Image Classification on CIFAR-10 (Accuracy, NLL, ECE, Brier, AUROC)
92.3AccuracyMC-D
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MC-D2025.11 | 92.3 | 99.8 | 0.23 | 1 | 11.3 | 92.3 | |
| SWAG2025.11 | 91.8 | 99.7 | 0.251 | 0.9 | 12.1 | 92 | |
| uCBOptϑ=→ 0+2025.11 | 91 | 99.6 | 0.296 | 3.8 | 13.5 | 91.8 | |
| SGD2025.11 | 90.9 | 99.7 | 0.308 | 3.9 | 13.9 | 91.4 | |
| Laplace2025.11 | 90.9 | 99.7 | 0.303 | 3.5 | 13.9 | 91.1 | |
| uCBOpt-adapt2025.11 | 90.9 | 99.7 | 0.3 | 3.8 | 13.8 | 91.7 | |
| lCBOpt-adapt2025.11 | 90.9 | 99.7 | 0.299 | 3.7 | 13.8 | 91.5 | |
| uCBOptϑ=5 ⋅ 10−52025.11 | 90.7 | 99.7 | 0.302 | 3.3 | 14 | 91.4 | |
| IVON2025.11 | 89.9 | 99.6 | 0.317 | 2.8 | 14.9 | 91 | |
| IVON@mean2025.11 | 89.2 | 99.6 | 0.358 | 4.5 | 16.2 | 90.8 | |
| AdamW2025.11 | 87.2 | 99.5 | 0.406 | 4.2 | 18.9 | 89.3 |