Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Preference Validation on RewardBench 2

74.5Accuracy

MRRG

19.17233.53647.962.264Jul 2, 2026
Updated 23d ago

Evaluation Results

MethodLinks
2026.07
74.5
2026.07
73.7
2026.07
65.3
2026.07
64.4
2026.07
62.4
2026.07
61.3
2026.07
60.3
2026.07
59.4
2026.07
59.2
2026.07
54.9
2026.07
52.3
2026.07
51.2
2026.07
51.2
2026.07
50.7
2026.07
42.3
2026.07
40.3
2026.07
37.6
2026.07
36.5
2026.07
29.4
2026.07
21.3