UneVEn: Universal Value Exploration for Multi-Agent Reinforcement Learning

Tarun Gupta, Anuj Mahajan, Bei Peng, Wendelin Böhmer, Shimon Whiteson

Research output: Chapter in Book/Conference proceedings/Edited volumeConference contributionScientificpeer-review


VDN and QMIX are two popular value-based algorithms for cooperative MARL that learn a centralized action value function as a monotonic mixing of per-agent utilities. While this enables easy decentralization of the learned policy, the restricted joint action value function can prevent them from solving tasks that require significant coordination between agents at a given timestep. We show that this problem can be overcome by improving the joint exploration of all agents during training. Specifically, we propose a novel MARL approach called Universal Value Exploration (UneVEn) that learns a set of related tasks simultaneously with a linear decomposition of universal successor features. With the policies of already solved related tasks, the joint exploration process of all agents can be improved to help them achieve better coordination. Empirical results on a set of exploration games, challenging cooperative predator-prey tasks requiring significant coordination among agents, and StarCraft II micromanagement benchmarks show that UneVEn can solve tasks where other state-of-the-art MARL methods fail.
Original languageEnglish
Title of host publicationProceedings of the International Conference on Machine Learning (ICML)
Number of pages19
Publication statusAccepted/In press - 2021
Event38th International Conference on Machine
- Virtual/online event
Duration: 18 Jul 202024 Jul 2020
Conference number: 38

Publication series

NameProceedings of Machine Learning Research
VolumePMLR 139
ISSN (Electronic)2640-3498


Conference38th International Conference on Machine
Abbreviated titleICML 2021


Dive into the research topics of 'UneVEn: Universal Value Exploration for Multi-Agent Reinforcement Learning'. Together they form a unique fingerprint.

Cite this