Fig. 3: Interaction between two encoded memories.
From: Explosive neural networks via higher-order interactions in curved statistical manifolds
a Values of φγ for different mean-field values m1, m2, indicating the attractor structure of the network for different values of β with J = 1, C = 0.2 for \({\gamma }^{{\prime} }=0\) (top row) and \({\gamma }^{{\prime} }=-1.2\) (bottom row). b Bifurcations of the order parameters m1, m2. For \({\gamma }^{{\prime} }=0\) we observe an attractor bifurcating into two and then into four. For \({\gamma }^{{\prime} }=-1.2\), we observe the same sequence, but with a coexistence hysteresis regime in which 7 attractors are possible.
