Attention-Enhancing Backdoor Attacks Against BERT-based Models.

Weimin Lyu, Songzhu Zheng, Lu Pang, Haibin Ling, Chao Chen

Browse the full EMNLP paper archive.