Multi-class Classification on MMLU HELM
59.48Balanced AccuracyBase-Best
Evaluation results
| Method | Links | |
|---|---|---|
| Base-Best2026.02 | 59.48 | |
| S-StackingNetMode=Supervised2026.02 | 58.8 | |
| Logistic RegressionMode=Supervised2026.02 | 57.73 | |
| MACE2026.02 | 55.1 | |
| PM2026.02 | 54.99 | |
| LAA2026.02 | 54.98 | |
| LA2026.02 | 54.97 | |
| EBCC2026.02 | 54.97 | |
| ZenCrowd2026.02 | 54.97 | |
| SML2026.02 | 54.9 | |
| WAwA2026.02 | 54.82 | |
| GLAD2026.02 | 54.74 | |
| Voting2026.02 | 54.6 | |
| M-MSR2026.02 | 54.58 | |
| U-StackingNetMode=Unsupervised2026.02 | 54.42 | |
| Dawid-Skene2026.02 | 53.9 | |
| Base-Worst2026.02 | 35.51 |