There is no hidden text
Artificial intelligence
Language models
Watermarking
Interactive
An interactive visual essay about how statistical watermarks can hide in the choices a language model makes.
k4v1t · MATHEMATICS · AI · SCIENCE
Working things out, visually
This site is a collection of visual explanations, interactive experiments, research notes and projects exploring mathematics, machine learning and scientific inference.
I am especially interested in what models are actually learning from data, where ambiguity remains, and how visualisation can make difficult ideas easier to reason about.
Some pieces explain ideas I have found useful; others are questions I am still working through.
Explore
Interactive essays
Visual explanations and experiments that let you adjust parameters, test assumptions and see how a system changes.
Research
PhD research at Queen Mary University of London on machine learning for autonomous optical networks.
Coming soon
Publications
Papers, preprints and other formally published work.
Coming soon
Projects & notes
Code, technical experiments and shorter explorations of ideas that are still developing.
Coming soon
Recent essays
No matching items
Current themes
The work here will mostly sit around scientific machine learning, inverse problems, mathematical intuition and the question of what can (or cannot) be learned from data.
I am more interested in developing better questions than presenting polished answers.