REVIEW 6 cited by
Binary perceptrons capacity via fully lifted random duality theory
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
Signed reviews
abstract
We study the statistical capacity of the classical binary perceptrons with general thresholds $\kappa$. After recognizing the connection between the capacity and the bilinearly indexed (bli) random processes, we utilize a recent progress in studying such processes to characterize the capacity. In particular, we rely on \emph{fully lifted} random duality theory (fl RDT) established in \cite{Stojnicflrdt23} to create a general framework for studying the perceptrons' capacities. Successful underlying numerical evaluations are required for the framework (and ultimately the entire fl RDT machinery) to become fully practically operational. We present results obtained in that directions and uncover that the capacity characterizations are achieved on the second (first non-trivial) level of \emph{stationarized} full lifting. The obtained results \emph{exactly} match the replica symmetry breaking predictions obtained through statistical physics replica methods in \cite{KraMez89}. Most notably, for the famous zero-threshold scenario, $\kappa=0$, we uncover the well known $\alpha\approx0.8330786$ scaled capacity.
Forward citations
Cited by 6 Pith papers
-
CLuP practically achieves $\sim 1.77$ positive and $\sim 0.33$ negative Hopfield model ground state free energy
CLuP±Hop approximates Hopfield ground state free energies to within about 0.3% using simple gradient descent, backed by the author's fully lifted random duality theory.
-
Deep ReLU networks -- injectivity capacity upper bounds
For deep ReLU networks with random Gaussian weights, the paper gives upper bounds on the layer expansion needed for injectivity and finds the expansion need saturates by four layers.
-
A CLuP algorithm to practically achieve $\sim 0.76$ SK--model ground state free energy
The authors propose a CLuP-SK barrier-descent algorithm and report it achieves approximately 0.76 of the SK ground state free energy for n around 2000 to 8000, approaching the theoretical Parisi limit of about 0.763.
-
Rare dense solutions clusters in asymmetric binary perceptrons -- local entropy via fully lifted RDT
For the asymmetric binary perceptron, the worst-case local entropy breaks down for constraint density alpha in (0.77, 0.78), matching replica predictions and the range where fast algorithms stop working.
-
Fully lifted \emph{blirp} interpolation -- a large deviation view
A large-deviation upgrade of fully lifted blirp interpolation is derived, yielding explicit derivative identities that the author links to local entropy and computational gaps in perceptron models.
-
A large deviation view of \emph{stationarized} fully lifted blirp interpolation
The paper derives new derivative identities for a stationarized fully lifted bilinearly indexed random process interpolator and states an equality between large deviation limits at the opposite ends of an interpolation path.
Discussion (0). Continue with ORCID to comment.