List your company
ResearchFormerly WizWand

Image Classification on ImageNet-S (test)

59.07Top-1 AccSS-CA
9.430822.317935.20548.0921May 29, 2023Dec 2, 2023Jun 6, 2024Dec 10, 2024Jun 15, 2025Dec 19, 2025Jun 24, 2026
Updated 4mo ago

Evaluation results

MethodLinks
SS-CAModel=CLIP (ViT /32b)2025.1159.07
Chen et al.Model=CLIP (ViT /32b)2025.1158.18
Xiao et al.Model=CLIP (ViT /32b)2025.1158.1
Conventional TrainingModel=CLIP (ViT /32b)2025.1157.56
GPUAEvaluation Protocol=Zero-shot2026.0656.9
SS-CABackbone=CLIP (ViT/32b), Evaluation Protocol=Linear Probing2025.1155.77
Chen et al.Backbone=CLIP (ViT/32b), Evaluation Protocol=Linear Probing2025.1155.12
Conventional TrainingBackbone=CLIP (ViT/32b), Evaluation Protocol=Linear Probing2025.1154.99
Xiao et al.Backbone=CLIP (ViT/32b), Evaluation Protocol=Linear Probing2025.1154.78
TACO2026.0650.52
WISE-FT2026.0649.72
SS-CAModel=ViT2025.1149.71
PromptSRC2026.0649.55
MMRL2026.0649.17
MaPLe2026.0649.15
MMA2026.0649.13
CoCoOp2026.0648.75
PromptSRC*2025.1148.6
MaPLe + ReBaPL2025.1148.5
MMRL + ReBaPL2025.1148.4
Chen et al.Model=ViT2025.1148.14
Xiao et al.Model=ViT2025.1148.12
Conventional TrainingModel=ViT2025.1147.81
MMRL*2025.1147.8
MaPLe*2025.1147.7
SS-CAModel=ResNet-1012025.1146.45
CLIPEvaluation Protocol=Zero-shot2026.0646.1
Chen et al.Model=ResNet-1012025.1145.99
Xiao et al.Model=ResNet-1012025.1145.94
Conventional TrainingModel=ResNet-1012025.1145.76
LaFTerBackbone=ViT-B/322023.0542.7
UPLBackbone=ViT-B/322023.0542.4
TRUST2025.0941.5
TRUST naive2025.0941.1
CLIPBackbone=ViT-B/322023.0540.6
CLIP-PRBackbone=ViT-B/322023.0538.6
SAR2025.0932.6
SHOT2025.0932.6
Tent2025.0932.5
SS-CABackbone=ViT, Evaluation Protocol=End-to-End Fine-Tuning2025.1131.56
Source only2025.0931.4
ETA2025.0931.4
LAME2025.0931.4
Chen et al.Backbone=ViT, Evaluation Protocol=End-to-End Fine-Tuning2025.1131.05
Xiao et al.Backbone=ViT, Evaluation Protocol=End-to-End Fine-Tuning2025.1130.92
Conventional TrainingBackbone=ViT, Evaluation Protocol=End-to-End Fine-Tuning2025.1130.9
SS-CABackbone=ResNet-101, Evaluation Protocol=End-to-End Fine-Tuning2025.1112.15
Chen et al.Backbone=ResNet-101, Evaluation Protocol=End-to-End Fine-Tuning2025.1111.54
Conventional TrainingBackbone=ResNet-101, Evaluation Protocol=End-to-End Fine-Tuning2025.1111.34
Xiao et al.Backbone=ResNet-101, Evaluation Protocol=End-to-End Fine-Tuning2025.1111.34