Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Hao, Zhezheng, Wang, Tianfu, Dong, Huanshuo, Liu, Ziyan, Wang, Hong, Lin, Xiankun, Lin, Qiang, Wang, Can, Dong, Hande, Chen, Jiawei
Natura: Preprint
Pubblicazione: 2026
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866910270084874240
author Hao, Zhezheng
Wang, Tianfu
Dong, Huanshuo
Liu, Ziyan
Wang, Hong
Lin, Xiankun
Lin, Qiang
Wang, Can
Dong, Hande
Chen, Jiawei
author_facet Hao, Zhezheng
Wang, Tianfu
Dong, Huanshuo
Liu, Ziyan
Wang, Hong
Lin, Xiankun
Lin, Qiang
Wang, Can
Dong, Hande
Chen, Jiawei
contents LLM-based multi-agent systems (MAS) have emerged as an effective paradigm for complex and long-horizon tasks. However, in real-world tasks, MAS often exhibit various failures during execution and such failures are difficult to eliminate during design. This motivates experience-driven MAS evolution, where a system improves based on its own execution experience. Yet such evolution is challenging because MAS experience is prolonged and intricate, interleaving multiple agents' execution chains and communication messages, which makes it difficult to identify what should be improved. To address this challenge, we propose Meta-Team, an experience-driven MAS evolution framework based on collaborative self-evolution. Meta-Team preserves the execution context of each agent and coordinates post-task communication, enabling agents to exchange distributed evidence for evolution. Building on this design, Meta-Team conducts multi-scale self-evolution, transforming execution experience into reusable improvements to agent behaviors, inter-agent coordination, and team-level organization. Across six long-horizon agent benchmarks, Meta-Team consistently outperforms single-agent systems, hand-crafted MAS, and prior MAS evolution methods; further analyses demonstrate that Meta-Team enables more reliable and scalable MAS self-evolution.
format Preprint
id arxiv_https___arxiv_org_abs_2605_29790
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems
Hao, Zhezheng
Wang, Tianfu
Dong, Huanshuo
Liu, Ziyan
Wang, Hong
Lin, Xiankun
Lin, Qiang
Wang, Can
Dong, Hande
Chen, Jiawei
Multiagent Systems
Artificial Intelligence
LLM-based multi-agent systems (MAS) have emerged as an effective paradigm for complex and long-horizon tasks. However, in real-world tasks, MAS often exhibit various failures during execution and such failures are difficult to eliminate during design. This motivates experience-driven MAS evolution, where a system improves based on its own execution experience. Yet such evolution is challenging because MAS experience is prolonged and intricate, interleaving multiple agents' execution chains and communication messages, which makes it difficult to identify what should be improved. To address this challenge, we propose Meta-Team, an experience-driven MAS evolution framework based on collaborative self-evolution. Meta-Team preserves the execution context of each agent and coordinates post-task communication, enabling agents to exchange distributed evidence for evolution. Building on this design, Meta-Team conducts multi-scale self-evolution, transforming execution experience into reusable improvements to agent behaviors, inter-agent coordination, and team-level organization. Across six long-horizon agent benchmarks, Meta-Team consistently outperforms single-agent systems, hand-crafted MAS, and prior MAS evolution methods; further analyses demonstrate that Meta-Team enables more reliable and scalable MAS self-evolution.
title Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems
topic Multiagent Systems
Artificial Intelligence
url https://arxiv.org/abs/2605.29790