Rethinking Backdoor Detection Evaluation for Language Models.

Jun Yan, Wenjie Jacky Mo, Xiang Ren, Robin Jia

Browse the full EMNLP paper archive.