GAVEL: Graph World Models for Verified and Efficient Long-Horizon LLM Task Planning

Researchers introduced GAVEL, a framework for verifying and repairing long-horizon LLM task plans that often fail to respect embodiment constraints, recover from planning errors, or reason under partial observability. GAVEL is described as a graph world model for verified and efficient long-horizon LLM task planning.

Large language models LLMs provide a flexible interface for long-horizon robot planning, but generated plans often fail to respect embodiment constraints, recover from planning errors, or reason effectively under partial observability. We present GAVEL, a framework for verifying and repairing long