Visual Grounded Reasoning on TRIG
38.69Chart ScorePreciseDoc-Reasoner
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| PreciseDoc-Reasoner2026.06 | 38.69 | 49.81 | 29.66 | 67.5 | 46.42 | |
| Gemini-3-ProVersion=3, Size=Pro2026.06 | 18.86 | 30.94 | 20.82 | 50.42 | 30.26 | |
| Qwen3VL-30B-a3b-thinkingParameters=30B, Reasoning Mode=a3b-thinking2026.06 | 12.05 | 39.07 | 15.49 | 68.34 | 33.74 | |
| Gemini-3-FlashVersion=3, Size=Flash2026.06 | 11.81 | 22.35 | 14.7 | 37.03 | 21.47 | |
| Qwen3VL-8BParameters=8B2026.06 | 8.27 | 31.19 | 9.3 | 65.16 | 28.48 | |
| GPT-4o2026.06 | 3.9 | 1.79 | 1.6 | 13.73 | 5.26 | |
| GPT-5.2Version=5.22026.06 | 3.09 | 1.87 | 2.17 | 11.98 | 4.78 | |
| Qwen3.5-27BParameters=27B2026.06 | 2.89 | 14.77 | 5.41 | 57.04 | 20.03 | |
| InternLM-XComposer2-4KHD-7BResolution=4KHD, Parameters=7B2026.06 | 1.04 | 0.1 | 0.9 | 0.14 | 0.55 | |
| Monkey-Chat2026.06 | 0.77 | 0.19 | 0.15 | 0.45 | 0.39 | |
| Qwen3.5-9BParameters=9B2026.06 | 0.73 | 20.43 | 5.08 | 44.64 | 17.72 | |
| InternVL3-8BParameters=8B2026.06 | 0.72 | 17.61 | 2.02 | 21.97 | 10.58 | |
| MiniCPM-Llama3-V 2.5Backbone=Llama32026.06 | 0.44 | 1.4 | 0.65 | 4.96 | 1.86 | |
| InternLM-XComposer2-VL-7BParameters=7B2026.06 | 0.15 | 0.2 | 0.13 | 0.57 | 0.26 | |
| LLaVA-v1.6-Vicuna-13BBackbone=Vicuna, Parameters=13B2026.06 | 0 | 0 | 0 | 0 | 0 | |
| Qwen2-VL-7BParameters=7B2026.06 | 0 | 0 | 0 | 0 | 0 | |
| Qwen2.5-VL-7BParameters=7B2026.06 | 0 | 0 | 0 | 0 | 0 | |
| InternVL2-8BParameters=8B2026.06 | 0 | 0 | 0 | 0 | 0 | |
| InternVL2.5-8BParameters=8B2026.06 | 0 | 0 | 0 | 0 | 0 | |
| Gemini-2.0-FlashVersion=2.0, Size=Flash2026.06 | 0 | 0 | 0 | 0 | 0 | |
| Gemini-2.5-FlashVersion=2.5, Size=Flash2026.06 | 0 | 0 | 0 | 0 | 0 |