List your company
ResearchFormerly WizWand

Reward Modeling on UltraFeedback Cleaned

92.36Total ScoreVRM
53.235263.392673.5583.7074Mar 5, 2026
Updated 8mo ago

Evaluation results

MethodLinks
VRM2026.0392.36
RM2026.0388.98
SELECTIVE DPO2026.0382.28
DPO2026.0375.71
KTO2026.0370.67
IPO2026.0360.69
SIMPO2026.0359.16
WPO2026.0354.74