Visual Instruction Following on RefCOCO
47.8AccuracyLIVE
Evaluation Results
| Method | Links | |
|---|---|---|
| LIVEViT Model=SigLip 2 So400m, Evaluation Protocol=Zero-Shot2026.06 | 47.8 | |
| LIVEViT Model=SigLip B/16, Evaluation Protocol=Zero-Shot2026.06 | 47.24 | |
| LIVEViT Model=SigLip 2 B/16, Evaluation Protocol=Zero-Shot2026.06 | 45.42 | |
| LIVEViT Model=SigLip T/14, Evaluation Protocol=Zero-Shot2026.06 | 33.52 | |
| Menon et al.ViT Model=SigLip 2 B/16, Evaluation Protocol=Zero-Shot2026.06 | 9.72 | |
| SigLipViT Model=SigLip 2 B/16, Evaluation Protocol=Zero-Shot2026.06 | 7.75 | |
| FusionViT Model=SigLip 2 B/16, Evaluation Protocol=Zero-Shot2026.06 | 7.75 | |
| Menon et al.ViT Model=SigLip 2 So400m, Evaluation Protocol=Zero-Shot2026.06 | 7.65 | |
| Menon et al.ViT Model=SigLip B/16, Evaluation Protocol=Zero-Shot2026.06 | 7.09 | |
| FusionViT Model=SigLip B/16, Evaluation Protocol=Zero-Shot2026.06 | 6.94 | |
| SigLipViT Model=SigLip B/16, Evaluation Protocol=Zero-Shot2026.06 | 6.93 | |
| FusionViT Model=SigLip 2 So400m, Evaluation Protocol=Zero-Shot2026.06 | 6.68 | |
| SigLipViT Model=SigLip 2 So400m, Evaluation Protocol=Zero-Shot2026.06 | 6.63 | |
| SigLipViT Model=SigLip T/14, Evaluation Protocol=Zero-Shot2026.06 | 5.72 | |
| FusionViT Model=SigLip T/14, Evaluation Protocol=Zero-Shot2026.06 | 5.72 | |
| Menon et al.ViT Model=SigLip T/14, Evaluation Protocol=Zero-Shot2026.06 | 5.72 |