Multimodal Assistant Evaluation on LLaVA-Bench GPT-4V-aided (full)
6.96AccuracyLTS-FS(Nullu)
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LTS-FS(Nullu)Model=LLaVA-1.52026.03 | 6.96 | 6.23 | |
| LTS-FS(Nullu)Model=Qwen-vl2.52026.03 | 6.59 | 6.07 | |
| NulluModel=LLaVA-1.52026.03 | 6.46 | 5.51 | |
| NulluModel=Qwen-vl2.52026.03 | 6.37 | 5.68 | |
| OriginalModel=Qwen-vl2.52026.03 | 6.06 | 5.54 | |
| OriginalModel=LLaVA-1.52026.03 | 5.74 | 5.23 |