Tag
The paper shows that diffusion training acceleration methods like REPA and dispersive regularization primarily reorganize the AdaLN-Zero residual gate α, and proposes a teacher-free radial objective that broadens gate participation to improve FID on ImageNet 256×256 with SiT-B/2.
该论文提出扩散Transformer的训练加速部分源于AdaLN-Zero残差门控α的重组，并引入归一化参与度比(nPR)和一种无教师的径向目标，在ImageNet 256x256上以SiT-B/2改进了FID。