Reproduce “Strategist: Evolving Reasoning with Agents” 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
Taiming Lu, Qi Xu, Zhigang Hua, Zhuang Liu
Strategist lets a frozen meta-agent search for model-calling strategies, test them, and save reusable blueprints for later tasks. Its models remain unchanged. Across 30 benchmarks, it reports 74.7% mean accuracy versus 74.0% for the strongest baseline and 0.61× weighted inference-token cost. It wins on 24 benchmarks, but does not improve both accuracy and cost universally.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.