Temperature
In one sentence
A parameter that controls the randomness of an LLM's responses. Lower values produce more stable answers, while higher values lead to more varied outputs.
At lower temperatures, the model leans toward selecting the most probable words, giving similar answers to the same query each time. This is ideal for tasks requiring high accuracy, such as customer support, data organization, and contract review.
When the temperature is high, the model is more likely to choose less common expressions, making it great for brainstorming headlines and creative ideas. However, a higher temperature also makes it easier to go off-topic or make mistakes.
Example
To write 20 different event headlines, turn the temperature up; to extract an amount from an invoice, turn the temperature down.
Related terms
- Large Language Model (LLM)
- AI models trained on massive volumes of text that can understand and generate text. Doubao, DeepSeek, ChatGPT, and Claude are all powered by large language models.
- Hallucination
- When an AI presents non-existent or incorrect information with full confidence, such as fabricating data, sources, legal clauses, or product features.
- Prompt
- The instructions and reference materials provided to an AI, specifying what it should do, what information to use, and what format to follow.
