2020
DOI: 10.1103/physrevx.10.031045
|View full text |Cite
|
Sign up to set email alerts
|

Rethinking Mean-Field Glassy Dynamics and Its Relation with the Energy Landscape: The Surprising Case of the Spherical Mixed p -Spin Model

Abstract: The spherical p-spin model is a fundamental model in statistical mechanics of a disordered system with a random first-order transition. The dynamics of this model is interesting both for the physics of glasses and for its implications on hard optimization problems. Here, we revisit the out-of-equilibrium dynamics of the spherical mixed p-spin model, which differs from the pure p-spin model by the fact that the Hamiltonian is not a homogeneous function of its variables. We consider quenches (gradient descent dy… Show more

Help me understand this report
View preprint versions

Search citation statements

Order By: Relevance

Paper Sections

Select...
1
1
1
1

Citation Types

11
104
1

Year Published

2020
2020
2024
2024

Publication Types

Select...
8

Relationship

1
7

Authors

Journals

citations
Cited by 49 publications
(116 citation statements)
references
References 52 publications
(83 reference statements)
11
104
1
Order By: Relevance
“…Therefore, the generalization of the calculation in ref. 50 to sphere systems would be very appealing.…”
Section: Discussionmentioning
confidence: 99%
“…Therefore, the generalization of the calculation in ref. 50 to sphere systems would be very appealing.…”
Section: Discussionmentioning
confidence: 99%
“…We set a = 0.3, as it is a common choice used in literature[36] 4. A similar phenomenology that has enabled the identification of the onset temperature via the overlap has recently been observed in diverse systems, ranging from the mixed p-spin model[37] to the 3D Heisenberg spin glass[38].…”
mentioning
confidence: 99%
“…Marginal states of different energy can also be reached asymptotically by the dynamics, as shown for instance in Ref. [9], provided that the system is initialized at lower temperatures.…”
Section: Introductionmentioning
confidence: 80%
“…Similarly, the integration on the elements m a N −1 N −1 in U 1 is free, while the elements m a N −1 N and m a N N are constrained to take a given value. To decouple the constrained matrix elements from the unconstrained ones, we make use of Gaussian conditioning 9 . Introducing the vector notation Ξ αβ = (Ξ 1 αβ , Ξ 2 αβ ) T and imposing m a αN = 0, we obtain:…”
Section: B Derivation Of Eq 23mentioning
confidence: 99%