Metraipolis › archive › dc_46802e29c116
Finite-Sample Convergence in Networked Average Reward MARL: Decentralization Pitfalls and Entropy Remedies — arXiv:2611.04072
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.04072
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-04T21:35:37.044Z
- SHA-256 of the archived text
9be4fc5b89c3bfc3e49f0dfb7b15f407492ed827e89adc729424bd4131298c80- Stable URL of this page
https://metraipolis.com/documento/dc_46802e29c116
How to cite
Finite-Sample Convergence in Networked Average Reward MARL: Decentralization Pitfalls and Entropy Remedies — arXiv:2611.04072 — arXiv (en), https://arxiv.org/abs/2611.04072 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_46802e29c116 (Metraipolis archive dc_46802e29c116, sha256 9be4fc5b89c3bfc3e49f0dfb7b15f407492ed827e89adc729424bd4131298c80)
BibTeX
@misc{dc46802e29c116,
title = {Finite-Sample Convergence in Networked Average Reward MARL: Decentralization Pitfalls and Entropy Remedies — arXiv:2611.04072},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.04072},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_46802e29c116; sha256 9be4fc5b89c3bfc3e49f0dfb7b15f407492ed827e89adc729424bd4131298c80},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.04072. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We study the problem of average reward Multi-Agent Reinforcement Learning (MARL) where agents interact within a network. Each agent in the network conducts local updates with information within its $k$-hop neighborhood and collaboratively maximizes the overall reward of the entire network. We first provide impossibility results in achieving convergence to the global optimal joint policies in our setting. These impossibility results highlight the challenges of decentralized learning with local information in multi-agent systems, namely the issues of multiagency, where each agent independently optimizes its policy, and partial observability, where agents have limited visibility into the global state. Given these challenges which indicate that decentralized policy optimization can generally be certified only up to stationarity, we provide finite-sample convergence guarantees for a decentralized actor-critic algorithm with linear function approximation that converges to an approximate stationary point with small gradient norm. To overcome such impossibility results, we further show that adding fixed-state entropy regularization reshapes the objective landscape which leads to even stronger global convergence guarantees for our proposed actor-critic algorithm. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.MA, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Multiagent Systems (cs.MA). Pagina del lavoro: https://arxiv.org/abs/2611.04072 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-11-04T21:35:34.328Z dalla pagina https://arxiv.org/list/cs.MA/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).