| visual-question-answering-vqa-on-hr-bench-4k | DeepEyes | #2 | Overall: 75.1FSP: 91.3FCP: 59.0 |
| visual-question-answering-vqa-on-hr-bench-4k | Qwen2.5-VL-32B | #3 | Overall: 73.9FSP: 89.8FCP: 58.0 |
| visual-question-answering-vqa-on-hr-bench-4k | Pixel-Reasoner | #4 | Overall: 72.9FSP: 86.0FCP: 60.3 |
| visual-question-answering-vqa-on-hr-bench-4k | Qwen2.5-VL-7B | #5 | Overall: 68.8FSP: 85.2FCP: 52.2 |
| visual-question-answering-vqa-on-hr-bench-4k | LLaVA-OneVision-7B | #6 | Overall: 63.0FSP: 72.0FCP: 54.0 |
| visual-question-answering-vqa-on-hr-bench-8k | DeepEyes | #2 | Overall: 72.6FSP: 86.8FCP: 58.5 |
| visual-question-answering-vqa-on-hr-bench-8k | Qwen2.5-VL-32B | #3 | Overall: 70.4FSP: 84.5FCP: 56.3 |
| visual-question-answering-vqa-on-hr-bench-8k | Pixel-Reasoner | #4 | Overall: 66.9FSP: 80.0FCP: 54.3 |
| visual-question-answering-vqa-on-hr-bench-8k | Qwen2.5-VL-7B | #5 | Overall: 65.3FSP: 78.8FCP: 51.8 |
| visual-question-answering-vqa-on-hr-bench-8k | LLaVA-OneVision-7B | #6 | Overall: 59.8FSP: 67.3FCP: 52.3 |