Multimodal Reasoning on Mantis-Eval
59.23AccuracyMantis
Evaluation Results
| Method | Links | |
|---|---|---|
| MantisSOFA=true2025.03 | 59.23 | |
| MantisSOFA=false2025.03 | 56.85 | |
| LLaVA-NeXTSOFA=true2025.03 | 54.51 | |
| VILASOFA=false2025.03 | 52.64 | |
| VILASOFA=true2025.03 | 52.13 | |
| InternVL2SOFA=true2025.03 | 51.11 | |
| LLaVA-NeXTSOFA=false2025.03 | 50.83 | |
| Idefics2SOFA=true2025.03 | 50.77 | |
| InternVL2SOFA=false2025.03 | 50.3 | |
| Idefics2SOFA=false2025.03 | 48.41 | |
| Random Guessing2025.03 | 25.31 |