# LLM Assisted Automated Item Generation - Challenges - Difficulty - Variety - Chan, X., Wang, X., Yu, D., Mi, H., and Yu, D. Scaling synthetic data creation with 1,000,000,000 personas. arXiv preprint arXiv:2406.20094, 2024. - Evaluation is saturated -- synthesized items can be used to elevate model performance, but may cause overfitting. - The current item generations are mostly used in [[fine-tune|fine-tuning]] rather than in evaluations.