Abstract
We analyze the thermal equilibrium distribution of 2p mean field variables for the Hopfield model withp stored patterns, in the case where 2p is small compared to the number of spins. In particular, we give a full description of the free energy density in the thermodynamic limit, and of the so-called “symmetric solutions” for the mean field equations.
Similar content being viewed by others
References
W. A. Little, The existence of persistent states in the brain,Math. Biosci 19:101 (1974).
J. J. Hopfield, Neural networks and physical systems with emergent collective computational abilities,Proc. Natl. Acad. Sci. USA 79:2554 (1982).
J. P. Provost and G. Vallée, Ergodicity of the coupling constants and the symmetricn-replicas trick for a class of mean-field spin-glass models,Phys. Rev. Lett. 50:598 (1983).
D. J. Amit, H. Gutfreund, and H. Sompolinsky, Spin-glass models of neural networks,Phys. Rev. A 32:1007 (1985).
R. J. McEliece, E. C. Posner, E. R. Rodemich, and S. S. Vankatesh, The capacity of the Hopfield associative memory,IEEE Trans. Information Theory 33:461 (1986).
D. J. Amit, The properties of models of simple neural networks, inHeidelberg Colloquium on Glassy Dynamics (1986), J. L. van Hemmen and I. Morgenstern, eds. (Lecture Notes in Physics, No. 275, 1987), p. 430.
D. J. Amit, H. Gutfreund, and H. Sompolinsky, Statistical mechanics of neural networks near saturation,Ann. Phys. 173:30 (1987).
J. L. van Hemman, Spin-glass model of a neural network,Phys. Rev. A 34:3435 (1986).
D. Grensing and R. Kühn, Random-site spin-glass models,J. Phys. A 19:L1153 (1986).
D. Grensing, J. L. van Hemmen, A. Huber, and R. Kühn, Nonlinear neural networks: I. General theory,J. Stat. Phys. 50:231 (1987); II. Information processing,J. Stat. Phys. 50:259 (1987).
J. L. van Hemmen and V. A. Zagrebnov, Storing extensively many weighted patterns in a saturated neural network,J. Phys. A 20:3989 (1987).
P. Baldi, Symmetries and learning in neural network models,Phys. Rev. Lett. 59:1976 (1987).
C. M. Newman, Memory capacity in neural network models: Rigorous lower bounds,Neural Networks 1 (1988).
J. Komlós and R. Paturi, Convergence results in the Hopfield model, Preprint, UC San Diego (1987).
C. Peterson and J. R. Anderson, A mean field theory learning algorithm for neural networks, Preprint MCC-EI-259-87, MCC Austin (1987).
Author information
Authors and Affiliations
Rights and permissions
About this article
Cite this article
Koch, H., Piasko, J. Some rigorous results on the Hopfield neural network model. J Stat Phys 55, 903–928 (1989). https://doi.org/10.1007/BF01041071
Received:
Revised:
Issue Date:
DOI: https://doi.org/10.1007/BF01041071