Reproduce “TeamFusion: Supporting Open-ended Teamwork with Multi-Agent Systems” on OpenResearch

Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.

Universal app for Apple silicon and Intel. Requires macOS 11 or later.

About this paper

Jiale Liu, Victor S. Bursztyn, Lin Ai, Haoliang Wang, Sunav Choudhary, Saayan Mitra, Qingyun Wu

TeamFusion lets teams turn differing preferences into revisable design proposals through personalized AI proxies that discuss trade-offs before an AI remixes their discussion. In DeliberationBank OpenQA, it scored 0.621 in representativeness and 0.622 in informativeness, versus 0.599 and 0.561 for Multi-Agent Debate. In asynchronous design evaluations, generated designs entered top-ranked selections in 42% of Full-Team and 50% of Small-Team settings; this was not live unanimous acceptance.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.