Sub-class Machine Unlearning on CIFAR-20 benchmark leaderboard · SOTA2 Research

Sub-class Machine Unlearning on CIFAR-20

83.1ArMeGU

Evaluation Results

Method	Links
MeGU2026.02		83.1	50.27	4.33
baseline2026.02		83	69.88	84.5
baseline2026.02		82.96	72.4	82.46
baseline2026.02		82.9	77.08	85.86
AMNC2026.02		82.8	9.9	10.05
baseline2026.02		82.8	90.36	91.49
SSD2026.02		82.77	6.68	4.8
MeGU2026.02		82.74	72.55	26.71
SSD2026.02		82.67	72.57	61.42
baseline2026.02		82.63	82.64	88.43
retrain2026.02		82.4	6.08	5.6
MeGU2026.02		82.37	70.4	6.26
retrain2026.02		82.35	67.71	20.26
AMNC2026.02		82.35	5.12	4.28
MeGU2026.02		82.32	13.12	3.7
MeGU2026.02		82.29	69.14	3.94
AMNC2026.02		82.2	40.97	9.4
retrain2026.02		82.09	67.27	36.47
AMNC2026.02		82.08	20.92	3.87
retrain2026.02		82.08	67.53	16.2
UNSIR2026.02		82.06	56.68	35.43
UNSIR2026.02		82	76.13	71
AMNC2026.02		81.98	4.34	14.61
retrain2026.02		81.97	37.59	9.47
UNSIR2026.02		81.66	79.08	77.8
SSD2026.02		81.62	7.46	5.45
UNSIR2026.02		81.39	82.12	56.02
UNSIR2026.02		80.98	55.3	47.22
SSD2026.02		80.1	0.78	8.81
SSD2026.02		79.9	0	21.8
FT2026.02		75.49	16.32	16.57
FT2026.02		74.81	61.37	54.39
FT2026.02		74	61.02	33
FT2026.02		73.23	72.05	58.64
FT2026.02		71.5	17.71	15.66