Glossary Entry

Transformer

A neural network architecture built around attention mechanisms that became the foundation for many modern language models.

Architecture LLMs

Seed source: Google ML Glossary

Transformers process relationships between tokens in parallel rather than strictly one step at a time. That design made them especially effective for large-scale language modeling and many downstream NLP tasks.

Much of the current machine learning tooling around chatbots, retrieval, summarization, and fine-tuning builds on transformer-based models.