HyperAI

Adversarial Attack On Cifar 10

Metrics

Attack: PGD20

Results

Performance results of various models on this benchmark

Comparison Table
Model NameAttack: PGD20
theoretically-principled-trade-off-between45.900
towards-deep-learning-models-resistant-to48.440
attention-masks-help-adversarial-attacks-to-
ensemble-everything-everywhere-multi-scale-
an-orthogonal-classifier-for-improving-the78.680
robust-representation-learning-via-asymmetric-