| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| UltraMedical | ProMedical-RAG | Q-Score0.8195 | 5 | 8d ago | |
| UltraFeedback 10% holdout (test) | BENCHALIGN | Pairwise Accuracy (RM1-Honest)86.3 | 4 | 1mo ago | |
| UltraFeedback 5% holdout (test) | BENCHALIGN | Pairwise Accuracy (RM1-Honest)87 | 4 | 1mo ago | |
| UltraFeedback 2% holdout (test) | BENCHALIGN | Pairwise Acc (RM1-Honest)89.1 | 4 | 1mo ago | |
| Helpsteer 10% holdout (test) | BENCHALIGN | Pairwise Acc (RM1-Helpful)85.5 | 4 | 1mo ago | |
| Helpsteer 5% holdout (test) | BENCHALIGN | Pairwise Accuracy (RM1-Helpful)84.9 | 4 | 1mo ago | |
| Helpsteer 2% holdout (test) | BENCHALIGN | Pairwise Acc (RM1)86.6 | 4 | 1mo ago |