Share your thoughts, 1 month free Claude Pro on usSee more

AI-generated text detection on Peer-Review GPT-4o generated

33.04TPR @ 0.1% FPR

Calibrated DMAP

Updated 26d ago

Evaluation Results

Method	Links
Calibrated DMAP 2026.05		33.04	63.43	0.9734
Calibrated Log-Surprisal 2026.05		32.71	64.35	0.9544
DALD 2026.05		8.51	27.42	0.8523
Calibrated Token Fast-Detect 2026.05		1.48	24.35	0.8623
Calibrated Token NPR 2026.05		1.21	14.84	0.8089
DMAP 2026.05		0.49	7.95	0.8337
Calibrated Log-Rank 2026.05		0.22	13.72	0.8262
Baseline NPR 2026.05		0.16	1.82	0.5996
Baseline Fast-DetectGPT 2026.05		0.14	2.06	0.6397
Token Fast-Detect 2026.05		0.13	1.97	0.6403
Token NPR 2026.05		0.13	1.8	0.6157
Baseline Binoculars 2026.05		0.02	0.31	0.2887
Baseline Log-Surprisal 2026.05		0	0.52	0.6354
Baseline Log-Rank 2026.05		0	0.32	0.6021