High storage capacity in the Hopfield model with auto-interactions—stability analysis. (23rd October 2017)

• Record Type: Journal Article
• Authors: Rocchi, Jacopo; Saad, David; Tantari, Daniele
• Is Part Of: Journal of physics. Volume 50:Number 46(2017)
• Publication Date: 2017-10-23
• Subjects: recurrent neural networks -- Hopfield model -- retrieval ability -- capacity; Mathematical physics -- Periodicals; Statistical physics -- Periodicals; Quantum theory -- Periodicals; Matter -- Properties -- Periodicals; 530.105
• DOI: 10.1088/1751-8121/aa8fd7
• Languages: English
• ISSNs: 1751-8113

Abstract

Abstract: Recent studies point to the potential storage of a large number of patterns in the celebrated Hopfield associative memory model, well beyond the limits obtained previously. We investigate the properties of new fixed points to discover that they exhibit instabilities for small perturbations and are therefore of limited value as associative memories. Moreover, a large deviations approach also shows that errors introduced to the original patterns induce additional errors and increased corruption with respect to the stored patterns.