Calibration
Vals MMLU-Pro → ARC-AGI-3
ARC-AGI-3 is estimated from Vals MMLU-Pro with a linear curve fitted on 11 models measured on both: y = 2.8724·x + -2.5025, R² = 0.38, cross-validated error 9.2 pp. It is used for 8 estimates.
| Estimated model | Vals MMLU-Pro | ARC-AGI-3 | Source |
|---|---|---|---|
| Claude Haiku 4.5 | 78.7% | 0.0% | estimated ± 9.2 pp, low confidence |
| Gemini 3.1 Flash-Lite | 86.2% | 0.0% | estimated ± 9.2 pp, low confidence |
| GLM-4.5 | 81.2% | 0.0% | estimated ± 9.2 pp, low confidence |
| GLM-5.3-Flash | 86.1% | 0.0% | estimated ± 9.2 pp, low confidence |
| Laguna M.1 | 68.8% | 0.0% | estimated ± 9.2 pp, low confidence |
| Laguna XS.2 | 69.1% | 0.0% | estimated ± 9.2 pp, low confidence |
| MiMo-V2.5 | 82.9% | 0.0% | estimated ± 9.2 pp, low confidence |
| Qwen3.5 Flash | 84.1% | 0.0% | estimated ± 9.2 pp, low confidence |