Memories from the ergodic phase: the awkward dynamics of spherical mixed p-spin models
arXiv: Disordered Systems and Neural Networks(2019)
摘要
We revisit the long-time limit of the out of equilibrium dynamics of mean-field spherical mixed p-spin models. We consider quenches (gradient descent dynamics) starting from initial conditions thermalized at some temperature in the ergodic phase. We perform numerical integration of the dynamical mean-field equations of the model and we find an unexpected dynamical phase transition. Below an onset temperature $T_{onset}$, higher than the dynamical transition temperature $T_{MCT} $, the asymptotic goes below the threshold energy of the dominant marginal minima of the function and memory of the initial condition is kept. This behavior, not present in the pure spherical p-spin , resembles closely the one observed in simulations of glass forming liquids. We then investigate the nature of asymptotic dynamics, finding an aging state that relaxes towards deep marginal minima. Careful analysis however rules out simple aging solutions, leaving open a full comprehension of the memory effect in these models.
更多查看译文
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要