Large Language Models (LLMs) are evaluated for table editing tasks using the WikiTableEdit dataset, showcasing challenges in editing irregular tables.