SAAD adaptively weights adversarial training samples by their transferability to the teacher, yielding higher AutoAttack robustness than prior distillation methods on CIFAR and Tiny-ImageNet without extra compute.
How and when adversarial robustness transfers in knowledge distillation?arXiv preprint arXiv:2110.12072, 2021
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
representative citing papers
Knowledge distillation evaluations must report lost teacher capabilities via a Distillation Loss Statement rather than relying solely on task scores.
citing papers explorer
-
Sample-wise Adaptive Weighting for Transfer Consistency in Adversarial Distillation
SAAD adaptively weights adversarial training samples by their transferability to the teacher, yielding higher AutoAttack robustness than prior distillation methods on CIFAR and Tiny-ImageNet without extra compute.
-
Knowledge Distillation Must Account for What It Loses
Knowledge distillation evaluations must report lost teacher capabilities via a Distillation Loss Statement rather than relying solely on task scores.