Ian Fischer, Alexander A. Alemi
Published 2020-02-13Version 1
We demonstrate that the Conditional Entropy Bottleneck (CEB) can improve model robustness. CEB is an easy strategy to implement and works in tandem with data augmentation procedures. We report results of a large scale adversarial robustness study on CIFAR-10, as well as the ImageNet-C Common Corruptions Benchmark, ImageNet-A, and PGD attacks.
The Conditional Entropy Bottleneck
Using Pre-Training Can Improve Model Robustness and Uncertainty
Using Self-Supervised Learning Can Improve Model Robustness and Uncertainty
![QR code for arXiv:2002.05380 [cs.LG]](http://oss.arxitics.com/images/favicon.svg)