Exploring valleys of aging systems: the spin glass case

摘要: We present a statistical method for complex energy landscape exploration which provides information on the metastable states--or valleys--actually explored by an unperturbed aging process following quench. Energy fluctuations of record size are identified as events move system from one valley to next. This allows semi-analytical description in terms log-Poisson statistics, whose main features briefly explained. The bulk paper is devoted thorough investigations Ising spin glasses with Gaussian interactions both short and long range, well established paradigm glassy dynamics. Simple scaling expressions universal exponents (a) barrier energies, (b) minima, (c) Hamming distance function index found. distribution residence time inside valleys entered at age t w investigated, along global minimum hit. Finally, correlations between minima presented. results fit into framework available knowledge about glass aging. At same they support novel interpretation thermal relaxation landscapes multiple states. marginal stability attractors selected emphasized explained geometrical properties landscape.