Stack Attention: Improving the Ability of Transformers to Model Hierarchical Patterns.

Brian DuSell, David Chiang

Browse the full ICLR paper archive.