The importance of being empty: a spectral approach to Hopfield neural networks with diluted examples

Fuente: arXiv
Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Agliari, Elena, Fachechi, Alberto, Luongo, Domenico
Format: Preprint
Veröffentlicht: 2025
Schlagworte:
Online-Zugang:
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
_version_ 1866909846595436544
author Agliari, Elena
Fachechi, Alberto
Luongo, Domenico
author_facet Agliari, Elena
Fachechi, Alberto
Luongo, Domenico
contents We consider Hopfield networks, where neurons interact pair-wise by Hebbian couplings built over $i$. a set of definite patterns (ground truths), $ii$. a sample of labeled examples (supervised setting), $iii$. a sample of unlabeled examples (unsupervised setting). We focus on the case where ground-truths are Rademacher vectors and examples are noisy versions of these ground-truths, possibly displaying some blank entries (e.g., mimicking missing or dropped data), and we determine the spectral distribution of the coupling matrices in the three scenarios, by exploiting and extending the Marchenko-Pastur theorem. By levering this knowledge, we are able to analytically inspect the stability and attractiveness of the ground truths, as well as the generalization capabilities of the networks. In particular, as corroborated by long-running Monte Carlo simulations, the presence of black entries can have benefits in some specific conditions, suggesting strategies based on data sparsification; the robustness of these results in structured datasets is confirmed numerically. Finally, we demonstrate that the Hebbian matrix, built on sparse examples, can be recovered as the fixed point of a gradient descent algorithm with dropout, over a suitable loss function.
format Preprint
id arxiv_https___arxiv_org_abs_2503_15353
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle The importance of being empty: a spectral approach to Hopfield neural networks with diluted examples
Agliari, Elena
Fachechi, Alberto
Luongo, Domenico
Disordered Systems and Neural Networks
Mathematical Physics
We consider Hopfield networks, where neurons interact pair-wise by Hebbian couplings built over $i$. a set of definite patterns (ground truths), $ii$. a sample of labeled examples (supervised setting), $iii$. a sample of unlabeled examples (unsupervised setting). We focus on the case where ground-truths are Rademacher vectors and examples are noisy versions of these ground-truths, possibly displaying some blank entries (e.g., mimicking missing or dropped data), and we determine the spectral distribution of the coupling matrices in the three scenarios, by exploiting and extending the Marchenko-Pastur theorem. By levering this knowledge, we are able to analytically inspect the stability and attractiveness of the ground truths, as well as the generalization capabilities of the networks. In particular, as corroborated by long-running Monte Carlo simulations, the presence of black entries can have benefits in some specific conditions, suggesting strategies based on data sparsification; the robustness of these results in structured datasets is confirmed numerically. Finally, we demonstrate that the Hebbian matrix, built on sparse examples, can be recovered as the fixed point of a gradient descent algorithm with dropout, over a suitable loss function.
title The importance of being empty: a spectral approach to Hopfield neural networks with diluted examples
topic Disordered Systems and Neural Networks
Mathematical Physics
url https://arxiv.org/abs/2503.15353