Technical GlossaryGenerative AI and LLM
Temperature Sampling
A parameter that adjusts output distribution sharpness to produce more controlled or more creative generation.
Temperature changes generation behavior by making the model probability distribution sharper or flatter. Lower values tend to produce safer and more repetitive outputs, while higher values increase diversity and risk. This makes temperature a critical control mechanism in scenarios such as creative writing, code generation, and idea exploration.
You Might Also Like
Explore these concepts to continue your artificial intelligence journey.
