Prompts revealed that make AI text score 100% human on detectors
paulnovosad · x · 2026-10-02
Paul Novosad shares the prompts behind his AI-detector experiments: T4 produced text flagged as 100% human, T5 scored 60% AI-assisted / 40% human, while T6 — which never saw the French source, only an English bulleted summary — was flagged 100% AI-generated, exposing how detectors weigh translation vs. rewriting.
More from Research
- Marin's 535B MoE hero run hits ~27% MFU on 11 NVL72 racks: expert parallelism deep dive — dlwh · 2026-10-03
- SWE-sweep benchmark tests whether AI agents can find bugs before users hit them — klieret · 2026-10-02
- Harvard-led team unveils brain imaging 60x faster than fMRI, tracking activity in ~100ms — melnykowycz · 2026-10-02
- RExBench: best coding agent implements research extensions only 33% of the time — najoungkim · 2026-10-02
- Retrieval-augmented episodic memory narrows LLMs' lexical frequency gap in syntax tests — najoungkim · 2026-10-02
- EMPIRIC teaches robots missing physics as code, solving all 25 tasks where baselines get 14-16 — tomssilver · 2026-10-02