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

Reward Modeling on PKU-SafeRLHF (test)

0.0871MAE

ILDE

0.0781240.1387120.19930.259888Mar 19, 2026Mar 20, 2026Mar 21, 2026Mar 22, 2026Mar 23, 2026Mar 24, 2026
Updated 24d ago

Evaluation Results

MethodLinks
2026.03
0.0871-0.71040.2681
2026.03
0.1053-0.78720.2294
2026.03
0.1190.0570.77-
2026.03
0.1251-0.6280.3039
2026.03
0.1279-0.64710.296
2026.03
0.1282-0.67420.2844
2026.03
0.1290.0550.779-
2026.03
0.129-0.60430.3134
2026.03
0.1422-0.73460.2567
2026.03
0.1423-0.63890.2994
2026.03
0.1465-0.7540.2472
2026.03
0.1710.070.72-
2026.03
0.1771-0.69480.2753
2026.03
0.1871-0.50110.352
2026.03
0.1899-0.70930.2687
2026.03
0.1959-0.68760.2785
2026.03
0.2423-0.59140.2718
2026.03
0.2887-0.55350.333
2026.03
0.3115-0.52280.3442