{"id":41980,"date":"2026-01-30T11:29:42","date_gmt":"2026-01-30T10:29:42","guid":{"rendered":"https:\/\/csw.agency\/glossar\/perplexity\/"},"modified":"2026-01-30T11:29:42","modified_gmt":"2026-01-30T10:29:42","slug":"perplexity","status":"publish","type":"glossar","link":"https:\/\/csw.agency\/en\/glossar\/perplexity\/","title":{"rendered":"Perplexity"},"content":{"rendered":"

Perplexity is a fundamental metric in the field of Artificial intelligence (AI)<\/a> and the Natural language processing (NLP)<\/a>, which evaluates the quality of a language model. It measures the uncertainty or \u201esurprise\u201c of a model when predicting a word sequence or the next word in a text. A low perplexity value means that the model can predict the text well and therefore has a high confidence in its predictions, resulting in more natural language generation.<\/p>\n

The importance of perplexity in language models<\/h2>\n

In the development and evaluation of Large Language Models (LLMs)<\/a> perplexity is a crucial indicator. It quantifies how well a model predicts a sample, in particular the order of words in a text. The metric is closely related to the concept of information theory, more specifically entropy. While entropy measures the average information per symbol, perplexity quantifies the uncertainty of a model in predicting the next word in a sequence.<\/p>\n