cs.LG(2026-07-16)

📊 共 15 篇论文 | 🔗 2 篇有代码

🎯 兴趣领域导航

支柱二:RL算法与架构 (RL & Architecture) (12 🔗2) 支柱九:具身大模型 (Embodied Foundation Models) (1) 支柱一:机器人控制 (Robot Control) (1) 支柱八:物理动画 (Physics-based Animation) (1)

🔬 支柱二:RL算法与架构 (RL & Architecture) (12 篇)

#题目一句话要点标签🔗
1 Multimodal Semantic-Aware Contrastive Learning For False Negative Mitigation in 3D Medical Imaging 提出多模态语义感知对比学习以解决医学影像中的假阴性问题 contrastive learning multimodal
2 A Continuous-Time Reinforcement Learning Framework for Fine-Tuning Discrete Diffusion Models 提出连续时间强化学习框架以优化离散扩散模型 reinforcement learning PPO MDM
3 MIDI-RAE-JEPA: Hierarchical Representation Learning and Generation for Symbolic Music 提出MIDI-RAE-JEPA以解决符号音乐表示的自监督学习问题 flow matching JEPA representation learning
4 Non-vacuous Generalization Bounds for Reinforcement Learning with Verifiable Rewards 提出非空泛的强化学习泛化界限以解决可验证奖励问题 reinforcement learning distillation large language model
5 On-Policy Delta Distillation 提出On-Policy Delta Distillation以提升强化学习中的推理能力 reinforcement learning distillation
6 Angular Gaussian Supervised Contrastive Learning for Long-Tailed Electrocardiogram Arrhythmia Diagnosis 提出Angular Gaussian监督对比学习以解决长尾心电图心律失常诊断问题 representation learning contrastive learning
7 Learning in Infinitesimal Non-Compositional Sketches 提出LINCS框架以修复非组合性问题 reinforcement learning large language model
8 Evaluating covariate balance for long time horizon Markov decision processes 提出协变量平衡诊断以解决离线强化学习中的偏差问题 reinforcement learning offline RL offline reinforcement learning
9 Optimal Self-Distillation for Rectified Flow via Linear Probing 提出最优自蒸馏方法以提升修正流模型性能 distillation
10 Trajectory-Aware Flow Matching for Topology Optimisation 提出基于流匹配的拓扑优化框架以提升设计效率 flow matching
11 TIDE: Trustworthy and Interpretable Battery Degradation Estimation with Contextual Learning and Symbolic Distillation 提出TIDE以解决电池健康估计的信任性与可解释性问题 distillation
12 Beyond Entropy: Correctness-Aware Advantage Shaping via Contrastive Policy Optimization 提出对比策略优化以解决奖励可验证性问题 reinforcement learning distillation

🔬 支柱九:具身大模型 (Embodied Foundation Models) (1 篇)

#题目一句话要点标签🔗
13 A Minimal Interpretable Architecture for Zero-Shot Reconstruction of Dynamical Systems 提出DynaBase以解决动态系统零-shot重建问题 foundation model

🔬 支柱一:机器人控制 (Robot Control) (1 篇)

#题目一句话要点标签🔗
14 Active Real-World Factor-Based Evaluation for Generalist Robot Policies 提出主动评估框架以解决通用机器人策略评估问题 manipulation

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

#题目一句话要点标签🔗
15 Interleaved Noise Injection Improves Clean, Corrupted, and OOD Performance 提出交错噪声注入以提升模型对干净、受损及OOD数据的性能 PULSE

⬅️ 返回 cs.LG 首页 · 🏠 返回主页