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

Reward Modeling on RM-Bench (full)

83Chat Score

CodeScaler

60.84866.59972.3578.101Jan 21, 2025Mar 25, 2025May 27, 2025Jul 29, 2025Sep 30, 2025Dec 2, 2025Feb 4, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2026.02
8379.976.996.492.587.971.884.1
2026.02
80.67573.696.591.885.56781.4
2026.02
73.770.572.18878.378.365.576.1
2025.01
71.259.859.487.58171.456.169.5
2025.01
71.261.854.193.18473.25370
69.560.654.595.78974.746.670.1
2026.02
66.765.366.989.979.974.462.272.2
2025.01
66.451.451.885.486.966.737.763.8
2025.01
65.555.951.793.887.571.347.468.8
2025.01
63.166.856.786.582.671.650.768.3
2025.01
61.771.449.785.585.470.745.167.1