Sequential Order Adjustment of Action Decisions for Multi-Agent Transformer (Student Abstract)
Abstract
Multi-agent reinforcement learning (MARL) trains multiple agents in shared environments. Recently, MARL models have significantly improved performance by leveraging sequential decision-making processes. Although these models can enhance performance, they do not explicitly con-sider the importance of the order in which agents make decisions. We propose AOAD-MAT, a novel model incorporating action decision sequence into learning. AOAD-MAT uses a Transformer-based actor-critic architecture to dynamically adjust agent action order. It introduces a subtask predicting the next agent to act, integrated into a PPO-based loss function. Experiments on StarCraft Multi-Agent Challenge and Multi-Agent MuJoCo benchmarks show AOAD-MAT out-performs existing models, demonstrating the effectiveness of adjusting agent order in MARL.
Cite
Text
Takayama and Fujita. "Sequential Order Adjustment of Action Decisions for Multi-Agent Transformer (Student Abstract)." AAAI Conference on Artificial Intelligence, 2025. doi:10.1609/AAAI.V39I28.35306Markdown
[Takayama and Fujita. "Sequential Order Adjustment of Action Decisions for Multi-Agent Transformer (Student Abstract)." AAAI Conference on Artificial Intelligence, 2025.](https://mlanthology.org/aaai/2025/takayama2025aaai-sequential/) doi:10.1609/AAAI.V39I28.35306BibTeX
@inproceedings{takayama2025aaai-sequential,
title = {{Sequential Order Adjustment of Action Decisions for Multi-Agent Transformer (Student Abstract)}},
author = {Takayama, Shota and Fujita, Katsuhide},
booktitle = {AAAI Conference on Artificial Intelligence},
year = {2025},
pages = {29509-29511},
doi = {10.1609/AAAI.V39I28.35306},
url = {https://mlanthology.org/aaai/2025/takayama2025aaai-sequential/}
}