2D Spatial Reasoning on CV-Bench (full)
78.2AccuracyArgos
Evaluation Results
| Method | Links | |
|---|---|---|
| Argos2025.12 | 78.2 | |
| Qwen2.5VL 7B2025.12 | 77 | |
| Qwen2.5VL 7BPrompting Strategy=CoT2025.12 | 75.6 | |
| Video-R1Training Protocol=SFT2025.12 | 75.1 | |
| Video-R1Training Protocol=RL2025.12 | 60.2 |