Is temperature the creativity parameter of large language models?

Large language models (LLMs) are applied to all sorts of creative tasks, and their outputs vary from beauti ful, to peculiar, to pastiche, into plain plagiarism. The temperature parameter of an LLM regulates the amount of randomness, leading to more diverse outputs; there fore, it is often claimed to be the creativity parameter. Here, we investigate this claim using a narrative genera tion task with a predetermined fixed context, model and prompt. Specifically, we present an empirical analysis of the LLM output for different temperature values using four necessary conditions for creativity in narrative gen eration: novelty, typicality, cohesion, and coherence. We find that temperature is weakly correlated with novelty, and unsurprisingly, moderately correlated with incoher ence, but there is no relationship with either cohesion or typicality. However, the influence of temperature on cre ativity is far more nuanced and weak than suggested by the “creativity parameter” claim; overall results suggest that the LLM generates slightly more novel outputs as temperatures get higher. Finally, we discuss ideas to al low more controlled LLM creativity, rather than relying on chance via changing the temperature parameter.