Large Language Models
Temperature
Large Language Models· Intermediate
Definition
A parameter controlling LLM output randomness by scaling the probability distribution over tokens before sampling. Low temperature (near 0) produces deterministic, focused output; high temperature produces more random, creative output.
Tags
#sampling#randomness#creativity#inference
MS
Maxx Stacks Editorial
Reviewed by enterprise AI practitioners
Maxx University
Keep learning. Keep building.
250+ terms. 5 learning paths. AI maturity assessment. Jargon translator. All free, always.