Ground state preparation in two-dimensional pure $\mathbb{Z}_2$ lattice gauge theory via deterministic quantum imaginary time evolution

This paper demonstrates that a generalized, gauge-invariant deterministic quantum imaginary time evolution (QITE) algorithm can efficiently prepare the ground state of two-dimensional pure $\mathbb{Z}_2$ lattice gauge theory with high accuracy (relative error < 0.1%) and reduced resource costs, as validated by classical tensor network simulations against DMRG results.

The Gist

Imagine you are trying to find the deepest, most stable valley in a vast, foggy mountain range. In the world of physics, this "valley" is called the ground state, and it represents the most stable, lowest-energy configuration of a system. Finding this state is crucial for understanding how materials behave, how magnets work, or even how the fundamental forces of the universe interact.

However, the "mountain range" we are looking at is a Lattice Gauge Theory—a complex mathematical model used to describe forces like the strong nuclear force. It's so complex that even the world's most powerful supercomputers struggle to find the bottom of the valley without getting lost in a "fog" of mathematical errors (known as the sign problem).

This paper presents a new, clever way to navigate this fog using a Quantum Computer. Here is the story of how they did it, explained simply.

1. The Problem: The "Imaginary" Hike

Physicists have a tool called Imaginary Time Evolution (ITE). Think of this as a magical hiking guide. If you start anywhere on the mountain and let this guide walk you forward in "imaginary time," it naturally pushes you downhill, smoothing out the bumps until you inevitably settle in the deepest valley (the ground state).

The problem? Real quantum computers can only do "real" time travel (moving forward or backward in time). They cannot naturally do "imaginary" time. It's like trying to drive a car that only goes forward, but you need to drive backward to get to your destination.

2. The Solution: The "Deterministic" Detour

The authors use a method called Deterministic Quantum Imaginary Time Evolution (QITE).

• The Metaphor: Imagine you need to walk downhill, but your car can only drive in straight lines. The QITE algorithm is like a GPS that breaks your downhill journey into thousands of tiny, straight-line segments. At each step, it calculates the best straight line to approximate the curve of the hill.
• The Catch: To calculate these straight lines, the computer has to ask a lot of questions (measurements) about the current state of the system. If the mountain is huge, the number of questions becomes so massive that the computer gets overwhelmed. It's like trying to map every single blade of grass in a forest just to find the path.

3. The Secret Weapon: The "Gauge" Filter

This is where the paper's main breakthrough shines. The system they are studying (the Z2 Lattice Gauge Theory) has a special rule called Gauss's Law.

• The Analogy: Imagine a strict bouncer at a club (Gauss's Law). The bouncer says, "No one enters unless they are wearing a specific badge." If you try to enter without the badge, you are rejected.
• The Innovation: The authors realized that the "straight lines" (the mathematical operators) their GPS was trying to calculate didn't all need to be checked. Many of them were "badged" incorrectly and would be rejected by the bouncer anyway.
• The Result: They built a filter that only lets the "badged" operators through. This is like telling the GPS: "Don't bother calculating paths that the bouncer will block."
  • Impact: This reduced the number of questions the computer had to ask by a huge factor (in some cases, from 255 questions down to just 8). It made the journey much faster and cheaper.

4. The Test Drive: A Digital Simulation

Since they didn't have a giant quantum computer ready to run this, they simulated the whole process on a classical supercomputer using a technique called Tensor Networks (think of this as a highly efficient way to compress the map of the mountain).

They tested their method on a "ladder-like" mountain range (a 2D grid) with varying sizes and difficulty levels (coupling strengths).

• The Results: They compared their "QITE GPS" against the gold standard of mountain mapping (called DMRG).
• The Verdict: Their method was incredibly accurate. For systems up to a certain size (12 "plaquettes" or small grid squares), their error was less than 0.1%. That's like finding the bottom of a valley and missing it by less than a millimeter.

