Image Classification on CIFAR-10 (Accuracy, AUC, NLL)
88.36AccuracyNesterov-SDEBNN
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| Nesterov-SDEBNNIntegration scheme=fixed-step2026.03 | 88.36 | 85.61 | 0.597 | |
| SDE-BNNIntegration scheme=fixed-step2026.03 | 87.56 | 83.05 | 0.989 | |
| Nesterov-SDEBNNIntegration scheme=adaptive-step2026.03 | 86.99 | 83.94 | 0.532 | |
| SDE-BNNIntegration scheme=adaptive-step2026.03 | 85.87 | 78.42 | 0.677 |