An Information-Theoretic Explanation for the Adversarial Fragility of AI Classifiers.

Hui Xie, Jirong Yi, Weiyu Xu, Raghu Mudumbai

Browse the full ISIT paper archive.