Non-equilibrium criticality and efficient exploration of glassy landscapes with memory dynamics
arXiv:2102.04557 · doi:10.1016/j.physa.2021.126727
Abstract
Spin glasses are notoriously difficult to study both analytically and numerically due to the presence of frustration and metastability. Their highly non-convex landscapes require collective updates to explore efficiently. Currently, most state-of-the-art algorithms rely on stochastic spin clusters to perform non-local updates, but such "cluster algorithms" lack general efficiency. Here, we introduce a non-equilibrium approach for simulating spin glasses based on classical dynamics with memory. By simulating various classes of 3d spin glasses (Edwards-Anderson, partially-frustrated, and fully-frustrated models), we find that memory dynamically promotes critical spin clusters during time evolution, in a self-organizing manner. This facilitates an efficient exploration of the low-temperature phases of spin glasses.
arXiv comment21 pages, 9 figures
References in corpus (21)
- Simulated Tempering: A New Monte Carlo Scheme
- Statistical mechanics for natural flocks of birds
- Spin-Glass Theory for Pedestrians
- Recent advances in percolation theory and its applications
- Memcomputing: a computing paradigm to store and process information on the same physical platform
- Cluster algorithms for general-S quantum spin systems
- Markov Chain Monte Carlo Method without Detailed Balance
- Low-temperature behavior of two-dimensional Gaussian Ising spin glasses
- An introduction to Monte Carlo methods
- Hiding solutions in random satisfiability problems: A statistical mechanics approach
- Polynomial-time solution of prime factorization and NP-hard problems with digital memcomputing machines
- Population Monte Carlo algorithms
- Efficient Solution of Boolean Satisfiability Problems with Digital MemComputing
- From Near to Eternity: Spin-glass planting, tiling puzzles, and constraint satisfaction problems
- Mode-Assisted Unsupervised Learning of Restricted Boltzmann Machines
- Percolation and cluster Monte Carlo dynamics for spin models
- On generalized cluster algorithms for frustrated spin models
- The ground state energy of the Edwards-Anderson spin glass model with a parallel tempering Monte Carlo algorithm
- The Fully Frustrated Hypercubic Model is Glassy and Aging at Large
- Taming a non-convex landscape with dynamical long-range order: memcomputing Ising benchmarks
- New universality class in percolation on multifractal scale-free planar stochastic lattice