Critical properties of the Anderson transition on random graphs: Two-parameter scaling theory, Kosterlitz-Thouless type flow, and many-body localization

The Anderson transition in random graphs has raised great interest, partly out of the hope that its analogy with the many-body localization (MBL) transition might lead to a better understanding of this hotly debated phenomenon. Unlike the latter, many results for random graphs are now well established, in particular, the existence and precise value of a critical disorder separating a localized from an ergodic delocalized phase. However, the renormalization group flow and the nature of the transition are not well understood. In turn, recent works on the MBL transition have made the remarkable prediction that the flow is of Kosterlitz-Thouless type. In this paper, we show that the Anderson transition on graphs displays the same type of flow. Our work attests to the importance of rare branches along which wave functions have a much larger localization length xi(parallel to) than the one in the transverse direction xi(perpendicular to). Importantly, these two lengths have different critical behaviors: xi(parallel to) diverges with a critical exponent v(parallel to) = 1, while xi(perpendicular to) reaches a finite universal value xi(c)(perpendicular to) at the transition point W-c. Indeed, xi(-1)(perpendicular to) approximate to xi(c-1)(perpendicular to), with xi similar to(W - W-c)(-v perpendicular to) associated with a new critical exponent v(perpendicular to) = 1/2, where exp(xi) controls finite-size effects. The delocalized phase inherits the strongly nonergodic properties of the critical regime at short scales, but is ergodic at large scales, with a unique critical exponent v = 1/2. This shows a very strong analogy with the MBL transition: the behavior of xi(perpendicular to) is identical to that recently predicted for the typical localization length of MBL in a phenomenological renormalization group flow. We demonstrate these important properties for a small-world complex network model and show the universality of our results by considering different network parameters and different key observables of Anderson localization.