cs.LG（2026-08-12）

📊 共 13 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) (8 🔗1) 支柱九：具身大模型 (Embodied Foundation Models) (2) 支柱一：机器人控制 (Robot Control) (2 🔗1) 支柱八：物理动画 (Physics-based Animation) (1)

🔬 支柱二：RL算法与架构 (RL & Architecture) (8 篇)

#题目一句话要点标签🔗⭐
1 Continuous-Latent Predictive Modeling with Semantic Alignment for EEG-Language Foundation Models 提出脑潜在预测模型以解决EEG与语言语义对齐问题 predictive model large language model foundation model
2 Distillation of Foundation Models for Time-dependent PDEs 提出TREX框架以高效蒸馏时间依赖PDE模型 distillation foundation model
3 ScreenShot: A Foundation Model for Few-Shot Combination Drug Screening 提出ScreenShot以解决组合药物筛选中的数据稀缺问题 predictive model foundation model ✅
4 LEMUR: Latent Entropy-aware Multimodal Unlearning via Visual-anchored Reasoning Redirection 提出LEMUR以解决多模态模型的隐私泄露问题 reinforcement learning multimodal
5 LoongReflect: Boosting Long-Horizon Reflection in Search Agents via Global Perspective Distillation 提出LoongReflect以解决长远反思在搜索代理中的挑战 reinforcement learning distillation large language model
6 Towards Understanding On-Policy Distillation through the Lens of Test-Time Scaling 通过测试时缩放理解在政策蒸馏中的作用 distillation
7 REOPD: Reliability-Adaptive Reward Extrapolation for On-Policy Distillation 提出REOPD以解决OPD中的奖励不稳定问题 distillation
8 FM-LLM: A frequency-enhanced mixture-of-experts framework for adapting LLMs to time series forecasting 提出FM-LLM框架以解决时间序列预测中的LLM适应问题 MAE large language model

🔬 支柱九：具身大模型 (Embodied Foundation Models) (2 篇)

#题目一句话要点标签🔗⭐
9 Chain-of-Thought Shows the Path to a Tree: Realizing Branching Complexity 提出深度优先搜索与Dijkstra算法的CoT实现以解决分支复杂性问题 chain-of-thought
10 Earth observation embeddings are effective sub-grid descriptors for probabilistic weather downscaling 提出地球观测嵌入以解决气象下分辨率问题 foundation model

🔬 支柱一：机器人控制 (Robot Control) (2 篇)

#题目一句话要点标签🔗⭐
11 Redistribution-based Cost Inference Improves Sparse Safe Offline RL 提出基于重分配的成本推断以改善稀疏安全离线强化学习 manipulation offline RL
12 MOON: Multi-Objective OrthoNormalized Updates for Multitask Learning 提出MOON以优化多任务学习中的梯度更新问题 manipulation ✅

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

#题目一句话要点标签🔗⭐
13 Attractor Image-Based Deep Learning of Arterial Pulse Waves for Age Classification 提出基于吸引子图像的深度学习方法以进行动脉脉搏波年龄分类 PULSE

⬅️ 返回 cs.LG 首页 · 🏠 返回主页