How a transformer works, with a real GPT you can run
Type a sentence and watch a real trained GPT predict the next word and show where each of its attention heads is looking.
Interactive explainers
See how AI works by using it
Each of these takes one idea and lets you change it and watch what happens. They run in your browser, with no account and nothing to install.
Neural network playground: train one in your browser
Pick a dataset and a network shape, press Train, and watch a real neural network learn to separate two classes, right on this page.
Attention in transformers, explained interactively
Pick a word and see how it shares its attention across a sentence, then see what multiple heads and the causal mask add.
Temperature, top-k and top-p, explained visually
Three settings decide how a language model turns probabilities into words. Move them, watch the odds change, then sample.
Gradient descent, visualized
Step down a curve yourself and see why the learning rate matters, how valleys trap you, and why plain gradient descent zig-zags.
Backpropagation, step by step
Run a tiny network forward, then watch blame flow backward through it, one multiplication at a time.
ROC curve, precision and recall, explained interactively
Slide a classifier's threshold and watch precision, recall and the ROC curve respond, then see why accuracy can tell you nothing.
Overfitting, visualized
Raise a model's complexity and watch training error keep falling while error on new data turns back up. Then see bias, variance and regularization at work.