General Utility on MMMU
69.1MMMU ScoreAOD
Evaluation Results
| Method | Links | |
|---|---|---|
| AODModel=InternVL3-8B2026.05 | 69.1 | |
| PruneHalModel=InternVL3-8B2026.05 | 68.9 | |
| TruthPrIntModel=InternVL3-8B2026.05 | 68 | |
| ASDModel=InternVL3-8B2026.05 | 66.5 | |
| VCDModel=InternVL3-8B2026.05 | 66.1 | |
| VASparseModel=InternVL3-8B2026.05 | 65.8 | |
| BaseModel=InternVL3-8B2026.05 | 65 | |
| AODModel=Qwen2.5-VL-7B2026.05 | 57.6 | |
| TruthPrIntModel=Qwen2.5-VL-7B2026.05 | 54.9 | |
| ASDModel=Qwen2.5-VL-7B2026.05 | 54 | |
| BaseModel=Qwen2.5-VL-7B2026.05 | 51.9 | |
| PruneHalModel=Qwen2.5-VL-7B2026.05 | 50.7 | |
| VASparseModel=Qwen2.5-VL-7B2026.05 | 50.2 | |
| VCDModel=Qwen2.5-VL-7B2026.05 | 49.8 | |
| AODModel=LLaVA-1.5-7B2026.05 | 38.3 | |
| TruthPrIntModel=LLaVA-1.5-7B2026.05 | 37.4 | |
| ASDModel=LLaVA-1.5-7B2026.05 | 36.7 | |
| BaseModel=LLaVA-1.5-7B2026.05 | 34.5 | |
| VCDModel=LLaVA-1.5-7B2026.05 | 34 | |
| PruneHalModel=LLaVA-1.5-7B2026.05 | 33.6 | |
| VASparseModel=LLaVA-1.5-7B2026.05 | 33.5 |