Jiang, P.; Lyu, X.; Li, Y.; Ma, J. Backdoor Token Unlearning: Exposing and Defending Backdoors in Pretrained Language Models. AAAI 2025, 39, 24285-24293.