Planning on Blocks (Blocksworld)
100AccuracyStick-Breaking Transformer
Evaluation Results
| Method | Links | |
|---|---|---|
| Stick-Breaking Transformerencoding=init-atoms2025.09 | 100 | |
| Sinusoidalencoding=init-atoms2025.09 | 100 | |
| RoPEencoding=init-atoms2025.09 | 100 | |
| o4-miniAlgorithm=API, Model=o4-mini2025.03 | 98.5 | |
| SEM-CTRLAlgorithm=SEM-CTRL, Model=Llama 70B2025.03 | 96.8 | |
| DeepSeek-R1Algorithm=API, Model=DeepSeek-R12025.03 | 96.5 | |
| o1-previewAlgorithm=API, Model=o1-preview2025.03 | 94.5 | |
| SEM-CTRLAlgorithm=SEM-CTRL, Model=Llama 1B2025.03 | 74 | |
| BoNAlgorithm=BoN, Model=Llama 70B2025.03 | 48.8 | |
| BaseAlgorithm=Base, Model=Llama 70B2025.03 | 23.2 | |
| BoNAlgorithm=BoN, Model=Llama 1B2025.03 | 4.3 | |
| BaseAlgorithm=Base, Model=Llama 1B2025.03 | 0 |