jevbooks

← All projects

jev-paint

achimala/jev-paint

Paint images by having Jev predict each pixel's colour.

Game & Simulation98%Runner-up: Other

What it is

A local browser app that turns Jev's pixel probability distributions into paintings. Aimed at people with a TypeSafe Jev API key who want to run a small creative experiment on their own machine.

How it uses Jev

For each pixel location on the chosen grid, the app sends typed questions about palette or HSL/RGB channels and asks Jev for calibrated probabilities. Four concurrent batches are validated, then spatially correlated categorical sampling picks stroke pigments, mean colour supplies the underpainting, and entropy controls relief.

Primitives:choicescore

Technique worth stealing

Use per-pixel Jev probabilities as sampling weights, letting predicted confidence decide stroke width, then combine independent channel probabilities into a painting.

Try it

Run `python3 server.py`, open http://127.0.0.1:8791, paste your Jev API key, pick a representation and grid, then enter a prompt.

View on GitHub

judged by Jevjev-1.13.0

Evidence

Each line is one question put to Jev about the README. ≥ 0.60 reads as yes, ≤ 0.40 as no; in between Jev is not making a call.

  • Jev-centricunclear0.51
  • Shows a System One patternyes0.65
  • Handles uncertaintyno0.06
  • Measuredno0.04
  • Runnableyes0.93
  • Worth recommendingunclear0.43
  • Model replicano0.03
  • Problem scopescore on a 0–2 scale0.00
  • About Jevyes0.98

Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.