Union of Experts: Adapting Hierarchical Routing to Equivalently Decomposed Transformer Paper • 2503.02495 • Published Mar 4 • 9
MUDDFormer: Breaking Residual Bottlenecks in Transformers via Multiway Dynamic Dense Connections Paper • 2502.12170 • Published Feb 13 • 12
Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling Paper • 2501.16975 • Published Jan 28 • 31