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.
Daydreaming Hopfield Networks and their surprising effectiveness on correlated data / Ludovica Serricchio, Dario Bocchi, Claudio Chilin, Raffaele Marino, Matteo Negri, Chiara Cammarota, Federico Ricci-Tersenghi. - In: NEURAL NETWORKS. - ISSN 0893-6080. - STAMPA. - 186:(2025), pp. 107216.1-107216.12. [10.1016/j.neunet.2025.107216]
Daydreaming Hopfield Networks and their surprising effectiveness on correlated data
Raffaele Marino;Federico Ricci-Tersenghi
2025
Abstract
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.File | Dimensione | Formato | |
---|---|---|---|
Daydreaming.pdf
accesso aperto
Tipologia:
Pdf editoriale (Version of record)
Licenza:
Creative commons
Dimensione
2.5 MB
Formato
Adobe PDF
|
2.5 MB | Adobe PDF |
I documenti in FLORE sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.