Pairwise Preference Ranking
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
1,187ELO Score
5
Apr 29, 2026
0.82Q-Score
5
Apr 10, 2026
86.3Pairwise Accuracy (RM1-Honest)
4
Feb 26, 2026
87Pairwise Accuracy (RM1-Honest)
4
Feb 26, 2026
89.1Pairwise Acc (RM1-Honest)
4
Feb 26, 2026
85.5Pairwise Acc (RM1-Helpful)
4
Feb 26, 2026
84.9Pairwise Accuracy (RM1-Helpful)
4
Feb 26, 2026
86.6Pairwise Acc (RM1)
4
Feb 26, 2026