{"type":"ScholarlyArticle","title":"From Traits to Roles: Consensus-Guided Composition of Orthogonal Experts for Cooperative MARL","authors":[{"name":"Yewei Zhou","slug":"yewei-zhou","given":"Yewei","family":"Zhou","affiliation":""},{"name":"Bin Zhang","slug":"bin-zhang","given":"Bin","family":"Zhang","affiliation":""},{"name":"Ying Zhou","slug":"ying-zhou","given":"Ying","family":"Zhou","affiliation":""},{"name":"Xuri Ge","slug":"xuri-ge","given":"Xuri","family":"Ge","affiliation":""},{"name":"Dapeng Li","slug":"dapeng-li","given":"Dapeng","family":"Li","affiliation":""},{"name":"Hangyu Mao","slug":"hangyu-mao","given":"Hangyu","family":"Mao","affiliation":""},{"name":"Pengjie Ren","slug":"pengjie-ren","given":"Pengjie","family":"Ren","affiliation":""},{"name":"Zhiwei Xu","slug":"zhiwei-xu","given":"Zhiwei","family":"Xu","affiliation":""}],"abstract":"Parameter sharing is a central design choice in cooperative multi-agent reinforcement learning, yet it fundamentally conflicts with the need for role specialization in heterogeneous cooperative environments. Existing role-based methods typically learn monolithic role representations, which often suffer from gradient interference and fail to capture the compositional structure of complex behaviors. Inspired by Trait Theory, we propose DEcompose and COnstruct Roles (DECOR), a framework that models agent roles as dynamic compositions of orthogonal behavioral traits. DECOR introduces an orthogonal Mixture-of-Experts architecture to decompose behaviors into independent traits, mitigating destructive gradient interference under parameter sharing, and a group-consensus guided mechanism to extract team-level tactical intents that guide role composition.Experiments on multiple benchmarks demonstrate that DECOR consistently improves sample efficiency and overall performance over existing related methods.","doi":"10.24963/ijcai.2026/55","url":"https://ijcai.org/proceedings/2026/55","pdf":"https://ijcai.org/proceedings/2026/0055.pdf","markdown":"https://ijcai.org/proceedings/2026/55.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/55","volume":{"slug":"2026","title":"Proceedings of the Thirty-Fifth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2026","url":"https://ijcai.org/proceedings/2026/","year":2026,"location":"Bremen, Germany","dates":"15-21 August 2026","editors":"Diego Calvanese","isbn":"978-1-956792-09-6","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2026"},"section":"Agent-based and Multi-agent Systems","track":"Main Track","number":55,"pages":{"first":481,"last":489},"keywords":["Agent-based and Multi-agent Systems: Multi-agent learning"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}