Daydreaming Hopfield Networks and their surprising effectiveness on correlated data

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Serricchio, Ludovica, Bocchi, Dario, Chilin, Claudio, Marino, Raffaele, Negri, Matteo, Cammarota, Chiara, Ricci-Tersenghi, Federico
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912404417282048
author Serricchio, Ludovica
Bocchi, Dario
Chilin, Claudio
Marino, Raffaele
Negri, Matteo
Cammarota, Chiara
Ricci-Tersenghi, Federico
author_facet Serricchio, Ludovica
Bocchi, Dario
Chilin, Claudio
Marino, Raffaele
Negri, Matteo
Cammarota, Chiara
Ricci-Tersenghi, Federico
contents To improve the storage capacity of the Hopfield model, we develop a version of the dreaming algorithm that perpetually reinforces the patterns to be stored (as in the Hebb rule), and erases the spurious memories (as in dreaming algorithms). For this reason, we called it Daydreaming. Daydreaming is not destructive and it converges asymptotically to stationary retrieval maps. When trained on random uncorrelated examples, the model shows optimal performance in terms of the size of the basins of attraction of stored examples and the quality of reconstruction. We also train the Daydreaming algorithm on correlated data obtained via the random-features model and argue that it spontaneously exploits the correlations thus increasing even further the storage capacity and the size of the basins of attraction. Moreover, the Daydreaming algorithm is also able to stabilize the features hidden in the data. Finally, we test Daydreaming on the MNIST dataset and show that it still works surprisingly well, producing attractors that are close to unseen examples and class prototypes.
format Preprint
id arxiv_https___arxiv_org_abs_2405_08777
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Daydreaming Hopfield Networks and their surprising effectiveness on correlated data
Serricchio, Ludovica
Bocchi, Dario
Chilin, Claudio
Marino, Raffaele
Negri, Matteo
Cammarota, Chiara
Ricci-Tersenghi, Federico
Disordered Systems and Neural Networks
To improve the storage capacity of the Hopfield model, we develop a version of the dreaming algorithm that perpetually reinforces the patterns to be stored (as in the Hebb rule), and erases the spurious memories (as in dreaming algorithms). For this reason, we called it Daydreaming. Daydreaming is not destructive and it converges asymptotically to stationary retrieval maps. When trained on random uncorrelated examples, the model shows optimal performance in terms of the size of the basins of attraction of stored examples and the quality of reconstruction. We also train the Daydreaming algorithm on correlated data obtained via the random-features model and argue that it spontaneously exploits the correlations thus increasing even further the storage capacity and the size of the basins of attraction. Moreover, the Daydreaming algorithm is also able to stabilize the features hidden in the data. Finally, we test Daydreaming on the MNIST dataset and show that it still works surprisingly well, producing attractors that are close to unseen examples and class prototypes.
title Daydreaming Hopfield Networks and their surprising effectiveness on correlated data
topic Disordered Systems and Neural Networks
url https://arxiv.org/abs/2405.08777