Robust Pre-Training by Adversarial Contrastive Learning

This fits particularly well with the goal of adversarial robustness, as one cause of adversarial fragility is the lack of feature invariance, i.e., small input perturbations can