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

Reward Model Transfer on PKU-SafeRLHF

2.46AOG

ANCHOR

0.5881.0741.562.046May 25, 2026
Updated 8d ago

Evaluation Results

MethodLinks
2026.05
2.46--2.46
2026.05
2.38--2.38
2026.05
2.37--2.37
2026.05
2.05--2.05
2026.05
2.04--2.04
2026.05
1.97--1.97
2026.05
1.72--1.21
2026.05
0.66--0.66
2026.05
-146.14.91-
2026.05
-163.45.49-
2026.05
-10.82-
2026.05
-2.72.21-