J. Phys. I France
Volume 4, Numéro 5, May 1994
Page(s) 629 - 633
DOI: 10.1051/jp1:1994166
J. Phys. I France 4 (1994) 629-633

Short Communication

Saddle point equations of a neural network with correlated attractors

M. Shcherbina1 and B. Tirozzi2

1  Institute for Low Temperature Physics and Engineering, 47 Lenin Avenue, Kharkov 310164, Ukraine
2  Department of Mathematics, University of Rome "La Sapienza", P.le Aldo Moro, 5 - 00185, Roma, Italy

(Received 17 January 1994, received in final form 7 February 1994, accepted 25 February 1994)

We derive in a rigorous way, using a new technique, the saddle-point equations of a modified Hopfield model which stores sequences of patterns . We obtain the same equations as those of the replica-symmetry approach without using the $n \rightarrow 0$ limit but assuming selfaveraging of the Edwards-Anderson order parameter.

© Les Editions de Physique 1994