RCTs & Human Uplift Studies: Methodological Challenges and Practical Solutions for Frontier AI Evaluation 文章

ArXiv CS.AI2026-05-26NEWSen作者: Patricia Paskov, Kevin Wei, Shen Zhou Hong, Dan Bateyko, Xavier Roberts-Gaal, Carson Ezell, Gailius Praninskas, Valerie Chen, Umang Bhatt, Ella Guest

查看原文 →

RCTs & Human Uplift Studies: Methodological Challenges and Practical Solutions for Frontier AI Evaluation · 相关技术

相关技术

iOS ORM PLA 研究法辛烷值 eval divide-and-conquer partitioning cybersecurity UCT Tera Straight-Through Estimator Stan SUL Referring expression comprehension (REC)For Effort Metric Attention CIS ANN