Increasing Confidence in Adversarial Robustness Evaluations


METADATA ONLY
Loading...

Date

2022

Publication Type

Conference Paper

ETH Bibliography

yes

Citations

Altmetric
METADATA ONLY

Data

Rights / License

Abstract

Hundreds of defenses have been proposed to make deep neural networks robust against minimal (adversarial) input perturbations. However, only a handful of these defenses held up their claims because correctly evaluating robustness is extremely challenging: Weak attacks often fail to find adversarial examples even if they unknowingly exist, thereby making a vulnerable network look robust. In this paper, we propose a test to identify weak attacks, and thus weak defense evaluations. Our test slightly modifies a neural network to guarantee the existence of an adversarial example for every sample. Consequentially, any correct attack must succeed in breaking this modified network. For eleven out of thirteen previously-published defenses, the original evaluation of the defense fails our test, while stronger attacks that break these defenses pass it. We hope that attack unit tests - such as ours - will be a major component in future robustness evaluations and increase confidence in an empirical field that is currently riddled with skepticism.

Publication status

published

Book title

Advances in Neural Information Processing Systems 35

Journal / series

Volume

Pages / Article No.

13174 - 13189

Publisher

Curran

Event

36th Annual Conference on Neural Information Processing Systems (NeurIPS 2022)

Edition / version

Methods

Software

Geographic location

Date collected

Date created

Subject

Machine Learning; Adversarial Examples

Organisational unit

09764 - Tramèr, Florian / Tramèr, Florian check_circle

Notes

Funding

Related publications and datasets