List your company
ResearchFormerly WizWand

Machine Unlearning on ImageNet-1K All Classes

54.33Zero-Shot AccuracyOriginal
-1.892412.703827.341.8962May 14, 2026
Updated 5mo ago

Evaluation results

MethodLinks
OriginalBackbone=RN1012026.0554.33—
CLIP-LIPBackbone=RN502026.0553.78100
ICEDBackbone=RN1012026.0553.4298.33
OriginalBackbone=RN502026.0551.94—
CLIP-LIPBackbone=RN1012026.0551.4994.77
ICEDBackbone=RN502026.0550.4397.09
TIFSBackbone=RN1012026.0549.5191.13
TIFSBackbone=RN502026.0547.3491.14
GABackbone=RN1012026.0538.7571.32
GABackbone=RN502026.0535.3968.14
EMMNBackbone=RN502026.0532.2562.09
FTBackbone=RN1012026.0527.6250.84
EMMNBackbone=RN1012026.0526.4348.65
FTBackbone=RN502026.0524.847.75
LIPBackbone=RN1012026.051.743.2
LIPBackbone=RN502026.051.142.19
FisherBackbone=RN502026.051.092.1
FisherBackbone=RN1012026.050.270.5