Moment-based method identifies regression coefficient beta and LDA concentration alpha_0 from response-weighted corrected word moments and commutativity, without per-document topic estimation.
Inference for regression with variables generated by ai or machine learning.arXiv preprint arXiv:2402.15585
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
representative citing papers
A coupled-label bootstrap provides valid inference for OLS regressions that use AI/ML-generated binary labels despite misclassification errors, unlike standard fixed-label bootstraps.
A framework that uses LLM sparse autoencoder features as high-dimensional outcome variables, with new k-FWER controlling bootstrap tests and formalized autointerp evaluation, to make interpretable discoveries from unstructured data.
A three-tower embedding model fine-tuned from Fashion CLIP combined with a latent-class deep demand system captures heterogeneous consumer aesthetics, price sensitivities, and substitution patterns from large-scale retail transaction data.
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
-
Moment-Based Inference for Regression with Latent Dirichlet Covariates
Moment-based method identifies regression coefficient beta and LDA concentration alpha_0 from response-weighted corrected word moments and commutativity, without per-document topic estimation.
-
Bootstrapping with AI/ML-generated labels
A coupled-label bootstrap provides valid inference for OLS regressions that use AI/ML-generated binary labels despite misclassification errors, unlike standard fixed-label bootstraps.
-
Making Interpretable Discoveries from Unstructured Data: A High-Dimensional Multiple Hypothesis Testing Approach
A framework that uses LLM sparse autoencoder features as high-dimensional outcome variables, with new k-FWER controlling bootstrap tests and formalized autointerp evaluation, to make interpretable discoveries from unstructured data.
-
A Deep Learning Approach to Heterogeneous Consumer Aesthetics in Fast Fashion
A three-tower embedding model fine-tuned from Fashion CLIP combined with a latent-class deep demand system captures heterogeneous consumer aesthetics, price sensitivities, and substitution patterns from large-scale retail transaction data.
-
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.