We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

stat.ML

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Statistics > Machine Learning

Title: Wasserstein GANs Work Because They Fail (to Approximate the Wasserstein Distance)

Abstract: Wasserstein GANs are based on the idea of minimising the Wasserstein distance between a real and a generated distribution. We provide an in-depth mathematical analysis of differences between the theoretical setup and the reality of training Wasserstein GANs. In this work, we gather both theoretical and empirical evidence that the WGAN loss is not a meaningful approximation of the Wasserstein distance. Moreover, we argue that the Wasserstein distance is not even a desirable loss function for deep generative models, and conclude that the success of Wasserstein GANs can in truth be attributed to a failure to approximate the Wasserstein distance.
Subjects: Machine Learning (stat.ML); Machine Learning (cs.LG)
Cite as: arXiv:2103.01678 [stat.ML]
  (or arXiv:2103.01678v3 [stat.ML] for this version)

Submission history

From: Jan Stanczuk [view email]
[v1] Tue, 2 Mar 2021 12:30:25 GMT (1741kb,D)
[v2] Wed, 3 Mar 2021 16:50:36 GMT (1741kb,D)
[v3] Fri, 5 Mar 2021 17:41:34 GMT (1741kb,D)

Link back to: arXiv, form interface, contact.