In the high-dimensional limit the spherical Boltzmann machine admits exact equations for training dynamics, Bayesian evidence, and cascades of phase transitions tied to mode alignment with data, which connect to generative phenomena including double descent and out-of-equilibrium biases.
hub Canonical reference
From quantum chaos and eigenstate thermaliza- tion to statistical mechanics and thermo- dynamics
Canonical reference. 71% of citing Pith papers cite this work as background.
22
Pith papers citing it
Background
71% of classified citations
hub tools
citation-role summary
background 6
method 1
citation-polarity summary
representative citing papers
Coherent-state propagation enables quasi-polynomial classical simulation of bosonic circuits with logarithmically many Kerr gates at exponentially small trace-distance error, with polynomial runtime in the weak-nonlinearity regime.
Experimental observation of coexisting extended, localized, and critical states in a quasiperiodic Floquet-modulated orbital optical lattice using ultracold atoms.
Introduces a TAP-motivated framework and constructs explicit parameter-free spectral algorithms that achieve strong detection and weak recovery thresholds in three canonical correlated two-view models with matching lower bounds.
Neural ODEs reproduce 2RDM dynamics from data only when three-particle cumulant correlations are strong, mapping the validity regime of cumulant expansions.
k-local quantum Hamiltonians admit system-size-independent spectral gap for Gibbs samplers at high temperature, enabling FPT quantum approximation algorithms for partition functions.
At special drive frequencies, the leading perturbative Floquet Hamiltonian of a driven Rydberg chain maps to the XXZ model, producing emergent prethermal integrability confirmed by level statistics and entanglement in exact diagonalization.
Double-scaled fermionic and bosonic embedded ensembles are equivalent to double-scaled complex SYK and solvable via the Wick product of non-commuting Gaussian random variables, yielding a duality to the chord Hilbert space.
Quantum thermalization occurs in chaotic and integrable regimes of a multispecies Bose-Josephson junction, with quantum scars remaining athermal in the chaotic regime.
Quantum systems reach a Maximal Entanglement Limit where entanglement geometry produces thermal reduced density matrices and probabilistic behavior in statistical and high-energy physics.
Microcanonical free cumulants of central-system observables scale universally with system-bath interaction strength in closed setups, connecting to thermalization dynamics.
Energy eigenstates in SU(2)-symmetric quantum many-body systems obey a KMS relation whose finite-size correction scales as usual or polynomially larger depending on circumstances, supported by numerics on small Heisenberg chains.
In the fully chaotic regime of the kicked top, long-time freeness is reached exponentially fast, accompanied by a hierarchy of time scales indicating a multifractal approach.
A framework detects chaos via divergence of speed-Fisher information under slow driving, controlled by low-frequency spectral weight and tied to entropy production, applying to classical, quantum, and non-Hamiltonian systems.
Polfed.jl provides an efficient implementation of polynomially filtered Lanczos diagonalization for mid-spectrum eigenpairs in quantum many-body systems, supporting larger sizes via on-the-fly polynomial transformations and GPU acceleration.
In the long-range Haldane-Shastry model, pristine Poisson level statistics emerge only with combined position disorder and random magnetic fields, with an approximate scaling collapse governed by the product αδ when SU(2) symmetry is broken.
Thermalization after a quench in the Hubbard-Holstein model occurs via sharp fronts in real time and DMFT iteration space, with electron fronts appearing earlier than phonon fronts at weak coupling.
In the disorder-free SYK model, off-diagonal matrix elements of operators built from n≥4 Majorana fermions follow a generalized inverse Gaussian distribution.
Authors characterize the MBL crossover via many-body quantum metric and localization parameter, extracting a localization length from wavefunction spread measurable by the metric.
Thesis uses statistical mechanics to study DAM and RBM models for understanding memorization, low-dimensional learning, and adversarial robustness in neural networks.
A review that contrasts common assumptions about the Lindblad equation with refined expectations drawn from examples, culminating in a checklist for assessing its breakdown.
Krylov complexity is a canonical, parameter-independent measure of operator spreading that probes chaotic dynamics to late times and admits a geometric interpretation in holographic duals.
citing papers explorer
-
Spherical Boltzmann machines: a solvable theory of learning and generation in energy-based models
In the high-dimensional limit the spherical Boltzmann machine admits exact equations for training dynamics, Bayesian evidence, and cascades of phase transitions tied to mode alignment with data, which connect to generative phenomena including double descent and out-of-equilibrium biases.