{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/provably-minimally-distorted-adversarial","title":"Provably Minimally-Distorted Adversarial Examples","arxiv_id":"1709.10207","date":"2017-09-29","proceeding":null,"authors":["Nicholas Carlini","Guy Katz","Clark Barrett","David L. Dill"],"abstract":"The ability to deploy neural networks in real-world, safety-critical systems\nis severely limited by the presence of adversarial examples: slightly perturbed\ninputs that are misclassified by the network. In recent years, several\ntechniques have been proposed for increasing robustness to adversarial examples\n--- and yet most of these have been quickly shown to be vulnerable to future\nattacks. For example, over half of the defenses proposed by papers accepted at\nICLR 2018 have already been broken. We propose to address this difficulty\nthrough formal verification techniques. We show how to construct provably\nminimally distorted adversarial examples: given an arbitrary neural network and\ninput sample, we can construct adversarial examples which we prove are of\nminimal distortion. Using this approach, we demonstrate that one of the recent\nICLR defense proposals, adversarial retraining, provably succeeds at increasing\nthe distortion required to construct adversarial examples by a factor of 4.2.","url_abs":"http://arxiv.org/abs/1709.10207v2","url_pdf":"http://arxiv.org/pdf/1709.10207v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"provably-minimally-distorted-adversarial","repo_url":"https://github.com/huanzhang12/ATLA_robust_RL","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1709.10207","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}