Masking as an Efficient Alternative to Finetuning for Pretrained Language Models.

Mengjie Zhao, Tao Lin, Fei Mi, Martin Jaggi, Hinrich Schtze

Browse the full EMNLP paper archive.