PROC2PDDL: Open-Domain Planning Representations from Texts

arXiv:2403.00092v1 Announce Type: new
Abstract: Planning in a text-based environment continues to be a major challenge for AI systems. Recent approaches have used language models to predict a planning domain definition (e.g., PDDL) but have only been evaluated in closed-domain simulated environments. To address this, we present Proc2PDDL , the first dataset containing open-domain procedural texts paired with expert-annotated PDDL representations. Using this dataset, we evaluate state-of-the-art models on defining the preconditions and effects of actions. We show that Proc2PDDL …

abstract ai systems arxiv challenge cs.cl dataset definition domain environment environments language language models major planning systems text type