Theoretical guarantees in KL for Diffusion Flow Matching

Neural Information Processing Systems 

These two ingredients define a path measure which can then be approximated by learning the drift of its Markovian projection.