郑重声明：原文参见标题，如有侵权，请联系作者，将会撤销发布！
同大组工作
Abstract
- Robustness-enhanced Adversarial Regularized Robustness enhancedrobustness-enhanced adversarial regularized robustness-enhanced reinforcement adversarial redefining robustness regularized factorization non-negative interactions regularized regularized networks dropout r-drop robustness translation augmentation robustness improving robustness similarity improving attention understanding transformers robustness vision