1-Form Symmetric Projected Entangled-Pair States

arXiv:2407.16531 · cond-mat.str-el, quant-ph · Submitted 2024-07-23 · Read on arXiv

Listen

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: We're looking at the paper: "1-Form Symmetric Projected Entangled-Pair States".

Mira: Here's the gist of the paper. The gist This work presents a novel framework for understanding 1-form symmetries within Projected Entangled-Pair States (PEPS) by deriving algebraic relations for symmetry matrices on virtual legs along loops, revealing that these symmetries impose stringent constraints leading to distinct anomalous braiding phases and enhancing ground state optimization efficiency.

Kai: Quantum many-body systems are challenging to study phases of matter due to complex interactions and entanglements Projected Entangled-Pair States (PEPS) offer a powerful two-dimensional tensornetwork framework for efficiently representing these states.

Mira: By encoding symmetries, tensor-network representations classify phases of matter, particularly symmetry-protected topological phases.

Kai: Higher-form symmetries, especially 1-form symmetries, have gained prominence in the study of quantum phases.

Title and authors: Kai: So we’re looking at a paper called "one-Form Symmetric Projected Entangled-Pair States <ref:2407.16531#pg1,1-Form Symmetric Projected Entangled-Pair States>." Mira, what’s the main thing you see in that title?

Mira: It sounds really dense, doesn't it? It connects this new idea of one-form symmetries with PEPS. Essentially, they are exploring how these higher-dimensional symmetries constrain the tensor network representations themselves.

Kai: Right. It suggests that just having a symmetry isn't enough; it forces a specific structure on the way we build the states in PEPS, especially when you look at those loops through string operators.

Lev: From an error correction standpoint, I wonder if these constraints translate into something practical for building fault-tolerant hardware. If these symmetries dictate the allowed representations, does that simplify or complicate finding a stable physical state?

Kai: Exactly. The paper shows that this framework lets them derive algebraic relations for those symmetry matrices on the virtual legs along loops. That’s the core mechanism they are proposing to understand how these symmetries actually work in practice within PEPS.

Mira: I think the authors are setting up a way to classify different phases of matter based on these loop symmetries, and showing that those symmetries impose some really strict rules we haven't fully seen before.

Lev: So, instead of just checking if a state has a symmetry, they’re deriving the actual mathematical constraints it must satisfy on the tensor network structure itself. That’s a different level of analysis than just looking at observables.

The paper's summary: Kai: So what’s the actual takeaway from this paper? It seems they formalize how one-form symmetries manifest as loop-like symmetries in PEPS and derive these algebraic relations for the symmetry matrices on those virtual legs <ref:2407.16531#pg1>.

Mira: They show that these symmetries impose stringent constraints on the tensor network representations, which then leads to some distinct anomalous braiding phases carried by those symmetry matrices. That’s a really important finding for classifying topological states.

Lev: Braiding phases sound like they are related to anyons, right? If the braiding phase is different—say, it's not just a simple two pi shift—that means you have something fundamentally non-trivial happening in the system's topology <ref:2407.16531#pg1>.

Kai: That’s what they’re pointing to. They introduce a "charge" or flux operator, ggamma, which has a commutation relation with the loop operator Wgamma that sets up this braiding phase omegaU.

Mira: And they show that this braiding phase omegaU can be expressed as omegaU = -u, where u is an integer. That integer u is what tells you whether you have a bosonic anyon related to the one-form symmetric operator or if you’re dealing with something else entirely <ref:2407.16531#pg1>.

Lev: If we’re talking about real hardware, knowing that the braiding phase is tied to an integer u, that gives us a concrete way to check for these specific topological properties in simulations or measurements.

Kai: So the paper provides this tool to classify states precisely using these braiding phases, which is a big step forward for understanding topological order in these models.

The paper's improvements: Mira: Beyond just showing the constraints, the authors highlight how their framework actually improves optimization efficiency when you’re trying to find a state. They look at how to use this one-form symmetric tangent space for variations and excitations around a given PEPS configuration <ref:2407.16531#pg1>.

Kai: They suggest that by symmetrizing that tangent vector using specific eigen-equations, W hat p Bi⟩ = wpNB Bi⟩ for the four plaquette operators, you can utilize this one-form symmetric tangent gradient <ref:2407.16531#pg1>.

