cs.AI（2026-06-16）

📊 共 40 篇论文 | 🔗 1 篇有代码

🎯 兴趣领域导航

支柱九：具身大模型 (Embodied Foundation Models) (28) 支柱二：RL算法与架构 (RL & Architecture) (7 🔗1) 支柱一：机器人控制 (Robot Control) (3) 支柱八：物理动画 (Physics-based Animation) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱九：具身大模型 (Embodied Foundation Models) (28 篇)

#题目一句话要点标签🔗⭐
1 MathVis-Fine: Aligning Visual Supervision with Necessity via Progressive Dependency-Guided Training for Multimodal Mathematical Reasoning 提出MathVis-Fine以解决多模态数学推理中的视觉依赖问题 multimodal visual grounding chain-of-thought
2 FllumaOne: A Code-Native Multimodal CAD Dataset with Executable Programs and Kernel-Validated Feature Histories 提出FllumaOne数据集以解决可编辑CAD研究中的数据不足问题 multimodal
3 Querying an astronomical database using large language models: the ALeRCE text-to-SQL system 提出基于大语言模型的文本到SQL系统以查询天文数据库 large language model
4 Small Initialization Matters for Large Language Models 提出小初始化策略以提升大语言模型的推理能力 large language model
5 FinAcumen: Financial Multimodal Reasoning via Self-Evolving Experience Memory Harness 提出FinAcumen以解决金融多模态推理中的状态不稳定问题 multimodal
6 Surrogate Assisted Pedestrian Protection Design via a Foundation Model Orchestrated Workflow 提出基于基础模型的工作流程以优化行人保护设计 foundation model
7 Structural Role Injection in Handlebars-Templated LLM Prompts: Triple-Brace Interpolation, Delimiter Family, and the Limits of HTML Auto-Escaping 提出结构角色注入防护机制以解决模板引擎安全问题 large language model
8 Beyond Domains: Reusing Web Skills via Transferable Interaction Patterns 提出SkillMigrator以提升网页技能重用效率 large language model
9 EvolveNav: Proactive Preflection and Self-Evolving Memory for Zero-Shot Object Goal Navigation 提出自演化框架以解决零-shot目标导航问题 foundation model
10 A Red-Team Study of Anthropic Fable 5 & Opus 4.8 Models 评估Anthropic Fable 5与Opus 4.8模型的对抗鲁棒性 large language model
11 IUU+DB: Tracking Illegal, Unreported, and Unregulated Fishing, Seafood Fraud, and Labor Abuse through LLM-driven Information Extraction 提出IUU+DB以解决非法渔业活动监测问题 large language model
12 The Measurement Gap in the Automation of EU Law: Benchmarking Doctrinal Legal Reasoning under the EU AI Act 提出法律推理基准以填补EU法律自动化中的测量空白 large language model
13 Learning Cardiac Electrophysiology Digital Twins Through Agentic Discovery of Hybrid Structure 提出LEADS框架以解决个性化心脏电生理数字双胞胎建模问题 large language model
14 A Unified Framework for Context-Aware and Relation-Aware Graph Retrieval-Augmented Generation 提出HyGRAG框架以解决图检索增强生成中的知识融合问题 large language model
15 PseudoBench: Measuring How Agentic Auto-Research Fuels Pseudoscience 提出PseudoBench以评估自动科研系统抵抗伪科学的能力 large language model
16 LLM Consumer Behavior Theory: Foundations of a Novel Research Field 提出LLM消费者行为理论以应对代理市场中的决策挑战 large language model
17 A Neuro-Symbolic Approach to Strategy Synthesis for Strategic Logics 提出神经符号方法以解决多智能体系统中的策略合成问题 large language model
18 DecoSearch: Complexity-Aware Routing and Plan-Level Repair for Text-to-SQL 提出DecoSearch以解决复杂文本到SQL查询的问题 large language model
19 A Framework for Evaluating Agentic Skills at Scale 提出评估智能技能的框架以解决跨领域应用不足问题 instruction following
20 MIVE: A Minimalist Integer Vector Engine for Softmax LayerNorm and RMSNorm Acceleration 提出MIVE以解决大语言模型推理瓶颈问题 large language model
21 FacProcessTwin: An LLM-Based System for Process Twin Development 提出FacProcessTwin以降低过程双胞胎开发成本 large language model
22 Brick-DICL: Dynamic In-Context Learning for Automated Brick Schema Classification 提出Brick-DICL以解决建筑管理系统中砖块模式分类问题 large language model
23 Understanding LLMs in Title-Abstract Screening: From Disagreements to Recommendations 提出定性分析以提升LLM在标题-摘要筛选中的可靠性 large language model
24 Unlocking LLM Code Correction with Iterative Feedback Loops 通过迭代反馈机制提升LLM代码修正能力 large language model
25 LLM-as-Judge in Education: A Curriculum-Grounded Marking Pipeline 提出课程基础的LLM评估管道以支持教育评估 large language model
26 Can LLMs Be CEOs? Benchmarking Strategic Resource Reallocation with Multi-Role Agent Simulation 提出CEO-Bench以评估LLMs在战略资源重新分配中的决策能力 large language model
27 MapSatisfyBench: Benchmarking Satisfaction-Aware Map Agents through Behavior-Grounded Implicit Decision Factors 提出MapSatisfyBench以解决地图代理满意度评估问题 large language model
28 SoK: AI-Augmented Binary Reversing 提出全面知识体系以推动AI增强的二进制逆向分析 large language model

