Abstract page for arXiv paper 2608.23705: The Limits of Automatic Evaluation of Creativity in Large Language Models