Paper Folding on Paper Folding
16.8AccuracyGemini-2.5-Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini-2.5-ProModel Type=Commercial2026.06 | 16.8 | |
| PERIA-8BModel Type=Proposed2026.06 | 15.7 | |
| VTool-R1-7BModel Type=Tool-Augmented2026.06 | 15.2 | |
| w/o ToolModel Type=Ablation2026.06 | 15 | |
| w/o RLModel Type=Ablation2026.06 | 14.6 | |
| InternVL3.5-8BModel Type=Open-Source2026.06 | 13.5 | |
| R1-OneVision-7BModel Type=Tool-Augmented2026.06 | 12.9 | |
| Mini-o3Model Type=Tool-Augmented2026.06 | 12.9 | |
| GPT-5Model Type=Commercial2026.06 | 11 | |
| InternVL3.5-14BModel Type=Open-Source2026.06 | 11 | |
| Qwen3-8BModel Type=Backbone Baseline2026.06 | 11 | |
| Qwen3-30BModel Type=Open-Source2026.06 | 8.5 | |
| Qwen3-32BModel Type=Open-Source2026.06 | 8.1 | |
| Gemini-2.5-FlashModel Type=Commercial2026.06 | 7.5 | |
| Qwen3-235BModel Type=Open-Source2026.06 | 6 | |
| w/o SFTModel Type=Ablation2026.06 | 6 |