Fetching the paper…

Transferability in Machine Learning: from Phenomena to Black-Box Attacks using Adversarial Samples · Around