Commonsense Reasoning on WorldSense
42.6Overall ScoreGPT-4o
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| GPT-4oModel=GPT-4o2025.10 | 42.6 | — | |
| InternVL-2.5-8BModel=InternVL-2.5-8B2025.10 | 39.6 | 38.5 | |
| VideoRFT-7BModel=VideoRFT-7B2025.10 | 38.2 | 36.6 | |
| Open-o3-Video-7B (Ours)Model=Open-o3-Video-7B (Ours)2025.10 | 37.5 | 36.8 | |
| VideoLLaMA3-7BModel=VideoLLaMA3-7B2025.10 | 37.3 | 38.1 | |
| Qwen2.5-VL-7B (Base)Model=Qwen2.5-VL-7B (Base)2025.10 | 36.1 | 33.7 | |
| VideoR1-7BModel=VideoR1-7B2025.10 | 35.5 | 32.8 |