JUMP TO ↑ ↓ to select · enter to open

Loading search

Loading the index…

Menu

Transformer Explainer

Tokens flow through embeddings, attention heads, and a transformer block.
Open standalone ↗Read the write-up →

GPT-2 · 626 MB model download on desktop · preset examples on mobile

Adapted from work by Polo Club of Data Science and collaborators · MIT.

← All labs