Zero-Shot Iterative Formalization and Planning in Partially Observable Environments
This addresses the challenge of planning in realistic, partially observable settings for AI systems, representing an incremental advance over prior work on fully observable environments.
The paper tackles the problem of formalizing partially observable environments into PDDL using LLMs in a zero-shot manner, resulting in improved goal-reaching success and robustness against complexity in simulated environments.
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. Existing work focuses on fully observable environments; we tackle the more realistic and challenging partially observable environments that lack of complete, reliable information. We propose PDDLego+, a framework to iteratively formalize, plan, grow, and refine PDDL representations in a zero-shot manner, without needing access to any existing trajectories. On two textual simulated environments, we show that PDDLego+ improves goal reaching success and exhibits robustness against problem complexity. We also show that the domain knowledge captured after a successful trial can benefit future tasks.