PROXIMA scores proxy reliability via a composite of effect correlation, directional accuracy, and segment fragility, achieving 98.4% decision agreement with an oracle on two public datasets.
From augmentation to decomposition: A new look at CUPED in 2023.arXiv preprint arXiv:2312.02935, 2023
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
years
2026 2verdicts
UNVERDICTED 2representative 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
-
PROXIMA: A Reliability Scoring Framework for Proxy Metrics in Online Controlled Experiments
PROXIMA scores proxy reliability via a composite of effect correlation, directional accuracy, and segment fragility, achieving 98.4% decision agreement with an oracle on two public datasets.
-
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.