[KongchangAI]
ModelBinoculars检测器

Binoculars

A zero-shot AI-generated text detector that identifies text provenance by comparing metrics such as model log-likelihoods. It is relatively sensitive to variation in text length.

Timeline (last 90 days)

Nov 10

Taking the Binoculars detector as an example, a single threshold set at a length of 200 tokens causes the false positive rate to soar to 29% when the text is shortened to 10 tokens.

Unverified50%
Nov 10

On three zero-shot detectors, three corpora and two open-weight generators, averaged over 300 random document-level repartitions, binning calibration stabilized the average false positive rate at 5.0% in every detector-length combination from 10 tokens to 200 tokens.

Unverified50%
Nov 10

In the experiments, no detector reached an 80% true positive rate; the best-performing one, Binoculars, reached only 59% at 200 tokens.

Unverified50%

All Facts (3)

Source Articles