Large language model
Also known as: LLM, LLMs, foundation model
What is large language model?
A large language model is a neural network trained on very large text corpora to predict the next token in a sequence. That single capability, at scale, produces summarisation, translation, classification, code generation and reasoning-like behaviour.
For business purposes an LLM is best understood as a probabilistic text transformer with broad general knowledge and no reliable memory of your data unless you supply it.
Practical implications
- Output varies between runs unless temperature is fixed
- Knowledge has a training cutoff
- Accuracy on your domain depends on the context you provide, not the model alone
Last reviewed by the gAIcko editorial team.