Multimodal Question Answering on CRIT Natural Image 1.0 (test)
58.6EMQwen2.5-VL_CRIT
Evaluation results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-VL_CRITModel Parameters=7B, Evaluation Setting=CoT, Fine-tuned on CRIT=Yes2026.04 | 58.6 | 59.5 | |
| Idefics2_CRITModel Parameters=8B, Evaluation Setting=CoT, Fine-tuned on CRIT=Yes2026.04 | 54.1 | 54.9 | |
| Qwen2.5-VLModel Parameters=72B, Evaluation Setting=CoT2026.04 | 38 | 39.4 | |
| GPT-4oEvaluation Setting=CoT2026.04 | 35.1 | 37.7 | |
| LLaVA-OnevisionModel Parameters=7B, Evaluation Setting=CoT2026.04 | 31 | 32 | |
| Gemini 2.0 FlashEvaluation Setting=CoT2026.04 | 30.4 | 32.7 | |
| Qwen2.5-VLModel Parameters=7B, Evaluation Setting=CoT2026.04 | 28.3 | 29.1 | |
| Intern2.5-VLModel Parameters=8B, Evaluation Setting=CoT2026.04 | 26.7 | 27.9 | |
| GPT-4o-miniEvaluation Setting=CoT2026.04 | 24.7 | 25.5 | |
| Phi3.5-VisionModel Parameters=4B, Evaluation Setting=CoT2026.04 | 21.5 | 23 | |
| Idefics2Model Parameters=8B, Evaluation Setting=CoT2026.04 | 18.3 | 18.9 |