Visual Language Model Evaluation on MM-Bench CN
57.5MMB (CN) ScoreSPpruner
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SPprunerModel=LLaVA-Next-7B, Token Budget=320, Reduction Ratio=88.9%2026.05 | 57.5 | 97.7 | |
| VanillaModel=LLaVA-Next-7B, Token Budget=2880, Reduction Ratio=0%2026.05 | 57.3 | 100 | |
| DARTModel=LLaVA-Next-7B, Token Budget=320, Reduction Ratio=88.9%2026.05 | 56.7 | 96.4 | |
| VisionZipModel=LLaVA-Next-7B, Token Budget=320, Reduction Ratio=88.9%2026.05 | 55.7 | 96.2 | |
| DivPruneModel=LLaVA-Next-7B, Token Budget=320, Reduction Ratio=88.9%2026.05 | 55.7 | 96.1 | |
| FastVModel=LLaVA-Next-7B, Token Budget=320, Reduction Ratio=88.9%2026.05 | 51.9 | 86.8 | |
| PyramidDropModel=LLaVA-Next-7B, Token Budget=320, Reduction Ratio=88.9%2026.05 | 49.5 | 90.1 |