Bert is Robust! A Case Against Word Substitution-Based Adversarial Attacks

Jens Hauser; Zhao Meng; Damian Pascual; Roger Wattenhofer

ICASSP 2023accepted0 citations

Bert is Robust! A Case Against Word Substitution-Based Adversarial Attacks

Jens Hauser, Zhao Meng, Damian Pascual, Roger Wattenhofer

Abstract

In this work, we investigate the robustness of BERT using four word substitution-based attacks. We combine a human evaluation of individual word substitutions and probabilistic analysis to show that most of the adversarial examples from the four studied attacks do not preserve enough semantics from the original examples, and can thus be easily recognized by human annotators. To further confirm that, we introduce an efficient adversarial defense consisting of a data augmentation step and a post-processing step. We show that many successful attacks can be defended using our defense method by including data similar to adversarial examples during training.

BibTeX

@inproceedings{icassp2023_bertisrobustacas,
  title = {Bert is Robust! A Case Against Word Substitution-Based Adversarial Attacks},
  author = {Jens Hauser and Zhao Meng and Damian Pascual and Roger Wattenhofer},
  booktitle = {ICASSP 2023},
  year = {2023}
}