summaryrefslogtreecommitdiff
path: root/marginal.tex
diff options
context:
space:
mode:
authorJaron Kent-Dobias <jaron@kent-dobias.com>2024-07-24 15:59:55 +0200
committerJaron Kent-Dobias <jaron@kent-dobias.com>2024-07-24 15:59:55 +0200
commit8296960face27afafe54b586a1e1451b3ee71bec (patch)
tree52d7b330dd5542df4d85bd4f3495a44546a873cb /marginal.tex
parent375a5d19d82ebf40ed6d3dcf11aae94ad70f1a03 (diff)
downloadmarginal-8296960face27afafe54b586a1e1451b3ee71bec.tar.gz
marginal-8296960face27afafe54b586a1e1451b3ee71bec.tar.bz2
marginal-8296960face27afafe54b586a1e1451b3ee71bec.zip
Added Fyodorov citations for also independence of Hessian.
Diffstat (limited to 'marginal.tex')
-rw-r--r--marginal.tex4
1 files changed, 2 insertions, 2 deletions
diff --git a/marginal.tex b/marginal.tex
index 8414310..508b674 100644
--- a/marginal.tex
+++ b/marginal.tex
@@ -95,7 +95,7 @@ Hessian are necessary to lead to marginal minima. This strategy is so
successful in the spherical spin glasses because it is straightforward to implement.
First, the shape of the Hessian's spectrum is independent of energy and even
whether one sits at a stationary point or not. This is a property of models
-whose energy is a Gaussian random variable \cite{Bray_2007_Statistics}.
+whose energy is a Gaussian random variable \cite{Fyodorov_2004_Complexity, Bray_2007_Statistics}.
Furthermore, a natural parameter in the analysis of these models linearly
shifts the spectrum of the Hessian. Therefore, tuning this parameter to a
specific constant value allows one to require that the Hessian spectrum have a
@@ -761,7 +761,7 @@ The marginal optima of these models can be studied without the methods
introduced in this paper, and have been in the past \cite{Folena_2020_Rethinking,
Kent-Dobias_2023_How}. First, these models are Gaussian, so at large $N$ the
Hessian is statistically independent of the gradient and energy
-\cite{Bray_2007_Statistics}. Therefore, conditioning the Hessian can be done
+\cite{Fyodorov_2004_Complexity, Bray_2007_Statistics}. Therefore, conditioning the Hessian can be done
mostly independently from the problem of counting stationary points. Second, in
these models the Hessian at every point in the landscape belongs to the GOE
class with the same width of the spectrum $\mu_\mathrm m=2\sqrt{f''(1)}$.