Complete Realization of Energy Landscape and Non-equilibrium Trapping Dynamics in Spin Glass and Optimization Problem

Ho Fai Po,Chi Ho Yeung
2024-05-23
Abstract:Energy landscapes are high-dimensional surfaces representing the dependence of system energy on variable configurations, which determine crucially the system's emergent behavior but are difficult to be analyzed due to their high-dimensional nature. In this article, we introduce an approach to reveal the complete energy landscapes of small spin glasses and Boolean satisfiability problems, which also unravels their non-equilibrium dynamics at an arbitrary temperature for an arbitrarily long time. In contrary to our common belief, our results show that it can be less likely to identify the ground states when temperature decreases, due to trapping in individual local minima, which ceases at different time, leading to multiple abrupt jumps with time in the ground-state probability. Simulations agree well with theoretical predictions on these remarkable phenomena. Finally, for large systems, we introduce a variant approach to extract partially the energy landscapes and observe both analytically and in simulations similar phenomena. This work introduces new methodology to unravel the non-equilibrium dynamics of glassy systems, and provides us with a clear, complete and new physical picture on their long-time behaviors inaccessible by modern numerics.
Disordered Systems and Neural Networks,Statistical Mechanics,Computational Physics
What problem does this paper attempt to address?