conftrace_

dense-to-moe conversion

1 paper

Explore in graph

Co-occurring keywords · papers mentioning both

expert expansion (1) parameter merging (1) expert specialization (1) genetic algorithm (1) mixture of expert (1) instruction tuning (1)

Papers

Upcycling Instruction Tuning from Dense to Mixture-of-Experts via Parameter Merging ACL 2025