Efficient Long-Range Transformers: You Need to Attend More, but Not Necessarily at Every Layer.

Published in: EMNLP (Findings) (2023)

Keyphrases