Randomized Smoothing with Masked Inference for Adversarially Robust Text Classifications

Han Cheol Moon, Shafiq Joty, Ruochen Zhao, Megh Thakkar, Chi Xu


Abstract
Large-scale pre-trained language models have shown outstanding performance in a variety of NLP tasks. However, they are also known to be significantly brittle against specifically crafted adversarial examples, leading to increasing interest in probing the adversarial robustness of NLP systems. We introduce RSMI, a novel two-stage framework that combines randomized smoothing (RS) with masked inference (MI) to improve the adversarial robustness of NLP systems. RS transforms a classifier into a smoothed classifier to obtain robust representations, whereas MI forces a model to exploit the surrounding context of a masked token in an input sequence. RSMI improves adversarial robustness by 2 to 3 times over existing state-of-the-art methods on benchmark datasets. We also perform in-depth qualitative analysis to validate the effectiveness of the different stages of RSMI and probe the impact of its components through extensive ablations. By empirically proving the stability of RSMI, we put it forward as a practical method to robustly train large-scale NLP models. Our code and datasets are available at https://github.com/Han8931/rsmi_nlp
Anthology ID:
2023.acl-long.282
Volume:
Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)
Month:
July
Year:
2023
Address:
Toronto, Canada
Editors:
Anna Rogers, Jordan Boyd-Graber, Naoaki Okazaki
Venue:
ACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
5145–5165
Language:
URL:
https://aclanthology.org/2023.acl-long.282
DOI:
10.18653/v1/2023.acl-long.282
Bibkey:
Cite (ACL):
Han Cheol Moon, Shafiq Joty, Ruochen Zhao, Megh Thakkar, and Chi Xu. 2023. Randomized Smoothing with Masked Inference for Adversarially Robust Text Classifications. In Proceedings of the 61st Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 5145–5165, Toronto, Canada. Association for Computational Linguistics.
Cite (Informal):
Randomized Smoothing with Masked Inference for Adversarially Robust Text Classifications (Moon et al., ACL 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-5/2023.acl-long.282.pdf
Video:
 https://preview.aclanthology.org/nschneid-patch-5/2023.acl-long.282.mp4