Sequential Planning on VSP-Super
99Success Rate (Length 16)DiffThinker++
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DiffThinker++Setting=Flow Matching2025.12 | 99 | 80 | |
| DiffThinkerSetting=Flow Matching2025.12 | 96 | 83 | |
| Qwen3-VL-32BSetting=SFT2025.12 | 85 | 21 | |
| Qwen3-VL-8BSetting=SFT2025.12 | 61 | 8 | |
| Gemini-3-FlashSetting=N/A2025.12 | 52 | 3 | |
| GPT-5Setting=N/A2025.12 | 3 | 0 | |
| Qwen3-VL-8BSetting=N/A2025.12 | 1 | 0 | |
| Qwen3-VL-32BSetting=GRPO2025.12 | 1 | 0 | |
| Qwen3-VL-8BSetting=GRPO2025.12 | 0 | 0 | |
| Qwen3-VL-32BSetting=N/A2025.12 | 0 | 0 | |
| Qwen-Image-Edit-2509Setting=N/A2025.12 | 0 | 0 | |
| Qwen-Image-Edit-2511Setting=N/A2025.12 | 0 | 0 |