MAC-Bench: Benchmarking Multi-Agent Collaboration via Interconnected Tasks

DOI10.48550/ARXIV.2611.01728MaRDI QIDQ7420481arXivFDO

Authors Zerui Chen, Jinsong Su, Zhi-Hong Zhang, Qinggang Zhang, Qichuan Liu, Zhishang Xiang, Zhimin Wei, Yunbo Tang, Zhuo Li, Zhihong Zhang

Publication date 1 November 2026

Published in arXiv

Tldr We introduce MAC-Bench, a novel benchmark for evaluating multi-agent collaboration through interconnected tasks by directly embedding collaboration requirements into task dependencies and shared-resource constraints; we mine reusable collaboration structures at both local and global levels to create diverse task graphs that can be used in various settings.

Full work available at URL https://doi.org/10.48550/ARXIV.2611.01728

arXiv classification

cs.AI cs.CL cs.MA

This page was built for publication: MAC-Bench: Benchmarking Multi-Agent Collaboration via Interconnected Tasks

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q7420481)