Chapter notesFunctional testingEvidently AI · 5:09 · EnglishWatch on YouTubeVideo overviewEdit on GitHub Checks whether the LLM output actually works for its specific task 4:02 Example: if the LLM generates SQL queries or Python code, test that the code is valid — and possibly even run it in a sandbox 4:02 Previous chapterRegular expressions and text statisticsNext chapterMatching evaluation methods to test datasets