NUAT-GAN: Generating Black-Box Natural Universal Adversarial Triggers for Text Classifiers Using Generative Adversarial Networks
- 19 June 2024
- journal article
- research article
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Information Forensics and Security
- Vol. 19 (15566013) , 6484-6498
Abstract
Recent works have demonstrated that text classifiers are vulnerable to universal adversarial triggers (UATs), which are concatenated to any original text from the dataset to mislead text classifiers. Existing methods for generating UATs are limited to a white-box setting, where the adversary needs access to the gradient information about the target model. In the more practical black-box setting, the adversary can only access the logit output of the target model, which increases the difficulty of crafting UATs. In this paper, we propose a framework for generating natural UATs using generative adversarial networks (NUAT-GAN) in the black-box setting. To update parameters of the generator in the black-box setting, we design a training generator algorithm with policy gradient (TGPG), in which the gradient of the target model is replaced with the policy gradient of reinforcement learning. On three text classification datasets, we evaluate the attack and the natural performance of UATs generated on LSTM, CNN and BERT models. Results show that UATs generated by NUAT-GAN can mislead the above models. The average values of the Attack Success Rate (ASR) and GPT-2 Loss of UATs are 0.81 and 9.10, respectively. The UATs can effectively attack online models, such as AllenNLP and ChatGPT. Moreover, we replace the reward given by the discriminator with GPT-2 Loss , the attack and the natural performance of UATs are close to those of NUAT-GAN. This shows that NUAT-GAN is extensible and can combined with the language model.Keywords
Funding Information
- National Natural Science Foundation of China (62072051)
- Beijing University of Posts and Telecommunications (BUPT) Excellent Ph.D. Students Foundation (CX2022229)
This publication has 36 references indexed in Scilit:
- Spam Review Detection Using Deep Learning2019
- TextBugger: Generating Adversarial Text Against Real-world Applications2019
- Generative Adversarial Perturbations2018
- Black-Box Generation of Adversarial Text Sequences to Evade Deep Learning Classifiers2018
- Mask R-CNN2017
- Towards Evaluating the Robustness of Neural Networks2017
- DeepFool: A Simple and Accurate Method to Fool Deep Neural Networks2016
- Deep Residual Learning for Image Recognition2016
- Hownet and the Computation of Meaning2006
- Simple statistical gradient-following algorithms for connectionist reinforcement learningMachine Learning, 1992