Interactive Labs

Run concepts in the browser.

WebAssembly notebooks plus graph tutorials — hands-on learning tied to papers and models.

Browser labs · 3

WASM

Self-attention from scratch

The mechanism behind every Transformer, in ~20 lines of NumPy — see exactly how query/key/value vectors turn into attention weights.

Open →
WASM

Cosine similarity playground

Embeddings work because similar meanings end up near each other in vector space. Edit the vectors and watch the similarity score move.

Open →
WASM

Tokenization explorer

Before a model sees any text, it gets chopped into tokens. This is a simplified byte-pair-style merge, showing the core idea behind real tokenizers like the one GPT/Claude use.

Open →