Video Reasoning on MMVU (test)
68.5AccuracyVideoRFT
Evaluation Results
| Method | Links | |
|---|---|---|
| VideoRFTBase Model=Qwen2.5-VL-7B2025.11 | 68.5 | |
| VIDEOP2RBase Model=Qwen2.5-VL-7B2025.11 | 65.4 | |
| VersaVid-R1Base Model=Qwen2.5-VL-7B2025.11 | 64.3 | |
| Video-R1Base Model=Qwen2.5-VL-7B2025.11 | 63.8 | |
| VideoChat-R1Base Model=Qwen2.5-VL-7B2025.11 | 63 | |
| Time-R1Base Model=Qwen2.5-VL-7B2025.11 | 62.9 | |
| Qwen2.5-VL-7BModel Scale=7B2025.11 | 60 | |
| LLaVA-OneVision-7BModel Scale=7B2025.11 | 49.2 | |
| VideoLLaMA2-7BModel Scale=7B2025.11 | 44.8 |