Statistical Guarantees for Variational Autoencoders using PAC-Bayesian Theory

Published: 21 Sept 2023, Last Modified: 13 Jan 2024NeurIPS 2023 spotlightEveryoneRevisionsBibTeX
Keywords: Variational Autoencoders, PAC-Bayes, Statistical Learning Theory
TL;DR: We provide reconstruction, regeneration and generation guarantees for VAEs using PAC-Bayesian Theory
Abstract: Since their inception, Variational Autoencoders (VAEs) have become central in machine learning. Despite their widespread use, numerous questions regarding their theoretical properties remain open. Using PAC-Bayesian theory, this work develops statistical guarantees for VAEs. First, we derive the first PAC-Bayesian bound for posterior distributions conditioned on individual samples from the data-generating distribution. Then, we utilize this result to develop generalization guarantees for the VAE's reconstruction loss, as well as upper bounds on the distance between the input and the regenerated distributions. More importantly, we provide upper bounds on the Wasserstein distance between the input distribution and the distribution defined by the VAE's generative model.
Supplementary Material: pdf
Submission Number: 1485