Multimodal Understanding on LLaVA-NeXT Multimodal Understanding Suite
83ChartQA AccuracyGenLIP
Evaluation Results
| Method | Links | |||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GenLIPPatches=729, Arch=So/16, Data=8.0B2026.05 | 83 | 76.9 | 69.6 | 64.7 | 50.4 | 79.1 | 78.1 | 54.5 | 80.1 | 89.4 | 65.1 | 53.2 | 70.3 | |
| RICE-ViTPatches=729, Arch=L/14, Data=13.0B2026.05 | 82.6 | 75.1 | 66.2 | 58.8 | 49.5 | 76.5 | 77.6 | 54.1 | 79 | 89.1 | 62.9 | 51.2 | 68.6 | |
| GenLIPPatches=576, Arch=So/16, Data=8.0B2026.05 | 79.3 | 75.2 | 68.5 | 59.7 | 48.4 | 78.6 | 77.7 | 48.6 | 78.2 | 89.2 | 65.9 | 53.1 | 68.5 | |
| RICE-ViTPatches=576, Arch=L/14, Data=13.0B2026.05 | 79.2 | 72.3 | 65.9 | 57.5 | 48.9 | 77.9 | 76.6 | 54.6 | 80.7 | 88.5 | 63.1 | 51.8 | 68.1 | |
| SigLIPv2Patches=729, Arch=So/14, Data=40.0B2026.05 | 79.1 | 70.2 | 66.2 | 58.7 | 48.6 | 77 | 77.1 | 46.6 | 80.4 | 89.3 | 63.4 | 52.8 | 67.5 | |
| AIMv2Patches=576, Arch=L/14, Data=12.8B2026.05 | 77.2 | 72.7 | 65.9 | 57.2 | 47.3 | 75.4 | 78.6 | 48.3 | 75 | 88.4 | 62.2 | 50.2 | 66.5 | |
| SigLIPPatches=729, Arch=So/14, Data=40.0B2026.05 | 76.7 | 69.3 | 64.7 | 55.4 | 48.4 | 76.2 | 77 | 46.1 | 79.9 | 88.8 | 63.7 | 47.3 | 66.1 | |
| MLCDPatches=576, Arch=L/14, Data=12.0B2026.05 | 76.5 | 67.8 | 61.7 | 53.1 | 48.4 | 77 | 76.5 | 54.1 | 79.9 | 88.7 | 61.1 | 51 | 66.3 | |
| CLIPPatches=576, Arch=L-14, Data=12.8B2026.05 | 75.2 | 66.5 | 62.5 | 52.5 | 47.4 | 73.2 | 74.6 | 48 | 75.6 | 88.8 | 63.7 | 49 | 64.8 |