Clifford Circuits Augmented Matrix Product States for Fermion Systems

Basic Information

• Paper ID: 2501.00413
• Title: Clifford Circuits Augmented Matrix Product States for Fermion Systems
• Authors: Jiale Huang, Xiangjian Qian, Mingpu Qin
• Classification: cond-mat.str-el quant-ph
• Publication Date: January 3, 2025
• Paper Link: https://arxiv.org/abs/2501.00413

Abstract

Clifford Circuits Augmented Matrix Product States (CAMPS) was recently proposed to leverage the advantages of both Clifford circuits and Matrix Product States (MPS). Clifford circuits can support large entanglement and can be efficiently simulated classically according to the Gottesman-Knill theorem. In CAMPS, MPS needs only to handle the so-called Non-stabilizerness Entanglement Entropy, which significantly improves the simulation accuracy for a given bond dimension. In this work, we generalize CAMPS to study fermion systems by taking advantage of the Jordan-Wigner transformation, which maps the studied fermion system to a spin system. We benchmark the method on both the spinless $t-V$ model and the spinful Hubbard model. Our test results show significant improvement of CAMPS accuracy over MPS, especially when interactions are strong. Fermionic CAMPS provides a useful tool for the accurate study of many-body fermion systems in the future and has the potential to help resolve long-standing issues.

Research Background and Motivation

Research Problem

The core problem addressed in this paper is how to efficiently and accurately simulate strongly correlated many-body fermion systems. Traditional numerical methods face significant computational challenges when dealing with such systems, particularly in the strong interaction regime.

Problem Significance

The study of strongly correlated many-body fermion systems is a cornerstone of modern condensed matter physics and is crucial for understanding phenomena such as superconductivity, magnetism, and quantum phase transitions. Accurate simulation of these systems is important for:

1. Understanding the physical mechanisms of high-temperature superconductors
2. Exploring novel quantum materials
3. Resolving long-standing theoretical problems

Limitations of Existing Methods

1. Quantum Monte Carlo Methods: Suffer from the sign problem, making them difficult to apply in certain parameter regimes
2. Density Matrix Renormalization Group (DMRG)/Matrix Product States (MPS): Limited by entanglement structure in high-dimensional systems, with prohibitively high computational costs
3. Two-Dimensional Tensor Network Methods: Methods such as PEPS, MERA, etc., can encode two-dimensional entanglement structures but typically have high computational costs and difficulty achieving large bond dimensions

Research Motivation

The CAMPS method has demonstrated significant advantages in spin systems, effectively reducing entanglement entropy and improving simulation accuracy. The motivation of this work is to extend this successful method to fermion systems by utilizing the Jordan-Wigner transformation to establish the mapping relationship between fermion and spin systems.

Core Contributions

1. Method Generalization: First generalization of the CAMPS method from spin systems to fermion systems through the Jordan-Wigner transformation, implementing the mapping from fermionic to spin operators
2. Algorithm Implementation: Proposes a CAMPS algorithmic framework applicable to fermion systems while maintaining the efficient classical simulation properties of Clifford circuits
3. Benchmark Testing: Systematic benchmark testing on both spinless $t-V$ model and spinful Hubbard model, validating the method's effectiveness
4. Accuracy Enhancement: Experimental results demonstrate significant accuracy improvement of CAMPS over traditional MPS, especially in the strong interaction regime
5. Theoretical Insight: Provides practical application of the Non-stabilizerness Entanglement Entropy concept for fermion systems, demonstrating its importance in classical simulation of quantum many-body systems

Methodology Details

Task Definition

The task in this paper is to develop an efficient numerical method to compute ground state properties of strongly correlated fermion systems, including ground state energy and entanglement structure. The input is the fermionic Hamiltonian parameters, and the output is a high-precision approximation of the ground state wavefunction and related physical quantities.

Model Architecture

Jordan-Wigner Transformation

Fermionic creation operators $c_i^†$ and annihilation operators $c_i$ satisfy anticommutation relations:

{c_i, c_j} = {c_i^†, c_j^†} = 0, {c_i, c_j^†} = δ_{ij}

The Jordan-Wigner transformation maps these operators to string operators:

c_i^† = σ_1^z σ_2^z ⋯ σ_{i-1}^z σ_i^+
c_i = σ_1^z σ_2^z ⋯ σ_{i-1}^z σ_i^-

where $σ_i^± = (σ_i^x ± iσ_i^y)/2$, and the number operator $n_i = c_i^†c_i$ maps to $(1 + σ_i^z)/2$.

