AI glossary
Large language model (LLM)
A large language model is a neural network trained on a great deal of text to predict the next token, which lets it write, summarise, translate and answer questions.
An LLM reads your text as tokens and, one token at a time, picks a likely next token until it stops. Everything it does, from answering a question to writing code, is that one step repeated.
Its size is counted in parameters, the numbers learned in training. More parameters usually means better answers, more memory and slower replies, which is why a small model on a laptop and a large one behind an API feel so different.
A common mistake is to treat it as a database. It has no list of facts to look up: it produces text that is likely, and likely is not the same as true, so anything that matters is checked.
Learn it hands-on
These free lessons teach it with commands to run in a lab and checked tasks. Read them whole with no account.