Visual Reasoning on SAT
76AccuracyQwen3-VL-32B
Evaluation Results
| Method | Links | |
|---|---|---|
| Qwen3-VL-32B3-run average=true2026.06 | 76 | |
| GPT-53-run average=true2026.06 | 73.3 | |
| Qwen3-VL-8B3-run average=true2026.06 | 54 | |
| Visual-OPSD3-run average=true, Latency (s)=10.02026.06 | 54 | |
| ThinkMorph3-run average=true, Latency (s)=142.82026.06 | 52.67 | |
| InternVL3.5-38B3-run average=true2026.06 | 49.33 | |
| Visual-OPSD-Noise3-run average=true, Latency (s)=14.92026.06 | 48 | |
| Text-only SFT3-run average=true, Latency (s)=28.52026.06 | 46.63 | |
| InternVL3.5-8B3-run average=true2026.06 | 45.33 | |
| BAGEL-7B3-run average=true2026.06 | 44.67 | |
| GPT-4o3-run average=true2026.06 | 28 | |
| Janus-Pro-7B3-run average=true2026.06 | 22 | |
| Chameleon-7B3-run average=true2026.06 | 10.67 |