CAMPS Architecture

The CAMPS wavefunction is defined as:

|CAMPS⟩ = C|MPS⟩

where C represents a Clifford circuit, and the MPS ansatz is:

|MPS⟩ = ∑_{σ_i} Tr(A_1^{σ_1} A_2^{σ_2} ⋯ A_N^{σ_N})|σ_1σ_2⋯σ_N⟩

Observable Calculation

The expectation value of any observable $O = ∑_i P_i$ is calculated as:

⟨O⟩ = ⟨CAMPS|O|CAMPS⟩ = ⟨MPS|O'|MPS⟩

where $O' = C^†OC$. Since Clifford circuits map Pauli strings to other Pauli strings, $O'$ maintains the same computational complexity.

Technical Innovations

1. Fermion-Spin Mapping: Cleverly utilizes the Jordan-Wigner transformation, enabling the CAMPS method, originally applicable only to spin systems, to handle fermion systems
2. Entanglement Separation: Clifford circuits handle the stabilizable entanglement portion, while MPS focuses on handling Non-stabilizerness Entanglement Entropy (NsEE), significantly improving simulation accuracy for a given bond dimension
3. Computational Efficiency: According to the Gottesman-Knill theorem, Clifford circuits can be efficiently simulated on classical computers, avoiding exponential computational complexity
4. Two-Dimensional Extension: Converts two-dimensional lattices to one-dimensional chains through snake-like mapping, enabling the method to handle high-dimensional fermion systems

Experimental Setup

Test Models

2D t-V Model

The Hamiltonian is:

H = -t ∑_{⟨i,j⟩} c_i^† c_j + h.c. + V ∑_{⟨i,j⟩} (n_i - 1/2)(n_j - 1/2)

Test parameters:

• System sizes: 6×6, 8×8
• Interaction strengths: V = 0.5, 1, 3
• Filling: Half-filling
• Boundary conditions: Open boundaries

Hubbard Model

The Hamiltonian is:

H = -t ∑_σ ∑_{⟨i,j⟩} c_{i,σ}^† c_{j,σ} + h.c. + U ∑_i n_{i,↑}n_{i,↓} - μ ∑_i n_i

Test parameters:

• System sizes: 1×32, 2×16, 4×8
• Interaction strength: U = 8
• Doping: 1/8 hole doping
• Boundary conditions: Open boundaries

Evaluation Metrics

1. Relative Error: Relative error in ground state energy $(E - E_{ref})/|E_{ref}|$
2. Entanglement Entropy: Entanglement entropy of the MPS portion at the central bond
3. Improvement Ratio: Defined as $(E_{MPS} - E_{ref})/(E_{CAMPS} - E_{ref})$

Reference Benchmark

High-precision ground state energy computed using large bond dimension DMRG is used as the reference value, with extrapolation ensuring reference reliability.

Experimental Results

Main Results

t-V Model Results

For the 6×6 system:

• At V = 0.5, CAMPS achieves an improvement ratio of approximately 4× at D = 2000
• At V = 1, the improvement ratio is approximately 3.5×
• At V = 3, the improvement ratio reaches approximately 4×, increasing with bond dimension

For the 8×8 system:

• At V = 0.5, the improvement ratio is approximately 2.5×
• At V = 1, the improvement ratio is approximately 3×
• At V = 3, the improvement ratio reaches 4×, showing more pronounced CAMPS advantages at stronger interactions

Hubbard Model Results

• 1×32 system: Improvement ratio reaches approximately 7.5×
• 2×16 system: Improvement ratio is approximately 5×
• 4×8 system: Improvement ratio is approximately 2.75×

Key Findings

1. Strong Interaction Advantage: Across all test cases, stronger interactions lead to more pronounced CAMPS advantages over MPS
2. Reduced Entanglement Entropy: CAMPS shows significantly lower entanglement entropy in the MPS portion compared to pure MPS
3. Scalability: The improvement ratio increases with bond dimension D, demonstrating good scalability
4. System Size Dependence: Consistent improvement effects are observed across different system sizes

Quantified Accuracy Enhancement

Taking the 8×8 t-V model with V=3 as an example:

• At the same bond dimension D=1000, CAMPS has approximately one order of magnitude lower relative error than MPS
• Entanglement entropy reduces from approximately 0.58 in MPS to approximately 0.48 in CAMPS

