Pixel-by-pixel Classification on MNIST Unpermuted (test)
99.2Accuracy (Test)Dilated GRU
Evaluation results
| Method | Links | |
|---|---|---|
| Dilated GRU# layers=9, hidden / layer=50, # parameters (≈, k)=130, Max dilations=2562017.10 | 99.2 | |
| GRU# layers=1, hidden / layer=256, # parameters (≈, k)=200, Max dilations=12017.10 | 99.1 | |
| Dilated GRU# layers=9, hidden / layer=20, # parameters (≈, k)=21, Max dilations=2562017.10 | 99 | |
| Dilated LSTM# layers=9, hidden / layer=20, # parameters (≈, k)=28, Max dilations=2562017.10 | 98.9 | |
| Dilated LSTM# layers=9, hidden / layer=50, # parameters (≈, k)=173, Max dilations=2562017.10 | 98.9 | |
| GRU# layers=9, hidden / layer=20, # parameters (≈, k)=21, Max dilations=12017.10 | 98.8 | |
| LSTM# layers=9, hidden / layer=20, # parameters (≈, k)=28, Max dilations=12017.10 | 98.7 | |
| Dilated CNN# layers=10, hidden / layer=50, # parameters (≈, k)=46, Max dilations=5122017.10 | 98.3 | |
| LSTM# layers=1, hidden / layer=256, # parameters (≈, k)=270, Max dilations=12017.10 | 98.2 | |
| Skipped RNN# layers=1, hidden / layer=95, # parameters (≈, k)=16, Max dilations=212017.10 | 98.1 | |
| Dilated CNN# layers=10, hidden / layer=20, # parameters (≈, k)=7, Max dilations=5122017.10 | 98 | |
| Dilated Vanilla# layers=9, hidden / layer=50, # parameters (≈, k)=44, Max dilations=2562017.10 | 98 | |
| Dilated Vanilla# layers=9, hidden / layer=20, # parameters (≈, k)=7, Max dilations=2562017.10 | 97.7 | |
| Full uRNN# layers=1, hidden / layer=512, # parameters (≈, k)=270, Max dilations=12017.10 | 97.5 | |
| IRNN# layers=1, hidden / layer=100, # parameters (≈, k)=12, Max dilations=12017.10 | 97 | |
| Zoneout# layers=1, hidden / layer=100, # parameters (≈, k)=42, Max dilations=12017.10 | 93.1 | |
| Skipped RNN# layers=9, hidden / layer=20, # parameters (≈, k)=11, Max dilations=2562017.10 | 85.4 | |
| Vanilla RNN# layers=9, hidden / layer=20, # parameters (≈, k)=7, Max dilations=12017.10 | 49.1 |