Learn by doing.
Eight ways to explore machine learning. Change a parameter, follow a pattern, and see the idea come to life.
K-Means
Move from scattered points to structure. Watch centroids and cluster assignments converge, one iteration at a time.
Launch playgroundPCA
Find the directions of maximum variance and see a 3D dataset projected into two dimensions.
Launch playgroundDecision trees
Build interpretable decision rules from CSV data. Explore splits, impurity, and the path an example takes through a tree.
Launch playgroundNeural networks
Draw a digit, inspect a prediction, and teach the network. See activations, weights, and gradients change through backpropagation.
Launch playgroundGradient descent
Follow an optimizer across a loss landscape. Compare gradient descent, Momentum, Nesterov, and Adam.
Launch playgroundManifold learning
Explore 3D shapes and illustrative 2D mappings. Use rotation and color to follow how a surface changes under projection.
Launch playgroundAttention
Follow tokens through a query, key, and value pipeline. Compare attention patterns and inspect a token-to-token matrix.
Launch playgroundObject detection
Explore the original browser-based object-detection demo, hosted on GitHub Pages.
Open external demo