Perplexity

In information theory, perplexity is a measurement of how well a probability distribution or probability model predicts a sample, and it is used to compare the predictive quality of different probability models. A lower perplexity score indicates that a model anticipates the observed data more accurately, so perplexity functions as a measure of the quality of a model's probabilistic predictions. This description is adapted from Wikipedia contributors under CC BY-SA 4.0; changes were made. https://creativecommons.org/licenses/by-sa/4.0/