WGAN with an Infinitely Wide Generator Has No Spurious Stationary Points

Author name not available (Why is that?)

Publication date: 15 February 2021

Abstract: Generative adversarial networks (GAN) are a widely used class of deep generative models, but their minimax training dynamics are not understood very well. In this work, we show that GANs with a 2-layer infinite-width generator and a 2-layer finite-width discriminator trained with stochastic gradient ascent-descent have no spurious stationary points. We then show that when the width of the generator is finite but wide, there are no spurious stationary points within a ball whose radius becomes arbitrarily large (to cover the entire parameter space) as the width goes to infinity.

Has companion code repository: https://github.com/sehyunkwon/Infinite-WGAN

This page was built for publication: WGAN with an Infinitely Wide Generator Has No Spurious Stationary Points

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6360648)