INR Classification on MNIST (test)
98.6AccuracyINR2JLS
Evaluation results
| Method | Links | |
|---|---|---|
| INR2JLS#Params=~6M2026.07 | 98.6 | |
| INR2ARRAY (NFT)#Params=~59M2026.07 | 98.5 | |
| NG-GNN#Params=~6M, Probe features=642026.07 | 97.3 | |
| NG-T#Params=~6M, Probe features=642026.07 | 96.83 | |
| WS-KAN2026.02 | 94.3 | |
| NFN#Params=~135M2026.07 | 92.9 | |
| SetTrans2026.02 | 87.5 | |
| MLPWeight Alignment=true2026.02 | 81 | |
| DMC2026.02 | 73.4 | |
| Monomial-NFN Quasiparameters=≈ 3%2026.04 | 70.21 | |
| NP2026.04 | 69.82 | |
| Monomial-NFN tunedparameters=≈ 3%2026.04 | 68.87 | |
| Monomial-NFN2026.04 | 68.43 | |
| HNP2026.04 | 66.02 | |
| MLPData Augmentation=true2026.02 | 62.7 | |
| DS2026.02 | 59.1 | |
| MLP2026.02 | 34.1 | |
| MLP2026.04 | 10.62 |