Glossary: Vision and Image Generation

Transformer

The network design behind today's language models, whose layers let every item in a sequence take every other into account.