Swenor, A. Using Random Perturbations to Mitigate Adversarial Attacks on NLP Models. AAAI 2022, 36, 13142-13143.