Abstract
Publications home
Catastrophic overfitting is a bug but also a feature
at
| Guillermo Ortiz-Jiménez, Pau De Jorge, Amartya Sanyal, Adel Bibi, Puneet K. Dokania, Pascal Frossard, Gregory Rogéz, Philip H.S. Torr |
| Workshop on New Frontiers in Adversarial Machine Learning (ADVML) at ICML 2022, 22 July 2022, Baltimore, Maryland, USA. |
| arXiv |
| Code |
Despite clear computational advantages in building robust neural networks, adversarial training (AT) using single-step methods is unstable as it suffers from catastrophic overfitting (CO): Networks gain non-trivial robustness during the first stages of adversarial training, but suddenly reach a breaking point where they quickly lose all robustness in just a few iterations. Although some works have succeeded at preventing CO, the different mechanisms that lead to this remarkable failure mode are still poorly understood. In this work, however, we find that the interplay between the structure of the data and the dynamics of AT plays a fundamental role in CO. Specifically, through active interventions on typical datasets of natural images, we establish a causal link between the structure of the data and the onset of CO in single-step AT methods. This new perspective provides important insights into the mechanisms that lead to CO and paves the way towards a better understanding of the general dynamics of robust model construction. The code to reproduce the experiments of this paper can be found at this https URL.
Related Content
This web site uses cookies for the site search, to display videos and for aggregate site analytics.
Learn more about these cookies in our privacy notice.
Cookie settings
You may choose which kind of cookies you allow when visiting this website. Click on "Save cookie settings" to apply your choice.
This content is currently blocked. To view the content please either 'Accept social media cookies' or 'Accept all cookies'.
For more information on cookies see our privacy notice.