Visual Question Answering on HRBench 8K (FSP/FCP/Overall)
88.5FSPZoomEye
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ZoomEye2026.02 | 88.5 | 50 | 69.3 | |
| ConFoThinking (Qwen3-VL-8B)Backbone=Qwen3-VL-8B2026.02 | 87.3 | 68.3 | 77.8 | |
| Qwen3-VL-8BBackbone=Qwen3-VL-8B2026.02 | 86.3 | 58 | 72.1 | |
| ConFoThinking (Qwen2.5-VL-7B)Backbone=Qwen2.5-VL-7B2026.02 | 85 | 57 | 71 | |
| InternVL3.5-8BBackbone=InternVL3.5-8B2026.02 | 84.3 | 57.3 | 70.8 | |
| Qwen3-VL-8B (with tools)Backbone=Qwen3-VL-8B, Thinking Tools=true2026.02 | 83.8 | 66 | 74.9 | |
| ConFoThinking (Qwen3-VL-4B)Backbone=Qwen3-VL-4B2026.02 | 83.5 | 66.8 | 75.1 | |
| Qwen3-VL-4BBackbone=Qwen3-VL-4B2026.02 | 82.3 | 58 | 70.1 | |
| LLaVA-OneVision-1.5-8BBackbone=LLaVA-OneVision-1.5-8B2026.02 | 82 | 56.5 | 69.3 | |
| Qwen3-VL-4B (with tools)Backbone=Qwen3-VL-4B, Thinking Tools=true2026.02 | 81.5 | 61 | 71.3 | |
| Pixel-Reasoner2026.02 | 80.1 | 59.8 | 66.9 | |
| Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B2026.02 | 78.8 | 51.8 | 65.3 | |
| Visual Sketchpad2026.02 | 75.3 | 53 | 64.1 | |
| SEAL2026.02 | 70.1 | 47.5 | 58.8 | |
| ICoT2026.02 | 58 | 48.5 | 53.3 |