neural visualiser

• 

  result

• 

  interface

Neural networks are often taught as equations, matrices, and abstract math on slides. For a long time, that made them feel distant and intimidating — like something only experts could truly understand.

This project started from a simple frustration: “I know the code runs, but I can’t see what’s actually happening.”

I wanted to break that barrier.

Instead of treating a neural network as a black box, this project turns it into something visual and alive. Every neuron, every connection, every weight update is shown in real time, making the learning process tangible. You don’t just train a model — you watch it learn.

The goal isn’t just accuracy or performance. The goal is intuition. To help students, beginners, and curious minds understand why neural networks behave the way they do, not just that they work.

This project is my attempt to bridge the gap between theory and understanding — turning complex machine learning concepts into something you can explore, pause, and play with.

Because once you can see learning happen, it stops being magic and starts making sense.

Built With

• api
• architecture:
• backend:-python-3.8+
• canvas
• css3
• flask
• javascript
• numpy-frontend:-html5
• rendering:
• restful
• vanilla