Related Work

Development of Tensor Network Methods

1. MPS/DMRG: Excellent performance in one-dimensional and quasi-one-dimensional systems, but limited by the area law
2. Two-Dimensional Tensor Networks: PEPS, MERA, PESS and other methods handle two-dimensional systems but with high computational costs
3. Enhanced MPS: Methods such as FAMPS enhance MPS entanglement encoding capability through disentanglers

Evolution of CAMPS Method

1. Original CAMPS: Initially developed for ground state calculations in spin systems
2. Time Evolution Extension: Combined with TDVP for time evolution simulation
3. Finite Temperature Generalization: Extended to finite temperature calculations
4. Critical System Application: Applied to disentangled critical quantum chains

Fermion-Spin Mapping

1. Jordan-Wigner Transformation: Classical fermion-spin mapping method
2. Bravyi-Kitaev Transformation: More efficient alternative in certain cases
3. Other Mapping Techniques: Various improved mapping methods developed recently

Conclusions and Discussion

Main Conclusions

1. Successful Generalization: Successfully extends the CAMPS method from spin systems to fermion systems, validating the method's universality
2. Significant Improvement: Achieves 2-7× accuracy improvement over traditional MPS in the strong interaction regime
3. Entanglement Optimization: Effectively handles stabilizable entanglement through Clifford circuits, allowing MPS to focus on NsEE
4. Practical Value: Provides a new efficient numerical tool for studying strongly correlated fermion systems

Limitations

1. Mapping Dependence: Method effectiveness depends on the Jordan-Wigner transformation, producing long-range string operators in two-dimensional systems
2. Symmetry Constraints: Current implementation does not enforce U(1) symmetry, requiring chemical potential adjustment to control filling factor
3. Boundary Conditions: Current testing limited to open boundary conditions; periodic boundary condition effects remain to be verified
4. System Size: Limited by computational resources, tested system sizes are relatively small

Future Directions

1. Symmetry Integration: Incorporate U(1) symmetry into Clifford circuits to improve computational efficiency
2. Mapping Scheme Comparison: Test effectiveness of alternative fermion-spin mapping schemes such as Bravyi-Kitaev
3. Method Extension: Generalize to real-time evolution and finite temperature calculations
4. Application Expansion: Apply to resolving long-standing physics problems such as high-temperature superconductivity

In-Depth Evaluation

Strengths

1. Method Innovation: First successful extension of CAMPS to fermion systems, opening new research directions
2. Solid Theoretical Foundation: Theoretical basis grounded in the Gottesman-Knill theorem ensures method feasibility
3. Comprehensive Experiments: Systematic benchmark testing across multiple models and parameter settings
4. Convincing Results: Numerical results clearly demonstrate method advantages, particularly in strong interaction regimes
5. Clear Presentation: Well-structured paper with accurate technical descriptions

Weaknesses

1. System Size Limitations: Tested system sizes are relatively small; large system performance remains to be verified
2. Single Boundary Condition: Only open boundary conditions tested; periodic boundary condition results are lacking
3. Symmetry Treatment: Insufficient exploitation of system symmetries, potentially affecting computational efficiency
4. Insufficient Theoretical Analysis: Lacks in-depth theoretical analysis of why the method performs better in strong interaction regimes

Impact

1. Academic Contribution: Provides new tools for numerical research on strongly correlated fermion systems
2. Practical Value: Promising applications to important physics problems such as high-temperature superconductivity
3. Methodological Significance: Demonstrates the application potential of quantum information methods in condensed matter physics
4. Reproducibility: Clear method description facilitates reproduction and extension by other researchers

Applicable Scenarios

1. Strongly Correlated Fermion Systems: Particularly suitable for calculations in strong interaction regimes
2. Two-Dimensional Quantum Materials: Can be used to study quantum phase transitions in two-dimensional electron systems
3. High-Temperature Superconductivity Research: Provides new computational tools for understanding cuprate superconductors
4. Quantum Simulation: Provides classical algorithm benchmarks for quantum simulators

References

This paper cites 49 important references covering key works in strongly correlated electron systems, tensor network methods, quantum information theory, and other relevant fields, providing solid theoretical foundation and methodological support for the research.

Overall Assessment: This is a high-quality methodological paper that successfully extends the CAMPS method to fermion systems, providing important new tools for numerical research on strongly correlated many-body systems. While there is room for improvement in system size and symmetry treatment, its innovation and practical value make it a significant contribution to the field.