Poster
Path-Gradient Estimators for Continuous Normalizing Flows
Lorenz Vaitl · Kim A. Nicoli · Shinichi Nakajima · Pan Kessel
Hall E #724
Keywords: [ DL: Generative Models and Autoencoders ] [ APP: Physics ] [ DL: Everything Else ] [ DL: Algorithms ] [ PM: Variational Inference ]
Recent work has established a path-gradient estimator for simple variational Gaussian distributions and has argued that the path-gradient is particularly beneficial in the regime in which the variational distribution approaches the exact target distribution. In many applications, this regime can however not be reached by a simple Gaussian variational distribution. In this work, we overcome this crucial limitation by proposing a path-gradient estimator for the considerably more expressive variational family of continuous normalizing flows. We outline an efficient algorithm to calculate this estimator and establish its superior performance empirically.