Xu, Yifan, Xichen Ye, Yifan Chen, and Qiaosheng Zhang. “When Human Preferences Flip: An Instance-Dependent Robust Loss for RLHF”. Proceedings of the AAAI Conference on Artificial Intelligence 40, no. 44 (March 14, 2026): 38057–38065. Accessed May 14, 2026. https://ojs.aaai.org/index.php/AAAI/article/view/41143.