Diagnosing phase transitions through time-scale entanglement

This paper introduces time-scale entanglement, a novel form of entanglement between imaginary time scales accessible via quantics tensor train diagnostics (QTTD), as a universal and unbiased indicator that is generically enhanced near phase transitions and becomes scale-invariant at quantum critical points.

The Gist

Imagine you are trying to understand a complex orchestra playing a piece of music. Usually, to figure out if the orchestra is about to switch from a slow, sad song to a fast, energetic one (a "phase transition"), you might listen for a specific instrument, like the drums or the violins, to change their rhythm. If you don't know which instrument to listen for, or if the change is subtle, you might miss it entirely.

This paper introduces a new way to listen to the "music" of quantum materials. Instead of focusing on specific instruments (like magnetic spins or electrical charges), the authors propose listening to the relationship between different speeds of time.

Here is the breakdown of their discovery using simple analogies:

1. The New Kind of "Entanglement"

In quantum physics, "entanglement" usually means two particles are linked so closely that what happens to one instantly affects the other, no matter the distance. We usually think of this as a link across space.

The authors discovered a different kind of link: Time-Scale Entanglement.

• The Analogy: Imagine a movie. You have the "wide shot" (the whole scene), the "medium shot" (a character talking), and the "close-up" (a flicker of an eye). Usually, these are just different views. But in this quantum world, the "wide shot" and the "close-up" are so deeply connected that you can't describe one without the other. They are "entangled" across different speeds of time.
• The Tool: To measure this, they use a mathematical tool called Quantics Tensor Train (QTT). Think of this as a super-smart compression algorithm (like a ZIP file for complex data). It breaks the quantum "movie" down into layers of time scales.

2. The "Bond Dimension" as a Stress Gauge

The QTT tool has a number called the bond dimension.

• The Analogy: Imagine the bond dimension is the width of a bridge connecting different time scales.
  • If the system is calm and stable, the bridge is narrow. The "wide shots" and "close-ups" of the movie don't need to talk to each other much.
  • If the system is about to undergo a dramatic change (like water turning to ice, or a metal turning into an insulator), the bridge suddenly gets massive. The different time scales become wildly entangled and dependent on each other.

3. The Main Discovery: The Bridge Spikes at Critical Moments

The paper claims that whenever a material is about to change its state (a phase transition) or is in a "crossover" (a smooth shift between states), this "bridge" (the bond dimension) gets huge.

• The "Universal Detector": The most exciting part is that you don't need to know what is changing. Whether it's a magnet losing its magnetism or an electron getting stuck, the bridge gets wide in all cases.
• The Metaphor: It's like having a single sensor that detects an earthquake. You don't need to know if the earthquake is caused by shifting tectonic plates or a volcanic eruption; the ground just shakes, and your sensor goes off. Similarly, this method detects the "shaking" of time scales without needing to know the specific physics of the transition beforehand.

4. What They Tested

The authors tested this idea on several different "orchestras" (quantum models):

• Tiny Rings of Electrons: They watched how the "bridge" widened exactly when the electrons changed their ground state.
• The Ising Model (Magnets): They found that at the exact moment the magnet switches from ordered to disordered, the time scales become perfectly balanced and uniform (scale-invariant). The bridge becomes a flat, wide plain where every time scale is equally important.
• Real Materials (NdNiO2): They applied this to a real chemical compound. Even though the data was noisy and complex, the "bridge" still widened, correctly identifying the moment the material switched from conducting electricity to blocking it (the Mott transition).

5. Why This Matters (According to the Paper)

Currently, scientists often have to guess which "instrument" (susceptibility) to measure to find a phase transition. If they guess wrong, they miss it.

• The Paper's Claim: This new method (called QTTD) is a "universal and unbiased" diagnostic. It doesn't care what specific property you are looking at. If you have the data for any correlation function (any interaction between particles), you can run it through this QTT tool. If the "bond dimension" spikes, you know a phase transition or crossover is happening, even if you didn't know it was coming.

Summary

The paper argues that phase transitions are not just about space; they are about how time scales talk to each other. When a quantum system is about to change its nature, all its different time scales get tangled up together, creating a massive "traffic jam" of information. By measuring the size of this jam (the bond dimension), we can detect these changes universally, without needing to know the specific details of the material beforehand.

Technical Summary

Technical Summary: Diagnosing Phase Transitions through Time-Scale Entanglement

Problem Statement
The detection and numerical treatment of phase transitions in extended many-body systems present significant challenges. Traditional methods often rely on calculating specific susceptibilities or order parameters, which requires a priori knowledge of the transition type and the relevant symmetry breaking. This approach can be brittle, particularly for exotic phases or first-order transitions where susceptibility signals may be weak or ambiguous. Furthermore, calculating entanglement in true many-electron systems is computationally prohibitive. While Matrix Product States (MPS) have successfully utilized bond dimensions to gauge spatial entanglement in quasi-one-dimensional systems, a corresponding physical measure for the entanglement between different temporal scales in generic correlation functions has been lacking.

