Fetching the paper…

Provable robustness against all adversarial $l_p$-perturbations for $p\geq 1$ · Around