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

Safety Unlearning on SAFEERASER (test)

4.9RA (ASR)

KL

-0.1961.1272.453.773Feb 18, 2025
Updated 1mo ago

Evaluation Results

MethodLinks
2025.02
4.93.84.23.7
2025.02
3.53.44.43.3
2025.02
1.72.12.90.2
2025.02
00.10.50