2025

An Extended Benchmarking of Multi-Agent Reinforcement Learning Algorithms in Complex Fully Cooperative Tasks

Paper page PDF
Year
2025
Authors
George Papadopoulos, Andreas Kontogiannis, Foteini Papadopoulou, Chaido Poulianou, Ioannis Koumentis, George Vouros
arXiv
2502.04773 [cs.LG]
Keywords
Fully Cooperative Multi-Agent Reinforcement Learning, Benchmarking, Image-based Observations, Open-Source Framework

Abstract

Benchmarking of Multi-Agent Reinforcement Learning Algorithms in Com- Multi-Agent Reinforcement Learning (MARL) has recently emerged plex Fully Cooperative Tasks. In Proc. of the 24th International Conference on Autonomous Agents and Multiagent Systems (AAMAS 2025), Detroit, Michi- as a significant area of research. However, MARL evaluation often gan, USA, May 19 – 23, 2025, IFAAMAS, 31 pages. lacks systematic diversity, hindering a comprehensive understand- ing of algorithms’ capabilities. In particular, cooperative MARL algorithms are predominantly evaluated on benchmarks such as