The stability of the equilibrium point (background activity) of oscillatory
neural networks is an important property for computational applications th
at explore the switching between background activity and oscillatory states
. Here we consider a general approach to this problem for networks of arbit
rary size. For symmetric coupling, often the case in associative learning a
lgorithms, we derive the stability constraints and establish explicit resul
ts for the coupling strengths to satisfy in order that the equilibrium stat
e is stable. (C) 2000 Elsevier Science B.V. All rights reserved.