用进化算法自动生成更可靠、高效的多智能体系统。
EvoMAS: Evolutionary Generation of Multi-Agent Systems
- 在配置空间中通过进化搜索优化多智能体架构。
- 在BBEH和WorkBench上分别提升10.5和7.1分，优于现有方法。
- 适合需要高可执行性与鲁棒性的复杂任务设计者。
基于大语言模型的多智能体系统（MAS）在复杂推理、规划和工具增强任务中展现出巨大潜力，但有效架构的设计仍费时费力、脆弱且难泛化。现有自动生成方法要么依赖代码生成，常导致可执行性与鲁棒性失败，要么强加僵化模板，限制表达力与适应性。我们提出进化生成多智能体系统（EvoMAS），将MAS生成建模为结构化配置生成问题。EvoMAS在配置空间中进行进化生成：从池中选择初始配置，依据执行轨迹反馈进行条件化变异与交叉，迭代优化候选池与经验记忆。我们在多个基准上评估，涵盖BBEH、SWE-Bench和WorkBench，涉及推理、软件工程与工具使用任务。EvoMAS持续优于人工设计的MAS及先前自动生成方法，在任务性能、可执行性与运行鲁棒性上均表现更优。在BBEH推理任务上较EvoAgent提升10.5分，在WorkBench上提升7.1分。使用Claude-4.5-Sonnet时，于SWE-Bench-Verified达到79.1%，匹配排行榜顶尖水平。代码已开源。
原文摘要 · Abstract (English)
Large language model (LLM)-based multi-agent systems (MAS) show strong promise for complex reasoning, planning, and tool-augmented tasks, but designing effective MAS architectures remains labor-intensive, brittle, and hard to generalize. Existing automatic MAS generation methods either rely on code generation, which often leads to executability and robustness failures, or impose rigid architectural templates that limit expressiveness and adaptability. We propose Evolutionary Generation of Multi-Agent Systems (EvoMAS), which formulates MAS generation as structured configuration generation. EvoMAS performs evolutionary generation in configuration space. Specifically, EvoMAS selects initial configurations from a pool, applies feedback-conditioned mutation and crossover guided by execution traces, and iteratively refines both the candidate pool and an experience memory. We evaluate EvoMAS on diverse benchmarks, including BBEH, SWE-Bench, and WorkBench, covering reasoning, software engineering, and tool-use tasks. EvoMAS consistently improves task performance over both human-designed MAS and prior automatic MAS generation methods, while producing generated systems with higher executability and runtime robustness. EvoMAS outperforms the agent evolution method EvoAgent by +10.5 points on BBEH reasoning and +7.1 points on WorkBench. With Claude-4.5-Sonnet, EvoMAS also reaches 79.1% on SWE-Bench-Verified, matching the top of the leaderboard. Code is available at https://github.com/amazon-science/EvoMAS
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。