Visual Reasoning on MME (test)
190ExistenceONLY
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| ONLYBackbone=LLaVA-1.52026.06 | 190 | 141.67 | 133.33 | 163.33 | 628.33 | |
| CAIBackbone=LLaVA-1.52026.06 | 190 | 143.33 | 148.33 | 178.33 | 660 | |
| RegularBackbone=LLaVA-1.52026.06 | 185 | 126.67 | 128.33 | 148.33 | 588.33 | |
| PAIBackbone=LLaVA-1.52026.06 | 185 | 131.67 | 133.33 | 153.33 | 603.33 | |
| VCDBackbone=LLaVA-1.52026.06 | 180 | 141.67 | 128.33 | 153.33 | 603.33 | |
| ONLYBackbone=Qwen-VL2026.06 | 180 | 136.67 | 153.33 | 180 | 650 | |
| CAIBackbone=Qwen-VL2026.06 | 180 | 146.67 | 163.33 | 185 | 675 | |
| CAIBackbone=InstructBLIP2026.06 | 175 | 100 | 70 | 165 | 510 | |
| PAIBackbone=Qwen-VL2026.06 | 175 | 135 | 163.33 | 175 | 648.33 | |
| RegularBackbone=InstructBLIP2026.06 | 170 | 75 | 68.33 | 140 | 453.33 | |
| ONLYBackbone=InstructBLIP2026.06 | 170 | 68.33 | 63.33 | 145 | 446.66 | |
| VCDBackbone=Qwen-VL2026.06 | 170 | 120 | 133.33 | 175 | 598.33 | |
| RegularBackbone=Qwen-VL2026.06 | 165 | 135 | 163.33 | 175 | 638.33 | |
| VCDBackbone=InstructBLIP2026.06 | 155 | 78.33 | 76.67 | 155 | 465 | |
| PAIBackbone=InstructBLIP2026.06 | 150 | 88.33 | 78.33 | 150 | 466.67 |