Monocular Spatial Understanding on OmniSpatial (test)
46.84AccuracyQwen2.5-VL-3B + SSL-R1
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen2.5-VL-3B + SSL-R1Model Scale=3B, Training Strategy=+ SSL-R12026.04 | 46.84 | |
| Qwen2.5-VL-3B + Visual JigsawModel Scale=3B, Training Strategy=+ Visual Jigsaw2026.04 | 46.12 | |
| Qwen2.5-VL-7B + SSL-R1Model Scale=7B, Training Strategy=+ SSL-R12026.04 | 45.4 | |
| Qwen2.5-VL-3BModel Scale=3B, Training Strategy=Baseline2026.04 | 45.27 | |
| Qwen2.5-VL-7B + Visual JigsawModel Scale=7B, Training Strategy=+ Visual Jigsaw2026.04 | 45.21 | |
| VL-Cogito-7BModel Scale=7B, Training Strategy=Representative Reasoning Model2026.04 | 44.29 | |
| LLaVA-Critic-R1-7BModel Scale=7B, Training Strategy=Representative Reasoning Model2026.04 | 42.73 | |
| Qwen2.5-VL-7BModel Scale=7B, Training Strategy=Baseline2026.04 | 42.66 | |
| ThinkLite-VL-7BModel Scale=7B, Training Strategy=Representative Reasoning Model2026.04 | 42.6 |