Preference Modeling
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
61.4Accuracy
30
May 27, 2026
87.6Accuracy
20
Mar 11, 2026
65.2Accuracy
20
Mar 11, 2026
72.2Accuracy
20
Mar 11, 2026
68.42Factuality
10
May 19, 2026
73.2Win Rate
9
Feb 26, 2026
1.09NLL
8
Jun 11, 2026
1.08NLL
8
Jun 11, 2026
68.4Preference Accuracy
7
Mar 4, 2026
66.4Preference Accuracy
6
Mar 4, 2026
0.319NDCG@5
4
Feb 26, 2026