Backdoor Token Unlearning: Exposing and Defending Backdoors in Pretrained Language Models.

Peihai Jiang, Xixiang Lyu, Yige Li, Jing Ma

Browse the full AAAI paper archive.