响应数据问题 large language model
4 Does Verbose Chain-of-Thought Really Help? In-Distribution Evidence that Content, Not Length, Matters 提出基于内容而非长度的推理链优化方法 chain-of-thought
5 Temporal Feature Extractors in EEG Foundation Models: A Controlled Comparison Including a Pretrained Time-Series Model 比较三种时序特征提取方法在EEG基础模型中的效果 foundation model
6 AlgoSkill: Learning to Design Algorithms by Scheduling Human-Like Skills 提出AlgoSkill以优化算法设计过程中的技能调度 large language model chain-of-thought
7 A Machine-Verified Proof of a Quantum-Optimization Conjecture 提出机器验证的量子优化猜想证明方法 large language model
8 Words Speak Louder Than Code: Investigating Cognitive Heuristics in LLM-Based Code Vulnerability Detection 系统探讨认知启发式在LLM代码漏洞检测中的影响 large language model
9 Situation Perception: A Necessary Primitive to Artificial Superintelligence 提出情境感知以解决人工超智能的关键能力缺失问题 large language model
10 Collective cooperation without individual fidelity in LLM agents 研究LLM代理在囚徒困境中的集体合作行为 large language model
11 The FIL Hypothesis: Inductive Biases Help with Kernel Engineering 提出基于人类启发的知识约束方法以解决反馈信息循环问题 large language model ✅
12 Translating Natural Language to Strategic Temporal Specifications via LLMs 提出基于大语言模型的自然语言到战略时序规范的翻译方法 large language model
13 BayesEvolve: Explicit Belief States for Autonomous Scientific Discovery 提出BayesEvolve以解决自主科学发现中的不确定性问题 large language model
14 MCP Server Architecture Patterns for LLM-Integrated Applications 提出MCP服务器架构模式以优化LLM集成应用 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (13 篇)

#题目一句话要点标签🔗⭐
15 FacePlex: Full-Duplex Joint Speech-Facial Motion Generation for Conversational Avatars 提出FacePlex以解决实时语音与面部动作同步生成问题 flow matching motion generation
16 Data-Efficient Multimodal Alignment for Histopathology-based Molecular Prediction 提出数据高效的多模态对齐方法以实现组织病理学分子预测 contrastive learning open-vocabulary open vocabulary
17 Exploration and Online Transfer with Behavioral Foundation Models 提出在线转移方法以解决零-shot强化学习中的探索问题 reinforcement learning foundation model zero-shot transfer
18 Toward Secure and Reliable PDDL Formalization of Large Language Models with Planner-in-the-Loop Feedback 提出NL-PDDL-Bench以解决大型语言模型的安全性与可靠性问题 direct preference optimization large language model ✅
19 GUICrafter: Weakly-Supervised GUI Agent Leveraging Massive Unannotated Screenshots 提出GUICrafter以解决GUI代理数据收集困难问题 reinforcement learning curriculum learning foundation model ✅
20 Self-Evolving World Models for LLM Agent Planning 提出自演化世界模型以提升LLM代理规划能力 world model world models
21 DOPD: Dual On-policy Distillation 提出DOPD以解决蒸馏过程中的特权幻觉问题 distillation privileged information large language model
22 The CRISTAL Method: Neurosymbolic analysis from AI-synthesized world models 提出CRISTAL方法以解决复杂分析工作流自动化问题 world model world models
23 Be Faithful When Response: Returning Fluent and Grounded Answers for Vision-Language Models Reinforcement Learning 提出Faithful Warm-Start策略以解决视觉语言模型的推理不稳定问题 reinforcement learning multimodal
24 Hierarchical Reinforcement Learning in StarCraft Micromanagement with Influence Maps and Cluster-based Scripts 提出HRL-IM/CBS框架以解决StarCraft微观管理中的决策透明性问题 reinforcement learning
25 ACPO: Agent-Chained Policy Optimization for Multi-Agent Reinforcement Learning 提出ACPO以解决多智能体强化学习中的策略优化问题 reinforcement learning
26 Semi-Supervised Sound Event Detection with Conditional Mixup and Embedding-Level Contrastive Loss 提出条件混合与嵌入级对比损失以解决声事件检测数据稀缺问题 contrastive learning foundation model
27 HERO: Improving the Reliability and Sensitivity of Generative Model Evaluation Using Historical Data 提出HERO框架以提高生成模型评估的可靠性与敏感性 world model world models

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
28 Linguistic Firewall: Geometry as Defense in Multi-Agent Systems Routing 提出ANTAP以解决多智能体系统中的路由安全问题 manipulation representation learning large language model
29 Domain Adaptation with Adaptive Imagination for Visual Reinforcement Learning under Limited Target Data 提出AIDA框架以解决有限目标数据下的视觉强化学习问题 sim-to-real reinforcement learning
30 AI Training Manager: Bounded Closed-Loop Control of Adaptive Training Recipes 提出AI训练管理器以解决自适应训练中的控制问题 manipulation reinforcement learning

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

#题目一句话要点标签🔗⭐
31 Mandol: An Agglomerative Agent Memory System for Long-Term Conversations 提出Mandol以解决长时间对话中的记忆碎片化问题 semantic map

⬅️ 返回 cs.AI 首页 · 🏠 返回主页