Neural Network
A web of tiny maths units, loosely inspired by brain cells, that pass signals to each other.
In everyday terms
Each unit takes some numbers, weighs them, and passes a result on. Stack millions in layers and the network can learn very complex patterns.
For professionals
Layers of affine transforms followed by non-linear activations, trained with backpropagation. The basis of nearly all modern AI.
Think of it like…
A huge office where each worker reads notes from the previous desk, decides how much each note matters, and passes a summary forward.
You've already seen it
Every major AI breakthrough since 2012 (image recognition, translation, chatbots) is a neural network.
Myth vs reality
Myth: Neural networks work like the human brain.
Reality: The inspiration is loose. Real neurons are far more complex.
Quick check
What does each "neuron" in a neural network do?
- Stores a memory
- Weighs its inputs and passes on a number
- Thinks a thought
- Connects to the internet
Show answer
Weighs its inputs and passes on a number: It is a small weighted-sum-and-squash operation.