Skip to content

Topic

LLM output variance

The tendency of large language models to return different answers, rankings and citations when the same prompt is run repeatedly, and the sampling methods used to quantify it.

Current clusters