| Benchmark | Model | Rank | Results |
|---|---|---|---|
| visual-reasoning-on-vsr | LXMERT | #1 | accuracy: 70.1 |
| visual-reasoning-on-vsr | ViLT | #2 | accuracy: 69.3 |
| visual-reasoning-on-vsr | CLIP (finetuned) | #3 | accuracy: 65.1 |
| visual-reasoning-on-vsr | CLIP (frozen) | #4 | accuracy: 56.0 |
| visual-reasoning-on-vsr | VisualBERT | #5 | accuracy: 55.2 |