General Preference Pairwise on LMArena Med
0.594Kendall's Tau_bIRT (pr.) Ranking
Evaluation Results
| Method | Links | |
|---|---|---|
| IRT (pr.) Rankingn=36, Domain (Format)=General Prefer (Pairwise)2025.09 | 0.594 | |
| IRT(All) Rankingn=36, Domain (Format)=General Prefer (Pairwise)2025.09 | 0.59 | |
| Accuracy Rankingn=36, Domain (Format)=General Prefer (Pairwise)2025.09 | 0.568 |