Lev: That sounds like it could speed things up significantly for iterative methods. If you can converge to the eigenstate of the one-form symmetry operator faster, that's a major win when dealing with large PEPS states on real hardware where computation time is limited <ref:2407.16531#pg1>.

Mira: And they claim this method successfully converged to the eigenstate of the one-form symmetry operator with eigenvalues like omega = one e±i2pi/three for all those local excited states <ref:2407.16531#pg1>.

Kai: So what this means for us is that these one-form symmetries impose very stringent constraints on how we should represent these tensor networks, and they can actually help us optimize the ground state search much more effectively than standard gradient methods <ref:2407.16531#pg1>.

Lev: It sounds like a way to bypass some of the convergence problems we see when using non-symmetric gradient optimization for complex quantum many-body systems. That’s exactly what we need in real hardware setups where you don't have infinite time to converge.

Conclusion: Mira: So, to wrap up on "one-Form Symmetric Projected Entangled-Pair States," the main point is that these one-form symmetries impose very strict rules on tensor network representations by constraining the symmetry matrices and leading to distinct anomalous braiding phases <ref:2407.16531#pg1,1-Form Symmetric Projected Entangled-Pair States>.

Kai: And these constraints are not just theoretical; they provide a way to improve optimization efficiency, allowing AI systems to reliably converge to the eigenstate of that one-form symmetry operator using those derived matrices <ref:2407.16531#pg1>.

Lev: For me, what this means is we have a new way to characterize topological order by analyzing the braiding phase omegaU = -u, where u is an integer. That gives us a concrete metric for distinguishing between different types of topological phases in the Kitaev models or other systems we look at.

Kai: Right. And they’ve shown this framework can be applied to other models, like the one-form symmetry in the Kitaev star lattice model, opening up new avenues for studying these strongly correlated systems <ref:2407.16531#pg1>.

Mira: It really opens up a door to characterizing the one-form symmetry structure not just in ground states but also in excited states and understanding how they influence them <ref:2407.16531#pg1>.

Lev: Yeah, it gives us better tools for analyzing those excited state properties, which is valuable because we need to understand the full spectrum of possibilities when designing and running quantum algorithms or hardware.

Kai: So that’s the picture for this work on "one-Form Symmetric Projected Entangled-Pair States <ref:2407.16531#pg1,1-Form Symmetric Projected Entangled-Pair States>." It’s a solid paper that connects abstract symmetry ideas to concrete computational benefits.

Yi Tan, Ji-Yao Chen, *Didier Poilblanc, Fei Ye, Jia-Wei Mei

Department of Physics, Southern University of Science and Technology, Shenzhen · Center for Neutron Science and Technology, Guangdong Provincial Key Laboratory of Magnetoelectric Physics and Devices, School of Physics, Sun Yat-sen University · Laboratoire de Physique Th´eorique, C.N.R.S. and Universit´e de Toulouse

cond-mat.str-el, quant-ph

Submitted: 2024-07-23

Updated: 2024-08-01

Comments: 6 pages, 1 figure. In memory of T. M. Rice for his invaluable teaching on the physics of strongly correlated systems

Journal ref: Commun Phys (2026)

DOI: 10.1038/s42005-026-02729-x

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 80/100

The gist: The gist This work presents a novel framework for understanding 1-form symmetries within Projected Entangled-Pair States (PEPS) by deriving algebraic relations for symmetry matrices on virtual legs

Key concepts

1-Form Symmetry: This is a higher-dimensional symmetry in quantum systems that manifests as loop-like symmetries rather than simple global ones. It extends standard symmetries to structures around loops, which are crucial for understanding topological phases in many-body systems.
Symmetry Matrices on Virtual Legs: These matrices are derived by applying the 1-form symmetry constraints to the virtual legs of a PEPS along specific loops. The algebraic relations governing these matrices dictate how the symmetry behaves across different parts of the tensor network structure.
Braiding Phase ($\omega_U$): The braiding phase is a critical indicator of whether a 1-form symmetry anomaly exists in the system. It quantifies how two loop operators interact, and its value (e.g., $\omega_U = -u$, where $u$ is an integer) helps classify the topological order of the state.
Ground State Optimization Efficiency: The framework provides a method to use 1-form symmetric tangent vectors for gradient-based optimization. This approach allows AI systems to converge reliably to the actual eigenstate of the symmetry operator, overcoming failures in standard optimization methods.

