Large Language Model (LLM): a neural network trained on enormous amounts of text to predict the next token. In practice we treat it as a powerful, general-purpose text-in / text-out engine that we ...