When is the average number of saddle points typical?

Abstract

A common measure of a function's complexity is the count of its stationary points. For complicated functions, this count grows exponentially with the volume and dimension of their domain. In practice, the count is averaged over a class of functions (the annealed average), but the large numbers involved can produce averages biased by extremely rare samples. Typical counts are reliably found by taking the average of the logarithm (the quenched average), which is more difficult and not often done in practice. When most stationary points are uncorrelated with each other, quenched and anneals averages are equal. Equilibrium heuristics can guarantee when most of the lowest minima will be uncorrelated. We show that these equilibrium heuristics cannot be used to draw conclusions about other minima and saddles by producing examples among Gaussian-correlated functions on the hypersphere where the count of certain saddles and minima has different quenched and annealed averages, despite being guaranteed `safe' in the equilibrium setting. We determine conditions for the emergence of nontrivial correlations between saddles, and discuss the implications for the geometry of those functions and what out-of-equilibrium settings might be affected.

References in corpus (18)

• Spin-Glass Theory for Pedestrians
• On the stationary points of the TAP free energy
• A Landscape Analysis of Constraint Satisfaction Problems
• The statistics of critical points of Gaussian fields on large-dimensional spaces
• Topological and Dynamical Complexity of Random Neural Networks
• Properties of equilibria and glassy phases of the random Lotka-Volterra model with demographic noise
• The spherical 2+p spin glass model: an analytically solvable model with a glass-to-glass transition
• The spherical $2+p$ spin glass model: an exactly solvable model for glass to spin-glass transition
• Generalized Lotka-Volterra equations with random, non-reciprocal interactions: the typical number of equilibria
• Quenched complexity of the p-spin spherical spin-glass with external magnetic field
• Gradient descent dynamics in the mixed $p$-spin spherical model: finite size simulation and comparison with mean-field integration
• How to count in hierarchical landscapes: a 'full' solution to mean-field complexity
• Quenched complexity of equilibria for asymmetric Generalized Lotka-Volterra equations
• Statistical mechanical approach to secondary processes and structural relaxation in glasses and glass formers
• Comment on ``Spherical 2 + p spin-glass model: An analytically solvable model with a glass-to-glass transition''
• The density of stationary points in a high-dimensional random energy landscape and the onset of glassy behaviour
• Complex complex landscapes
• On-Site Potential Creates Complexity in Systems with Disordered Coupling

Cited by in corpus (2)

• Arrangement of nearby minima and saddles in the mixed spherical energy landscapes
• Landscape Complexity for the Empirical Risk of Generalized Linear Models: Discrimination between Structured Data