Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2610.36041

Phase Transitions in Mismatched Bayesian Inference for the Ising Model

Xinyu Sun, Zhou‐Quan Wan, Shao-Kai Jian, Hong Yao

Published Oct 26, 2026• 0 citations
arXivHTMLPDF

Abstract

Bayesian inference with a mismatched model can develop an overconfident phase, characterized by posterior agreement without faithful reconstruction of the true state. We study Bayesian inference from noisy bond measurements of an Ising model in its paramagnetic phase, with true inverse temperature $β$ and measurement-noise strength $Δ$, while inference assumes corresponding parameters $β_Q$ and $Δ_Q$. Using replica field theory, we analytically show that the overconfidence phase can occur when inference underestimates the measurement noise ($Δ_Q < Δ$) or assumes a lower prior temperature than the true one ($β_Q > β$). An exact change of variables maps the problem to a correlated random-bond Ising model, where the informative, uninformative, and overconfident phases correspond to ferromagnetic, paramagnetic, and spin-glass phases, respectively. We further perform Monte Carlo simulations of the mapped model in three dimensions and establish a finite overconfident region between the informative and uninformative phases, consistent with our analytical prediction. Finally, we show that the true model parameters can be inferred directly from the measurement records, providing a natural route to calibrating the inference model.

DOI
10.48550/arxiv.2610.36041

Related papers

Bayesian phase transition for the critical Ising model: Enlarged replica symmetry in the epsilon expansion and in 2D

Kay Joerg Wiese, Alapan Das, Adam Nahum

A process that images or measures bond energies in the critical Ising model can be in distinct measurement 'phases', depending on the precision of measurement. We study the transition into the strong-measurement phase using replica field theory (an epsilon expansion around six dimensions), a symmetry analysis, and numerical simulations in two dimensions. The results show that certain inference problems possess unexpected structure associated with an enlarged symmetry of the replica description. The symmetry imposes constraints on correlation functions, and is an analog of the Nishimori phenomenon in the Ising spin glass, but in a distinct replica limit. It determines the exact critical exponent for an Edwards-Anderson-like correlator in the measured Ising model, both in 2D and near the upper critical dimension. The symmetry is present microscopically for certain measurement protocols, and it emerges in the infrared for generic protocols at the critical measurement strength. We also examine the epsilon expansion for models with power-law interactions and/or long-range measurement. Finally, we describe an analogous replica symmetry enlargement for measurement in Wegner's lattice gauge theory.

cond-mat.stat-mechCross-listedarXiv:2604.23346•Apr 28, 2026•8 citations

Inferring hidden states in a random kinetic Ising model: replica analysis

Ludovica Bachschmid-Romano, Manfred Opper

We consider the problem of predicting the spin states in a kinetic Ising model when spin trajectories are observed for only a finite fraction of sites. In a Bayesian setting, where the probabilistic model of the spin dynamics is assumed to be known, the optimal prediction can be computed from the conditional (posterior) distribution of unobserved spins given the observed ones. Using the replica method, we compute the error of the Bayes optimal predictor for parallel discrete time dynamics in a fully connected spin system with non-symmetric random couplings. The results, exact in the thermodynamic limit, agree very well with simulations of finite spin systems.

arXiv:1405.4164•Journal of Statistical Mechanics Theory and Experiment•Jun 19, 2014•18 citations

Magnetization-resolved density of states and mixed-order transition in the two-dimensional random bond Ising model: an entropic sampling study

Yi Liu, Ding Wang, Xin Wang +2 more

Abstract Systems with quenched disorder possess complex energy landscapes that are challenging to explore under conventional Monte Carlo methods. In this work, we implement an efficient entropy sampling scheme for accurate computation of the entropy function in low-energy regions. The method is applied to the two-dimensional ± J random-bond Ising model, where frustration is controlled by the fraction p of ferromagnetic bonds. We investigate the low-temperature paramagnetic–ferromagnetic phase boundary below the multicritical point at T N = 0.9530(4), P N = 0.89078(8), as well as the zero-temperature ferromagnetic–spin-glass transition. Finite-size scaling analysis reveals that the phase boundary for T < T N exhibits reentrant behavior. By analyzing the evolution of the magnetization-resolved density of states g ( E , M ) and ground-state spin configurations against increasing frustration, we provide strong evidence that the zero-temperature transition is a mixed-order. Finite-size scaling conducted on the spin-glass side supports the validity of β = 0, where β is the magnetization exponent, with a correlation length exponent ν = 1.50(8). Our results provide new insights into the nature of the ferromagnetic-to-spin-glass phase transition in an extensively degenerate ground state.

