A data-driven method adaptively selects the number of LLM-simulated responses to form confidence sets with nominal coverage for human survey parameters and equates that number to the LLM's effective human-equivalent sample size.
write newline
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
stat.ME 3verdicts
UNVERDICTED 3representative citing papers
Links LDA discriminant directions to multivariate regression coefficients to enable a new framework for multi-class classification with excess risk bounds for regularized and non-parametric methods.
A model-free method builds confidence sets for latent parameters to proxy sim-to-real discrepancies and estimates the quantile function of that proxy to produce a distribution-level fidelity profile for simulators.
citing papers explorer
-
How Many Human Survey Respondents is a Large Language Model Worth? An Uncertainty Quantification Perspective
A data-driven method adaptively selects the number of LLM-simulated responses to form confidence sets with nominal coverage for human survey parameters and equates that number to the LLM's effective human-equivalent sample size.
-
A New Regression Lens on Multi-Class Classification
Links LDA discriminant directions to multivariate regression coefficients to enable a new framework for multi-class classification with excess risk bounds for regularized and non-parametric methods.
-
Model-Free Assessment of Simulator Fidelity via Quantile Curves
A model-free method builds confidence sets for latent parameters to proxy sim-to-real discrepancies and estimates the quantile function of that proxy to produce a distribution-level fidelity profile for simulators.