ResearchBenchmarksImage Classification on ImageNet-100 (40-20 split)For AI assistantsMarkdownContent indexImage Classification on ImageNet-100 (40-20 split)Follow82.7Accuracy (R)GS-LORA44.63654.51864.474.282Mar 18, 2024Evaluation resultsMethodMethodLinksAccuracy (R)H ScoreF1 ScoreAccuracy (Overall)GS-LORABackbone=ViTBackbone=ViT2024.0382.773.84.10MAS*Backbone=ViTBackbone=ViT2024.0380.262.219.88.8EWC*Backbone=ViTBackbone=ViT2024.03806220.19.1Pre-trainBackbone=ViTBackbone=ViT2024.0376.5—70.7—L2*Backbone=ViTBackbone=ViT2024.0374.752.729.923.5RetrainBackbone=ViTBackbone=ViT2024.0346.154.73.40