arXiv:2505.04298•Communications in Theoretical Physics•Jun 14, 2025•5 citations

Ferromagnetic-glassy transitions in three-dimensional Ising spin glasses

Giacomo Ceccarelli, Andrea Pelissetto C, Ettore Vicari

We investigate the ferromagnetic-glassy transitions that separate the low-temperature ferromagnetic and spin-glass phases in the temperature-disorder phase diagram of three-dimensional Ising spin-glass models. For this purpose, we consider the cubic-lattice $\ifmmode\pm\else\textpm\fi{}J$ (Edwards-Anderson) Ising model with bond distribution $P(J)=p\ensuremath{\delta}(J\ensuremath{-}1)+(1\ensuremath{-}p)\ensuremath{\delta}(J+1)$, and present a numerical Monte Carlo study of the critical behavior along the line that marks the onset of ferromagnetism. The finite-size scaling analysis of the Monte Carlo data shows that the ferromagnetic-glassy transition line is slightly reentrant. As a consequence, for an interval of the disorder parameter $p$, around $p\ensuremath{\approx}0.77$, the system presents a low-temperature glassy phase, an intermediate ferromagnetic phase, and a high-temperature paramagnetic phase. Along the ferromagnetic-glassy transition line, magnetic correlations show a universal critical behavior with critical exponents $\ensuremath{\nu}=0.96(2)$ and $\ensuremath{\eta}=\ensuremath{-}0.39(2)$. The hyperscaling relation $\ensuremath{\beta}/\ensuremath{\nu}=(1+\ensuremath{\eta})/2$ is satisfied at the transitions, so that $\ensuremath{\beta}/\ensuremath{\nu}=0.305(10)$. This magnetic critical behavior represents a new universality class for ferromagnetic transitions in Ising-type disordered systems. Overlap correlations are apparently not critical and show a smooth behavior across the transition.

arXiv:1107.3005•Physical Review B•Oct 7, 2011•29 citations

Measurement-induced phase transitions in quantum inference problems and quantum hidden Markov models

Sunwoo Kim, Curt W. von Keyserlingk, Austen Lamacraft

Recently, there is interest in coincident “sharpening” and “learnability” transitions in monitored quantum systems. In the latter, an outside observer's ability to infer properties of a quantum system from measurements undergoes a phase transition. Such transitions appear to be related to the decodability transition in quantum error correction, but the precise connection is not clear. Here, we study these problems under one framework we call the general quantum inference problem. In cases as above where the system has a Markov structure, we say that the inference is on a quantum hidden Markov model. We show a formal connection to classical hidden Markov models and that they coincide for certain setups. For example, we prove this for those involving Haar-random unitaries and measurements. We introduce the notion of Bayes nonoptimality, where parameters used for inference differs from true ones. This allows us to expand the phase diagrams of above models. At Bayes optimality, we obtain an explicit relation between “sharpening” and “learnability” order parameters, explicitly showing that the two transitions coincide. Next, we study concrete examples. We review quantum error correction on the toric and repetition code and their mapping to two-dimensional (2D) random-bond Ising model (RBIM) through our framework. We study the Haar-random U ( 1 ) -symmetric monitored quantum circuit and tree, mapping each to inference models that we call the planted simple symmetric exclusion process and planted , respectively, and expanding the phase diagram to Bayes nonoptimality. For the circuit, we deduce the phase boundary numerically and analytically argue that it is of a single universality class. For the tree, we present an exact solution of the entire phase boundary, which displays re-entrance as does the 2D RBIM. We discuss these phase diagrams, with their interpretations for quantum inference problems and rigorous arguments on their shapes.

arXiv:2504.08888•Physical Review Research•Jan 15, 2026•15 citations

Low-temperature transition of 2d random-bond Ising model and quantum infinite randomness

