Setting the Trap: Capturing and Defeating Backdoors in Pretrained Language Models through Honeypots Ruixiang T ang

Open in new window