Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2610.01414

On thermalization in random free fermions: Statistical origins and energy-dependent ETH structure

Xiao-Kan Guo·Zhiqiang Huang
View original source ↗Read source document ↗Export metadata ↓

Abstract

We study the thermalization in the random free fermion model by a detailed random-matrix analysis. By computing the ensemble average and fluctuations of $\operatorname{Tr}(Aρ(t))$ for a single-particle observable $A$, we derive the thermalization function $g^2(t/τ_λ)$ with $τ_λ= \hbar/(2η\sqrt{N})$, prove its $|t|^{-3}$ asymptotic decay, and show that the variance of the fluctuations vanishes as $O(1/N)$ in the thermodynamic limit.Particle number conservation is also incorporated into the model, by which we further study the energy-shell eigenstate statistics and prove a factorization theorem of the eigenstate statistics. We further show that under a controlled eigenvector-eigenvalue correlation deformation, the diagonal energy-resolved slope and the off-diagonal eigenstate thermalization hypothesis spectral function acquire an energy dependence.Finally, we compute the fluctuations of correlation functions and compare the fluctuation scales with those of a fully chaotic system, revealing quantitative differences rooted in the Gaussian nature of the eigenstates.Our work establishes the random free fermion model as an analytically solvable realization of weak eigenstate thermalization hypothesis type self-averaging, and as a controllable setting in which eigenbasis chaos can be separated from spectral chaos.

Explore related subjects

cond-mat.stat-mech

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Xiao-Kan Guo, Zhiqiang Huang. 2026-10-01. On thermalization in random free fermions: Statistical origins and energy-dependent ETH structure. https://arxiv.org/abs/2610.01414

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

KEEP EXPLORING

Related papers

arXiv2026

Stochastic Thermodynamics for Autoregressive Generative Models: A Non-Markovian Perspective

Autoregressive generative models -- including Transformers, recurrent neural networks, classical Kalman filters, state space models, and Mamba -- all generate sequences by sampling each output from a deterministic summary of the past, producing genuinely non-Markovian observed processes. We develop a general theoretical framework based on stochastic thermodynamics for this class of architectures and introduce the entropy production, which can be efficiently estimated from sampled trajectories without exponential overhead, despite the non-Markovian nature of the observed dynamics. As a proof-of-concept experiment with a large language model (LLM), we evaluate the entropy production for a pre-trained Transformer-based model, GPT-2. We find that the token-level entropy production is dominated by a syntactic artifact, while the sentence-level entropy production tends to be larger for causally ordered than for non-causal text sets. This observation is supported by a re-evaluation with a substantially larger model, Qwen3-4B-Base. We also demonstrate the framework in the linear Gaussian case, where the model reduces to the Kalman innovation representation and the entropy production admits an analytical expression. We also show that the entropy production decomposes exactly into non-negative per-step contributions in terms of retrospective inference, and each of those terms further splits into information-theoretically meaningful terms: a compression loss and a model mismatch. Our results establish a bridge between stochastic thermodynamics and modern generative models, and provide a starting point for using irreversibility as a quantitative probe of the highly non-Markovian processes generated by models such as LLMs.

cond-mat.stat-mech↗
arXiv2026

Beyond average: heterogeneous first-passage dynamics in many-particle systems with resetting

Stochastic resetting is well understood for single-particle first-passage processes, but its consequences for collective first-passage behavior remain less clear. We address this problem in a many-particle system where all particles reset to the position of the rightmost particle, a protocol motivated by problems in artificial selection and avoidance. We use stochastic simulations of particles diffusing in a confining potential with an adsorbing boundary to examine two notions of group arrival: the first group-hitting time, when the first particle reaches the boundary, and the median group-hitting time, when half of them have reached it. We find that resetting produces broad hitting-time distributions with extended plateaus spanning several orders of magnitude. As the resetting rate increases, these plateaus extend and the mean group-hitting times grow rapidly. At the same time, the first-passage dynamics become increasingly heterogeneous. Consequently, the most probable and mean hitting times become widely separated, indicating the absence of a single characteristic time scale. These results demonstrate that the definition of group arrival is crucial for understanding and controlling collective first-passage behavior under resetting.

cond-mat.stat-mech↗
arXiv2026

An agitated oscillator chain

Reduced dynamical descriptions for spatially extended systems in contact with a nonequilibrium medium give essential information about extensions of fluctuation-dissipation relations and, at the same time, are relevant for many natural systems. Here we study a chain of harmonic oscillators coupled to fast run-and-tumble particles, discretizing earlier work on an elastic string. We derive the effective Langevin dynamics, including the explicit discrete counterparts for the nonequilibrium-induced streaming term, friction coefficient, and noise amplitude. At sufficiently high bath persistence, the linear friction becomes negative, destabilizing the chain dynamics. However, numerical simulations confirm that the anti-damping is eventually arrested and results in a nonlinear stationary condition, whose properties are the central topic of this work. We characterize the long-time phase-space trajectories and their Rayleigh-like structure, identify multistage relaxation of the mean-squared displacement and mean-squared velocity, and establish a non-monotonic dependence of stationary fluctuations on bath persistence. We further analyze the stationary distributions of the displacement and velocity, as well as the spatial correlations, which develop damped oscillations in the strongly persistent regime. The transfer of activity thus transforms a passive harmonic chain into a self-sustained fluctuating medium with many-body Rayleigh-like dynamics, pulsating displacements, and persistent velocities.

cond-mat.stat-mech↗