AI.info

Sujie Hu

Explore Sujie Hu on AI.info.

• Taming Preference Mode Collapse via Directional Decoupling Alignment in Diffusion Reinforcement Learning