| chart-question-answering-on-charxiv-dq | GAS (4B) | – | Accuracy: 78.4 |
| chart-question-answering-on-charxiv-rq | GAS (4B) | – | Accuracy: 40.5 |
| logic-puzzle-solving-on-zebralogic | GAS (2B) | – | Accuracy: 20.8 |
| multimodal-reasoning-on-mathvista | GAS (2B after pretraining) | – | Accuracy: 71.2 |
| multimodal-reasoning-on-mmmu | GAS (4B) | – | Accuracy: 57.0 |
| multimodal-reasoning-on-visulogic | GAS (2B) | – | Accuracy: 28.6 |
| video-question-answering-on-mvbench | GAS (2B after pretraining) | – | Avg.: 54.6 |
| visual-question-answering-vqa-on-blink | GAS (4B) | – | Accuracy: 56.9 |
| visual-question-answering-vqa-on-countbenchqa | GAS (4B) | – | Accuracy: 90.8 |
| visual-question-answering-vqa-on-realworldqa | GAS (4B) | – | Accuracy: 70.3 |
| zero-shot-video-question-answer-on-video-mme-w-o-subs | GAS (4B) | – | Accuracy (%): 54.4 |