Large language model
- In French:
- Grand modèle de langage
- Abbreviation:
- LLM
Definition
A neural network of the Transformer type, pretrained on very large volumes of text to model language. Beyond a certain size, these models show abilities that smaller models lack. In robotics, they are used in particular to generate code or to provide common-sense reasoning for planning a task.
Checked on against the references below.
This term belongs to the software that drives robots: our AI agents section follows how it is used.