cs.LG（2026-07-23）

📊 共 22 篇论文

🎯 兴趣领域导航

支柱二：RL算法与架构 (RL & Architecture) (9) 支柱九：具身大模型 (Embodied Foundation Models) (7) 支柱一：机器人控制 (Robot Control) (2) 支柱七：动作重定向 (Motion Retargeting) (1) 支柱五：交互与反应 (Interaction & Reaction) (1) 支柱八：物理动画 (Physics-based Animation) (1) 支柱三：空间感知与语义 (Perception & Semantics) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (9 篇)

#题目一句话要点标签🔗⭐
1 Toward Generalizable Cognitive Impairment Detection with Speech-Based Multimodal Large Language Models 提出多模态框架以解决认知障碍检测问题 representation learning large language model multimodal
2 M$^3$-Gen: Interpretable Multimodal Generation of Gene Expression Profiles Using Clinical and Imaging Data 提出M$^3$-Gen以解决基因表达数据生成问题 contrastive learning multimodal
3 Training Large Language Models for Self-Explanation Faithfulness 提出强化学习方法以优化自我解释的可信度 reinforcement learning large language model
4 X$^3$-OPD: Distilling Reasoning into Large Audio-Language Models via On-Policy Alignment 提出X$^3$-OPD以解决音频语言模型推理能力不足问题 distillation large language model chain-of-thought
5 Compact Latent Coordination for Autonomous Vehicles at Unsignalized Intersections 提出MAPS以解决无信号交叉口自主车辆协调问题 reinforcement learning deep reinforcement learning DRL
6 Context-weighted Discrete Flow Matching 提出上下文加权离散流匹配以提升生成模型性能 flow matching
7 Windowed-MTP: Removing the Full-Context Draft-KV Tax at Million-Token Context 提出Windowed-MTP以解决百万标记上下文中的全上下文草稿KV成本问题 Mamba linear attention
8 Approximate Quantum State Preparation Through Proximal Policy Optimization 提出基于近端策略优化的量子态近似制备框架 reinforcement learning deep reinforcement learning
9 Relative Value Learning 提出相对价值学习框架以优化强化学习控制 reinforcement learning PPO

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

#题目一句话要点标签🔗⭐
10 Test-Time Scaling via Error Localization 提出测试时错误定位的缩放方法以提升推理效率 large language model
11 AI Assistants Overassist 提出Int-Bench以优化AI助手的干预策略 large language model
12 Multi-Task Learning for Heterogeneous Prediction from Video Game State with Transfer Learning 提出多任务学习方法以提升视频游戏状态预测能力 multimodal
13 LeAct: Learning to Reason from Expert Actions 提出LeAct以从专家行动中学习推理 foundation model chain-of-thought
14 How Do AI Coding Agents Contribute to Software Development? an Empirical Study of Agentic Pull Requests 通过实证研究揭示AI编码代理在软件开发中的贡献 large language model
15 Prompt as a Data Type: In-Database LLM Prompt Management and Rewriting 提出PromptDB以解决数据库中提示管理与重写问题 large language model
16 Test-Time Scaling via Error Localization 提出测试时错误定位的缩放方法以提升推理性能 large language model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
17 TOUR: A Trajectory-Level Unlearning Benchmark for Offline Reinforcement Learning 提出TOUR基准以解决离线强化学习中的轨迹级别遗忘问题 locomotion reinforcement learning offline RL
18 Offline RL with Hierarchical Action Chunking 提出Hierarchical Implicit Q-Chunking以解决离线强化学习中的长时间任务问题 humanoid reinforcement learning offline RL

🔬 支柱七：动作重定向 (Motion Retargeting) (1 篇)

#题目一句话要点标签🔗⭐
19 CASC: Causal Adversarial Subspace Clustering for Multivariate Spatiotemporal Data 提出CASC框架以解决多变量时空数据聚类中的因果依赖问题 spatial relationship spatiotemporal

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

#题目一句话要点标签🔗⭐
20 Agree on the Model, Verify the Inference: GKR Protocols for HND-Based Transformer Inference 提出GKR-HND协议以解决Transformer推理的安全性问题 OMOMO

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

#题目一句话要点标签🔗⭐
21 Climate-resilient electric vehicle charging infrastructure for sustainable cities: An interpretable causal-ensemble framework for preventive maintenance and low-carbon mobility 提出FGDSE框架以提升电动车充电基础设施的气候韧性 PULSE

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

#题目一句话要点标签🔗⭐
22 Quantifying Political Partisanship for Cross-Platform Analyses 提出一种跨平台的政治偏见量化方法以解决社交媒体分析问题 affordance

⬅️ 返回 cs.LG 首页 · 🏠 返回主页