Skip to content
BENIBIX Learn

LLM Temperature

AI💬 LLMsText Generation

Also known as Temperature, Sampling Temperature

A setting for how random the model's word picks are. Low is steady, high is wild.

Think of it like

Think of it like a dial from calm to wild. Turn it down for safe words and up for surprise.

Example

Emma sets temperature to 0 to pull dates out of invoices, so she gets the same result each time. For slogan ideas she uses 0.9 and gets more variety.

Why it matters

It is the first knob to turn when answers are too dull or too wild for the job.

Definition

Temperature is a setting that controls how random a model's word choices are. A low value, near 0, makes the most likely token win almost every time, so answers are steady and plain. A value of 1 keeps the chances as the model gave them. Higher values flatten the chances, so rare tokens get picked more and the text gets more varied, then strange. Most tools let you set it between 0 and 1 or 2.
The same chances at three temperatures. Low makes one word tower; high makes the bars nearly flat.

Common mistake

Thinking 0 makes the model correct. It makes it consistent, which is not the same; a wrong answer just repeats.

Key takeaways

  1. 1
    Low is steady. The top token wins.
  2. 2
    High is varied. Rare tokens get picked.
  3. 3
    Pick per job. Facts low, ideas higher.

Compare

Learn it properly