Methodology: Quantics Tensor Train Diagnostics (QTTD)
The authors propose a novel diagnostic framework termed Quantics Tensor Train Diagnostics (QTTD). This method leverages the Quantics Tensor Train (QTT) ansatz, a tensor network structure originally developed for compressing high-dimensional functions by representing variables in binary (quantic) form.

1. QTT Representation: In this framework, an imaginary time variable $\tau$ (discretized on a grid of size $M=2^R$) is expressed via its binary representation $\tau = \sum_{\ell=1}^R 2^{R-\ell}\sigma_\ell$. Each binary index $\sigma_\ell$ corresponds to an exponentially distinct time scale (from coarse to fine).
2. Tensor Train Factorization: The correlation function (e.g., a Green's function $G(\tau)$) is factorized into a tensor train (or Matrix Product State) where each tensor $M_\ell$ connects adjacent time scales. The "virtual" indices connecting these tensors are the bonds.
3. Bond Dimension as Entanglement Measure: The authors define the bond dimension $D_\ell$ (and its maximum $D_{\max}$ or sum $D_{\text{sum}}$) not merely as a compression metric, but as a physical measure of time-scale entanglement. Analogous to how MPS bond dimensions quantify spatial entanglement, QTT bond dimensions quantify the coupling and information flow between different imaginary time scales.
4. Diagnostic Procedure: The method involves calculating a system correlator, compressing it into a QTT, and analyzing the bond dimension profile. The authors hypothesize that the bond dimension will exhibit maxima at phase transitions and crossovers due to the system-inherent enhancement of time-scale entanglement.

Key Contributions and Results
The paper demonstrates the efficacy of QTTD across several distinct models, showing that enhanced time-scale entanglement is a universal signature of criticality and crossovers, largely independent of the specific observable chosen.

• Hubbard Dimer and Rings: In finite Hubbard systems, the QTT bond dimension of the four-point Green's function exhibits sharp peaks precisely at ground state crossings (phase transitions between different electron fillings). The method successfully identifies these crossings without prior knowledge of the specific susceptibility. The peaks sharpen as temperature decreases, converging to the zero-temperature ground state crossings.
• Transverse-Field Ising Model (TFIM): At the quantum critical point (QCP) of the 1D TFIM, the system exhibits scale invariance. The QTT analysis reveals that the bond dimension profile becomes flat (uniform) across all logarithmic time scales near the critical point. This indicates that all time scales contribute equally to the correlation function, providing a direct numerical manifestation of scale invariance.
• Mott Transition (DMFT): Applied to the first-order metal-insulator transition in the Hubbard model on a Bethe lattice (via Dynamical Mean-Field Theory), QTTD detects pronounced peaks in the sum of bond dimensions ($D_{\text{sum}}$) as the system approaches the coexistence region from both metallic and insulating sides. In the case of the 3D material NdNiO$_2$, the method identifies a crossover from metallic to insulating behavior via a peak in $D_{\text{sum}}$ and a broadening of the bond dimension profile, even in the absence of a diverging correlation length typical of second-order transitions.
• Single-Impurity Anderson Model (SIAM): The method diagnoses the crossover between the local moment and Kondo regimes. The maximum bond dimension shows a broad peak in the vicinity of the Kondo temperature ($T_K$), effectively mapping the crossover without requiring specific order parameters.
• Universality and Robustness: The results hold for both single- and two-particle propagators. Crucially, the enhanced entanglement is observed even in observables (like the one-particle Green's function) that do not show singular behavior in their static, time-independent parts near a transition. The method is robust against noise; by adjusting the QTT compression cutoff, artifacts from low-accuracy data can be truncated, leaving stable peaks associated with physical transitions.

Significance and Claims
The paper claims that time-scale entanglement is a fundamental, system-inherent property that becomes maximal at phase transitions and crossovers. The primary significance of this work is the establishment of QTTD as a universal and unbiased diagnostic for criticality.

• Unbiased Detection: Unlike traditional methods that require selecting specific susceptibilities based on theoretical expectations, QTTD can utilize any available correlator (one- or two-particle) to detect transitions.
• Scale Invariance Signature: The method provides a new way to observe scale invariance numerically: the flattening of the bond dimension profile across time scales at a quantum critical point.
• Applicability: The approach is applicable to both zero and finite temperatures and does not rely on ground-state-only methods. It offers a pathway to probe how criticality and scale invariance manifest in generic observables through the coupling of imaginary time scales.

The authors conclude that this perspective offers a fresh understanding of phase transitions, suggesting that the "complexity" of correlation functions at critical points is encoded in the entanglement between temporal scales, which is directly accessible through the bond dimensions of QTT representations.