R-AT: Regularized Adversarial Training for Natural Language Understanding

Shiwen Ni, Jiawen Li, Hung-Yu Kao. R-AT: Regularized Adversarial Training for Natural Language Understanding. In Yoav Goldberg, Zornitsa Kozareva, Yue Zhang, editors, Findings of the Association for Computational Linguistics: EMNLP 2022, Abu Dhabi, United Arab Emirates, December 7-11, 2022. pages 6427-6440, Association for Computational Linguistics, 2022. [doi]

Abstract

Abstract is missing.