Mini toolsDefinition
A parameter that controls the randomness and creativity in AI model outputs with higher values producing more diverse but potentially less focused responses.
Detailed Explanation
Temperature is a hyperparameter that affects the probability distribution during text generation. At low temperatures (near 0) the model becomes more deterministic always choosing the most likely next token. At higher temperatures the model becomes more random sampling from a broader range of possibilities. This parameter effectively controls the trade-off between consistency and creativity.
Use Cases
Creative writing applications conversational diversity brainstorming sessions and variable response generation
