Preference Prediction
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
66.62Accuracy
51
Feb 26, 2026
70.5Accuracy
42
Jun 1, 2026
74.8Positional Consistent Accuracy
30
Jul 3, 2026
92.2Accuracy
25
Jun 4, 2026
79.8Preference Score
24
Mar 4, 2026
70.7EvalMuse Score
22
May 19, 2026
61Accuracy (Q=0.25)
20
May 28, 2026
73.4Accuracy
19
May 26, 2026
76.9Accuracy
14
May 26, 2026
80.25Preference Prediction Score (k=99)
12
Jun 19, 2026
76.81Score @ Threshold 99
12
Jun 19, 2026
67Accuracy (0.25 Quantile)
10
May 28, 2026
61Accuracy (0.25 Quantile)
10
May 28, 2026
62Accuracy (0.25 Quantile)
10
May 28, 2026
59Accuracy (0.25 Quantile)
10
May 28, 2026
62Accuracy (0.25 Quantile)
10
May 28, 2026
73.9Accuracy
10
Apr 16, 2026
87.8Accuracy
10
Apr 16, 2026
100Accuracy
8
Mar 16, 2026
70.53Test Accuracy
7
Feb 26, 2026
68Accuracy (Q0.25)
5
May 28, 2026
61Accuracy (0.25 Q)
5
May 28, 2026
74Accuracy
5
May 19, 2026
66.3Reward Accuracy
4
Mar 6, 2026
65.7Reward Accuracy
4
Mar 6, 2026