Fetching the paper…

Provably Robust Deep Learning via Adversarially Trained Smoothed Classifiers · Around