Visual Spatial Reasoning on VSP
85.83AccuracyVisual-OPSD
Evaluation Results
| Method | Links | |
|---|---|---|
| Visual-OPSD3-run average=true, Latency (s)=10.02026.06 | 85.83 | |
| ThinkMorph3-run average=true, Latency (s)=142.82026.06 | 75.83 | |
| GPT-53-run average=true2026.06 | 57.33 | |
| Visual-OPSD-Noise3-run average=true, Latency (s)=14.92026.06 | 50.83 | |
| Text-only SFT3-run average=true, Latency (s)=28.52026.06 | 49.17 | |
| Qwen3-VL-32B3-run average=true2026.06 | 43.5 | |
| GPT-4o3-run average=true2026.06 | 33.5 | |
| Qwen3-VL-8B3-run average=true2026.06 | 22 | |
| InternVL3.5-38B3-run average=true2026.06 | 20.16 | |
| InternVL3.5-8B3-run average=true2026.06 | 8.17 | |
| Chameleon-7B3-run average=true2026.06 | 0.83 | |
| BAGEL-7B3-run average=true2026.06 | 0.83 | |
| Janus-Pro-7B3-run average=true2026.06 | 0 |