@article{aievaluationshouldrequirestandardizedite, title = {AI Evaluation Should Require Standardized Item-Level Data Releases}, author = {Han Jiang and Susu Zhang and Dongyao Zhu and Yuzhuo Bai and Sang T. Truong and Xiaoyuan Yi and Sanmi Koyejo and Xing Xie and Ziang Xiao}, year = {2026}, eprint = {2604.03244}, archivePrefix = {arXiv}, url = {https://arxiv.org/abs/2604.03244}, }