ResearchBenchmarksImage Classification on ImageNet (unseen)For AI assistantsMarkdownContent indexImage Classification on ImageNet (unseen)Follow70.43AccuracyCoCoOp67.77868.466569.15569.8435Nov 26, 2023Evaluation resultsMethodMethodLinksAccuracySuccess RateCoCoOpBackbone=ViT-B/16, Sho...Backbone=ViT-B/16, Shots=162023.1170.43—BadCLIPBackbone=ViT-B/16, Sho...Backbone=ViT-B/16, Shots=16, Context length=4, Maximum noise strength (epsilon)=42023.1170.3399.4CLIPBackbone=ViT-B/16, mod...Backbone=ViT-B/16, mode=zero-shot2023.1168.14—CoOpBackbone=ViT-B/16, Sho...Backbone=ViT-B/16, Shots=162023.1167.88—