• search hit 38 of 38
Back to Result List

Emergent Cooperation from Mutual Acknowledgment Exchange in Multi-Agent Reinforcement Learning

  • Peer incentivization (PI) is a recent approach, where all agents learn to reward or to penalize each other in a distributed fashion which often leads to emergent cooperation. Current PI mechanisms implicitly assume a flawless communication channel in order to exchange rewards. These rewards are directly integrated into the learning process without any chance to respond with feedback. Furthermore, most PI approaches rely on global information which limits scalability and applicability to real-world scenarios, where only local information is accessible. In this paper, we propose Mutual Acknowledgment Token Exchange (MATE), a PI approach defined by a two-phase communication protocol to mutually exchange acknowledgment tokens to shape individual rewards. Each agent evaluates the monotonic improvement of its individual situation in order to accept or reject acknowledgment requests from other agents. MATE is completely decentralized and only requires local communication and information. We evaluate MATE in three social dilemma domains. OurPeer incentivization (PI) is a recent approach, where all agents learn to reward or to penalize each other in a distributed fashion which often leads to emergent cooperation. Current PI mechanisms implicitly assume a flawless communication channel in order to exchange rewards. These rewards are directly integrated into the learning process without any chance to respond with feedback. Furthermore, most PI approaches rely on global information which limits scalability and applicability to real-world scenarios, where only local information is accessible. In this paper, we propose Mutual Acknowledgment Token Exchange (MATE), a PI approach defined by a two-phase communication protocol to mutually exchange acknowledgment tokens to shape individual rewards. Each agent evaluates the monotonic improvement of its individual situation in order to accept or reject acknowledgment requests from other agents. MATE is completely decentralized and only requires local communication and information. We evaluate MATE in three social dilemma domains. Our results show that MATE is able to achieve and maintain significantly higher levels of cooperation than previous PI approaches. In addition, we evaluate the robustness of MATE in more realistic scenarios, where agents can defect from the protocol and where communication failures can occur. We also evaluate the sensitivity of MATE w.r.t. the choice of token values.show moreshow less

Download full text files

Export metadata

Additional Services

Search Google Scholar
Metadaten
Author:Thomy Phan, Felix Sommer, Fabian Ritz, Philipp Altmann, Jonas Nüßlein, Michael Kölle, Lenz BelznerORCiD, Claudia Linnhoff-Popien
Language:English
Document Type:Preprint
Year of first Publication:2022
published in (English):Research Square
Publisher:Research Square
Place of publication:Durham
ISSN:2693-5015
Pages:52
Review:nein
Open Access:ja
URN:urn:nbn:de:bvb:573-45471
Related Identifier:https://doi.org/10.21203/rs.3.rs-2315844/v1
Faculties / Institutes / Organizations:Fakultät Elektro- und Informationstechnik
AImotion Bavaria
Licence (German):License Logo Creative Commons BY 4.0
Release Date:2024/02/27