List your company
ResearchFormerly WizWand

INR classification on F-MNIST (test)

84.6AccuracyWS-KAN
6.96427.119547.27567.4305Jun 15, 2024Oct 6, 2024Jan 27, 2025May 21, 2025Sep 11, 2025Jan 2, 2026Apr 26, 2026
Updated 6mo ago

Evaluation results

MethodLinks
WS-KAN2026.0284.6
ScaleGMN-Btraining_seeds=32024.0680.78
ScaleGMNtraining_seeds=32024.0680.46
SetTrans2026.0280.2
MLPWeight Alignment=true2026.0273.6
DMC2026.0273.1
NFNH_NPbaseline_source=re-trained, training_seeds=32024.0668.94
NFNN_Pbaseline_source=re-trained, training_seeds=32024.0668.19
NG-GNNbaseline_source=reported, training_seeds=32024.0668
DWSbaseline_source=reported, training_seeds=32024.0667.06
DS2026.0265.9
MLPData Augmentation=true2026.0263
Monomial-NFN Quasiparameters=≈ 3%2026.0462.11
Monomial-NFN tunedparameters=≈ 3%2026.0461.44
Monomial-NFN2026.0461.15
NP2026.0458.21
HNP2026.0457.43
MLP2026.0241.3
Inr2Vecbaseline_source=reported, training_seeds=32024.0622.33
MLPbaseline_source=reported, training_seeds=32024.0619.91
MLP2026.049.95