Adv-OLM: Generating Textual Adversaries via OLM

Vijit Malik, Ashwani Bhat, Ashutosh Modi


Abstract
Deep learning models are susceptible to adversarial examples that have imperceptible perturbations in the original input, resulting in adversarial attacks against these models. Analysis of these attacks on the state of the art transformers in NLP can help improve the robustness of these models against such adversarial inputs. In this paper, we present Adv-OLM, a black-box attack method that adapts the idea of Occlusion and Language Models (OLM) to the current state of the art attack methods. OLM is used to rank words of a sentence, which are later substituted using word replacement strategies. We experimentally show that our approach outperforms other attack methods for several text classification tasks.
Anthology ID:
2021.eacl-main.71
Volume:
Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume
Month:
April
Year:
2021
Address:
Online
Editors:
Paola Merlo, Jorg Tiedemann, Reut Tsarfaty
Venue:
EACL
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
841–849
Language:
URL:
https://aclanthology.org/2021.eacl-main.71
DOI:
10.18653/v1/2021.eacl-main.71
Bibkey:
Cite (ACL):
Vijit Malik, Ashwani Bhat, and Ashutosh Modi. 2021. Adv-OLM: Generating Textual Adversaries via OLM. In Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume, pages 841–849, Online. Association for Computational Linguistics.
Cite (Informal):
Adv-OLM: Generating Textual Adversaries via OLM (Malik et al., EACL 2021)
Copy Citation:
PDF:
https://aclanthology.org/2021.eacl-main.71.pdf
Code
 vijit-m/Adv-OLM