Gracefully Filtering Backdoor Samples for Generative Large Language Models without Retraining.

Zongru Wu, Pengzhou Cheng, Lingyong Fang, Zhuosheng Zhang, Gongshen Liu

Browse the full COLING paper archive.