6 posts in total

---

2025

09-26

The relationship between prompt and acc expected values

07-01

REINFORCE / RLOO / REINFORCE++ / GRPO

06-18

KL Divergence的三种估计方法

06-18

LLM Reasoning Model - Math 训练记录

05-28

Entroy的数值稳定计算方法

2024

10-29

A unified perspective of RLHF