Image Classification on 8 computer vision tasks
95.9Average AccuracyFine-tuned
Evaluation results
| Method | Links | |
|---|---|---|
| Fine-tunedBackbone=ViT-L/142026.06 | 95.9 | |
| TSV-C + SiMBackbone=ViT-L/142026.06 | 95.4 | |
| EMR + SiMBackbone=ViT-L/142026.06 | 95 | |
| TM-TIES + SiMBackbone=ViT-L/142026.06 | 95 | |
| MoW-MergingBackbone=ViT-L/142026.06 | 94.9 | |
| WEMoEBackbone=ViT-L/142026.06 | 94.8 | |
| Fine-tunedBackbone=ViT-B/162026.06 | 94.7 | |
| TM-TIES + SiMBackbone=ViT-B/162026.06 | 94.3 | |
| TM-TA + SiMBackbone=ViT-B/162026.06 | 94 | |
| TSV-C + SiMBackbone=ViT-B/162026.06 | 94 | |
| TWIN-MergingBackbone=ViT-L/142026.06 | 93.7 | |
| MoW-MergingBackbone=ViT-B/162026.06 | 93.7 | |
| Iso-CBackbone=ViT-L/142026.06 | 93.3 | |
| WEMoEBackbone=ViT-B/162026.06 | 93.1 | |
| TSV-MBackbone=ViT-L/142026.06 | 93 | |
| EMR + SiMBackbone=ViT-B/162026.06 | 92.9 | |
| Fine-tunedBackbone=ViT-B/322026.06 | 92.8 | |
| TM-TIES + SiMBackbone=ViT-B/322026.06 | 92.7 | |
| TM-TA + SiMBackbone=ViT-B/322026.06 | 92.3 | |
| TSV-C + SiMBackbone=ViT-B/322026.06 | 92.1 | |
| TM-TA + SiMBackbone=ViT-L/142026.06 | 92 | |
| DaWinBackbone=ViT-L/142026.06 | 91.6 | |
| TWIN-MergingBackbone=ViT-B/162026.06 | 91.4 | |
| EMR + SiMBackbone=ViT-B/322026.06 | 90.7 | |
| WEMoEBackbone=ViT-B/322026.06 | 90.4 | |
| Iso-CBackbone=ViT-B/162026.06 | 89.2 | |
| TSV-MBackbone=ViT-B/162026.06 | 89 | |
| DaWinBackbone=ViT-B/322026.06 | 89 | |
| MoW-MergingBackbone=ViT-B/322026.06 | 88.1 | |
| DaWinBackbone=ViT-B/162026.06 | 87.1 | |
| TIES-MergingBackbone=ViT-L/142026.06 | 86.9 | |
| TSV-MBackbone=ViT-B/322026.06 | 85.7 | |
| Task ArithmeticBackbone=ViT-L/142026.06 | 84.9 | |
| Iso-CBackbone=ViT-B/322026.06 | 84.7 | |
| TWIN-MergingBackbone=ViT-B/322026.06 | 84 | |
| TIES-MergingBackbone=ViT-B/162026.06 | 79.7 | |
| Task ArithmeticBackbone=ViT-B/162026.06 | 75.4 | |
| TIES-MergingBackbone=ViT-B/322026.06 | 75.1 | |
| Task ArithmeticBackbone=ViT-B/322026.06 | 70.8 |