Reproduce “Evolve as a Team: Collaborative Self-Evolution for LLM-based 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
Zhezheng Hao, Tianfu Wang, Huanshuo Dong, Ziyan Liu, Hong Wang, Xiankun Lin, Qiang Lin, Can Wang and 2 more
Meta-Team lets multi-agent teams learn from task failures as agents retain local context, exchange evidence, and update behavior, collaboration, and organization. Unlike centralized analysis of the interleaved trace, this approach changes prompts, skills, profiles, and team rules without updating model parameters or infrastructure. Across nine benchmark columns it averages 62.7 versus MASFly’s 56.4, with results limited to selected long-horizon benchmarks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.