@kaedeshioharacs
Project Assistant Professor at the University of Tokyo
AI
Its introduction says the added layer predicts the next prior, without VQVAEs or diffusion models.
AI
X
