A method estimates unsigned CATE from residual outcome covariances to assist randomization tests without sample splitting, establishing identification, consistency, and validity while showing higher power in simulations.
arXiv preprint arXiv:2501.07722 , year=
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 2verdicts
UNVERDICTED 2roles
background 1polarities
background 1representative citing papers
Including LLM predictions as covariates in standard regression adjustment for randomized experiments reduces variance with a do-no-harm property that reverts to the unadjusted estimator when predictions are uninformative.
citing papers explorer
-
Fit CATE Once: Model-Assisted Randomization Tests Without Sample Splitting
A method estimates unsigned CATE from residual outcome covariances to assist randomization tests without sample splitting, establishing identification, consistency, and validity while showing higher power in simulations.
-
AI-Assisted Variance Reduction in Randomized Experiments
Including LLM predictions as covariates in standard regression adjustment for randomized experiments reduces variance with a do-no-harm property that reverts to the unadjusted estimator when predictions are uninformative.