Making Teams and Influencing Agents: Efficiently Coordinating Decision Trees for Interpretable Multi-Agent Reinforcement Learning
Open Access
- 15 October 2025
- journal article
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society
- Vol. 8 (1) , 567-578
Abstract
Poor interpretability hinders the practical applicability of multi-agent reinforcement learning (MARL) policies. Deploying interpretable surrogates of uninterpretable policies enhances the safety and verifiability of MARL for real-world applications. However, if these surrogates are to interact directly with the environment within human supervisory frameworks, they must be both performant and computationally efficient. Prior work on interpretable MARL has either sacrificed performance for computational efficiency or computational efficiency for performance. To address this issue, we propose HYDRAVIPER, a decision tree-based interpretable MARL algorithm. HYDRAVIPER coordinates training between agents based on expected team performance, and adaptively allocates budgets for environment interaction to improve computational efficiency. Experiments on standard benchmark environments for multi-agent coordination and traffic signal control show that HYDRAVIPER matches the performance of state-of-the-art methods using a fraction of the runtime, and that it maintains a Pareto frontier of performance for different interaction budgets.Keywords
All Related Versions
This publication has 13 references indexed in Scilit:
- MIXRTs: Toward Interpretable Multi-Agent Reinforcement Learning via Mixing Recurrent Soft Decision TreesIEEE Transactions on Pattern Analysis and Machine Intelligence, 2025
- Explainable Reinforcement Learning: A Survey and Comparative ReviewACM Computing Surveys, 2024
- Enhancing the Interpretability of Deep Multi-agent Reinforcement Learning via Neural Logic Reasoning2023
- Extracting Decision Tree From Trained Deep Reinforcement Learning in Traffic Signal ControlIEEE Transactions on Computational Social Systems, 2022
- Toward Human-in-the-Loop AI: Enhancing Deep Reinforcement Learning via Real-Time Human Guidance for Autonomous DrivingEngineering, 2022
- MoËT: Mixture of Expert Trees and its application to verifiable reinforcement learningNeural Networks, 2022
- Reinforcement learning in urban network traffic signal control: A systematic literature reviewExpert Systems with Applications, 2022
- Collective eXplainable AI: Explaining Cooperative Strategies and Agent Contribution in Multiagent Reinforcement Learning With Shapley ValuesIEEE Computational Intelligence Magazine, 2022
- Mastering the game of Go with deep neural networks and tree searchNature, 2016
- A Fast and High Quality Multilevel Scheme for Partitioning Irregular GraphsSIAM Journal on Scientific Computing, 1998