Sokoban on Sokoban (Standard)
98.83AccuracyOpenAI o3
Evaluation Results
| Method | Links | |
|---|---|---|
| OpenAI o3category=Proprietary Models2025.11 | 98.83 | |
| GPT-5category=Proprietary Models2025.11 | 96.88 | |
| OpenAI o4-minicategory=Proprietary Models2025.11 | 83.2 | |
| WMAct2025.11 | 78.57 | |
| PPO - Interactivemode=interactive2025.11 | 64.21 | |
| Claude 4.5 Sonnetcategory=Proprietary Models2025.11 | 60.94 | |
| PPO - EntirePlanmode=single-turn output2025.11 | 49.12 | |
| Gemini 2.5 Procategory=Proprietary Models2025.11 | 36.72 | |
| Qwen3-14Bcategory=Opensource Models2025.11 | 18.75 | |
| Qwen3-8Bcategory=Opensource Models2025.11 | 16.41 | |
| Qwen3-8B-Ownbackbone=Qwen3-8B2025.11 | 3.29 | |
| GPT-4ocategory=Proprietary Models2025.11 | 1.95 | |
| Qwen2.5-7B-Instructcategory=Opensource Models2025.11 | 0 | |
| Qwen2.5-32B-Instructcategory=Opensource Models2025.11 | 0 |