Convergence of Continuous Normalizing Flows for Learning Probability Distributions

arXiv:2404.00551v1 Announce Type: cross
Abstract: Continuous normalizing flows (CNFs) are a generative method for learning probability distributions, which is based on ordinary differential equations. This method has shown remarkable empirical success across various applications, including large-scale image synthesis, protein structure prediction, and molecule generation. In this work, we study the theoretical properties of CNFs with linear interpolation in learning probability distributions from a finite random sample, using a flow matching objective function. We establish non-asymptotic error bounds for the distribution …

abstract applications arxiv continuous continuous normalizing flows convergence cs.lg differential generative image ordinary prediction probability protein protein structure protein structure prediction scale stat.ml study success synthesis type work