Demo only, not production detectors. Both detectors below are transparent keyword heuristics, run entirely in your browser. Detector A is the repo's real weak baseline. Detector B is an illustrative stricter heuristic, added only to show the detection vs false-positive tradeoff.
Dataset: huggingface.co/datasets/RishabhPande/fraud-eval (v0.1.0 smoke test, 20 cases) · Code: github.com/RishabhPande/fraud-eval
Which detector handled this message better?
Votes in this browser session: A 0, B 0, Tie 0. Votes are local only — this is a free Static Space with no backend, so votes are NOT counted globally and do NOT affect the leaderboard.