Reproduce “Letting AI Formulate the Research Question: Autonomous Research 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
Ko Fujimura
The authors test whether a coding agent can choose research questions and report every result within a human-set research area. Across five runs, it completed question formulation, experiments, AI review, and revised papers without recorded human research advice or permission requests. AI reviewers rated papers favorably to borderline for workshops, but rejected all for main-track acceptance; human review was absent.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.