ICML 2021spotlight21 citations
Provable Lipschitz Certification for Generative Models
Abstract
We present a scalable technique for upper bounding the Lipschitz constant of generative models. We relate this quantity to the maximal norm over the set of attainable vector-Jacobian products of a given generative model. We approximate this set by layerwise convex approximations using zonotopes. Our approach generalizes and improves upon prior work using zonotope transformers and we extend to Lipschitz estimation of neural networks with large output dimension. This provides efficient and tight bounds on small networks and can scale to generative models on VAE and DCGAN architectures.
BibTeX
@InProceedings{pmlr-v139-jordan21a,
title = {Provable Lipschitz Certification for Generative Models},
author = {Jordan, Matt and Dimakis, Alex},
booktitle = {Proceedings of the 38th International Conference on Machine Learning},
pages = {5118--5126},
year = {2021},
editor = {Meila, Marina and Zhang, Tong},
volume = {139},
series = {Proceedings of Machine Learning Research},
month = {18--24 Jul},
publisher = {PMLR},
pdf = {http://proceedings.mlr.press/v139/jordan21a/jordan21a.pdf},
url = {https://proceedings.mlr.press/v139/jordan21a.html},
abstract = {We present a scalable technique for upper bounding the Lipschitz constant of generative models. We relate this quantity to the maximal norm over the set of attainable vector-Jacobian products of a given generative model. We approximate this set by layerwise convex approximations using zonotopes. Our approach generalizes and improves upon prior work using zonotope transformers and we extend to Lipschitz estimation of neural networks with large output dimension. This provides efficient and tight bounds on small networks and can scale to generative models on VAE and DCGAN architectures.}
}