Video Reasoning on VCR (test)
51AccuracyVIDEOP2R
Evaluation results
| Method | Links | |
|---|---|---|
| VIDEOP2RBase Model=Qwen2.5-VL-7B2025.11 | 51 | |
| VersaVid-R1Base Model=Qwen2.5-VL-7B2025.11 | 49.8 | |
| Time-R1Base Model=Qwen2.5-VL-7B2025.11 | 49.6 | |
| VideoRFTBase Model=Qwen2.5-VL-7B2025.11 | 49.6 | |
| Video-R1Base Model=Qwen2.5-VL-7B2025.11 | 49 | |
| VideoChat-R1Base Model=Qwen2.5-VL-7B2025.11 | 49 | |
| Qwen2.5-VL-7BModel Scale=7B2025.11 | 44.3 |