| Benchmark | Model | Rank | Results |
|---|---|---|---|
| vision-language-understanding-on-mmstar | Scaffolding Encoder+RL | – | Accuracy: 73.5 |
| visual-question-answering-on-v-bench | Scaffolding Encoder+RL | – | Accuracy: 90.6 |
| visual-question-answering-vqa-on-blink | Scaffolding Encoder+RL | – | Accuracy: 67.2 |