On-Policy Diffusion Reinforcement Learning Meets Off-Policy Quality Anchoring

Open paper
Benchmark Model Rank Results
text-to-image-generation-on-genevalDiffusionCompass#1Overall: 0.97
text-to-video-generation-on-vbenchDiffusionCompass (Wan2.1-14B)+RP#1VBench Total Score: 87.16

Rank counts only results with a code link.