Fetching the paper…

On Equivariance and Fast Sampling in Video Diffusion Models Trained with Warped Noise · Around