Paper ID: 2210.11082
Apple of Sodom: Hidden Backdoors in Superior Sentence Embeddings via Contrastive Learning
Xiaoyi Chen, Baisong Xin, Shengfang Zhai, Shiqing Ma, Qingni Shen, Zhonghai Wu
This paper finds that contrastive learning can produce superior sentence embeddings for pre-trained models but is also vulnerable to backdoor attacks. We present the first backdoor attack framework, BadCSE, for state-of-the-art sentence embeddings under supervised and unsupervised learning settings. The attack manipulates the construction of positive and negative pairs so that the backdoored samples have a similar embedding with the target sample (targeted attack) or the negative embedding of its clean version (non-targeted attack). By injecting the backdoor in sentence embeddings, BadCSE is resistant against downstream fine-tuning. We evaluate BadCSE on both STS tasks and other downstream tasks. The supervised non-targeted attack obtains a performance degradation of 194.86%, and the targeted attack maps the backdoored samples to the target embedding with a 97.70% success rate while maintaining the model utility.
Submitted: Oct 20, 2022