← Back to Glossary
Temperature
A parameter controlling how randomly an LLM selects its next word — low values produce deterministic facts, high values produce creative outputs.
Why it exists
LLMs output probability distributions for the next token. Temperature scales these probabilities prior to sampling.
How it works
At temperature 0.0, the model always picks the single highest-probability token. At higher values (0.8+), lower probability tokens get a boost, introducing randomness.
Related terms
Where you'll use this
roadmap