Calibration
DeepSWE → CursorBench 3.2
CursorBench 3.2 is estimated from DeepSWE with a Hill curve fitted on 10 models measured on both: y = 0.0000 + (0.6928 − 0.0000)·x^6.00 / (0.37845^6.00 + x^6.00), R² = 0.39, cross-validated error 6.4 pp. It is used for 3 estimates.
| Estimated model | DeepSWE | CursorBench 3.2 | Source |
|---|---|---|---|
| Ember-1 | 75.2% | 68.2% | estimated ± 6.4 pp, low confidence |
| Pareto 26.10 Preview | 69.9% | 67.6% | estimated ± 6.4 pp, low confidence |
| Pareto 26.9 | 74.0% | 68.1% | estimated ± 6.4 pp, low confidence |