Watermarking Pre-trained Language Models with Backdooring

Gu, Chenxi, Huang, Chengsong, Zheng, Xiaoqing · arXiv · 2022