arXiv:2604. 14575v2 Announce Type: replace-cross Abstract: Large language models enable inexpensive AI-generated annotations, but using them reliably for causal inference remains challenging.
By Cheng Lu, Mengxin Wang, Dennis J. Zhang, Heng Zhang
The paper introduces Debiased Inference with Multiple Imperfect Measurements (DMM), a framework that uses several error‑prone AI measurements to perform valid downstream statistical inference without requiring costly gold‑standard labels. By assuming conditional independence of the measurements given the true label and unit‑level features, DMM leverages CP decomposition and semiparametric theory to prove consistency and asymptotic normality of its estimator. Simulations demonstrate that DMM yields valid inference and can improve efficiency when additional imperfect measurements are available, and the authors provide diagnostics for the key independence assumption.
By Naoki Egami, Sooahn Shin
arXiv:2607. 16239v1 Announce Type: new Abstract: AI judges offer a scalable, low-cost alternative to human evaluation, but their outputs can be biased relative to human preferences and highly item-dependent, varying across judges, tasks, and domains.
By Lei Shi, Anlan Zhang, Rita Lyu, Zhengmian Hu, Tong Yu, David Arbour, Avi Feller, Saayan Mitra, Ritwik Sinha
arXiv:2604. 17267v2 Announce Type: replace Abstract: Large Language Models can generate synthetic survey responses at low cost, but their accuracy varies unpredictably across questions.
By Zikun Ye, Hema Yoganarasimhan
The paper introduces an assumption‑lean framework that uses AI‑generated measurements as weak shadow variables to identify and infer population quantities when data are missing not at random. Weak shadow variables are outcome‑informative proxies that are conditionally independent of missingness given the true outcome and covariates, and they do not need to predict missing outcomes accurately. The authors derive sharp bounds via linear programs and propose a localized penalized estimator with a subsampling algorithm for confidence intervals, demonstrating in semi‑synthetic experiments that the resulting intervals are substantially narrower and more accurate than classical MNAR methods.
By Hongyu Chen, David Simchi-Levi, Ruoxuan Xiong
arXiv:2601. 20819v2 Announce Type: replace-cross Abstract: Machine learning predictions are increasingly used to supplement incomplete or costly-to-measure outcomes in fields such as biomedical research, environmental science, and social science.
By Yilin Song, Dan M. Kluger, Harsh Parikh, Tian Gu