AI.info

Haibin Lin

Explore Haibin Lin on AI.info.

• Towards Full Pipeline FP8 Reinforcement Learning for LLMs
• FAPO: Flawed-Aware Policy Optimization for Efficient and Reliable Reasoning