Learning a Diffusion Model Policy from Rewards via Q-Score Matching.

Michael Psenka, Alejandro Escontrela, Pieter Abbeel, Yi Ma

Browse the full ICML paper archive.