5. Why This Matters

• Efficiency: By using the "bouncer" (symmetry) to filter out useless calculations, they made the algorithm much lighter. This is essential for running these simulations on actual quantum hardware in the near future, which has limited memory and battery life.
• Accuracy: They proved that even with these shortcuts, the results are trustworthy.
• Future: This paves the way for simulating more complex physics, like the forces inside a proton or the behavior of exotic materials, which are currently impossible to calculate with classical computers.

Summary

The authors took a difficult quantum algorithm (finding the ground state), realized it was asking too many unnecessary questions, and built a smart filter based on the system's own rules (symmetry) to cut out the noise. They tested it in a simulation and found it works with near-perfect accuracy. It's like turning a slow, confused hiker into a guided tour that knows exactly which paths are open and which are dead ends.

Technical Summary

1. Problem Statement

The preparation of ground states for Lattice Gauge Theories (LGTs) is a critical task for probing quantum phases and properties of high-energy physics. However, classical Monte Carlo methods suffer from the "sign problem," and naive Hamiltonian simulations scale exponentially with system size. While Quantum Imaginary Time Evolution (QITE) offers a pathway to prepare ground states on quantum computers, the deterministic QITE algorithm faces two major bottlenecks:

1. Resource Cost: The algorithm requires solving a linear equation to determine coefficients for a "Pauli pool" (a set of Pauli strings). The size of this pool grows exponentially with the support (number of qubits), leading to an intractable number of measurements and gate operations.
2. Gauge Invariance: Standard implementations do not inherently preserve the local gauge symmetry (Gauss's law constraints), potentially leading to unphysical states or requiring expensive post-selection.

The authors aim to apply deterministic QITE to a two-dimensional (2D) pure $Z_2$ lattice gauge theory, addressing these bottlenecks by constructing a gauge-invariant Pauli pool and evaluating the algorithm's accuracy and resource efficiency via classical tensor network simulations.

2. Methodology

A. Theoretical Framework: Deterministic QITE

The authors utilize the deterministic QITE algorithm (originally proposed in Ref. [9]), which approximates non-unitary imaginary time evolution ($e^{-\tau H}$) using a sequence of unitary real-time evolutions ($e^{-i\Delta\tau A}$).

• Decomposition: The total imaginary time $\tau$ is divided into small steps $\Delta\tau$. The Hamiltonian $H$ is decomposed into local terms $h^{(m)}$.
• Unitary Approximation: For each step, the algorithm finds a unitary operator $A_P = \sum a_\sigma \sigma$ (where $\sigma$ are Pauli strings from a pool $P$) that minimizes the distance between the target imaginary-time state and the unitary-evolved state.
• Linear System: The coefficients $a_\sigma$ are determined by solving a linear equation $S a = b$, where matrix elements $S$ and vector $b$ are expectation values calculated on the current state.

B. Key Innovation: Symmetry-Reduced Pauli Pool

The core contribution is the generalization of symmetry reduction techniques to the 2D $Z_2$ LGT to drastically reduce the Pauli pool size while maintaining gauge invariance.

1. Reality Condition Reduction: By ensuring the initial state and Hamiltonian terms are real in the computational basis, the pool is reduced to only Pauli strings with an odd number of $Y$ operators ($P_{odd}$).
2. Gauge Symmetry Reduction: The authors construct a pool of Pauli operators that commute with Gauss's law constraints ($g_n$).
   • They define the symmetrized pool $P_G$ where the support of non-$X$ operators forms closed loops (or disjoint loops) on the lattice.
   • This ensures the resulting state remains in the physical subspace (gauge invariant) throughout the evolution.
3. Quotient Group Reduction: They further reduce the pool by identifying equivalent Pauli strings under the action of the Gauss's law operators. Instead of treating every string individually, they select representatives from equivalence classes defined by the quotient group $P_{G,odd}/G$.

