Human Preferences on WildVision 0617
89.4ScoreGPT-4o
Evaluation Results
| Method | Links | |
|---|---|---|
| GPT-4o2024.12 | 89.4 | |
| LLaVA-OV-7BModel Scale=7B2024.12 | 53.8 | |
| Qwen2-VL-72BModel Scale=72B2024.12 | 52.3 | |
| LLaVA-OV-72BModel Scale=72B2024.12 | 52.3 | |
| MAmmoTH-VL-8BModel Scale=8B, Single-image variant (SI)=true2024.12 | 51.9 | |
| InternVL-2-8BModel Scale=8B2024.12 | 51.5 | |
| MAmmoTH-VL-8BModel Scale=8B2024.12 | 51.1 | |
| Claude-3.5-Sonnet2024.12 | 50 | |
| Llama-3.2-11B-Vision-InsModel Scale=11B2024.12 | 49.7 | |
| LLaVA-OV-72BModel Scale=72B, Single-image variant (SI)=true2024.12 | 49.5 | |
| Qwen2-VL-7B-InsModel Scale=7B2024.12 | 44 | |
| Molmo-7B-DModel Scale=7B2024.12 | 40 | |
| LLaVA-OV-7BModel Scale=7B, Single-image variant (SI)=true2024.12 | 39.2 | |
| MiniCPM-V-2.6-7BModel Scale=7B2024.12 | 11.7 |