Adversarial Examples Are Not Bugs, They Are Features
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.
Authors: Andrew Ilyas, Shibani Santurkar, Dimitris Tsipras, Logan Engstrom, Brandon Tran, Aleksander Madry
https://arxiv.org/abs/1905.02175
Видео Adversarial Examples Are Not Bugs, They Are Features канала Yannic Kilcher
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.
Authors: Andrew Ilyas, Shibani Santurkar, Dimitris Tsipras, Logan Engstrom, Brandon Tran, Aleksander Madry
https://arxiv.org/abs/1905.02175
Видео Adversarial Examples Are Not Bugs, They Are Features канала Yannic Kilcher
Показать
Комментарии отсутствуют
Информация о видео
Другие видео канала
Learning To Classify Images Without Labels (Paper Explained)J. Z. Kolter and A. Madry: Adversarial Robustness - Theory and Practice (NeurIPS 2018 Tutorial)Adversarial Attacks on Neural Networks - Bug or Feature?My GitHub (Trash code I wrote during PhD)Adversarial Examples Are Not Bugs, They Are Features: NeurIPS 2019 Video[ML News] GitHub Copilot - Copyright, GPL, Patents & more | Brickit LEGO app | Distill goes on break#040 - Adversarial Examples (Dr. Nicholas Carlini, Dr. Wieland Brendel, Florian Tramèr)[Rant] The Male Only History of Deep LearningYoshua Bengio Extra Footage 1: Brainstorm with students 🔴Towards Evaluating the Robustness of Neural NetworksBeyond "provable" robustness: new directions in adversarial robustnessOn Evaluating Adversarial Robustness[Trash] Automated Inference on Criminality using Face ImagesIan Goodfellow: Adversarial Machine Learning (ICLR 2019 invited talk)Deep Networks Are Kernel Machines (Paper Explained)Variational Autoencoders - EXPLAINED!Intriguing Properties of Adversarial ML Attacks in the Problem SpaceIntroduction to Deep Learning, Keras, and TensorFlowAdversarial Machine Learning explained! | With examples.