A pretty network visualization inspired by a codepen post I stumbled upon.
3D-network-vis/
│
├── index.html
├── README.md
├── package.json (optional later)
│
├── assets/
│ ├── css/
│ │ └── styles.css
│ └── shaders/
│ ├── nodeVertex.glsl
│ ├── nodeFragment.glsl
│ ├── connectionVertex.glsl
│ └── connectionFragment.glsl
│
├── src/
│ ├── main.js ← entry point
│ ├── scene/
│ │ ├── sceneSetup.js
│ │ ├── camera.js
│ │ └── renderer.js
│ │
│ ├── network/
│ │ ├── Node.js
│ │ ├── generator.js ← YOUR math lives here
│ │ └── formations.js
│ │
│ ├── visualization/
│ │ ├── nodes.js
│ │ ├── connections.js
│ │ └── materials.js
│ │
│ ├── interaction/
│ │ ├── controls.js
│ │ └── pulses.js
│ │
│ └── utils/
│ ├── noise.js
│ └── math.jsThis mirros you you'd structure:
- MATLAB projects
- Python scientific code
Paste shader code in:
assets/shaders/nodeVertex.glsl
assets/shaders/nodeFragment.glslPaste shader code here.
Then load them:
javascript async function loadShader(path) { const res = await fetch(path); return await res.text(); }
Why this matters for me: This separation lets me think in layers
| Layer | Function |
|---|---|
generator.js |
math/physics |
| shaders | field dynamics |
| visualization | mapping |
main.js |
orchestration |
@misc{3D_nn_visualization,
author = {Techartist},
title = {3D Quantum Neural Network},
year = {2025},
howpublished = {\url{https://codepen.io/VoXelo/pen/dPMeGze}},
note = {One single HTML File}
}