Concentration of the complexity of spherical pure $p$-spin models at arbitrary energies

Abstract

We consider critical points of the spherical pure $p$-spin spin glass model with Hamiltonian $H_N\left(\mathbf{\sigma }\right)=\frac{1}{N^{\left(p-1\right)/2}}{\sum }_{i_1,...,i_p=1}^N{J}_{i_1,...,i_p}{\sigma }_{i_1}\cdots {\sigma }_{i_p}$, where $\mathbf{\sigma }=\left({\sigma }_1,...,{\sigma }_N\right)\in {\mathbb{S}}^{N-1}:=\left\{\mathbf{\sigma }\in {\mathbb{R}}^N:{\Vert \mathbf{\sigma }\Vert }_2=\sqrt{N}\right\}$ and $J_{i_1,...,i_p}$ are i.i.d standard normal variables. Using a second moment analysis, we prove that for $p\ge 32$ and any $E>-E_{\infty }$, where $E_{\infty }$ is the (normalized) ground state, the ratio of the number of critical points $\mathbf{\sigma }$ with $H_N(\mathbf{\sigma })\le NE$ and its expectation asymptotically concentrates at $1$. This extends to arbitrary $E$ a similar conclusion of [Sub17a].

References in corpus (10)

• Complexity of random smooth functions on the high-dimensional sphere
• Characteristic polynomials of real symmetric random matrices
• The geometry of the Gibbs measure of pure spherical spin glasses
• Parisi formula, disorder chaos and fluctuation for the ground state energy in the spherical mixed p-spin models
• Free energy and complexity of spherical bipartite models
• Low Temperature Asymptotics in Spherical Mean Field Spin Glasses
• The Complexity of the Spherical $p$-spin spin glass model, revisited
• Exponential growth of random determinants beyond invariance
• The generalized TAP free energy II
• The number of saddles of the spherical $p$-spin model

Cited by in corpus (3)

• Counting equilibria in a random non-gradient dynamics with heterogeneous relaxation rates
• Superposition of Random Plane Waves in High Spatial Dimensions: Random Matrix Approach to Landscape Complexity
• Tight Lipschitz Hardness for Optimizing Mean Field Spin Glasses