Resulting Pool Size:
For a support $D$ containing $n_{link}$ links and $n_{plaq}$ plaquettes, the size of the reduced pool is:
$$|P_{G,odd}| = 2^{n_{link}-1} \times (2^{n_{plaq}} - 1)$$
With quotient group reduction, this is further divided by $2^{n_G}$ (where $n_G$ is the number of bulk Gauss's law operators). For a single plaquette, the pool size drops from 255 (full) to 8 (reduced).

C. Numerical Simulation

Since actual quantum hardware is not used, the authors perform classical noiseless simulations using Tensor Networks:

• State Representation: Matrix Product States (MPS) with large bond dimensions to minimize systematic errors.
• Algorithm: Time-Evolving Block Decimation (TEBD) is used to implement the unitary steps.
• Comparisons: The deterministic QITE results are compared against:
  1. Suzuki-Trotterized ITE (ITE): The "exact" imaginary time evolution (without unitary approximation) to isolate QITE-specific errors.
  2. Density Matrix Renormalization Group (DMRG): Used as the ground truth for ground state energy.

3. Key Contributions

1. Generalized Gauge-Invariant Pauli Pool: The paper provides a general formula for constructing Pauli operators that commute with Gauss's law for arbitrary supports in 2D $Z_2$ LGT, extending previous work limited to single plaquettes.
2. Resource Efficiency: Demonstrates that combining reality conditions, gauge symmetry, and quotient group reductions reduces the measurement and gate costs by orders of magnitude (e.g., from $O(2^{N})$ to a manageable polynomial or smaller exponential factor) without introducing algorithmic bias.
3. Accuracy Benchmarking: Provides a rigorous error analysis of deterministic QITE in a 2D gauge theory context, separating Trotter errors, unitary approximation errors, and system-size scaling.

4. Results

• Accuracy: The deterministic QITE achieves a relative error of less than 0.1% compared to DMRG results for system sizes up to 12 plaquettes (ladder geometry with $N_y=3$ and $N_x$ up to 7) and coupling values $0.5 \leq \lambda \leq 5.0$.
• Coupling Dependence:
  • Accuracy is highest for weak coupling ($\lambda \approx 0.5$) because the initial state (ground state of the electric term) is close to the true ground state.
  • Error increases slightly with stronger coupling ($\lambda$), but remains within the 0.1% threshold for the studied regime.
• Time Step Dependence:
  • For weak coupling ($\lambda=0.5$), errors decrease as the time step $\Delta\tau$ decreases, indicating that both Trotter and unitary approximation errors are controllable.
  • For stronger coupling ($\lambda=2.0$), QITE errors saturate at a certain $\Delta\tau$ threshold, suggesting the error is dominated by the unitary approximation (limited Pauli pool support) rather than the time discretization.
• System Size Scaling: The error in QITE increases slowly with system size (number of links), whereas the error in the standard ITE (Trotterized) remains constant. This confirms that the unitary approximation error scales with system size, likely due to the fixed support size of the Pauli pool.

5. Significance and Future Directions

• Feasibility for NISQ: The study demonstrates that deterministic QITE, when combined with symmetry reduction, is a viable and resource-efficient algorithm for preparing ground states in 2D gauge theories on near-term quantum devices. The reduction in measurement overhead is critical for current hardware limitations.
• Gauge Invariance: The method ensures that the quantum state remains in the physical Hilbert space without the need for error correction or post-selection, a significant advantage for simulating gauge theories.
• Future Work: The authors suggest extending the method to:
  • Larger system sizes and genuine 2D geometries (beyond ladder-like structures).
  • Non-Abelian gauge theories.
  • Investigating the robustness of symmetry preservation in the presence of quantum noise.
  • Optimizing initial states for strong-coupling regimes to reduce unitary approximation errors.

In conclusion, this paper establishes a robust framework for applying deterministic QITE to 2D lattice gauge theories, proving that symmetry-aware resource reduction makes the algorithm accurate and practical for current and near-future quantum simulation efforts.