Preference Alignment
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
2.328Borda Avg
5
Feb 26, 2026
0.384BT Score
5
Feb 26, 2026
2.459Borda Avg Score
5
Feb 26, 2026
70.6GRA
4
Mar 6, 2026
64.4GRA (%)
4
Mar 6, 2026
72.7GRA (%)
4
Mar 6, 2026
72.9GRA (%)
4
Mar 6, 2026
71.25Win Rate (RM Evaluator)
3
May 11, 2026
77.75Win Rate (Reward Model)
3
May 11, 2026
72.14Helpfulness (Reward)
3
Apr 3, 2026
28.69Win Rate
3
Feb 26, 2026