Overall Performance Comparison on Aggregate Performance
100.2RelVisNec
Evaluation results
| Method | Links | |
|---|---|---|
| VisNecSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 100.2 | |
| Full-Data(665K)Sampling Ratio=100% (665K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 100 | |
| PreSelSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 97.7 | |
| CoIDOSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 97.7 | |
| XMASSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 97.3 | |
| ICONSSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 97.1 | |
| TypiClustSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 96.9 | |
| COINCIDESampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 95.8 | |
| RandomSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 94.2 | |
| IFDSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 92.1 | |
| CLIP-ScoreSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 92 | |
| EL2NSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 91.9 | |
| Self-FilterSampling Ratio=15% (98K samples), Base Model=LLaVA-v1.5-7B, Training Dataset=LLaVA-665K2026.03 | 89.3 |