Science Reasoning on GPQA-Diamond (test)
0.7513HypervolumeRADAR
Evaluation Results
| Method | Links | |
|---|---|---|
| RADAREvaluation Protocol=ID (In-distribution)2025.09 | 0.7513 | |
| IRT-RouterEvaluation Protocol=ID (In-distribution)2025.09 | 0.6942 | |
| RouterBenchEvaluation Protocol=ID (In-distribution)2025.09 | 0.6866 | |
| Random-PairEvaluation Protocol=ID (In-distribution)2025.09 | 0.5545 |