Skip to main content In just 5 minutes help us improve arXiv: Annual Global Survey We gratefully acknowledge support from the Simons Foundation, member institutions , and all contributors. Donate > stat > arXiv:1905.02175 Help | Advanced Search All fields Title Author Abstract Comments Journal reference ACM classification MSC classification Report number arXiv identifier DOI ORCID arXiv author ID Help pages Full text Search open search GO open navigation menu quick links Login Help Pages About --> Statistics > Machine Learning arXiv:1905.02175 (stat) [Submitted on 6 May 2019 ( v1 ), last revised 12 Aug 2019 (this version, v4)] Title: Adversarial Examples Are Not Bugs, They Are Features Authors: Andrew Ilyas , Shibani Santurkar , Dimitris Tsipras , Logan Engstrom , Brandon Tran , Aleksander Madry View a PDF of the paper titled Adversarial Examples Are Not Bugs, They Are Features, by Andrew Ilyas and 5 other authors View PDF Abstract: 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 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. Subjects: Machine Learning (stat.ML) ; Cryptography and Security (cs.CR); Computer Vision and Pattern Recognition (cs.CV); Machine Learning (cs.LG) Cite as: arXiv:1905.02175 [stat.ML] (or arXiv:1905.02175v4 [stat.ML] for this version) https://doi.org/10.48550/arXiv.1905.02175 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Dimitris Tsipras [ view email ] [v1] Mon, 6 May 2019 17:45:05 UTC (1,701 KB) [v2] Tue, 7 May 2019 02:01:14 UTC (1,772 KB) [v3] Wed, 19 Jun 2019 00:25:20 UTC (1,785 KB) [v4] Mon, 12 Aug 2019 14:36:10 UTC (1,787 KB) Full-text links: Access Paper: View a PDF of the paper titled Adversarial Examples Are Not Bugs, They Are Features, by Andrew Ilyas and 5 other authors View PDF TeX Source view license Current browse context: stat.ML < prev | next > new | recent | 2019-05 Change to browse by: cs cs.CR cs.CV cs.LG stat References & Citations NASA ADS Google Scholar Semantic Scholar 4 blog links ( what is this? ) export BibTeX citation Loading... BibTeX formatted citation &times; loading... Data provided by: Bookmark Bibliographic Tools Bibliographic and Citation Tools Bibliographic Explorer Toggle Bibliographic Explorer ( What is the Explorer? ) Connected Papers Toggle Connected Papers ( What is Connected Papers? ) Litmaps Toggle Litmaps ( What is Litmaps? ) scite.ai Toggle scite Smart Citations ( What are Smart Citations? ) Code, Data, Media Code, Data and Media Associated with this Article alphaXiv Toggle alphaXiv ( What is alphaXiv? ) Links to Code Toggle CatalyzeX Code Finder for Papers ( What is CatalyzeX? ) DagsHub Toggle DagsHub ( What is DagsHub? ) GotitPub Toggle Gotit.pub ( What is GotitPub? ) Huggingface Toggle Hugging Face ( What is Huggingface? ) Links to Code Toggle Papers with Code ( What is Papers with Code? ) ScienceCast Toggle ScienceCast ( What is ScienceCast? ) Demos Demos Replicate Toggle Replicate ( What is Replicate? ) Spaces Toggle Hugging Face Spaces ( What is Spaces? ) Spaces Toggle TXYZ.AI ( What is TXYZ.AI? ) Related Papers Recommenders and Search Tools Link to Influence Flower Influence Flower ( What are Influence Flowers? ) Core recommender toggle CORE Recommender ( What is CORE? ) Author Venue Institution Topic About arXivLabs arXivLabs: experimental projects with community collaborators arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website. Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them. Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs . Which authors of this paper are endorsers? | Disable MathJax ( What is MathJax? ) About Help contact arXiv Click here to contact arXiv Contact subscribe to arXiv mailings Click here to subscribe Subscribe Copyright Privacy Policy Web Accessibility Assistance arXiv Operational Status