Image Classification on CIFAR-100 (Corruption Robustness)
71.54Accuracy (Original)Zubov-Net
Evaluation results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| Zubov-NetActivation=Tanh2025.09 | 71.54 | 32.11 | 26.22 | 24.98 | 31.03 | 42.38 | 52.37 | 39.85 | 20.3 | 33.66 | |
| Neural ODEActivation=Tanh2025.09 | 70.81 | 23.65 | 21.26 | 18.58 | 23.92 | 34.49 | 49.21 | 38.2 | 19.61 | 28.62 | |
| ResNet182025.09 | 70.44 | 23.45 | 20.18 | 18.05 | 23.73 | 33.59 | 49.39 | 37.62 | 18.98 | 28.12 | |
| SODEFActivation=sin2025.09 | 70.38 | 28.75 | 23.8 | 22.38 | 28.63 | 38.66 | 51.6 | 39.8 | 22.98 | 32.08 | |
| LyaNet2025.09 | 70.24 | 25.6 | 22.53 | 20.01 | 24.5 | 36.22 | 49.61 | 39.76 | 19.87 | 29.76 | |
| FxTS-Net2025.09 | 70.2 | 26.49 | 24.42 | 21.5 | 25.93 | 37.48 | 50.01 | 40.23 | 20.15 | 30.78 | |
| Proj-NODEActivation=Tanh2025.09 | 69.55 | 27.94 | 24.37 | 22.81 | 28.94 | 40.02 | 51.43 | 34.59 | 19.92 | 31.25 |