{"type":"ScholarlyArticle","title":"Mitigating Collaboration Degeneration in Multi-Agent Code Generation via a Controllable Competitive Collaboration Approach","authors":[{"name":"Shanzhi Gu","slug":"shanzhi-gu","given":"Shanzhi","family":"Gu","affiliation":""},{"name":"Mingyang Geng","slug":"mingyang-geng","given":"Mingyang","family":"Geng","affiliation":""},{"name":"Yihong Dong","slug":"yihong-dong","given":"Yihong","family":"Dong","affiliation":""},{"name":"Yunxin Mao","slug":"yunxin-mao","given":"Yunxin","family":"Mao","affiliation":""},{"name":"Zhaoyang Qu","slug":"zhaoyang-qu","given":"Zhaoyang","family":"Qu","affiliation":""},{"name":"Hao Zou","slug":"hao-zou","given":"Hao","family":"Zou","affiliation":""},{"name":"Ruochun Jin","slug":"ruochun-jin","given":"Ruochun","family":"Jin","affiliation":""},{"name":"Zhipeng Liu","slug":"zhipeng-liu","given":"Zhipeng","family":"Liu","affiliation":""},{"name":"Chuanfu Xu","slug":"chuanfu-xu","given":"Chuanfu","family":"Xu","affiliation":""},{"name":"Haotian Wang","slug":"haotian-wang","given":"Haotian","family":"Wang","affiliation":""}],"abstract":"Empowered by large language models (LLMs), multi-agent systems (MAS) have shown significant potential in code generation by simulating collaborative workflows. However, we identify a collaboration degeneration phenomenon, where one agent dominates while others remain disengaged, occurring in 38.4% of non-routine HumanEval tasks. Surprisingly, introducing competition among LLM agents eliminates this degeneration and sparks innovation in generated code. Inspired by this catfish effect, we posit that competition helps prevent collaboration degeneration. We thus propose C3, a Controllable Competitive Collaboration framework featuring two novel mechanisms: Centralized Auction-Based Competition (CAB) for role allocation via bidding and\r\nDecentralized Communication-Aware Competition (DCC) for solution refinement through opponent-aware adaptation.\r\nEvaluated on 70 complex projects from extended SoftwareDev, C3 reduces collaboration degeneration from 32.4% to 12.8%, while enhancing code innovation and diversity. Human evaluations further confirm C3’s superiority in functionality, maintainability, and robustness over purely collaborative or competitive baselines.","doi":"10.24963/ijcai.2026/15","url":"https://ijcai.org/proceedings/2026/15","pdf":"https://ijcai.org/proceedings/2026/0015.pdf","markdown":"https://ijcai.org/proceedings/2026/15.md","bibtex":"https://ijcai.org/proceedings/2026/bibtex/15","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":15,"pages":{"first":127,"last":135},"keywords":["Agent-based and Multi-agent Systems: Multi-agent learning","Agent-based and Multi-agent Systems: Coordination and cooperation"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}