LEED: A Highly Efficient and Scalable LLM-Empowered Expert Demonstrations Framework for Multi-Agent Reinforcement Learning

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Duan, Tianyang, Zhang, Zongyuan, Guo, Songxiao, Huang, Dong, Zhao, Yuanye, Lin, Zheng, Fang, Zihan, Luan, Dianxin, Cui, Heming, Cui, Yong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918143439405056
author Duan, Tianyang
Zhang, Zongyuan
Guo, Songxiao
Huang, Dong
Zhao, Yuanye
Lin, Zheng
Fang, Zihan
Luan, Dianxin
Cui, Heming
Cui, Yong
author_facet Duan, Tianyang
Zhang, Zongyuan
Guo, Songxiao
Huang, Dong
Zhao, Yuanye
Lin, Zheng
Fang, Zihan
Luan, Dianxin
Cui, Heming
Cui, Yong
contents Multi-agent reinforcement learning (MARL) holds substantial promise for intelligent decision-making in complex environments. However, it suffers from a coordination and scalability bottleneck as the number of agents increases. To address these issues, we propose the LLM-empowered expert demonstrations framework for multi-agent reinforcement learning (LEED). LEED consists of two components: a demonstration generation (DG) module and a policy optimization (PO) module. Specifically, the DG module leverages large language models to generate instructions for interacting with the environment, thereby producing high-quality demonstrations. The PO module adopts a decentralized training paradigm, where each agent utilizes the generated demonstrations to construct an expert policy loss, which is then integrated with its own policy loss. This enables each agent to effectively personalize and optimize its local policy based on both expert knowledge and individual experience. Experimental results show that LEED achieves superior sample efficiency, time efficiency, and robust scalability compared to state-of-the-art baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2509_14680
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle LEED: A Highly Efficient and Scalable LLM-Empowered Expert Demonstrations Framework for Multi-Agent Reinforcement Learning
Duan, Tianyang
Zhang, Zongyuan
Guo, Songxiao
Huang, Dong
Zhao, Yuanye
Lin, Zheng
Fang, Zihan
Luan, Dianxin
Cui, Heming
Cui, Yong
Multiagent Systems
Machine Learning
Multi-agent reinforcement learning (MARL) holds substantial promise for intelligent decision-making in complex environments. However, it suffers from a coordination and scalability bottleneck as the number of agents increases. To address these issues, we propose the LLM-empowered expert demonstrations framework for multi-agent reinforcement learning (LEED). LEED consists of two components: a demonstration generation (DG) module and a policy optimization (PO) module. Specifically, the DG module leverages large language models to generate instructions for interacting with the environment, thereby producing high-quality demonstrations. The PO module adopts a decentralized training paradigm, where each agent utilizes the generated demonstrations to construct an expert policy loss, which is then integrated with its own policy loss. This enables each agent to effectively personalize and optimize its local policy based on both expert knowledge and individual experience. Experimental results show that LEED achieves superior sample efficiency, time efficiency, and robust scalability compared to state-of-the-art baselines.
title LEED: A Highly Efficient and Scalable LLM-Empowered Expert Demonstrations Framework for Multi-Agent Reinforcement Learning
topic Multiagent Systems
Machine Learning
url https://arxiv.org/abs/2509.14680