Pinned Loading

1. DeepEnlighten DeepEnlighten Public

   Pure RL to post-train base models for social reasoning capabilities. Lightweight replication of DeepSeek-R1-Zero with Social IQa dataset.

   Python 42

2. Logic-RL-Lite Logic-RL-Lite Public

   Lightweight replication study of DeepSeek-R1-Zero. Interesting findings include "No Aha Moment", "Longer CoT ≠ Accuracy", and "Language Mixing in Instruct Models".

   Python 50

3. Awesome-LLM-Interview-Questions-and-Answers Awesome-LLM-Interview-Questions-and-Answers Public

   大模型算法工程师、大模型Agent开发工程师面试：常见题目和答案

   45 2