List your company

Reward Modeling on PKU-SafeRLHF

0.985R2@50UARM
0.84460.881050.91750.95395Jun 18, 2026

Evaluation results

MethodLinks
UARMMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.9850.0130.016
MCNFMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.9550.0420.059
ClearMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.940.060.096
TorchNautMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.9330.050.052
SCCPMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.9250.0640.09
Packed EnsembleMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.9050.1030.147
ACIMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.9050.1130.256
WCPMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.8830.1410.192
Deep EnsembleMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.8810.110.167
DERMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.8810.0980.091
CQRMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.8810.1470.305
SCPMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.8660.1590.2
MC-DropoutMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.8630.0730.135
NaiveMiscoverage rate (alpha)=0.1, Confidence threshold=50% most confident samples2026.060.850.1730.206