Transformer

The neural network architecture, based on attention, behind modern large language models.

Transformers process sequences in parallel, unlike older recurrent networks.

Related terms