🔬 支柱二：RL算法与架构 (RL & Architecture) (7 篇)

#题目一句话要点标签🔗⭐
29 Shattering the Autoregressive Curse: Dynamic Epistemic Entropy Orchestrated Erasable Reinforcement Learning for LLMs 提出动态认知熵驱动的可擦除强化学习以解决自回归困境 reinforcement learning large language model
30 Temporal Preference Optimization for Unsupervised Retrieval 提出TPOUR以解决无监督检索中的时间相关性问题 preference learning contrastive learning TAMP ✅
31 Learning Red Agent Policy from Observations for Neurosymbolic Autonomous Cyber Agents 提出模仿学习技术以解决网络攻击下的自主防御问题 reinforcement learning policy learning imitation learning
32 Model Validation of Agentic AI Systems: A POMDP-Based Framework for Belief-State, Forecast, and Policy Validation 提出基于POMDP的框架以验证自主AI系统的模型风险 predictive model large language model
33 Trust the Right Teacher: Quality-Aware Self-Distillation for GUI Grounding 提出质量感知自蒸馏方法以提升GUI定位精度 distillation
34 MoCo-AIS: A Contrastive Learning Framework for Similarity Computation of Vessel Trajectories 提出MoCo-AIS框架以解决船舶轨迹相似性计算问题 contrastive learning
35 Using Cognitive Models to Improve Language Model Simulation of Human Persuasion Games 提出方程到行为提示以改善语言模型的人类说服游戏模拟 reinforcement learning large language model

🔬 支柱一：机器人控制 (Robot Control) (3 篇)

#题目一句话要点标签🔗⭐
36 Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems 研究品牌偏见与认知操控在LLM推荐系统中的动态影响 manipulation large language model
37 DeepInsight: A Unified Evaluation Infrastructure Across the Physical AI Stack 提出DeepInsight以解决物理AI栈评估的统一性问题 humanoid whole-body control foundation model
38 Knowledge Reutilization in Meta-Reinforcement Learning 提出元知识再利用框架以解决快速适应问题 locomotion reinforcement learning

🔬 支柱八：物理动画 (Physics-based Animation) (1 篇)

#题目一句话要点标签🔗⭐
39 STAR: SpatioTemporal Adaptive Reward Allocation for Text-to-Image RL Post-Training 提出STAR以解决文本到图像生成中的奖励分配问题 spatiotemporal

🔬 支柱五：交互与反应 (Interaction & Reaction) (1 篇)

#题目一句话要点标签🔗⭐
40 Structural Preservation and the Logical Expressiveness of Graph Neural Networks 提出图神经网络的逻辑表达能力与结构保持的新方法 OMOMO

⬅️ 返回 cs.AI 首页 · 🏠 返回主页