On Achieving Optimal Adversarial Test Error.

Justin D. Li, Matus Telgarsky

Browse the full ICLR paper archive.