Adversarial Examples Are Not Bugs, They Are Features



  • Andrew Ilyas
  • Shibani Santurkar
  • Dimitris Tsipras
  • Logan Engstrom
  • Brandon Tran
  • Aleksander Madry

Published on




Adversarial examples have attracted significant attention in machine learning, but the reasons for their existence and pervasiveness remain unclear. We demonstrate that adversarial examples can be directly attributed to the presence of non-robust features: features (derived from patterns in the data distribution) that are highly predictive, yet brittle and (thus) incomprehensible to humans. After capturing these features within a theoretical framework, we establish their widespread existence in standard datasets. Finally, we present a simple setting where we can rigorously tie the phenomena we observe in practice to a misalignment between the (human-specified) notion of robustness and the inherent geometry of the data.

This work was published in NeurIPS 2019.

Please cite our work using the BibTeX below.

 author = {Ilyas, Andrew and Santurkar, Shibani and Tsipras, Dimitris and Engstrom, Logan and Tran, Brandon and Madry, Aleksander},
 booktitle = {Advances in Neural Information Processing Systems},
 editor = {H. Wallach and H. Larochelle and A. Beygelzimer and F. d\textquotesingle Alch\'{e}-Buc and E. Fox and R. Garnett},
 pages = {},
 publisher = {Curran Associates, Inc.},
 title = {Adversarial Examples Are Not Bugs, They Are Features},
 url = {},
 volume = {32},
 year = {2019}

Close Modal