Dimensional reduction and its breakdown in the 3-dimensional long-range random field Ising model
Maxime Baczyk, Matthieu Tissier, Gilles Tarjus, Yoshinori Sakamoto

TL;DR
This paper explores the conditions under which dimensional reduction holds or breaks down in a 3D long-range random field Ising model, revealing a critical point dependent on decay exponents of interactions and disorder.
Contribution
It introduces a nonperturbative functional renormalization group approach combined with supersymmetry to identify a critical value separating dimensional reduction from its breakdown.
Findings
Existence of a critical value for decay exponents where dimensional reduction breaks down.
The transition depends on specific relations between interaction and disorder decay exponents.
The results are relevant for future simulations and experimental observations.
Abstract
We investigate dimensional reduction, the property that the critical behavior of a system in the presence of quenched disorder in dimension d is the same as that of its pure counterpart in d-2, and its breakdown in the case of the random-field Ising model in which both the interactions and the correlations of the disorder are long-ranged, i.e. power-law decaying. To some extent the power-law exponents play the role of spatial dimension in a short-range model, which allows us to probe the theoretically predicted existence of a nontrivial critical value separating a region where dimensional reduction holds from one where it is broken, while still considering the physical dimension d=3. By extending our recently developed approach based on a nonperturbative functional renormalization group combined with a supersymmetric formalism, we find that such a critical value indeed exists, provided…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
