cs.AI（2025-10-09）
📊 共 4 篇论文 | 🔗 1 篇有代码
🎯 兴趣领域导航
🔬 支柱九：具身大模型 (Embodied Foundation Models) (2 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 1 | Test-Time Matching: Unlocking Compositional Reasoning in Multimodal Models | 提出测试时匹配(TTM)算法，提升多模态模型在组合推理任务上的性能 | large language model multimodal | ||
| 2 | How to Teach Large Multimodal Models New Skills | 提出两种高效微调策略，提升大型多模态模型新技能学习能力并缓解灾难性遗忘 | multimodal | ✅ |
🔬 支柱二：RL算法与架构 (RL & Architecture) (2 篇)
| # | 题目 | 一句话要点 | 标签 | 🔗 | ⭐ |
|---|---|---|---|---|---|
| 3 | Unified World Models: Memory-Augmented Planning and Foresight for Visual Navigation | UniWM：统一的、记忆增强的世界模型，用于视觉导航中的规划与预测 | world model egocentric multimodal | ||
| 4 | Curriculum Learning with Synthetic Data for Enhanced Pulmonary Nodule Detection in Chest Radiographs | 结合课程学习与合成数据增强，提升胸部X光片肺结节检测性能 | curriculum learning |