Reproduce “ReDence: A Hybrid Agentic and Decision-Model Framework for Rapid, Evidence-Grounded Detection of Modern Social Engineering Scams” 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
Jaishnoor Kaur, Josef Horáček, Ian Waskom, Sayak Saha Roy
ReDence investigates scam websites by exploring their workflows and documenting evidence, rather than relying only on landing-page text. Its Investigator plans and interprets checks, while a Decider selects among bounded actions. On a test set of 1,551 scam and 1,500 benign sites, it reached 0.96 accuracy and 43.8-second median time, outperforming Mentor; the comparison does not isolate the Decider’s effect.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.