Built for scientific writing
Our own model, made for the way scientists write: introductions, methods, results, discussions and reviews.
Scientific prose is dense, formal and full of set phrases, which makes it hard to judge. Witfire™ AIDetects is our own model, built and tested on scientific writing, and it will be free to use.
Witfire™ AIDetects looks at how a passage is written: its rhythm, its habits and its choices, the things that separate a scientist's paragraph from a machine-drafted one.
Our own model, made for the way scientists write: introductions, methods, results, discussions and reviews.
Each paragraph gets its own reading, so one machine-drafted paragraph does not colour a whole document.
Short passages carry less evidence. When a passage is too short to judge, it says so instead of guessing.
Witfire™ AIDetects and Witfire™ Scientific Humanizer are kept apart: neither is built or tuned against the other.
When Witfire™ AIDetects opens, we will publish how it was tested: on which writing, how often it wrongly flags human text, and where it is weaker. Until then, we would rather say nothing than say something we have not measured.
An AI detector is one reading of a text, not a judgement on its author. We will say so on every result.
Yes. It will be free to use.
Soon. Write to support@kriyacortex.com and we will tell you the day it opens.
Scientific writing: introductions, methods, results, discussions and reviews. That is the writing it is built and tested on.
No. Each passage is read in memory and is not stored once you have the result.
No. No AI detector can prove authorship. Witfire™ AIDetects gives a reading of how a passage is written, which is one piece of evidence and never a verdict on a person.