One Bit Matters: Understanding Adversarial Examples as the Abuse of RedundancyDownload PDF

27 Sep 2018 (modified: 26 Nov 2018)ICLR 2019 Conference Withdrawn SubmissionReaders: Everyone
  • Abstract: Adversarial examples have somewhat disrupted the enormous success of machine learning (ML) and are causing concern with regards to its trustworthiness: A small perturbation of an input results in an arbitrary failure of an otherwise seemingly well-trained ML system. While studies are being conducted to discover the intrinsic properties of adversarial examples, such as their transferability and universality, there is insufficient theoretic analysis to help understand the phenomenon in a way that can influence the design process of ML experiments. In this paper, we deduce an information-theoretic model which explains adversarial attacks universally as the abuse of feature redundancies in ML algorithms. We prove that feature redundancy is a necessary condition for the existence of adversarial examples. Our model helps to explain the major questions raised in many anecdotal studies on adversarial examples. Our theory is backed up by empirical measurements of the information content of benign and adversarial examples on both image and text datasets. Our measurements show that typical adversarial examples introduce just enough redundancy to overflow the decision making of a machine learner trained on corresponding benign examples. We conclude with actionable recommendations to improve the robustness of machine learners against adversarial examples.
  • Keywords: adversarial examples, information theory, robust neural networks
  • TL;DR: A new theoretical explanation for the existence of adversarial examples
7 Replies