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

arXiv: https://arxiv.org/abs/2509.14680 · HTML: https://mlsanity.com/paper/2509.14680/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in Reinforcement learning for LLM agents, one of 16 directions in LLM agents and tool use.
- Builds on 35 of its 45 references that are in the map.
- First posted on arXiv 18 Sep 2025.

## Placed in

- LLM agents and tool use / [Reinforcement learning for LLM agents](https://mlsanity.com/agents/reinforcement-learning-for-llm-agents/)

## Builds on (most cited first, 30 of 35)

- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Multi-Agent Actor-Critic for Mixed Cooperative-Competitive Environments](https://mlsanity.com/paper/1706.02275/) (2017)
- [The Surprising Effectiveness of PPO in Cooperative, Multi-Agent Games](https://mlsanity.com/paper/2103.01955/) (2021)
- [Multi-Agent Reinforcement Learning: A Selective Overview of Theories and Algorithms](https://mlsanity.com/paper/1911.10635/) (2019)
- [Value-Decomposition Networks For Cooperative Multi-Agent Learning](https://mlsanity.com/paper/1706.05296/) (2017)
- [Monotonic Value Function Factorisation for Deep Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2003.08839/) (2020)
- [Describe, Explain, Plan and Select: Interactive Planning with Large Language Models Enables Open-World Multi-Task Agents](https://mlsanity.com/paper/2302.01560/) (2023)
- [Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models](https://mlsanity.com/paper/2503.09567/) (2025)
- [Is Independent Learning All You Need in the StarCraft Multi-Agent Challenge?](https://mlsanity.com/paper/2011.09533/) (2020)
- [Reward Design with Language Models](https://mlsanity.com/paper/2303.00001/) (2023)
- [Trust Region Policy Optimisation in Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2109.11251/) (2021)
- [Efficient Parallel Split Learning over Resource-constrained Wireless Edge Networks](https://mlsanity.com/paper/2303.15991/) (2023)
- [Pushing Large Language Models to the 6G Edge: Vision, Challenges, and Opportunities](https://mlsanity.com/paper/2309.16739/) (2023)
- [AdaptSFL: Adaptive Split Federated Learning in Resource-constrained Edge Networks](https://mlsanity.com/paper/2403.13101/) (2024)
- [SplitLoRA: A Split Parameter-Efficient Fine-Tuning Framework for Large Language Models](https://mlsanity.com/paper/2407.00952/) (2024)
- [Automated Federated Pipeline for Parameter-Efficient Fine-Tuning of Large Language Models](https://mlsanity.com/paper/2404.06448/) (2024)
- [Reducing Overestimation Bias in Multi-Agent Domains Using Double Centralized Critics](https://mlsanity.com/paper/1910.01465/) (2019)
- [FedSN: A Federated Learning Framework over Heterogeneous LEO Satellite Networks](https://mlsanity.com/paper/2311.01483/) (2023)
- [Hierarchical Split Federated Learning: Convergence Analysis and System Optimization](https://mlsanity.com/paper/2412.07197/) (2024)
- [Multi-Agent Constrained Policy Optimisation](https://mlsanity.com/paper/2110.02793/) (2021)
- [HSplitLoRA: A Heterogeneous Split Parameter-Efficient Fine-Tuning Framework for Large Language Models](https://mlsanity.com/paper/2505.02795/) (2025)
- [LLM-Empowered State Representation for Reinforcement Learning](https://mlsanity.com/paper/2407.13237/) (2024)
- [FedAC: An Adaptive Clustered Federated Learning Framework for Heterogeneous Data](https://mlsanity.com/paper/2403.16460/) (2024)
- [LEO-Split: A Semi-Supervised Split Learning Framework over LEO Satellite Networks](https://mlsanity.com/paper/2501.01293/) (2025)
- [IC3M: In-Car Multimodal Multi-object Monitoring for Abnormal Status of Both Driver and Passengers](https://mlsanity.com/paper/2410.02592/) (2024)
- [MERIT: Multimodal Wearable Vital Sign Waveform Monitoring](https://mlsanity.com/paper/2410.00392/) (2024)
- [LCFed: An Efficient Clustered Federated Learning Framework for Heterogeneous Data](https://mlsanity.com/paper/2501.01850/) (2025)
- [Multi-Agent Reinforcement Learning for Resources Allocation Optimization: A Survey](https://mlsanity.com/paper/2504.21048/) (2025)
- [Dynamic Uncertainty-aware Multimodal Fusion for Outdoor Health Monitoring](https://mlsanity.com/paper/2508.09085/) (2025)
- [Exploiting Fast Decaying and Locality in Multi-Agent MDP with Tree Dependence Structure](https://mlsanity.com/paper/1909.06900/) (2019)

## Built on by (most cited first, 6 of 6)

- [HFedMoE: Resource-aware Heterogeneous Federated Learning with Mixture-of-Experts](https://mlsanity.com/paper/2601.00583/) (2026)
- [LLM-Driven Stationarity-Aware Expert Demonstrations for Multi-Agent Reinforcement Learning in Mobile Systems](https://mlsanity.com/paper/2511.19368/) (2025)
- [Conflict-Aware Client Selection for Multi-Server Federated Learning](https://mlsanity.com/paper/2602.02458/) (2026)
- [SL-FAC: A Communication-Efficient Split Learning Framework with Frequency-Aware Compression](https://mlsanity.com/paper/2604.07316/) (2026)
- [SIDeR: Semantic Identity Decoupling for Unrestricted Face Privacy](https://mlsanity.com/paper/2602.04994/) (2026)
- [Large Language Model Enhanced Differentiable Trajectory Planning for IoT-Enabled Autonomous Driving](https://mlsanity.com/paper/2607.10438/) (2026)

## Abstract (arXiv)

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.