AI Glossary

The arrangement of words and phrases in a specific order to create meaning in language. If you change the position of one word, it is possible to change the context and meaning.

A parameter that controls the degree of randomness or unpredictability of the LLM output. A higher value means greater deviation from the input; a lower value means the output is more deterministic.

The use of AI to create coherent and contextually relevant written content. This can include anything from auto-completing sentences to writing entire paragraphs or articles.

The process of breaking down text into smaller pieces, like words, phrases, or symbols, to make it easier for machine learning models to analyse.

A unit of content corresponding to a subset of a word. Tokens are processed internally by LLMs and can also be used as metrics for usage and billing.

A collection of data used to train machine learning models. Training data helps the model learn patterns and make accurate predictions.

The collection of data used to train an AI model.

A model architecture at the core of most state of the art (SOTA) ML research. It is composed of multiple “attention” layers which learn which parts of the input data are the most important for a given task. Transformers started in NLP (specifically machine translation) and subsequently were expanded into computer vision, audio, and other modalities.

The use of AI to generate human-like speech from text input. This technology powers virtual assistants and text-to-speech applications.