Notwithstanding the promise of Lipschitz-based approaches to
determinist...
Certified defenses against small-norm adversarial examples have received...
Ensembling certifiably robust neural networks has been shown to be a
pro...
Recent work has shown that models trained to the same objective, and whi...
Neural networks are increasingly being deployed in contexts where safety...
Certifiable local robustness, which rigorously precludes small-norm
adve...
The threat of adversarial examples has motivated work on training certif...
LSTM-based recurrent neural networks are the state-of-the-art for many
n...
Local robustness ensures that a model classifies all inputs within an
ϵ-...
Membership inference (MI) attacks exploit a learned model's lack of
gene...
We study the phenomenon of bias amplification in classifiers, wherein a
...
We study the problem of explaining a rich class of behavioral properties...