Large language model
A neural network trained on very large amounts of text to predict the next token, and so to read and write language.
Also: LLM
A large language model, or LLM, is a neural network with billions of adjustable numbers, called parameters, trained on a large share of the written web, books and code. Its training task is simple: given some text, guess the next token. Doing this well across trillions of examples forces the model to pick up grammar, facts, styles and some reasoning. After that pre-training, most models are fine-tuned to follow instructions and hold a conversation. GPT-3 in 2020 is often taken as the point where models became large enough to be called LLMs.