AI Paper Atlas

Research map / Architectures and efficiency / Mixture of experts

Teacher-Guided Routing for Sparse Vision Mixture-of-Experts

Masahiro Kada, Ryota Yoshihashi, Satoshi Ikehata, Rei Kawakami, Ikuro Sato

CVPR 2026 · Architectures and efficiency › Mixture of experts · approach: Expert routing and specialization · 0 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add teacher-guided routingVMoE Tiny · ImageNet-1Ktop-1 accuracy: 77.85 -> 78.78Improved
add teacher-guided routingVMoE Small · ImageNet-1Ktop-1 accuracy: 82.63 -> 83.34Improved
add teacher-guided routingVMoE Base · ImageNet-1Ktop-1 accuracy: 83.97 -> 85.46Improved
add z-loss regularizationVMoE Tiny · ImageNet-1Ktop-1 accuracy: 77.85 -> 77.99No clear effect
keep TGR-MoE during fine-tuning (not only pre-training) (ablation)TGR-MoE Tiny · CIFAR-100top-1 accuracy: 86.15 -> 86.95Improved
keep TGR-MoE during fine-tuning (not only pre-training) (ablation)TGR-MoE Small · CIFAR-100top-1 accuracy: 89.18 -> 90.26Improved
keep TGR-MoE during fine-tuning (not only pre-training) (ablation)TGR-MoE Base · CIFAR-100top-1 accuracy: 90.05 -> 91.07Improved
add teacher-guided routing at 64 expertsVMoE Tiny · ImageNet-1K (64 experts)top-1 accuracy: 78.74% -> 79.95%Improved
add teacher-guided routing at 128 expertsVMoE Tiny · ImageNet-1K (128 experts)top-1 accuracy: 79.38% -> 80.36%Improved
apply distillation only in first half of training instead of throughout (ablation)TGR-MoE Tiny 8 experts · ImageNet-1Ktop-1 accuracy: 77.81 -> 78.13Improved
remove task loss from router training (distillation only) (ablation)TGR-MoE Tiny 8 experts · ImageNet-1Ktop-1 accuracy: 77.81 -> 77.83No clear effect
use teacher final-layer features instead of intermediate features for routing signal (ablation)TGR-MoE Tiny 8 experts · ImageNet-1Ktop-1 accuracy: 77.81 -> 75.83Hurt

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy