R-AT: Regularized Adversarial Training for Natural Language Understanding.

Shiwen Ni, Jiawen Li, Hung-Yu Kao

Browse the full EMNLP paper archive.