WATCH INTELLIGENCE UNFOLD
Every concept — from Alan Turing's test to a Transformer's attention head —
stepped through one frame at a time. Press play and watch the idea unfold.
↑ Click to place data points. Decision boundary updates in real time.
HOW IT WORKS
AI Visualised teaches each concept the same way — by animating it. Not text-book definitions. Live, interactive visualizers that let you poke, drag, and configure the math.
Start from History, jump to Neural Networks, or go straight to Modern AI. Learn at your own order.
Every module has a live sandbox. Click, drag sliders, add layers, type text — the model reacts.
See the math in motion. K selects neighbors. Signals pulse through neurons. Attention weights glow.
PICK A MODULE
AI Visualised covers the full arc — from its philosophical birth to the transformer architecture powering today's LLMs.
Mathematical roots to the GenAI era
Walk an interactive timeline spanning 8 landmark epochs — from Turing's 1950 paper to ChatGPT. Each milestone reveals who made it happen, the Fun Fact you won't find in textbooks, and the long-term ripple that shaped modern AI. Includes the AI Winters, the Expert Systems boom, the Deep Learning revolution, and the Transformer breakthrough.
K-Nearest Neighbors — live decision boundary
Place pink and blue training points on a 2D canvas. Tune the K slider from 1 to 15. Watch the decision boundary repaint in real time as the algorithm recalculates the majority vote for every pixel. Understand why small K overfits, large K over-smooths, and how Euclidean distance drives it all.
Build layers. Fire signals. Watch the math move.
Assemble a deep learning network from scratch — add hidden layers, resize neuron counts from 2 to 6 per layer, and switch between ReLU, Sigmoid, and Tanh activation functions. Hit "Run Forward Pass" and watch weighted signals cascade layer by layer, color-coded by activation strength. Understand backpropagation, vanishing gradients, and why ReLU dominates modern AI.
Transformers, self-attention & the LLM revolution
Type any sentence and watch how each word attends to every other word through a glowing attention heatmap. Switch between 3 presets (river bank vs financial bank) to see how the same word "bank" builds completely different context vectors. Learn Q/K/V projections, Multi-Head Attention, positional encoding, and the scaling laws that power GPT-4 and Gemini.
START NOW · FREE
No login. No paywall. Just open the playground and start clicking. The best way to understand AI is to interact with it.