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

Reward Modeling Evaluation on Reward Bench Math 2

72.3Pairwise Accuracy

Distribution-Calibrated Aggregation

60.02863.21466.469.586Dec 2, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
72.3
2025.12
70.9
2025.12
65.8
2025.12
65.4
2025.12
63.5
2025.12
63.4
2025.12
63.2
2025.12
62.6
2025.12
61.9
2025.12
61.6
2025.12
60.9
2025.12
60.5