RobustEmbed: Robust Sentence Embeddings Using Self-Supervised Contrastive Pre-Training

Javad Asl, Eduardo Blanco, Daniel Takabi


Abstract
Pre-trained language models (PLMs) have demonstrated their exceptional performance across a wide range of natural language processing tasks. The utilization of PLM-based sentence embeddings enables the generation of contextual representations that capture rich semantic information. However, despite their success with unseen samples, current PLM-based representations suffer from poor robustness in adversarial scenarios. In this paper, we propose RobustEmbed, a self-supervised sentence embedding framework that enhances both generalization and robustness in various text representation tasks and against diverse adversarial attacks. By generating high-risk adversarial perturbations to promote higher invariance in the embedding space and leveraging the perturbation within a novel contrastive objective approach, RobustEmbed effectively learns high-quality sentence embeddings. Our extensive experiments validate the superiority of RobustEmbed over previous state-of-the-art self-supervised representations in adversarial settings, while also showcasing relative improvements in seven semantic textual similarity (STS) tasks and six transfer tasks. Specifically, our framework achieves a significant reduction in attack success rate from 75.51% to 39.62% for the BERTAttack attack technique, along with enhancements of 1.20% and 0.40% in STS tasks and transfer tasks, respectively.
Anthology ID:
2023.findings-emnlp.305
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:
4587–4603
Language:
URL:
https://aclanthology.org/2023.findings-emnlp.305
DOI:
10.18653/v1/2023.findings-emnlp.305
Bibkey:
Cite (ACL):
Javad Asl, Eduardo Blanco, and Daniel Takabi. 2023. RobustEmbed: Robust Sentence Embeddings Using Self-Supervised Contrastive Pre-Training. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 4587–4603, Singapore. Association for Computational Linguistics.
Cite (Informal):
RobustEmbed: Robust Sentence Embeddings Using Self-Supervised Contrastive Pre-Training (Asl et al., Findings 2023)
Copy Citation:
PDF:
https://preview.aclanthology.org/nschneid-patch-5/2023.findings-emnlp.305.pdf