Large Vision-Language Model Evaluation on MME
1,711.44Perception ScoreHTDC
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| HTDCModel=Qwen3-VL-4B2026.04 | 1,711.44 | — | 610 | |
| RegularModel=Qwen3-VL-4B2026.04 | 1,701.93 | — | 630.71 | |
| VCDModel=Qwen3-VL-4B2026.04 | 1,695.82 | — | 607.86 | |
| DAMOModel=Qwen3-VL-4B2026.04 | 1,663.72 | — | 628.57 | |
| ICDModel=Qwen3-VL-4B2026.04 | 1,660.76 | — | 597.14 | |
| DoLaModel=Qwen3-VL-4B2026.04 | 1,640.93 | — | 601.79 | |
| HTDCModel=LLaVA-v1.52026.04 | 1,515.89 | — | 367.5 | |
| DAMOModel=LLaVA-v1.52026.04 | 1,512.01 | — | 297.86 | |
| RegularModel=LLaVA-v1.52026.04 | 1,508.38 | — | 357.86 | |
| VCDModel=LLaVA-v1.52026.04 | 1,477.74 | — | 357.86 | |
| ICDModel=LLaVA-v1.52026.04 | 1,441.19 | — | 333.93 | |
| DoLaModel=LLaVA-v1.52026.04 | 1,412.13 | — | 298.57 | |
| CNSbase_method=Vanilla2026.01 | — | 1,883.57 | — | |
| CNSbase_method=VCD2026.01 | — | 1,892.68 | — | |
| Vanilla2026.01 | — | 1,864.68 | — | |
| VCD2026.01 | — | 1,873.82 | — |