An Exact Conjugation Identity for the Many-Body Wilson-Loop Beyond Quantization

This paper establishes and numerically verifies an exact conjugation identity, $W(-\delta)=W(\delta)^*$, for the many-body Wilson loop in dimerized lattice models, demonstrating that the associated Berry phase satisfies $\gamma(-\delta)=-\gamma(\delta)$ even in regimes where the phase is not symmetry-quantized.

The Gist

The Big Idea: A Mirror in the Quantum World

Imagine you are walking through a strange, magical forest (the quantum system) where the trees are arranged in a specific pattern. Some trees are close together, and some are far apart. This pattern is called dimerization (represented by the symbol $\delta$).

In this forest, there is a special "magic wind" (a magnetic flux) that you can blow through the trees. As you walk in a circle around the forest, the wind changes direction. The paper is about a special rule that governs how the forest "feels" when you change the tree pattern or the wind direction.

The authors discovered a perfect mirror rule:

If you flip the tree pattern (make the close trees far and vice versa) AND reverse the wind direction, the "memory" of your walk through the forest becomes the exact mirror image (complex conjugate) of the original walk.

In math terms, if $W$ is the "memory" of the walk, then:
$$W(\text{flipped pattern}) = \text{Mirror}(W(\text{original pattern}))$$

The Characters in the Story

To understand how this works, let's meet the three "magicians" (mathematical operations) who team up to create this mirror effect:

1. The Translator (Shift): Imagine you take a step forward. Every tree moves one spot to the left. This changes which trees are "close" and which are "far."
2. The Swapper (Particle-Hole): Imagine swapping every tree with a hole where a tree could be. This flips the nature of the forest.
3. The Time-Reverser (Complex Conjugation): Imagine playing a movie of your walk backward. This turns the "memory" of the walk into its mirror image.

The Magic Trick:
The paper proves that if you do all three of these things at once (Translate + Swap + Reverse Time), the forest looks exactly the same as it did before, except that the tree pattern is flipped ($\delta \to -\delta$) and the wind is blowing the other way ($\theta \to -\theta$).

Because the forest looks the same, the "memory" of your walk must also be related. Specifically, the memory of the flipped forest is just the mirror image of the original memory.

Why This Is a Big Deal (The "Aha!" Moment)

Usually, in physics, we only see these perfect mirror rules when things are "locked" or "quantized." Think of a combination lock: the numbers can only be 0 or 1. In those cases, the rules are strict and easy to predict.

The breakthrough here is that this rule works even when the lock is unlocked.

• The Old View: "We can only predict the mirror rule if the system is stuck at a specific value (like 0 or $\pi$)."
• The New View: "No! The mirror rule works even if the system is free to slide anywhere between 0 and 1. It works in the 'in-between' zones."

The authors tested this using a supercomputer simulation (called DMRG, which is like a very smart way of solving puzzles with millions of pieces). They built a model of the forest, changed the tree patterns, and watched the "memory" of the walk.

The Result:
Even when the "memory" (called the Berry Phase) was floating freely and not stuck to a specific number, the mirror rule held true perfectly.

• If the memory was $+0.8$ for one pattern, it was $-0.8$ for the flipped pattern.
• If the memory was a complex number (like a direction on a compass), the flipped pattern pointed in the exact opposite direction.

The Practical Superpower

Why should a regular person care? Because this rule acts like a spell-checker for scientists.

When scientists use computers to simulate these quantum forests, they often make tiny mistakes or the computer gets "noisy."

• The Check: If you calculate the walk for the original pattern and the flipped pattern, and they don't look like perfect mirror images, you know something is wrong.
  • Maybe your computer simulation is inaccurate.
  • Maybe your model of the forest is broken (e.g., you forgot to keep the number of trees balanced).

By forcing the results to obey this mirror rule, scientists can clean up their data and get much more accurate answers. It's like having a built-in "error detector" for quantum experiments.

Summary Analogy

Imagine you are recording a song ($W$).

• Standard Physics: You can only predict what the song sounds like if you play it at a specific volume (Quantization).
• This Paper: The authors found a rule that says: "If you flip the lyrics (change the pattern) and play the tape backward (reverse flux), the new song is the exact echo of the old one."
• The Benefit: This rule works whether the song is loud, quiet, or anywhere in between. And if your recording studio doesn't follow this echo rule, you know your microphone is broken.

In short: The paper found a universal "mirror law" for quantum systems that works even when the system is free to move, providing a powerful new tool to check the accuracy of quantum simulations.

Technical Summary

1. Problem Statement

The paper addresses the characterization of one-dimensional correlated quantum systems using the many-body Wilson-loop ($W$), defined as the product of ground-state overlaps along a $U(1)$ flux-insertion (twist) cycle.

• Context: Traditionally, constraints on Wilson-loops and the associated Berry phase ($\gamma \equiv -\arg W$) have been discussed primarily in symmetry-quantized settings, where $\gamma$ is pinned to discrete values (e.g., $0$ or $\pi$) due to specific symmetries.
• Gap: There is a lack of rigorous, exact constraints on Wilson-loops in gapped but non-quantized (de-pinned) regimes, where the Berry phase varies continuously and is not fixed to discrete values by symmetry.
• Goal: To establish an exact, gauge-invariant identity for the Wilson-loop that holds even when the Berry phase is not quantized, relying on microscopic symmetries rather than global topological quantization.

2. Methodology

The author employs a combination of analytical derivation and numerical verification using the Density-Matrix Renormalization Group (DMRG).

