Rare find

convex_adversarial. A method for training neural networks that are provably robust to adversarial attacks.

github.com/locuslab/convex_adversarial
by @locuslab

Problem

Audience

Keep exploring

Updates

No founder updates yet. Check back soon.