Imaginary-time evolution of interacting spin systems in the truncated Wigner approximation

This paper introduces the imaginary-time truncated Wigner approximation (iTWA), a semiclassical phase-space method that maps the evolution of the density matrix to stochastic differential equations, enabling efficient and accurate calculation of ground states and thermal properties for large interacting spin systems, including frustrated models and those exhibiting quantum phase transitions.

The Gist

Imagine you are trying to find the lowest point in a massive, foggy mountain range. This isn't just any mountain range; it's a landscape made of billions of tiny magnets (spins) that are all trying to point in different directions, pulling and pushing against each other. Finding the absolute lowest valley (the "ground state") where the system is most stable is incredibly difficult. In fact, for certain types of mountains, finding the perfect lowest point is a math problem so hard that even the world's fastest supercomputers would take longer than the age of the universe to solve it exactly.

This is the problem physicists face when studying complex magnetic materials. The paper you shared introduces a new, clever shortcut to solve this problem, called iTWA (imaginary-time Truncated Wigner Approximation).

Here is how it works, explained through simple analogies:

1. The Problem: The Foggy Mountain

In the quantum world, particles don't just sit still; they jitter and fluctuate. To find the most stable state of a group of these particles, scientists usually try to simulate them step-by-step.

• The Old Way (Quantum Monte Carlo): Imagine trying to map the mountain by throwing thousands of darts at a map. For some mountains, the map has "negative areas" (like holes in the ground that don't make sense physically). When this happens, the dart-throwing method breaks down, and the errors grow so fast that the result becomes useless. This is known as the "sign problem."
• The New Way (iTWA): Instead of throwing darts, imagine you are a hiker with a special pair of glasses. These glasses let you see the mountain not as a jagged, impossible-to-solve puzzle, but as a smooth, rolling landscape that you can walk down.

2. The Magic Trick: "Imaginary Time"

The paper's biggest innovation is using "imaginary time."

• Real Time: If you watch a movie of a ball bouncing, it goes up and down forever. This is hard to predict because it keeps moving.
• Imaginary Time: Now, imagine the movie is played in a world where friction is super strong. If you drop a ball in this world, it doesn't bounce; it just rolls down the hill, loses energy, and eventually settles at the very bottom.
• The Analogy: The researchers created a mathematical "friction" that forces the quantum system to "cool down" and roll into its most stable state (the ground state) without needing to solve the impossible math of the exact quantum jitters.

3. The "Truncated Wigner" Map

To make this rolling hill work, they use a method called the Truncated Wigner Approximation (TWA).

• The Phase Space: Think of the quantum system not as a single point, but as a cloud of possibilities. The "Wigner function" is a map of this cloud.
• The Truncation: The map is incredibly detailed and complex. To make it usable, the researchers "blur" the map slightly. They throw away the tiny, microscopic details that are too hard to calculate, keeping only the big, smooth features.
• The Result: This blurring turns a terrifyingly complex quantum equation into a set of Stochastic Differential Equations.
  • Translation: This is just a fancy way of saying they turned the problem into a random walk. Imagine a drunk person walking down a hill. They have a general direction (downhill), but they also take random, jittery steps (quantum noise). By simulating thousands of these "drunk walkers" on a computer, the average path they take reveals the true bottom of the valley.

4. Testing the Method

The authors tested their "hiking glasses" on two very different challenges:

Challenge A: The Frustrated Maze (3-regular Graphs)

• The Setup: Imagine a network of 100 spins where every spin is connected to exactly three others, but they are all fighting to point in opposite directions (Anti-ferromagnetic). This is a classic "frustrated" system, like a group of friends trying to sit at a round table where everyone hates their two neighbors.
• The Result: For small groups, the method matched the perfect, exact answer. For the huge group of 100 spins (which is too big for exact math), their method found a solution that was almost as good as the best classical supercomputer algorithms, but much faster. It successfully navigated the "NP-hard" maze.

Challenge B: The Quantum Phase Transition (The Switch)

• The Setup: Imagine a row of magnets that can be flipped by a magnetic field. At a certain point, the whole row suddenly flips from one state to another. This is a "Quantum Phase Transition."
• The Result: The method correctly predicted exactly when this switch happens and how the system behaves right at the tipping point. This proves that even though they "blurred" the map, they didn't blur away the most important quantum effects.

Why This Matters

This paper is like giving physicists a new, powerful tool to explore the quantum world.

• Before: If a problem was too complex or "frustrated," we often had to give up or use approximations that were too rough to be useful.
• Now: With iTWA, we can simulate large, complex systems of interacting spins on standard computers (or GPUs) to find their stable states. It bridges the gap between the messy, probabilistic quantum world and the smooth, predictable world of classical physics.

In a nutshell: The authors built a "quantum friction" machine. By simulating how a system "cools down" in imaginary time using random walks, they can find the most stable state of complex magnetic systems that were previously too hard to solve. It's a bit like finding the bottom of a foggy valley by letting a thousand hikers wander down until they all settle in the same spot.

Technical Summary

Here is a detailed technical summary of the paper "Imaginary-time evolution of interacting spin systems in the truncated Wigner approximation" by Schlegel, Breu, and Fleischhauer.

1. Problem Statement

Calculating the thermal and ground states of large interacting spin systems is a fundamental challenge in many-body physics.

• Limitations of Existing Methods: Quantum Monte Carlo (QMC) techniques are efficient for non-frustrated systems but fail for frustrated models due to the "sign problem," where negative Boltzmann weights cause exponential growth in statistical errors.
• Computational Complexity: Finding the exact ground state of frustrated spin Hamiltonians (e.g., anti-ferromagnetic Ising models on general graphs) is an NP-hard problem. Even approximating the ground state within a specific error margin (e.g., $< 1/17$ for 3-regular graphs) is NP-hard.
• Need for Approximation: Since exact solutions are often computationally intractable for large systems, there is a need for efficient approximate simulation techniques that can capture essential quantum effects beyond mean-field theory.

