Impractical
Visual essays about the clever machinery hiding inside technology, mathematics, and everyday things.
Latest explainers
View allWhat actually runs an ML model?
Model files, runtimes, tensor kernels, GPUs, memory bandwidth, quantization, and serving.
How an LLM produces a sentence
Tokens, attention, transformers, sampling, and the one-token-at-a-time generation loop.
How a neural network learns
Weighted sums, activations, loss, backpropagation, gradients, and repeated updates.
How video formats actually work
Frames, chroma, motion prediction, bitrate, codecs, and containers.
Interactive atlases
The Epics
Explore characters, families, alliances, and story arcs across the Ramayana and Mahabharata.