cs.LG（2026-07-08）

📊 共 23 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) (10 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (9) 支柱一：机器人控制 (Robot Control) (1) 支柱八：物理动画 (Physics-based Animation) (1) 支柱四：生成式动作 (Generative Motion) (1) 支柱五：交互与反应 (Interaction & Reaction) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (10 篇)

#题目一句话要点标签🔗⭐
1 Entropy Pacing Policy Optimization for Multi-Task Agentic Reinforcement Learning 提出熵节奏策略优化以解决多任务强化学习中的探索-利用不匹配问题 reinforcement learning generalist agent large language model
2 Guidance Breaks the Fitted Operator: A Terminal-Fitted Repair for Classifier-Free Guidance 提出终端拟合修复方法以解决分类器无关引导的过饱和问题 flow matching classifier-free guidance
3 Selective Timestep Weighting and Advantage-Based Replay for Sample-Efficient Diffusion RLHF 提出选择性时间步加权和基于优势的重放以提高扩散RLHF的样本效率 reinforcement learning PPO RLHF
4 Single-Rollout Asynchronous Optimization for Agentic Reinforcement Learning 提出单次回合异步优化以解决强化学习稳定性问题 reinforcement learning large language model
5 UP: Unbounded Positive Asymmetric Optimization for Breaking the Exploration-Stability Dilemma 提出无界正向非对称优化以解决探索与稳定性困境 reinforcement learning large language model multimodal
6 TimEE: End-to-end Time Series Classification via In-Context Learning 提出TimEE以解决时间序列分类中的训练效率问题 representation learning foundation model ✅
7 FMMVCC: Fuzzy Mamba-based Multi-View Contrastive Clustering for Univariate Time Series 提出FMMVCC以解决时间序列聚类中的长程依赖问题 Mamba
8 Gimitest: A Comprehensive Tool for Testing Reinforcement Learning Policies 提出Gimitest以解决强化学习策略测试的可靠性问题 reinforcement learning
9 Mathematical methods of reinforcement learning 系统化数学方法以推动强化学习算法设计与分析 reinforcement learning
10 Weight-Space Physics: Interpretable Hypernetworks for Lattice Quantum Field Theories 提出JEPAWG以解决量子场论中的可解释性问题 Joint-Embedding Predictive Architecture joint-embedding predictive architecture

🔬 支柱九：具身大模型 (Embodied Foundation Models) (9 篇)

#题目一句话要点标签🔗⭐
11 Latent graph encoding of multimodal neuroimaging features with generative AI architectures 提出多模态图变分自编码器以提升神经影像分析精度 multimodal
12 The Optimal Sample Complexity of Learning Autoregressive Chain-of-Thought 提出最优样本复杂度以学习自回归思维链 chain-of-thought
13 Rethinking Multimodal Time-Series Forecasting Evaluation 提出TimesX基准以解决多模态时间序列预测评估问题 multimodal
14 Physical activities enable scalable foundation modelling for broad-spectrum health prediction 提出StepFM以解决健康预测模型的可扩展性问题 foundation model
15 A Unified Detection Framework for AI-Related Content and Artifacts 提出统一检测框架以识别AI相关内容与伪造物 large language model
16 Gradient-free Riemannian Langevin Sampler 提出无梯度黎曼朗之万采样器以解决多模态分布采样问题 multimodal
17 GIFT: Geometry-Informed Low-precision Gradient Communication for LLM Pretraining 提出GIFT以解决大语言模型预训练中的梯度通信瓶颈问题 large language model
18 Dissociating the Internal Representations of Sycophancy in LLMs 提出区分LLMs中谄媚行为内部表征的方法 large language model
19 Best-Arm Identification with Generative Proxy 提出PROBE算法以解决最佳臂识别中的样本效率问题 large language model

🔬 支柱一：机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗⭐
20 Safe Reinforcement Learning using Ideas from Model Predictive Control 提出一种结合深度强化学习与模型预测控制的安全强化学习框架 MPC model predictive control reinforcement learning

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

#题目一句话要点标签🔗⭐
21 Multimodal Spatiotemporal-Frequency Fusion with Peak Enhancement for Cellular Traffic Forecasting 提出MSPF-Net以解决蜂窝网络流量预测中的多模态融合问题 spatiotemporal multimodal

🔬 支柱四：生成式动作 (Generative Motion) (1 篇)

#题目一句话要点标签🔗⭐
22 Physics-guided spatiotemporal neural models for fuel density prediction 提出物理引导的时空神经模型以预测燃料密度 physically plausible spatiotemporal

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

#题目一句话要点标签🔗⭐
23 Any-Dimensional Learning by Sampling 提出随机采样映射以解决多维输入学习问题 OMOMO

⬅️ 返回 cs.LG 首页 · 🏠 返回主页