Saved in:
Bibliographic Details
Main Authors: Li, William, Hamilton, Lei, Al-natour, Kaise, Mohindra, Sanjeev
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2507.06157
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866911045623218176
author Li, William
Hamilton, Lei
Al-natour, Kaise
Mohindra, Sanjeev
author_facet Li, William
Hamilton, Lei
Al-natour, Kaise
Mohindra, Sanjeev
contents This paper focuses on evaluating the effectiveness of Large Language Models at solving embodied robotic tasks using the Meta PARTNER benchmark. Meta PARTNR provides simplified environments and robotic interactions within randomized indoor kitchen scenes. Each randomized kitchen scene is given a task where two robotic agents cooperatively work together to solve the task. We evaluated multiple frontier models on Meta PARTNER environments. Our results indicate that reasoning models like OpenAI o3-mini outperform non-reasoning models like OpenAI GPT-4o and Llama 3 when operating in PARTNR's robotic embodied environments. o3-mini displayed outperform across centralized, decentralized, full observability, and partial observability configurations. This provides a promising avenue of research for embodied robotic development.
format Preprint
id arxiv_https___arxiv_org_abs_2507_06157
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Evaluation of Habitat Robotics using Large Language Models
Li, William
Hamilton, Lei
Al-natour, Kaise
Mohindra, Sanjeev
Robotics
Computation and Language
This paper focuses on evaluating the effectiveness of Large Language Models at solving embodied robotic tasks using the Meta PARTNER benchmark. Meta PARTNR provides simplified environments and robotic interactions within randomized indoor kitchen scenes. Each randomized kitchen scene is given a task where two robotic agents cooperatively work together to solve the task. We evaluated multiple frontier models on Meta PARTNER environments. Our results indicate that reasoning models like OpenAI o3-mini outperform non-reasoning models like OpenAI GPT-4o and Llama 3 when operating in PARTNR's robotic embodied environments. o3-mini displayed outperform across centralized, decentralized, full observability, and partial observability configurations. This provides a promising avenue of research for embodied robotic development.
title Evaluation of Habitat Robotics using Large Language Models
topic Robotics
Computation and Language
url https://arxiv.org/abs/2507.06157