Toward Stronger Textual Attack Detectors

Pierre Colombo, Marine Picot, Nathan Noiry, Guillaume Staerman, Pablo Piantanida


Abstract
The landscape of available textual adversarial attacks keeps growing, posing severe threats and raising concerns regarding deep NLP systems integrity. However, the crucial problem of defending against malicious attacks has only drawn few attention in the NLP community. The latter is nonetheless instrumental to develop robust and trustworthy systems. This paper makes two important contributions in this line of search: (i) we introduce LAROUSSE, a new framework to detect textual adversarial attacks and (ii) we introduce STAKEOUT, an extended benchmark composed of nine popular attack methods, three datasets and two pre-trained models. LAROUSSE is ready-to-use in production as it is unsupervised, hyperparameter free and non-differentiable, protecting it against gradient-based methods. Our new benchmark STAKEOUT allows for a robust evaluation framework: we conduct extensive numerical experiments which demonstrate that LAROUSSE outperforms previous methods, and which allows to identify interesting factor of detection rate variations.
Anthology ID:
2023.findings-emnlp.35
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2023
Month:
December
Year:
2023
Address:
Singapore
Editors:
Houda Bouamor, Juan Pino, Kalika Bali
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
484–505
Language:
URL:
https://aclanthology.org/2023.findings-emnlp.35
DOI:
10.18653/v1/2023.findings-emnlp.35
Bibkey:
Cite (ACL):
Pierre Colombo, Marine Picot, Nathan Noiry, Guillaume Staerman, and Pablo Piantanida. 2023. Toward Stronger Textual Attack Detectors. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 484–505, Singapore. Association for Computational Linguistics.
Cite (Informal):
Toward Stronger Textual Attack Detectors (Colombo et al., Findings 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-2/2023.findings-emnlp.35.pdf