Akshat Pandey, Aditya Mahadevan, A. Alan Middleton +1 more

At low temperatures, the classical two-dimensional random bond Ising model undergoes a frustration-driven ferromagnet-to-paramagnet transition controlled by a zero-temperature fixed point separating ferromagnet and spin glass phases. We show that this critical point can be understood through a renormalization group transformation that constructs the ground state of the Ising model through a sequence of Hamiltonians that, starting with an unfrustrated model, iteratively adds in frustration until the target Hamiltonian is reached. Via a mapping of the thermodynamics of the 2d Ising model to the spectral properties of a related Hermitian matrix -- the Hamiltonian of a noninteracting quantum problem -- this RG procedure corresponds to an iterative diagonalization of the quantum Hamiltonian. The flow toward zero temperature in the Ising picture manifests as a flow toward infinite randomness in the spectrum of the quantum Hamiltonian, with the log gap of the Hamiltonian scaling as a power of the system size: $\log \varepsilon_{\it min}^{-1} \sim L^ψ$. The tunneling exponent $ψ$ is equal to the spin stiffness exponent $θ_c$ characterizing the zero-temperature fixed point.

cond-mat.stat-mechCross-listedarXiv:2603.02308•Mar 4, 2026•4 citations
Bayesian inference with a mismatched model can develop an overconfident phase, characterized by posterior agreement without faithful reconstruction of the true state. We study Bayesian inference from noisy bond measurements of an Ising model in its paramagnetic phase, with true inverse temperature βββ and measurement-noise strength ΔΔΔ, while inference assumes corresponding parameters βQβ_QβQ​ and ΔQΔ_QΔQ​. Using replica field theory, we analytically show that the overconfidence phase can occur when inference underestimates the measurement noise (ΔQ<ΔΔ_Q < ΔΔQ​<Δ) or assumes a lower prior temperature than the true one (βQ>ββ_Q > ββQ​>β). An exact change of variables maps the problem to a correlated random-bond Ising model, where the informative, uninformative, and overconfident phases correspond to ferromagnetic, paramagnetic, and spin-glass phases, respectively. We further perform Monte Carlo simulations of the mapped model in three dimensions and establish a finite overconfident region between the informative and uninformative phases, consistent with our analytical prediction. Finally, we show that the true model parameters can be inferred directly from the measurement records, providing a natural route to calibrating the inference model.
A process that images or measures bond energies in the critical Ising model can be in distinct measurement 'phases', depending on the precision of measurement. We study the transition into the strong-measurement phase using replica field theory (an epsilon expansion around six dimensions), a symmetry analysis, and numerical simulations in two dimensions. The results show that certain inference problems possess unexpected structure associated with an enlarged symmetry of the replica description. The symmetry imposes constraints on correlation functions, and is an analog of the Nishimori phenomenon in the Ising spin glass, but in a distinct replica limit. It determines the exact critical exponent for an Edwards-Anderson-like correlator in the measured Ising model, both in 2D and near the upper critical dimension. The symmetry is present microscopically for certain measurement protocols, and it emerges in the infrared for generic protocols at the critical measurement strength. We also examine the epsilon expansion for models with power-law interactions and/or long-range measurement. Finally, we describe an analogous replica symmetry enlargement for measurement in Wegner's lattice gauge theory.
We consider the problem of predicting the spin states in a kinetic Ising model when spin trajectories are observed for only a finite fraction of sites. In a Bayesian setting, where the probabilistic model of the spin dynamics is assumed to be known, the optimal prediction can be computed from the conditional (posterior) distribution of unobserved spins given the observed ones. Using the replica method, we compute the error of the Bayes optimal predictor for parallel discrete time dynamics in a fully connected spin system with non-symmetric random couplings. The results, exact in the thermodynamic limit, agree very well with simulations of finite spin systems.
Abstract Systems with quenched disorder possess complex energy landscapes that are challenging to explore under conventional Monte Carlo methods. In this work, we implement an efficient entropy sampling scheme for accurate computation of the entropy function in low-energy regions. The method is applied to the two-dimensional ± J random-bond Ising model, where frustration is controlled by the fraction p of ferromagnetic bonds. We investigate the low-temperature paramagnetic–ferromagnetic phase boundary below the multicritical point at T N = 0.9530(4), P N = 0.89078(8), as well as the zero-temperature ferromagnetic–spin-glass transition. Finite-size scaling analysis reveals that the phase boundary for T < T N exhibits reentrant behavior. By analyzing the evolution of the magnetization-resolved density of states g ( E , M ) and ground-state spin configurations against increasing frustration, we provide strong evidence that the zero-temperature transition is a mixed-order. Finite-size scaling conducted on the spin-glass side supports the validity of β = 0, where β is the magnetization exponent, with a correlation length exponent ν = 1.50(8). Our results provide new insights into the nature of the ferromagnetic-to-spin-glass phase transition in an extensively degenerate ground state.
