Robustness May Be at Odds with Accuracy 2: The Unit-Norm Soft-Margin SVM Is >99% Accurate but <1% Robust at ε ≥ 2η, While Adversarial Training Gives Accuracy pResearch Paper
Adversarial training asks a classifier to work even after an input has been changed within a prescribed perturbation budget. Tsipras, Santurkar, Engstrom, Turner, and Mądry studied why the classifier that performs best on ordinary inputs can behave very differently under these changes. Their explicit model separates one reliably predictive feature from many…