Temperature
Controls how creative or deterministic AI responses are (low = predictable, high = creative).
Why it matters
Low temperature for code and data extraction. Higher temperature for creative content.
In practice
We use low temperature (0.1-0.3) for structured tasks, higher (0.7-0.9) for blog content suggestions.
Related terms
Inference
The process of an AI model generating a response or prediction from input data.
LLM (Large Language Model)
A large language model like Claude, GPT, or Gemini. The "brain" that understands and generates language.
Prompt
An instruction or question given to an AI model. Quality determines output quality.