Sparse MoEs meet Efficient Ensembles.
James Urquhart AllinghamFlorian WenzelZelda E. MarietBasil MustafaJoan PuigcerverNeil HoulsbyGhassen JerfelVincent FortuinBalaji LakshminarayananJasper SnoekDustin TranCarlos Riquelme RuizRodolphe JenattonPublished in: Trans. Mach. Learn. Res. (2022)