ResearchTrend.AI
  • Papers
  • Communities
  • Events
  • Blog
  • Pricing
Papers
Communities
Social Events
Terms and Conditions
Pricing
Parameter LabParameter LabTwitterGitHubLinkedInBlueskyYoutube

© 2025 ResearchTrend.AI, All rights reserved.

  1. Home
  2. Papers
  3. 1912.00666
18
6

Interpolating between boolean and extremely high noisy patterns through Minimal Dense Associative Memories

2 December 2019
Francesco Alemanno
M. Centonze
A. Fachechi
ArXiv (abs)PDFHTML
Abstract

Recently, Hopfield and Krotov introduced the concept of {\em dense associative memories} [DAM] (close to spin-glasses with PPP-wise interactions in a disordered statistical mechanical jargon): they proved a number of remarkable features these networks share and suggested their use to (partially) explain the success of the new generation of Artificial Intelligence. Thanks to a remarkable ante-litteram analysis by Baldi \& Venkatesh, among these properties, it is known these networks can handle a maximal amount of stored patterns KKK scaling as K∼NP−1K \sim N^{P-1}K∼NP−1.\\ In this paper, once introduced a {\em minimal dense associative network} as one of the most elementary cost-functions falling in this class of DAM, we sacrifice this high-load regime -namely we force the storage of {\em solely} a linear amount of patterns, i.e. K=αNK = \alpha NK=αN (with α>0\alpha>0α>0)- to prove that, in this regime, these networks can correctly perform pattern recognition even if pattern signal is O(1)O(1)O(1) and is embedded in a sea of noise O(N)O(\sqrt{N})O(N​), also in the large NNN limit. To prove this statement, by extremizing the quenched free-energy of the model over its natural order-parameters (the various magnetizations and overlaps), we derived its phase diagram, at the replica symmetric level of description and in the thermodynamic limit: as a sideline, we stress that, to achieve this task, aiming at cross-fertilization among disciplines, we pave two hegemon routes in the statistical mechanics of spin glasses, namely the replica trick and the interpolation technique.\\ Both the approaches reach the same conclusion: there is a not-empty region, in the noise-TTT vs load-α\alphaα phase diagram plane, where these networks can actually work in this challenging regime; in particular we obtained a quite high critical (linear) load in the (fast) noiseless case resulting in lim⁡β→∞αc(β)=0.65\lim_{\beta \to \infty}\alpha_c(\beta)=0.65limβ→∞​αc​(β)=0.65.

View on arXiv
Comments on this paper