Image Classification on Flowers (CA, BA)
98.49Accuracy (BA)Attentive Probe
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Attentive ProbeLayers=All, Tokens=CLS + AP, Probing=attentive2026.01 | 98.49 | — | |
| AATLayers=Last, Tokens=all tokens, Probing=attentive2026.01 | 98.44 | — | |
| Linear ProbeLayers=Last, Tokens=CLS + AP, Probing=linear2026.01 | 98.43 | — | |
| Attentive ProbeLayers=Last, Tokens=CLS + AP, Probing=attentive2026.01 | 98.09 | — | |
| Linear ProbeLayers=Last, Tokens=CLS, Probing=linear2026.01 | 98.03 | — | |
| Linear ProbeLayers=All, Tokens=CLS + AP, Probing=linear2026.01 | 97.78 | — | |
| CLOPBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 97.18 | — | |
| SimCLRBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 97 | — | |
| FixMatchBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 96.69 | — | |
| SimMatchBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 96.46 | — | |
| SimMatch-V2Backbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 96.13 | — | |
| SupConBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 96 | — | |
| SsCLBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 95.74 | — | |
| CCSSLBackbone=ResNet-50 (4x), Pre-trained=ImageNet, Evaluation Protocol=Fine-tune2025.11 | 94.56 | — | |
| CorruptEncoderBackbone=ResNet-18, CL Algorithm=MoCo-v2, Evaluation Protocol=Linear evaluation2022.11 | 69.7 | — | |
| No AttackBackbone=ResNet-18, CL Algorithm=MoCo-v2, Evaluation Protocol=Linear evaluation2022.11 | — | 70.8 |