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

Reward Modeling on RewardBench 2 (test)

76.3RWBench2 Score

CE-RM-4B

55.70861.05466.471.746Jan 28, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.01
76.3
2026.01
75.9
2026.01
75.9
2026.01
74.6
2026.01
73.4
2026.01
71.6
2026.01
68.3
2026.01
67.3
2026.01
56.5