EUVIMEDOtkrijte EUVIMED
Open access · CC BY 4.0

Language-Driven Coordination in Multi-Agent Reinforcement Learning with Large Language Models

Benjamin J. Wilson, Oliver Smith, Chloe A. Thompson

Computer Science Bulletin · 2026-01-30 · 9 (2) · 290-296

DOI: 10.71465/csb224 ↗

At a glance

Journal article in Computer Science Bulletin, published online on 2026-01-30. Written by Benjamin J. Wilson and 2 co-authors. The full text is freely available (open access, CC BY 4.0).

SažetakBibliografski podaciCiteRelated studies

Abstract (original language)

This study integrates LLMs into MARL through language-based coordination, communication, and memory modules. The framework combines proximal policy optimization (PPO) with language-conditioned subgoal generation. Experiments across three multi-agent environments involving over 10,000 training episodes show a 14.7–26.3% increase in win rate compared with MAPPO and QMIX baselines, with ablation results confirming the contribution of languageguided coordination to sample efficiency and zero-shot generalization.

Cite

APA

Wilson, B. J., Smith, O., & Thompson, C. A. (2026). Language-Driven Coordination in Multi-Agent Reinforcement Learning with Large Language Models. Computer Science Bulletin, 9(2), 290-296. https://doi.org/10.71465/csb224

Vancouver

Wilson BJ, Smith O, Thompson CA. Language-Driven Coordination in Multi-Agent Reinforcement Learning with Large Language Models. Computer Science Bulletin. 2026;9(2):290-296. doi:10.71465/csb224

BibTeX
@article{Wilson2026,
  author = {Wilson, Benjamin J. and Smith, Oliver and Thompson, Chloe A.},
  title = {Language-Driven Coordination in Multi-Agent Reinforcement Learning with Large Language Models},
  journal = {Computer Science Bulletin},
  year = {2026},
  volume = {9},
  number = {2},
  pages = {290--296},
  doi = {10.71465/csb224}
}
Start for free →