Federated Unlearning on CIFAR-100 Non-IID benchmark leaderboard · SOTA2 Research

Federated Unlearning on CIFAR-100 Non-IID

56.09Accuracy (Group 95-100%)FedOSD

Evaluation Results

Method	Links
FedOSD2026.05		56.09	43.19	42.46	38.26	62.3	61.53	59.59	0.0881
FedAU2026.05		52.46	47.54	41.3	39.42	57.43	56.75	46.68	0.0719
Halimi et al.2026.05		50.87	43.04	45.8	46.23	64.16	64.08	61.92	0.0871
NoT2026.05		46.96	40.58	45.94	40.14	64.74	63.4	57.06	0.0501
Liu et al.2026.05		43.33	36.09	34.35	33.48	64.07	61.73	58.02	0.055
FedRecovery2026.05		42.9	42.17	42.61	42.17	61.72	60.93	48.02	0.111
FedMemPrune2026.05		27.25	27.39	29.71	30.43	63.44	60.37	62.99	0.0428
Retrained Baseline2026.05		21.74	22.9	26.09	29.13	65.28	61.3	63.89	0.0239