Massive Multitask Language Understanding on MMLU Pro (Accuracy, Gain, Wins vs AIM)
67.84AccuracyGCWM
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| GCWMScale=14B2026.05 | 67.84 | 1.23 | 9 | |
| OPCMScale=14B2026.05 | 66.61 | — | — | |
| GCWMScale=8B2026.05 | 63.66 | 0.74 | 8 | |
| AIMMergingScale=8B2026.05 | 62.92 | — | — | |
| GCWMScale=4B2026.05 | 59.43 | 1.19 | 14 | |
| OPCMScale=4B2026.05 | 58.24 | — | — | |
| GCWMScale=1.7B2026.05 | 43.45 | 1.61 | 12 | |
| AIMMergingScale=1.7B2026.05 | 41.84 | — | — | |
| GCWMScale=0.6B2026.05 | 27.14 | 0.3 | 7 | |
| OPCMScale=0.6B2026.05 | 26.84 | — | — |