Backdoor Attacks on Pre-trained Models by Layerwise Weight Poisoning
Linyang Li, Demin Song, Xiaonan Li, Jiehang Zeng, Ruotian Ma, Xipeng Qiu · Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing · 2021
Pre-Trained Models have been widely applied and recently proved vulnerable under backdoor attacks: the released pre-trained weights can be maliciously poisoned with certain triggers.When the triggers are activated, even the fine-tuned model will predict pre-defined labels, causing a security threat.These backdoors generated by the poisoning methods can be erased by changing hyper-parameters during fine-tuning or detected by finding the triggers.In this paper, we propose a stronger weight-poisoning attack method that introduces a layerwise weight poisoning strategy to plant deeper backdoors; we also introduce a combinatorial trigger that cannot be easily detected.The experiments on text classification tasks show that previous defense methods cannot resist our weight-poisoning method, which indicates that our method can be widely applied and may provide hints for future model robustness studies.