马东锡 NLP 🇸🇪(@dongxi

马东锡 NLP 🇸🇪

@dongxi_nlp

PhD in NLP, Senior Machine Learning Expert. Sharing insights on AI, autonomous agents, and large language & reasoning models.

523 Following 8.7K Followers

马东锡 NLP 🇸🇪@dongxi_nlp

2025.05.02 20:55

根据宝玉老师@dotey 关于notebooklm的反向工程获得的系统提示词，用自己最新分享的论文内容和读书笔记，生成了一个播客节目，不想用这个词，但效果只能说太炸裂了。「LLM, Reasoning」论文 Reinforcement Learning for Reasoning in Large Language Models with One Training Example 心有灵犀一点通，只用1 个训练样本去打动模型，让LLM的reasoning能力巨大提升。