Calibration
ResearchClawBench → GDPval-AA
GDPval-AA is estimated from ResearchClawBench with a Michaelis–Menten curve fitted on 13 models measured on both: y = 2.0000·x / (0.86900 + x), R² = 0.39, cross-validated error 8.3 pp. It is used for 3 estimates.
| Estimated model | ResearchClawBench | GDPval-AA | Source |
|---|---|---|---|
| Grok 4.1 | 13.5% | 26.9% | estimated ± 8.3 pp, low confidence |
| MiMo-V2.5 | 16.9% | 32.6% | estimated ± 8.3 pp, low confidence |
| MiMo-V2-Pro | 15.3% | 29.9% | estimated ± 8.3 pp, low confidence |