The Capacity of Modern Hopfield Networks under the Data Manifold Hypothesis

2025/03/12 by Achilli, Beatrice, Ambrogioni, Luca, Lucibello, Carlo +2 · 4 citations
#Disordered Systems and Neural Networks (cond-mat.dis-nn) #FOS: Physical sciences

paper · doi:10.48550/arxiv.2503.09518

Abstract

We generalize the computation of the capacity of exponential Hopfield model from Lucibello and Mézard (2024) to more generic pattern ensembles, including binary patterns and patterns generated from a hidden manifold model.

Cited by

• Asymptotically Stable Quaternion-valued Hopfield-structured Neural Network with Periodic Projection-based Supervised Learning Rules
• Phase diagram and eigenvalue dynamics of stochastic gradient descent in multilayer neural networks
• Transient dynamics of associative memory models
• Dynamic Manifold Hopfield Networks for Context-Dependent Associative Memory

Related

• Modern Hopfield Networks Require Chain-of-Thought to Solve NC¹-Hard Problems
• Dynamic Capacity Estimation in Hopfield Networks
• Capacity of the Hebbian-Hopfield network associative memory
• Generalization in the Hopfield Model
• Hopfield Networks is All You Need