Training
Neural Network
A computing system inspired by the brain, made up of layers of connected "neurons" that learn patterns from data — the building block of modern AI.
01 ——
In plain English
A neural network is a mathematical model loosely inspired by how brain cells connect. It consists of layers of artificial "neurons" (simple calculations) connected by weighted links. By adjusting those weights based on training data, the network learns to map inputs (an image, a sentence) to outputs (a label, a translation, a generated response).
Key concepts:
- Layers — input layer, hidden layers (where the magic happens), output layer
- Weights — the numbers that get tuned during training; modern models have billions
- Activation function — decides whether each neuron "fires"
- Backpropagation — the algorithm that updates weights based on errors
Modern variants:
- CNNs — for images
- RNNs / LSTMs — for sequences (largely replaced by transformers)
- Transformers — the architecture behind nearly every modern AI breakthrough
Every LLM, image generator, and recommendation engine you use is a neural network.
02 ——
Related terms
Deep Learning
A type of machine learning that uses layered neural networks to learn complex patterns — the foundation of modern AI.
Machine Learning
A type of AI where systems learn patterns from data rather than being explicitly programmed with rules.
Transformer
The neural network architecture introduced in 2017 that powers nearly every modern LLM, image generator, and AI breakthrough.
AI
Artificial Intelligence — software that performs tasks normally requiring human intelligence, like understanding language, recognising images, or generating content.
Embeddings
A way of converting text (or images) into lists of numbers so an AI can measure how similar two pieces of content are.
Back to glossaryLast reviewed May 2026