Socratic Planner: Inquiry-Based Zero-Shot Planning for Embodied Instruction Following

arXiv:2404.15190v1 Announce Type: cross
Abstract: Embodied Instruction Following (EIF) is the task of executing natural language instructions by navigating and interacting with objects in 3D environments. One of the primary challenges in EIF is compositional task planning, which is often addressed with supervised or in-context learning with labeled data. To this end, we introduce the Socratic Planner, the first zero-shot planning method that infers without the need for any training data. Socratic Planner first decomposes the instructions into substructural information …

abstract arxiv challenges context cs.ai cs.cl cs.cv cs.ro data embodied environments in-context learning language natural natural language objects planning type zero-shot