Evolve as a Team: Collaborative Self-Evolution for LLM-based Multi-Agent Systems
Zhezheng Hao, Tianfu Wang, Huanshuo Dong, Ziyan Liu, Hong Wang, Xiankun Lin, Qiang Lin, Can Wang, Hande Dong, Jiawei Chen
Why It Matters
What makes this one worth your time
This research addresses the challenges of improving multi-agent systems in real-world applications, potentially leading to more reliable and scalable solutions.
Meta-Team enables multi-agent systems to evolve collaboratively based on their execution experiences.
Summary
The paper introduces Meta-Team, a framework for collaborative self-evolution in multi-agent systems that leverages execution experience to enhance agent behaviors and coordination.
Key contributions
- Introduction of the Meta-Team framework for experience-driven evolution in MAS.
- Development of a method for agents to exchange distributed evidence for evolution.
- Demonstration of consistent performance improvements across multiple long-horizon benchmarks.
Notable insights
- The framework's focus on preserving execution context allows for nuanced improvements in agent interactions.
- The multi-scale self-evolution approach may provide a systematic way to enhance both individual and team-level performance.
Possible limitations
- Not stated in the abstract.
Abstract
arXiv:2605.29790v1 Announce Type: cross Abstract: 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.