What an LLM Actually Is (and Isn't)

Automated excerpt

A large language model is a function that takes in some text and outputs a probability distribution over what token is most likely to come next. The model outputs a probability for every token in its vocabulary being the next one. "The capital of France is": { Paris: 0. 91, Lyon: 0. 03, France: 0. 02, "? ": 0. 01 }, "2 + 2 =": { "4": 0. 97, "5": 0. 01, "22": 0. 01 }, console. log(predictNext("The capital of France is")); // "Paris" — highest probability, not "known fact" console.

Selected automatically from source text; not independently written or fact-checked. Read the original for full context.