A. Analytical Framework

The paper derives an exact conjugation identity based on two minimal conditions:

1. Condition 1 (Antiunitary Closure): There must exist a composite antiunitary mapping $\Xi$ acting on the flux-threaded Hamiltonian family $\hat{H}(\delta, \theta)$ such that it implements the transformation:
   $$(\delta, \theta) \mapsto (-\delta, -\theta)$$
   Here, $\delta$ is the bond-dimerization parameter, and $\theta$ is the $U(1)$ flux. This mapping involves reversing the bond pattern (via permutation of hopping parameters) and reversing the flux orientation.
2. Condition 2 (Gapped Cycle): The ground state must remain non-degenerate (gapped) along the entire twist cycle $\theta \in [0, 2\pi]$.

The Derivation:

• The mapping $\Xi$ relates ground states at opposite points in the parameter space: $|\Psi_{-\delta}(-\theta)\rangle = e^{i\phi(\theta)} \Xi |\Psi_{\delta}(\theta)\rangle$.
• Because $\Xi$ is antiunitary, it induces complex conjugation on overlaps.
• When calculating the Wilson-loop $W(\delta) = \prod \langle \Psi(\theta_j) | \Psi(\theta_{j+1}) \rangle$ over a closed cycle, the spurious phase factors $e^{i\phi}$ cancel out.
• This leads to the exact identity:
  $$W(-\delta) = W(\delta)^*$$
• Consequently, the Berry phase satisfies: $\gamma(-\delta) = -\gamma(\delta) \pmod{2\pi}$.

B. Microscopic Realization

The author constructs the mapping $\Xi$ explicitly for a dimerized staggered Hubbard ring at half-filling:

• Hamiltonian: Includes kinetic energy with dimerization $\delta$, a staggered potential $\lambda$, and on-site interaction $U$.
• Mapping Construction: $\Xi$ is defined as the product of three operations:
  1. One-site translation ($T_1$): Shifts the lattice, flipping the sign of the staggered factor and dimerization ($\delta \to -\delta$).
  2. Particle-hole transformation ($C_{ph}$): Flips the sign of the staggered potential term.
  3. Complex conjugation ($K$): Reverses the flux ($\theta \to -\theta$).
• The combined operation satisfies $\Xi \hat{H}(\delta, \theta) \Xi^{-1} = \hat{H}(-\delta, -\theta)$.

C. Numerical Verification

• Model: Dimerized staggered Hubbard model with $U=6.0$ and $7.0$, $\lambda=0.67$, and dimerization $\delta \in \{\pm 0.005, \pm 0.01\}$.
• Technique: DMRG was used to compute ground states $|\Psi_\delta(\theta)\rangle$ along the twist cycle.
• Regime: Parameters were chosen to ensure the system is gapped but the Berry phase is de-pinned (continuously varying), avoiding symmetry-enforced quantization.
• Diagnostics: The authors analyzed the stability of overlaps and the convergence of the Berry phase with respect to the twist-grid size ($N_\theta$).

3. Key Contributions

1. Exact Conjugation Identity: The paper establishes $W(-\delta) = W(\delta)^*$ as a rigorous constraint for interacting lattice models, valid beyond the regime of symmetry-quantized Berry phases.
2. Generalization of Symmetry: It demonstrates that "bond-pattern reversal" combined with "flux reversal" constitutes a composite symmetry sufficient to enforce this identity, even in the absence of quantization.
3. Diagnostic Tool: The identity serves as a stringent internal check for numerical calculations of many-body holonomies. Deviations from $W(-\delta) = W(\delta)^*$ indicate either broken symmetry (e.g., away from half-filling) or numerical inaccuracies in the overlap calculation.
4. Symmetrization Protocol: The author proposes a method to reduce statistical noise in numerical estimators by symmetrizing the Wilson-loop:
   $$W_{\text{sym}}(\delta) = \frac{W(\delta) + W(-\delta)^*}{2}$$

4. Results

• Numerical Confirmation: DMRG simulations confirmed that for gapped, non-degenerate ground states, the relation $W(-\delta) \approx W(\delta)^*$ holds within numerical precision for various interaction strengths ($U$) and dimerization values ($\delta$).
• Berry Phase Behavior: The results verified that $\gamma(-\delta) \approx -\gamma(\delta)$ in the de-pinned regime. The Berry phase varied smoothly with $\delta$ and was not pinned to $0$ or $\pi$.
• Finite-Size Analysis: The study analyzed the dependence of $\gamma$ on the twist-grid resolution ($N_\theta$). The data fit well to a quadratic expansion in $1/N_\theta$, confirming that the observed oddness in $\gamma$ is a physical property of the system, not a discretization artifact.
• Closing Overlap: The authors noted that the "closing link" overlap (connecting $\theta=2\pi$ to $\theta=0$) is the dominant source of suppression in $|W|$, but this behavior is consistent with discretization artifacts and improves with larger $N_\theta$.

5. Significance

• Beyond Quantization: This work fundamentally expands the understanding of topological invariants in 1D systems. It proves that useful, exact constraints exist even when the system is not in a topological phase characterized by quantized invariants.
• Robustness: The identity relies on the existence of a specific composite antiunitary mapping, making it robust against continuous deformations of the Hamiltonian as long as the gap remains open and the symmetry is preserved.
• Practical Utility: For computational physicists, this provides a powerful "sanity check" for DMRG and other tensor network methods. If the conjugation identity fails in a simulation where it should hold, it signals a problem with the simulation (e.g., insufficient bond dimension or failure to maintain the gap) or a breakdown of the assumed symmetry.
• Theoretical Bridge: It connects microscopic lattice symmetries (bond permutations) directly to macroscopic geometric phases (Wilson-loops) without requiring the intermediate step of topological quantization.