Paul Liang


2023

pdf bib
Difference-Masking: Choosing What to Mask in Continued Pretraining
Alex Wilf | Syeda Akter | Leena Mathur | Paul Liang | Sheryl Mathew | Mengrou Shou | Eric Nyberg | Louis-Philippe Morency
Findings of the Association for Computational Linguistics: EMNLP 2023

The self-supervised objective of masked prediction has led to promising performance gains on a variety of downstream tasks. However, while most approaches randomly mask tokens, there is strong intuition that deciding what to mask can substantially improve learning outcomes. We investigate this in continued pretraining setting in which pretrained models continue to pretrain on domain-specific data before performing some downstream task. We introduce Difference-Masking, a masking strategy that automatically chooses what to mask during continued pretraining by considering what makes a task domain different from the pretraining domain. Empirically, we find that Difference-Masking outperforms baselines on continued pretraining settings across four diverse language-only and multimodal video tasks.