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

Reward Modeling on RewardBench v2 (test)

86.5Average Score

Skywork-Reward-V2-Llama-3.1-8B-40M

24.0781640.2838356.489572.69517Jul 2, 2025Aug 8, 2025Sep 14, 2025Oct 22, 2025Nov 28, 2025Jan 4, 2026Feb 11, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
2025.07
86.587.967.883.197.399.283.9
2025.07
86.5------
2025.07
84.184.666.277.696.798.481.2
2025.07
84.1------
2025.07
82.187.254.472.791.396.890.1
2025.07
80.584.648.871.690.79790.6
2025.07
78.279.849.1779496.472.9
2025.07
78.2------
2025.07
77.265.755.381.190.986.783.4
2025.07
76.778.537.269.995.895.483.2
2025.07
76.7------
2025.07
76.574.141.969.996.490.386.2
2025.07
76.582.741.974.989.586.283.7
2025.07
76.5------
2025.07
7681.341.969.988.486.588.3
2025.07
75.873.740.370.594.293.282.6
2025.07
75.577.346.273.292.296.667.4
2025.07
75.5------
2025.07
75.473.354.47590.392.167.2
2025.07
75.376.737.567.296.991.781.8
2025.07
75.3------
2025.07
74.776.245.669.493.19667.7
2025.07
74.7------
2025.07
73.968.84563.991.897.676.5
2025.07
73.169.942.562.893.396.274.1
2025.07
72.5------
2025.07
71.8------
2025.07
68.365.84572.789.188.548.7
2025.07
68.3------
2025.07
66.5------
2025.07
64.9------
2025.07
64.8------
2025.07
64.7------
2026.02
64.376------
2025.07
64.360.945.659.687.389.343.1
2025.07
64.3------
2025.07
61.358.24071.684.479.434
2025.07
61.3------
2026.02
61.25------
2025.07
56.3------
2026.02
33.2------
2026.02
26.479------