MMODELYST
Papers/Semiparametric Efficient Bilevel Gradient Estimation
PAP

Semiparametric Efficient Bilevel Gradient Estimation

May 20, 2026

arXiv
Abstract

Functional bilevel methods estimate a lower-level function and plug it into a hypergradient, but this plug-in gradient can retain first-order bias when the lower-level problem is learned nonparametrically. To remove this bias, we develop a semiparametric debiasing theory for population bilevel gradients based on the efficient influence function. This perspective leads to a cross-fitted orthogonal hypergradient estimator for which we establish asymptotic normality together with uniform control over the outer parameter. Under quadratic losses, the estimator reduces to a simple doubly robust score based on conditional mean nuisances. On synthetic bilevel benchmarks with known ground truth, the method tracks the oracle efficient-gradient benchmark and improves over plug-in functional hypergradients and regularized kernel bilevel baselines.

Select text to highlight · click a highlight to remove · saved in this browser only
Authors
Fares El Khoury, Houssam Zenati, Nathan Kallus, Michael Arbel, Aurélien Bibaut
Your notes (browser-local)
saved
arXiv:2605.21341