Iterative Formalization and Planning in Partially Observable Environments
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915924113620992 |
|---|---|
| author | Gong, Liancheng Zhu, Wang Thomason, Jesse Zhang, Li |
| author_facet | Gong, Liancheng Zhu, Wang Thomason, Jesse Zhang, Li |
| contents | Using LLMs not to predict plans but to formalize an environment into the Planning Domain Definition Language (PDDL) has been shown to improve performance and control. While most existing methodology only applies to fully observable environments, we adapt to the more realistic and challenging partially observable environments without sufficient information to make a complete plan. We propose PDDLego, a framework to iteratively formalize, plan, grow, and refine PDDL representations by decomposing the environment and the goal into fully observable episodes. Without finetuning, in-context exemplars, or trajectories, PDDLego improves planning success and exhibits robustness against problem complexity compared to end-to-end approaches. We also show that the domain knowledge captured after a successful trial can benefit future tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_13126 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Iterative Formalization and Planning in Partially Observable Environments Gong, Liancheng Zhu, Wang Thomason, Jesse Zhang, Li Artificial Intelligence Computation and Language Using LLMs not to predict plans but to formalize an environment into the Planning Domain Definition Language (PDDL) has been shown to improve performance and control. While most existing methodology only applies to fully observable environments, we adapt to the more realistic and challenging partially observable environments without sufficient information to make a complete plan. We propose PDDLego, a framework to iteratively formalize, plan, grow, and refine PDDL representations by decomposing the environment and the goal into fully observable episodes. Without finetuning, in-context exemplars, or trajectories, PDDLego improves planning success and exhibits robustness against problem complexity compared to end-to-end approaches. We also show that the domain knowledge captured after a successful trial can benefit future tasks. |
| title | Iterative Formalization and Planning in Partially Observable Environments |
| topic | Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2505.13126 |