MAC-Bench：通过互联任务评测多智能体协作
原标题：MAC-Bench: Benchmarking Multi-Agent Collaboration via Interconnected Tasks
AI 摘要
该论文提出 MAC-Bench，一个通过互联任务评估多智能体协作的基准。它从已有基准任务中挖掘可复用的局部协作模式与全局任务结构，经粗到细组合生成任务图，并在可执行环境中实例化、自动验证部分进展与完整任务。作者在两种 LLM 骨干上评测三个多智能体系统与单智能体基线，发现多智能体设计在已验证进展上收益不一，且没有任何配置能解决 48 个测试任务中的 6.25% 以上。
正文节选
MAC-Bench: Benchmarking Multi-Agent Collaboration via Interconnected Tasks Abstract Large language model (LLM)-based multi-agent systems (MAS) tackle challenging tasks by dividing work among agents and coordinating their contributions toward shared objectives. Evaluating these systems therefore requires benchmarks that assess not only task-solving performance but also the ability to handle interdependent work. However, existing benchmarks often emphasize task difficulty without substantial colla