Brief

Researchers have identified a phenomenon called 'copying' in high-dimensional distillation of diffusion models. This occurs when a distilled student model replicates the original noise-data pairings of the teacher model, a behavior not observed in lower-dimensional settings. The study suggests this copying is an emergent property due to the student model's limited geometric freedom during distillation, rather than adversarial objectives or teacher memorization. AI

Researchers have introduced Continuous-Time Distribution Matching (CDM), a novel method for accelerating diffusion models. This approach moves beyond discrete-time distillation by employing a dynamic, continuous schedule and an off-trajectory matching objective. CDM aims to improve image generation fidelity and detail preservation in few-step diffusion processes without requiring complex auxiliary modules like GANs. AI