Signal Check Research Prototype

Statistical analysis of machine-generated text

Paste a passage below. Our experimental model returns a score from 0–100 reflecting how strongly the text carries statistical patterns associated with machine generation.

This score is not proof of anything. It is an experimental signal with an unmeasured error rate. It cannot establish who wrote a passage, and it should never be used on its own to accuse a student, employee, or writer of misconduct.

Text to analyze

0 characters · 0 words

What this tool can and can't do

The score is a signal, not a verdict. A high number means the text resembles patterns the model was built around — not that a machine wrote it.
False positives are real and unevenly distributed. Tools of this kind have repeatedly misfired on non-native English writers, technical and formulaic prose, and heavily edited drafts.
Short passages are close to noise. Below roughly 300 characters, treat any result as meaningless.
Light editing can change the outcome. Paraphrasing, translation, or a grammar checker can move a score substantially in either direction.
Accuracy has not been independently validated. No published false-positive rate accompanies this prototype. Weigh it accordingly.
Don't make consequential decisions on this alone. Grades, jobs, publication, and accusations require evidence this tool does not provide.