We investigate the ferromagnetic-glassy transitions that separate the low-temperature ferromagnetic and spin-glass phases in the temperature-disorder phase diagram of three-dimensional Ising spin-glass models. For this purpose, we consider the cubic-lattice $\ifmmode\pm\else\textpm\fi{}J$ (Edwards-Anderson) Ising model with bond distribution $P(J)=p\ensuremath{\delta}(J\ensuremath{-}1)+(1\ensuremath{-}p)\ensuremath{\delta}(J+1)$, and present a numerical Monte Carlo study of the critical behavior along the line that marks the onset of ferromagnetism. The finite-size scaling analysis of the Monte Carlo data shows that the ferromagnetic-glassy transition line is slightly reentrant. As a consequence, for an interval of the disorder parameter ppp, around $p\ensuremath{\approx}0.77$, the system presents a low-temperature glassy phase, an intermediate ferromagnetic phase, and a high-temperature paramagnetic phase. Along the ferromagnetic-glassy transition line, magnetic correlations show a universal critical behavior with critical exponents $\ensuremath{\nu}=0.96(2)$ and $\ensuremath{\eta}=\ensuremath{-}0.39(2)$. The hyperscaling relation $\ensuremath{\beta}/\ensuremath{\nu}=(1+\ensuremath{\eta})/2$ is satisfied at the transitions, so that $\ensuremath{\beta}/\ensuremath{\nu}=0.305(10)$. This magnetic critical behavior represents a new universality class for ferromagnetic transitions in Ising-type disordered systems. Overlap correlations are apparently not critical and show a smooth behavior across the transition.
Recently, there is interest in coincident “sharpening” and “learnability” transitions in monitored quantum systems. In the latter, an outside observer's ability to infer properties of a quantum system from measurements undergoes a phase transition. Such transitions appear to be related to the decodability transition in quantum error correction, but the precise connection is not clear. Here, we study these problems under one framework we call the general quantum inference problem. In cases as above where the system has a Markov structure, we say that the inference is on a quantum hidden Markov model. We show a formal connection to classical hidden Markov models and that they coincide for certain setups. For example, we prove this for those involving Haar-random unitaries and measurements. We introduce the notion of Bayes nonoptimality, where parameters used for inference differs from true ones. This allows us to expand the phase diagrams of above models. At Bayes optimality, we obtain an explicit relation between “sharpening” and “learnability” order parameters, explicitly showing that the two transitions coincide. Next, we study concrete examples. We review quantum error correction on the toric and repetition code and their mapping to two-dimensional (2D) random-bond Ising model (RBIM) through our framework. We study the Haar-random U ( 1 ) -symmetric monitored quantum circuit and tree, mapping each to inference models that we call the planted simple symmetric exclusion process and planted , respectively, and expanding the phase diagram to Bayes nonoptimality. For the circuit, we deduce the phase boundary numerically and analytically argue that it is of a single universality class. For the tree, we present an exact solution of the entire phase boundary, which displays re-entrance as does the 2D RBIM. We discuss these phase diagrams, with their interpretations for quantum inference problems and rigorous arguments on their shapes.
At low temperatures, the classical two-dimensional random bond Ising model undergoes a frustration-driven ferromagnet-to-paramagnet transition controlled by a zero-temperature fixed point separating ferromagnet and spin glass phases. We show that this critical point can be understood through a renormalization group transformation that constructs the ground state of the Ising model through a sequence of Hamiltonians that, starting with an unfrustrated model, iteratively adds in frustration until the target Hamiltonian is reached. Via a mapping of the thermodynamics of the 2d Ising model to the spectral properties of a related Hermitian matrix -- the Hamiltonian of a noninteracting quantum problem -- this RG procedure corresponds to an iterative diagonalization of the quantum Hamiltonian. The flow toward zero temperature in the Ising picture manifests as a flow toward infinite randomness in the spectrum of the quantum Hamiltonian, with the log gap of the Hamiltonian scaling as a power of the system size: log⁡εmin−1∼Lψ\log \varepsilon_{\it min}^{-1} \sim L^ψlogεmin−1​∼Lψ. The tunneling exponent ψψψ is equal to the spin stiffness exponent θcθ_cθc​ characterizing the zero-temperature fixed point.