Towards A Unified View of Sparse Feed-Forward Network in Pretraining Large Language Model.

Zeyu Liu, Tim Dettmers, Xi Lin, Veselin Stoyanov, Xian Li

Browse the full EMNLP paper archive.