Abstract
Neural network models are vulnerable to adversarial examples, and adversarial transferability further increases the risk of adversarial attacks. Current methods based on transferability often rely on substitute models, which can be impractical and costly in real-world scenarios due to the unavailability of training data and the victim model’s structural details. In this paper, we propose a novel approach that directly constructs adversarial examples by extracting transferable features across various tasks. Our key insight is that adversarial transferability can extend across different tasks. Specifically, we train a sequence-to-sequence generative model named CT-GAT (Cross-Task Generative Adversarial Attack) using adversarial sample data collected from multiple tasks to acquire universal adversarial features and generate adversarial examples for different tasks.We conduct experiments on ten distinct datasets, and the results demonstrate that our method achieves superior attack performance with small cost.- Anthology ID:
- 2023.emnlp-main.340
- Volume:
- Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing
- Month:
- December
- Year:
- 2023
- Address:
- Singapore
- Editors:
- Houda Bouamor, Juan Pino, Kalika Bali
- Venue:
- EMNLP
- SIG:
- Publisher:
- Association for Computational Linguistics
- Note:
- Pages:
- 5581–5591
- Language:
- URL:
- https://aclanthology.org/2023.emnlp-main.340
- DOI:
- 10.18653/v1/2023.emnlp-main.340
- Cite (ACL):
- Minxuan Lv, Chengwei Dai, Kun Li, Wei Zhou, and Songlin Hu. 2023. CT-GAT: Cross-Task Generative Adversarial Attack based on Transferability. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 5581–5591, Singapore. Association for Computational Linguistics.
- Cite (Informal):
- CT-GAT: Cross-Task Generative Adversarial Attack based on Transferability (Lv et al., EMNLP 2023)
- PDF:
- https://preview.aclanthology.org/nschneid-patch-1/2023.emnlp-main.340.pdf