List your company
ResearchFormerly WizWand

Machine Unlearning on ImageNet-1K Target Class

80Zero-Shot AccuracyOriginal
-3.218.44061.6May 14, 2026
Updated 5mo ago

Evaluation results

MethodLinks
OriginalBackbone=RN1012026.0580—
OriginalBackbone=RN502026.0568—
EMMNBackbone=RN1012026.056885
CLIP-LIPBackbone=RN502026.053754.41
GABackbone=RN1012026.053240
GABackbone=RN502026.0514.521.32
EMMNBackbone=RN502026.051014.71
FTBackbone=RN502026.0545.88
CLIP-LIPBackbone=RN1012026.0522.5
ICEDBackbone=RN1012026.0522.5
ICEDBackbone=RN502026.051.21.76
FisherBackbone=RN1012026.051.121.4
LIPBackbone=RN502026.050.951.4
LIPBackbone=RN1012026.050.240.3
FisherBackbone=RN502026.050.20.29
TIFSBackbone=RN502026.0500
FTBackbone=RN1012026.0500
TIFSBackbone=RN1012026.0500