Image Classification on Aggregate
94.74AccuracyθB fine-tune
Evaluation results
| Method | Links | ||
|---|---|---|---|
| θB fine-tuneTarget Backbone=ViT-L/14, Target Pre-training=Datacomp-XL, Evaluation Protocol=fine-tune2026.05 | 94.74 | 24.45 | |
| θA fine-tuneSource Backbone=ViT-B/16, Source Pre-training=LAION-2B, Evaluation Protocol=fine-tune2026.05 | 92.88 | 22.59 | |
| BiCoSource Backbone=ViT-B/16, Source Pre-training=LAION-2B, Target Backbone=ViT-L/14, Target Pre-training=Datacomp-XL, N=100, Evaluation Protocol=task-vector transfer2026.05 | 83.45 | 13.16 | |
| THESEUSSource Backbone=ViT-B/16, Source Pre-training=LAION-2B, Target Backbone=ViT-L/14, Target Pre-training=Datacomp-XL, N=100, Evaluation Protocol=task-vector transfer2026.05 | 73.88 | 3.59 | |
| θB + τA paddedSource Backbone=ViT-B/16, Source Pre-training=LAION-2B, Target Backbone=ViT-L/14, Target Pre-training=Datacomp-XL, Evaluation Protocol=task-vector transfer2026.05 | 72.96 | 2.67 | |
| θB zero-shotTarget Backbone=ViT-L/14, Target Pre-training=Datacomp-XL, Evaluation Protocol=zero-shot2026.05 | 70.29 | 0 |