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

Reward Model Evaluation on RewardBench (test)

1.65Kuiper

Consistency

0.82286.406411.9917.5736Dec 23, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.12
1.652.48
2025.12
1.662.82
2025.12
1.981.91
2025.12
6.838.79
2025.12
7.426.71
2025.12
14.0419.17
2025.12
21.7819.96
2025.12
22.3320.37