| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| MMVet turbo | D2Djus | Overall Score69.7 | 16 | 6d ago | |
| MMMU (val) | GPT-4o | Accuracy69.1 | 14 | 6d ago | |
| Macro-average of HallusionBench, AMBER, CRPE, R-Bench, and BLINK | IC-VCO | Overall Score63.35 | 13 | 2d ago | |
| LLaVA-Bench Wild | TroL-7B | Relative Score92.8 | 12 | 3mo ago | |
| MMBench CN (dev) | LLaVA-1.5 + FG-CLIP 2 | Accuracy60.5 | 4 | 8d ago | |
| MMBench EN (dev) | LLaVA-1.5 + FG-CLIP 2 | Accuracy67.6 | 4 | 8d ago |