Centralized Permutation Equivariant Policy for Cooperative Multi-Agent Reinforcement Learning
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866912539598651392 |
|---|---|
| author | Xu, Zhuofan Bollig, Benedikt Függer, Matthias Nowak, Thomas Dréau, Vincent Le |
| author_facet | Xu, Zhuofan Bollig, Benedikt Függer, Matthias Nowak, Thomas Dréau, Vincent Le |
| contents | The Centralized Training with Decentralized Execution (CTDE) paradigm has gained significant attention in multi-agent reinforcement learning (MARL) and is the foundation of many recent algorithms. However, decentralized policies operate under partial observability and often yield suboptimal performance compared to centralized policies, while fully centralized approaches typically face scalability challenges as the number of agents increases.
We propose Centralized Permutation Equivariant (CPE) learning, a centralized training and execution framework that employs a fully centralized policy to overcome these limitations. Our approach leverages a novel permutation equivariant architecture, Global-Local Permutation Equivariant (GLPE) networks, that is lightweight, scalable, and easy to implement. Experiments show that CPE integrates seamlessly with both value decomposition and actor-critic methods, substantially improving the performance of standard CTDE algorithms across cooperative benchmarks including MPE, SMAC, and RWARE, and matching the performance of state-of-the-art RWARE implementations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2508_11706 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Centralized Permutation Equivariant Policy for Cooperative Multi-Agent Reinforcement Learning Xu, Zhuofan Bollig, Benedikt Függer, Matthias Nowak, Thomas Dréau, Vincent Le Multiagent Systems Artificial Intelligence Machine Learning The Centralized Training with Decentralized Execution (CTDE) paradigm has gained significant attention in multi-agent reinforcement learning (MARL) and is the foundation of many recent algorithms. However, decentralized policies operate under partial observability and often yield suboptimal performance compared to centralized policies, while fully centralized approaches typically face scalability challenges as the number of agents increases. We propose Centralized Permutation Equivariant (CPE) learning, a centralized training and execution framework that employs a fully centralized policy to overcome these limitations. Our approach leverages a novel permutation equivariant architecture, Global-Local Permutation Equivariant (GLPE) networks, that is lightweight, scalable, and easy to implement. Experiments show that CPE integrates seamlessly with both value decomposition and actor-critic methods, substantially improving the performance of standard CTDE algorithms across cooperative benchmarks including MPE, SMAC, and RWARE, and matching the performance of state-of-the-art RWARE implementations. |
| title | Centralized Permutation Equivariant Policy for Cooperative Multi-Agent Reinforcement Learning |
| topic | Multiagent Systems Artificial Intelligence Machine Learning |
| url | https://arxiv.org/abs/2508.11706 |