1 paper
Yiqin Wang, Nuri Cingillioglu, Charles Pert
Language modeling using Transformers has become commonplace despite their fixed computational depth and quadratic runtime with respect to input tokens. Recurrent models on the othe…