Massive Multitask Language Understanding on MMLU (200 samples)
0.45MMLU Log-likelihood ScoreGemma 3 12B IT (Base)
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Gemma 3 12B IT (Base)Mode=Base, Quantization=4-bit NF4, Max sequence length=512, Zero-shot=true2026.04 | 0.45 | — | |
| Brainstacks (Routed)Mode=Routed, Quantization=4-bit NF4, Max sequence length=512, Zero-shot=true, Routing mechanism=meta-router selective gating, Chat floor=0.20, Gating=sigmoid2026.04 | 0.435 | 0.015 |