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

Reward Modeling on RewardBench (full)

99.2Chat Score

HyRe (best weight oracle)* + Skywork-Llama-3.1-8B

77.058482.806788.55594.3033Dec 11, 2024Mar 2, 2025May 23, 2025Aug 13, 2025Nov 2, 2025Jan 23, 2026Apr 15, 2026
Updated 14d ago

Evaluation Results

MethodLinks
99.29396.598.897.2-
2024.12
98.383.496.794.993.1-
2024.12
98.391.295.798.496.4-
2026.04
97.7754.8283.2466.1875.572.46
2024.12
97.560.584.574.179.1-
2024.12
97.28096.294.291.8-
2026.04
96.9358.9979.0571.5676.6374.54
2024.12
96.655.578.189.479.9-
2024.12
96.670.486.584.984.6-
2024.12
96.69193.699.195.1-
2024.12
96.457.574.980.277.2-
96.47481.684.784.2-
2024.12
96.473.187.489.887.1-
2024.12
96.474.792.493.589.7-
2026.04
96.3757.2468.9279.5275.5175.51
2024.12
96.176.188.186.686.7-
96.189.99398.194.3-
2024.12
95.989.393.597.995-
2024.12
95.858.379.573.276.7-
2024.12
95.588.693.297.894.7-
2024.12
95.287.89397.594.3-
956472.678.777.6-
2024.12
9572.590.393.188.5-
2024.12
9587.29396.894-
2024.12
94.788.692.796.794-
2024.12
94.17785.890.286.8-
2024.12
9377.292.291.288.4-
2024.12
92.471.585.192.586.5-
2024.12
92.381.892.593.190-
2024.12
91.178.191.992.388.6-
2024.12
90.472.59092.387.2-
2024.12
90.372.689.191.486.4-
2024.12
89.475.284.588.884.5-
2024.12
88.672.983.789.884.5-
2024.12
88.572.785.591.485.3-
2025.08
81.9860.3979.0656.24-70.08
2025.08
80.7363.4280.5955.31-70.56
2025.08
79.3357.3479.2657.64-70.12
2025.08
78.7451.5576.9855.92-68.82
2025.08
78.655.9378.0255.08-69.23
2025.08
77.9153.6478.6456.31-69.54