⚛️ quantum physics

Universal Quantum Criticality from a Gaussian Variational Manifold

This paper introduces a systematic Gaussian variational procedure to derive accurate noninteracting fermionic descriptions of quantum phase transitions, demonstrating that while the Gaussian wavefunction approximation is weakest near the critical point, it still successfully predicts universal critical properties and critical lines for systems like the long-range transverse-field Ising chain.

Original authors: Reyhaneh Khasseh, Fabian Ballar Trigueros, Markus Heyl, M. A. Rajabpour

Published 2026-10-20

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Reyhaneh Khasseh, Fabian Ballar Trigueros, Markus Heyl, M. A. Rajabpour

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the world of quantum physics, materials often undergo dramatic transformations known as phase transitions. Think of water turning to ice or a magnet losing its pull as it heats up. At the precise moment of these changes, the behavior of the system becomes "universal," meaning the specific details of the atoms matter less than the overall shape and symmetry of the material. For decades, physicists have sought a simple way to describe these complex, interacting systems. The most straightforward descriptions involve particles that do not talk to one another, a scenario that is easy to calculate but often fails to capture the messy reality of interacting matter. The central puzzle has been whether a simplified, non-interacting model could ever accurately predict the behavior of a truly interacting system right at the point of change, even if the model's description of the individual particles was imperfect.

A team of researchers has now provided a systematic method to answer this question, focusing on a specific quantum system: a chain of magnetic spins that influence each other over long distances. In this setup, the magnetic influence between two spins weakens as they get farther apart, following a specific mathematical rule. While the simplest version of this chain behaves like non-interacting particles, the long-range version creates a complex web of interactions that usually requires heavy computer simulations to solve. The researchers developed a new approach where they kept the full, complex rules of the magnetic chain but restricted their search for the solution to a specific, simpler type of quantum state known as a Gaussian state. These states are mathematically manageable and can be described by a single set of correlation data, making them much easier to work with than the full, messy reality.

By applying this restriction, the team reduced the problem to a single, self-consistent equation that could be solved to find the exact point where the phase transition occurs. Their calculations produced a critical line—a boundary separating different magnetic behaviors—that matched independent, high-precision numerical benchmarks almost perfectly. This was a significant success, suggesting that their simplified model could indeed predict the universal properties of the complex system. However, a deeper look revealed a surprising twist. When the researchers compared their simplified Gaussian state to the exact, true ground state of the system using a measure of similarity called fidelity, they found the two were least alike right at the transition point. In fact, the exact state was most "non-Gaussian," or furthest from their simplified model, precisely where the model's predictions for the transition were most accurate.

This counterintuitive result suggests that capturing the universal physics of a phase transition does not require a model to perfectly mimic the microscopic details of every particle. Instead, the accuracy comes from the model retaining the correct "singular structure" or the specific way the system's correlations behave at large distances. The researchers found that their simplified description successfully preserved the long-wavelength patterns that govern critical behavior, even though it failed to describe the short-range, microscopic details of the wave function. They used this framework to derive analytic formulas for critical properties, such as how magnetic correlations decay and how entanglement grows, which agreed closely with numerical data across a wide range of interaction strengths.

The study demonstrates that a variational theory, which seeks the best possible approximation within a restricted set of states, can remain effective even when the approximation itself is weakest at the microscopic level. The key is not whether the critical state looks like a simple, non-interacting system, but whether the simplified model can represent the specific infrared information—the large-scale, low-energy data—that controls the observable properties of interest. By showing that a Gaussian manifold can capture the essential physics of an interacting critical point, the work offers a new perspective on how to understand complex quantum materials. It implies that the reach of such theories is determined by their ability to encode the correct large-scale structure, rather than their ability to perfectly reconstruct the microscopic wave function. This insight could guide future efforts to simulate and understand other interacting quantum systems where exact solutions are out of reach.

Problem Statement
Continuous phase transitions are characterized by universal behavior encoded in long-wavelength fixed-point theories determined by dimensionality and symmetries. A central question in quantum many-body physics is whether this universal behavior, particularly for interacting systems, can be captured by a noninteracting (free) theory that allows for analytical and accurate predictions. Specifically, it is unclear if a variational manifold based on noninteracting fermionic Gaussian states can faithfully approximate an interacting critical wave function sufficiently to reproduce its universal long-distance physics, even if the microscopic wave-function fidelity is poor near the transition. The ferromagnetic long-range transverse-field Ising chain serves as a stringent testbed for this question. While its nearest-neighbor limit maps to an exactly solvable free-fermion model, the introduction of algebraically decaying spin couplings generates nonlocal Jordan–Wigner strings, creating an interacting fermionic problem with a rich phase diagram spanning long-range mean-field scaling, a regime with continuously varying critical exponents, and a crossover to short-range Ising criticality.