2. Methodology: Imaginary-Time Truncated Wigner Approximation (iTWA)

The authors propose a semiclassical phase-space method called iTWA, which extends the Truncated Wigner Approximation (TWA) from real-time dynamics to imaginary-time evolution.

• Phase-Space Mapping: The method maps the canonical density operator $\hat{\rho}$ to a Wigner function $W(\Omega)$ in a continuous phase space defined by angles $(\theta, \phi)$ for each spin. The mapping utilizes an overcomplete basis of phase-point operators $\hat{\Delta}(\Omega)$.
• Imaginary-Time Evolution: Instead of unitary evolution ($\partial_t \hat{\rho} = -i[\hat{H}, \hat{\rho}]$), the system evolves in imaginary time $\tau$ according to $\partial_\tau \hat{\rho} = -(\hat{H} - \langle\hat{H}\rangle)\hat{\rho}$.
• Derivation of Stochastic Equations:
  1. The evolution of the Wigner function is derived from the master equation, resulting in a partial differential equation (PDE).
  2. Unlike real-time TWA, the imaginary-time equation contains a term proportional to the Wigner function itself (due to the normalization factor).
  3. By exploiting the non-uniqueness of the Hilbert-to-phase-space mapping and gauge freedom, the authors truncate the PDE at the Fokker-Planck level (second-order derivatives).
  4. This truncation yields a set of Stochastic Differential Equations (SDEs):
     $$dx_j(\tau) = A_j(\Omega) d\tau + \sum_l B_{jl}(\Omega) dW_l(\tau)$$
     where $x_j$ represents the phase-space coordinates, $A_j$ is the drift term, $B_{jl}$ is the diffusion matrix, and $dW_l$ are Wiener processes.
• Handling Quantum Fluctuations: The iTWA incorporates quantum effects through two mechanisms:
  1. Sampling initial conditions from a distribution.
  2. Crucially, adding stochastic noise terms (diffusion) in the imaginary-time evolution, which allows the method to capture quantum fluctuations relevant to ground states and phase transitions, unlike real-time TWA which is often limited to short times.
• Expectation Values: Observables are calculated by averaging over stochastic trajectories, weighted by an exponential factor involving the integrated Hamiltonian along the trajectory:
  $$\langle \hat{A} \rangle(\tau) = \overline{A(\Omega(\tau)) e^{-\int_0^\tau d\tau' H(\Omega(\tau'))}}$$
  To mitigate large fluctuations caused by negative Hamiltonian values, a shift $E(\tau)$ is applied to the energy term in the exponent.

3. Key Contributions

• Formalism Development: Successfully extended the TWA formalism to imaginary time, deriving the specific SDEs required to simulate the canonical density matrix evolution.
• Handling NP-Hard Problems: Demonstrated the application of iTWA to frustrated anti-ferromagnetic Ising models on random 3-regular graphs, a class of problems known to be NP-hard.
• Quantum Phase Transition (QPT) Benchmarking: Validated the method's ability to describe critical quantum phenomena by analyzing the transverse-field Ising model (TFIM) in 1D and 2D, showing agreement with exact solutions and quantum-classical correspondence.
• Algorithmic Efficiency: Showed that the method can be efficiently simulated using GPUs and scales to system sizes ($N=100$) where exact diagonalization is impossible.

4. Results

A. Anti-Ferromagnetic Ising Model on 3-Regular Graphs

• Small Systems ($N=22$): The iTWA results for average energy $\langle \hat{H} \rangle$ across a range of inverse temperatures ($\tau$) showed very good agreement with Exact Diagonalization (ED). The ground state energies were predicted with near-perfect accuracy.
• Large Systems ($N=100$): Since ED is intractable for $N=100$, the authors compared iTWA results against the best-known classical algorithm, GUROBI.
  • The iTWA simulations approached the GUROBI estimates with high accuracy.
  • With as few as $10^2$trajectories, the method crossed the theoretical NP-hardness limit for approximation error ($\Delta \epsilon = 1/17$), demonstrating its utility as a heuristic solver for combinatorial optimization problems (like MaxCut).

B. Transverse-Field Ising Model (TFIM)

• Setup: Investigated the QPT in 1D and 2D lattices by varying the ratio of transverse field $h$ to interaction strength $J$.
• Critical Behavior: The iTWA correctly reproduced the quantum phase transition points ($h/J = 1$ in 1D and $h/J \approx 3.044$ in 2D).
• Order Parameter: The calculated squared magnetization $\langle m^2 \rangle$ matched exact finite-size and thermodynamic limit solutions.
• Significance: This confirms that iTWA, despite being a semiclassical method, can capture the quantum-classical correspondence and universal critical behavior of quantum phase transitions, provided the correspondence rules are optimized to minimize approximations in the diffusion matrix.

5. Significance and Conclusion

• Beyond Mean-Field: The iTWA provides a powerful tool that goes beyond standard mean-field descriptions by incorporating quantum fluctuations via stochastic noise in imaginary time.
• Versatility: It bridges the gap between classical optimization heuristics and quantum many-body physics, offering a semiclassical approach to solve NP-hard ground state problems and simulate quantum criticality.
• Limitations: The authors note that while effective for non-frustrated or weakly frustrated systems and specific phase transitions, the method may struggle with highly entangled ground states (e.g., spin liquids) where the semiclassical approximation breaks down.
• Impact: The work establishes iTWA as a viable, efficient numerical technique for studying thermal and ground states of large interacting spin systems, particularly where traditional QMC fails due to frustration or where exact diagonalization is computationally prohibitive.