Machine Unlearning on CIFAR-10 (100 In-Class Random Forgetting) benchmark leaderboard · SOTA2 Research

Machine Unlearning on CIFAR-10 (100 In-Class Random Forgetting)

99.8Retention Accuracy (RA)SalUn

Evaluation Results

Method	Links
SalUn2026.04		99.8	0.33	0.12	1	0.14	0.1	0.79
Retrained2026.04		99.56	92.02	0.37	50	—	—	—
Fisher2026.04		99.32	3.81	0.1	2	1.07	0.1	0.78
CF-k2026.04		99.03	6.95	0.33	5	0.37	0.1	0.79
GA2026.04		99.02	6.94	0.33	4	0	0.1	0.91
EU-k2026.04		99.02	6.96	0.33	5	0.37	0.1	0.78
FT2026.04		98.72	3.5	0.23	2	0.27	0.08	0.65
RELOAD2026.04		98.57	1.88	0.14	1	0.15	0.1	0.57
SCRUB2026.04		97.31	5.79	0.14	4	0.03	1.37	1.75
SSD2026.04		9.99	81.88	2.12	1	0.01	10.88	10.25