Machine Unlearning on MUFAC (10% train split) benchmark leaderboard · SOTA2 Research

Machine Unlearning on MUFAC (10% train split)

0.175Average GapBadT

Evaluation Results

Method	Links
BadT2025.03		0.175	1.89	2.09
SCRUB2025.03		0.1625	10.53	1.2
SSD2025.03		0.16	10.3	1.07
NegGrad+2025.03		0.155	19.16	0.91
Finetuning2025.03		0.1525	19.52	0.76
UNSIR2025.03		0.145	16.32	1.68
SalUn2025.03		0.14	15.25	2.72
RndLbl2025.03		0.13	9.51	1.06
LoTUS2025.03		0.125	6.9	0.62
BadT2025.03		0.1025	9.41	0.66
NegGrad+2025.03		0.0475	0.39	1.67
UNSIR2025.03		0.0475	0.85	3.21
Finetuning2025.03		0.04	0.27	1.4
RndLbl2025.03		0.02	0.35	1.89
SCRUB2025.03		0.02	0.05	2.21
SSD2025.03		0.02	0.27	3.47
SalUn2025.03		0.02	0.29	8.15
LoTUS2025.03		0.02	0.05	1.09
Gold Std2025.03		0	0	13.83
Gold Std2025.03		0	0	7.34