Information Analysis and Processing on MARS Level 2 (Evaluation Set)
2.2Average Ranking (AI GPT-5)Ours
Evaluation results
| Method | Links | ||
|---|---|---|---|
| OursParam.=1800B2025.11 | 2.2 | 1.5 | |
| Qwen3-235B-AParam.=22B2025.11 | 2.6 | 4 | |
| Llama-VisionParam.=11B2025.11 | 4.8 | 5.5 | |
| Llama-3.2-11BParam.=11B2025.11 | 5.8 | 6.7 | |
| Llama-3.2-90BParam.=90B2025.11 | 6.1 | 6.1 | |
| GPT-ossParam.=20B2025.11 | 6.7 | 5.3 | |
| Qwen-VLParam.=7B2025.11 | 6.78 | 7.9 | |
| Llama-4-ScoutParam.=17B2025.11 | 6.9 | 5.9 | |
| Qwen2-VLParam.=72B2025.11 | 7.3 | 7.9 | |
| Qwen2.5-VLParam.=72B2025.11 | 8.2 | 8.8 | |
| LLaVA-1.5Param.=7B2025.11 | 8.9 | 6.4 |