Prompts pretrained LLMs to create QNP abstractions for generalized planning from domains and tasks. Automated debugging detects/fixes errors iteratively. Guided LLMs produce useful abstractions for qualitative numerical planning.
Key Points
- 1.Prompt protocol for features/states/actions
- 2.Automated error debugging
- 3.Supports GP via QNP
Impact Analysis
Advances generalized planning by leveraging LLMs for scalable abstractions.
Technical Details
Inputs GP domain/training tasks; iterative LLM-guided fixes.