arXiv Science⌕ Search
Home/Archive/arXiv

arXiv · 2610.27872

Moments of Random Permanents and Hafnians: A Replica Approach

Paolo Facchi·Angelo Mariano·Giorgio Parisi·Saverio Pascazio·Antonello Scardicchio
View original source ↗Read source document ↗Export metadata ↓

Abstract

Using contour-integral representations of the permanent and of the Hafnian, we turn the computation of their moments for random matrices into a replica calculation, analogous to that of the moments of the partition function of a mean-field spin glass. For the $2k$-th moment of $N\times N$ matrices with independent entries, the dominant saddle point is replica symmetric for $k>2$, and we obtain in closed form the exponential growth rate, the prefactor and the first $1/N$ correction, in agreement with exact enumerations. The growth rate depends only on whether the entries are real or complex; higher cumulants affect only the prefactor and the $1/N$ term. At $k=2$ the replica-symmetric solution becomes unstable, separating a power-law bulk of the distribution from a large-deviation tail, which we compute. The results apply to boson sampling with single photons and with squeezed light.

Explore related subjects

cond-mat.dis-nncond-mat.stat-mechquant-ph

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Paolo Facchi, Angelo Mariano, Giorgio Parisi, Saverio Pascazio, Antonello Scardicchio. 2026-10-20. Moments of Random Permanents and Hafnians: A Replica Approach. https://arxiv.org/abs/2610.27872

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Resonance Proliferation Across Localization Transitions

Models of many-body localization (MBL) exhibit slow numerical drifts towards delocalization with increasing system size, for which no satisfactory theory exists. Numerical results indicate that these drifts are driven by the proliferation of many-body resonances at intermediate disorder strengths. We develop a statistical method to predict the distribution of resonance oscillation frequencies, accounting for how resonances formed at larger frequency scales affect the subsequent formation of resonances at lower frequencies. Working within the statistical Jacobi approximation (SJA), we derive a flow equation for a power-law exponent $θ(w)$ characterizing the density of resonances at frequency scale $w$. A localized phase is described by a line of fixed points with $θ(w)>0$, while an instability of the flow signals resonance proliferation and the onset of thermalization. The predicted $θ(w)$ agrees with numerical results for the Anderson model on random regular graphs and the Lévy--Rosenzweig--Porter random matrix ensemble, both of which host resonance-driven delocalization transitions. We further connect the flow to eigenstate properties such as the participation ratio and to dynamical observables such as the return probability. The predicted $θ(w)$ also agrees with numerical measurements in real-space models of MBL at intermediate disorder strengths, providing a step towards explaining the finite-size drifts observed in MBL.

cond-mat.dis-nn↗
arXiv2026

Ubiquitous Flat Bands in Edge-Inflated Random Graphs

Flat bands are usually associated with periodic lattice geometries, where destructive interference or chiral-symmetry constraints produce macroscopically degenerate eigenstates. Here we show that flat bands arise generically in a broad class of nonperiodic random networks through a simple edge-inflation procedure. Starting from random regular, Voronoi tessellation, Delaunay triangulation, scale-free, small-world, and Erdös-Rényi seed graphs, each edge is replaced by a finite one-dimensional tight-binding chain, either with fixed length $L$ or with random lengths of mean $\bar L$. Although the original random graphs typically have adjacency matrices of full rank and therefore few or no zero eigenstates, their inflated counterparts exhibit extensive flat bands. The flat-band energies are determined by the eigenstates residing on the chains or, more generally, by the eigenvalues of identical loopless motifs replacing the seed bonds. The corresponding eigenstates are states supported on subsets of the inflated chains formed by destructive interference at the original seed vertices. For zero-energy flat bands, we relate the degeneracy to the graph nullity $η(G)$ and the maximum matching number $ν(G)$. The numerical results support the correspondence $η(G)=N-2ν(G)$, up to small deviations, across the random graph ensembles studied. The fraction of degenerate zero-energy states in inflated random graphs is controlled primarily by the average degree of the seed graph. We further show that the flat bands exhibit characteristic robustness to disorder: disorder on the seed sites and random magnetic flux leave the flat bands essentially unchanged, while bond disorder and disorder on the chain or motif sites broaden them. These results show that flat bands do not require crystalline order, but can emerge as robust consequences of local graph inflation and motif substitution.

cond-mat.dis-nn↗
arXiv2026

Dynamical mean field approach to associative memory model with non-monotonic transfer functions

The Hopfield associative memory model stores random patterns in synaptic couplings according to Hebb's rule and retrieves them through gradient descent on an energy function. This conventional setting, where neurons are assumed to have monotonic transfer functions, has been central to understanding associative memory. Morita (1993, Neural Netw. 6 115), however, showed that introducing non-monotonic transfer functions can dramatically enhance retrieval performance. While this phenomenon has been qualitatively examined, a full quantitative theory remains elusive due to the difficulty of analysis in the absence of an underlying energy function. In this work, we apply dynamical mean-field theory to the discrete-time synchronous retrieval dynamics of the non-monotonic model, which succeeds in accurately characterizing its macroscopic dynamical properties. We also derive conditions for retrieval states, and clarify their relation to previous studies. Our results provide new insights into the non-equilibrium retrieval dynamics of associative memory models.

cond-mat.dis-nn↗