AI.info

Shengjun Zhang

Explore Shengjun Zhang on AI.info.

• E-GRPO: High Entropy Steps Drive Effective Reinforcement Learning for Flow Models