Spectral Estimators for Multi-Index Models: Precise Asymptotics and Optimal Weak Recovery

TL;DR

This paper provides a detailed asymptotic analysis of spectral estimators in multi-index models, revealing phase transitions and enabling optimal data preprocessing for weak subspace recovery.

Contribution

It offers the first precise asymptotic characterization of spectral estimator performance in multi-index models, identifying phase transitions and optimizing sample complexity.

Findings

Eigenvalues separate from the bulk at a critical sample size

Eigenvector overlaps indicate successful subspace recovery

Optimal preprocessing reduces the required sample size for weak recovery

Abstract

Multi-index models provide a popular framework to investigate the learnability of functions with low-dimensional structure and, also due to their connections with neural networks, they have been object of recent intensive study. In this paper, we focus on recovering the subspace spanned by the signals via spectral estimators -- a family of methods routinely used in practice, often as a warm-start for iterative algorithms. Our main technical contribution is a precise asymptotic characterization of the performance of spectral methods, when sample size and input dimension grow proportionally and the dimension $p$ of the space to recover is fixed. Specifically, we locate the top-$p$ eigenvalues of the spectral matrix and establish the overlaps between the corresponding eigenvectors (which give the spectral estimators) and a basis of the signal subspace. Our analysis unveils a phase transition phenomenon in which, as the sample complexity grows, eigenvalues escape from the bulk of the spectrum and, when that happens, eigenvectors recover directions of the desired subspace. The precise characterization we put forward enables the optimization of the data preprocessing, thus allowing to identify the spectral estimator that requires the minimal sample size for weak recovery.