A Phase Transition in Diffusion Models Reveals the Hierarchical Nature of Data

arXiv:2402.16991v1 Announce Type: cross
Abstract: Understanding the structure of real data is paramount in advancing modern deep-learning methodologies. Natural data such as images are believed to be composed of features organised in a hierarchical and combinatorial manner, which neural networks capture during learning. Recent advancements show that diffusion models can generate high-quality images, hinting at their ability to capture this underlying structure. We study this phenomenon in a hierarchical generative model of data. We find that the backward diffusion process …

abstract arxiv cond-mat.dis-nn cs.cv cs.lg data diffusion diffusion models features hierarchical images modern natural nature networks neural networks real data show stat.ml transition type understanding