Benign Overfitting in Adversarial Training of Neural Networks.

Yunjuan Wang, Kaibo Zhang, Raman Arora

Browse the full ICML paper archive.