Reproduce “Epstein Files Engine: Agentic Search for Investigative Journalism” 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
Duy K. Nguyen, Teresa Mondría Terol, Dylan Freedman, Zach Seward
The Epstein Files Engine turns plain-language questions into planned searches across released records, the Times archive, and outside headlines, using an LLM to assemble cited results for journalists to verify. Diff combines semantic text embeddings and visual fingerprints to identify duplicate or similar pages. At threshold 0.92, it reached 0.28 precision and 0.86 recall retrospectively, without a controlled ablation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.