Prompt Makes mask Language Models Better Adversarial Attackers.

He Zhu, Ce Li, Haitian Yang, Yan Wang, Weiqing Huang

Browse the full ICASSP paper archive.