pith. sign in

M o E fication: Transformer Feed-forward Layers are Mixtures of Experts

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

fields

cs.LG 1

years

2026 1

verdicts

UNVERDICTED 1

clear filters

representative citing papers

DOT-MoE: Differentiable Optimal Transport for MoEfication

cs.LG · 2026-06-01 · unverdicted · novelty 6.0

DOT-MoE uses differentiable optimal transport and straight-through estimators to partition FFN layers into capacity-constrained experts, outperforming heuristic baselines in retaining 90% performance at 50% active parameters.

citing papers explorer

Showing 1 of 1 citing paper after filters.

  • DOT-MoE: Differentiable Optimal Transport for MoEfication cs.LG · 2026-06-01 · unverdicted · none · ref 7

    DOT-MoE uses differentiable optimal transport and straight-through estimators to partition FFN layers into capacity-constrained experts, outperforming heuristic baselines in retaining 90% performance at 50% active parameters.