Fetching the paper…

Bridging SFT and DPO for Diffusion Model Alignment with Self-Sampling Preference Optimization · Around