Methodology
The authors develop a thermodynamic Gaussian variational description for the ferromagnetic long-range transverse-field Ising chain. The core strategy involves retaining the full interacting spin Hamiltonian while restricting the variational state to the manifold of fermionic Gaussian states.

Variational Formulation: The trial state is a fermionic Gaussian state, efficiently specified by a covariance matrix. Translation invariance and Gaussian purity constrain the mixed Majorana contraction to a momentum-space symbol $G(q) = -e^{i\phi_q}$ , parametrized by a single real phase profile $\phi_q$ .
Energy Functional: The thermodynamic variational energy density is expressed in terms of this phase profile. Crucially, while the state is Gaussian, the long-range spin interactions enter through Toeplitz determinants of the correlation functions, preserving the non-quadratic character of the original spin problem.
Euler Equation: Minimizing the energy functional leads to a self-consistent Euler equation. The stationarity condition aligns the phase of the Gaussian symbol with a complex "Euler field" $F_E(q)$ .
Critical Analysis:
- Critical Line: Criticality is identified as the point where the Euler field vanishes at the Brillouin-zone edge ( $q=\pi$ ). This yields a coupled nonlinear thermodynamic problem for the critical symbol $G_c$ and the critical field $h_c$ .
- Analytic Exponents: The authors analyze the singular momentum structure of the Euler field near the critical point. By combining the asymptotic behavior of the Euler field's magnitude (characterized by an exponent $\zeta_E$ ) with the Fisher–Hartwig asymptotics of Toeplitz determinants, they derive closed-form analytic expressions for the magnetic correlation exponent $x_\sigma$ and the dynamical exponent $z$ .
- Entanglement: The logarithmic entanglement coefficient $c_{\text{eff}}$ is derived by identifying the singular part of the critical covariance operator with the spectral class of a single-jump fermionic interface problem, allowing the use of known exact results for free fermions.

Key Results

Critical Line Accuracy: The numerical solution of the self-consistent Euler equation yields a critical line ( $h_c$ vs. interaction range $\alpha$ ) that agrees closely with independent finite-size benchmarks, including entanglement-entropy peaks, Binder-cumulant crossings, and Quantum Monte Carlo (QMC) results.
Paradox of Fidelity vs. Universality: Exact diagonalization reveals that the optimized Gaussian state is least faithful to the exact ground state (lowest fidelity) and that the exact state exhibits its strongest non-Gaussianity precisely near the critical transition. Despite this microscopic inaccuracy, the Gaussian variational description successfully reproduces the universal critical properties.
Analytic Predictions:
- Magnetic Exponent: The paper derives an analytic expression for the magnetic exponent $x_\sigma$ that captures the continuously varying behavior in the long-range regime ( $1 < \alpha < 11/4$ ) and the crossover to the short-range Ising value ( $x_\sigma = 1/4$ ) for $\alpha > 11/4$ . The crossover point $\alpha^* = 11/4$ emerges naturally from the theory without being inserted as an external assumption.
- Dynamical Exponent: The static Euler exponent $\zeta_E$ is shown to satisfy the relation $z = \zeta_E$ on the long-range branch, matching numerical estimates for the dynamical exponent.
- Entanglement: The analytic prediction for the logarithmic entanglement coefficient $c_{\text{eff}}$ matches numerical estimates from optimized Gaussian states, approaching the Ising value $c_{\text{eff}} = 1/2$ in the short-range limit.

Significance and Claims
The paper claims that universal critical information can survive a severe restriction of the variational manifold even when the wave-function description deteriorates near criticality. The key finding is that the accuracy of critical predictions does not track microscopic state fidelity. Instead, the relevant information is encoded in the singular structure retained by the optimized Gaussian solution (specifically, the zeros and phase singularities of the Euler field and the associated Toeplitz asymptotics).

The authors conclude that the reach of a variational theory of criticality is determined by two factors: whether the manifold can represent the infrared structure controlling a universal quantity, and whether variational optimization selects that structure. Gaussian manifolds may remain effective not only at Gaussian or mean-field fixed points but also at interacting critical points, provided the essential infrared data can be encoded through mass zeros, soft modes, or singular covariance structures. The limitations of such approaches arise only when universal physics depends on additional infrared parameters (e.g., Luttinger parameters) that are not generically encoded by the Gaussian covariance structure. Thus, the critical state need not be close to Gaussian for a Gaussian manifold to capture its critical properties.

More like this