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

Model Alignment Evaluation on Re:Form SFT Evaluation Set (val)

94.3Validation Rate

14B SFT

-0.13224.38448.973.416Jul 22, 2025
Updated 19d ago

Evaluation Results

MethodLinks
2025.07
94.362.934.2
2025.07
90.853.327.9
2025.07
88.74826.6
2025.07
84.241.622.1
2025.07
80.133.618
2025.07
47.712.17
2025.07
17.63.70
2025.07
6.62.30.2
2025.07
5.92.50.4
2025.07
5.51.20
2025.07
3.51.60