Score as Action: Fine Tuning Diffusion Generative Models by Continuous-time Reinforcement Learning.

Hanyang Zhao, Haoxian Chen, Ji Zhang, David D. Yao, Wenpin Tang

Browse the full ICML paper archive.