Multimodal Hallucination Evaluation on MME
720Total ScoreCAI
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CAI2026.06 | 720 | — | — | — | — | |
| CAIGLA-gated2026.06 | 700 | — | — | — | — | |
| Active-LookModel=Qwen3-VL-8B2026.04 | 698.33 | 195 | 170 | 148.33 | 185 | |
| PAI2026.06 | 691.66 | — | — | — | — | |
| Qwen-3.52026.06 | 685 | — | — | — | — | |
| BaselineModel=Qwen3-VL-8B2026.04 | 673.33 | 190 | 165 | 143.33 | 175 | |
| ILVADBase Model=LLaVA-1.5-7B2026.05 | 641.66 | 195 | 153.33 | 133.33 | 160 | |
| VARBase Model=LLaVA-1.5-7B2026.05 | 636.66 | 185 | 143.33 | 143.33 | 165 | |
| GreedyBase Model=LLaVA-1.5-7B2026.05 | 631.66 | 195 | 158.33 | 123.33 | 155 | |
| SPARCBase Model=LLaVA-1.5-7B2026.05 | 629.99 | 195 | 158.33 | 123.33 | 153.33 | |
| VHRBase Model=LLaVA-1.5-7B2026.05 | 629.99 | 195 | 158.33 | 123.33 | 153.33 | |
| Active-LookModel=InternVL2-8B2026.04 | 626.66 | 185 | 130 | 131.66 | 180 | |
| ONLYBase Model=LLaVA-1.5-7B2026.05 | 626.66 | 190 | 150 | 121.66 | 165 | |
| VAFBase Model=LLaVA-1.5-7B2026.05 | 624.99 | 185 | 151.66 | 133.33 | 155 | |
| AGLABase Model=LLaVA-1.5-7B2026.05 | 621.66 | 190 | 153.33 | 118.33 | 160 | |
| VCDBase Model=LLaVA-1.5-7B2026.05 | 614.99 | 190 | 163.33 | 118.33 | 143.33 | |
| CODEBase Model=LLaVA-1.5-7B2026.05 | 609.99 | 195 | 131.66 | 133.33 | 150 | |
| BeamBase Model=LLaVA-1.5-7B2026.05 | 599.99 | 195 | 136.66 | 123.33 | 145 | |
| BaselineModel=InternVL2-8B2026.04 | 568.33 | 180 | 128.33 | 120 | 140 | |
| Active-LookModel=LLaVA-1.5-7B2026.04 | 516.66 | 168.33 | 121.66 | 116.66 | 110 | |
| OPERAModel=LLaVA-1.5-7B2026.04 | 497.33 | 168.33 | 110 | 113.66 | 105.33 | |
| VAFModel=LLaVA-1.5-7B2026.04 | 485 | 158.33 | 111.33 | 113.66 | 101.66 | |
| VCDModel=LLaVA-1.5-7B2026.04 | 446 | 150 | 98.33 | 93.33 | 104.33 | |
| BaselineModel=LLaVA-1.5-7B2026.04 | 431.33 | 158.33 | 91.33 | 90 | 91.66 |