cs.LG（2025-06-23）
📊 共 6 篇论文 | 🔗 2 篇有代码
🎯 兴趣领域导航
🔬 支柱九：具身大模型 (Embodied Foundation Models) (4 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 1 | Thought Anchors: Which LLM Reasoning Steps Matter? | 提出思维锚点方法以解析大型语言模型的推理过程 | chain-of-thought | ||
| 2 | ReDit: Reward Dithering for Improved LLM Policy Optimization | 提出ReDit以解决LLM优化中的离散奖励问题 | large language model | ||
| 3 | No Training Wheels: Steering Vectors for Bias Correction at Inference Time | 提出无训练方法以解决分类模型偏差问题 | large language model | ||
| 4 | LLMs on a Budget? Say HOLA | 提出HOLA框架以高效部署大型语言模型 | large language model |
🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 5 | SlimMoE: Structured Compression of Large MoE Models via Expert Slimming and Distillation | 提出SlimMoE以解决大规模MoE模型的压缩与部署问题 | distillation large language model | ✅ | |
| 6 | Confucius3-Math: A Lightweight High-Performance Reasoning LLM for Chinese K-12 Mathematics Learning | 提出Confucius3-Math以解决中国K-12数学学习问题 | reinforcement learning large language model | ✅ |