Terminology

Projected Entangled-Pair States — 2× in this paper

Summary

The gist

This work presents a novel framework for understanding 1-form symmetries within Projected Entangled-Pair States (PEPS) by deriving algebraic relations for symmetry matrices on virtual legs along loops, revealing that these symmetries impose stringent constraints leading to distinct anomalous braiding phases and enhancing ground state optimization efficiency.

Introduction and Motivation

Quantum many-body systems are challenging to study phases of matter due to complex interactions and entanglements Projected Entangled-Pair States (PEPS) offer a powerful two-dimensional tensornetwork framework for efficiently representing these states. By encoding symmetries, tensor-network representations classify phases of matter, particularly symmetry-protected topological phases. Higher-form symmetries, especially 1-form symmetries, have gained prominence in the study of quantum phases. These symmetries manifest as loop-like symmetries through string operators, extending global (0-form) symmetries to higher-dimensional objects.

Framework for 1-Form Symmetry

The paper formalizes the framework of 1-form symmetric PEPS by deriving algebraic relations for symmetry matrices acting on the virtual legs of PEPS along loops. The loop operator is defined as Wγ = U(1)γ U(2)γ U(3)γ U(4)γ U(5)γ, where each Ui γ represents the physical operator at the i-th site along the loop γ. The loop symmetry operators can be projectively represented with the projective phase θU(i)γ,ss'A(i)s'abcd = e iθA(i)s ab'cd'V(il)bb' V(ir)dd'. The gauge consistency condition required for the virtual symmetry matrices is (3): (V t)(ir)V(i+1 l) = 1.

Flux Operators and Braiding Phases

To construct a faithful representation of Wγ, a “charge” (flux) operator gγ that satisfies the commutation relation gγWγ = wWγgγ with wN = 1 is introduced. The flux operator can be represented by multiplying a matrix gflux on any bond along the γ-loop. The consistency condition for the flux state requires (V t)(ir)g(i)fluxV(i+1 l) = wg(i)flux. The commutation relation [Wγ, Wδ] = 0 leads to the braiding phase ωU V(i)δ = ωU V(i)δ. This braiding phase is a critical indicator of the 1-form symmetry anomaly.

Solutions for Kitaev Models

The paper presents concrete solutions for the necessary algebraic equations for the symmetry matrices of the Z[N] 1-form symmetry on the honeycomb lattice for the ZN Kitaev model. For a translational PEPS, Eq.(9) is required to ensure that loop symmetries are accurately represented. The N-dimensional virtual symmetry matrices are chosen from Sylvester’s generalized P N-dimensional Pauli matrices. For the integer spin-S Kitaev model where N = 2 and u = 2, the corresponding solution is Xy = Yz = Zx = Xz = Yx = Zy = σz. For the Z3 Kitaev model where N = 3 and u = 1, the corresponding solution is Xy and Xz exhibit a non-trivial commutation braiding phase.

Implications for Optimization

The 1-form symmetric tangent space describes variations and excitations around a given PEPS configuration. The tangent vector can be symmetrized using the eigen-equations WˆpBi⟩ = wpNBBi⟩ for the surrounding four plaquette operators. Utilizing this 1-form symmetric tangent gradient significantly improved optimization efficiency and successfully converged to the eigenstate of the 1-form symmetry operator, achieving the anticipated eigenvalues ω = 1, e±i2π/3 for all the local excited states. This demonstrates that 1-form symmetries impose stringent constraints on tensor network representations.

Conclusion

In this study, we have introduced a novel framework for understanding 1-form symmetries within tensor networks, with a specific focus on Projected Entangled-Pair States (PEPS). Through the derivation of algebraic relations for symmetry matrices on the PEPS virtual legs along loops associated with 1-form symmetry, we have demonstrated that 1-form symmetries impose stringent constraints on tensor network representations. This results in distinct anomalous braiding phases carried by the symmetry matrices, enabling 1-form symmetric PEPS to effectively capture the anomaly in the 1-form symmetry. We have underscored the potential for enhancing ground state optimization efficiency and characterizing the 1-form symmetry structure in excited states. This work can be straightforwardly applied to other systems, such as the 1-form symmetry in the Kitaev star lattice model. Previous numerical studies of the Kitaev models have found significant entanglement spectrum degeneracy. The 1-form symmetry opens new avenues for the PEPS study of topological phases in strongly correlated systems.

ACKNOWLEDGMENT

This work is supported by the National Key Research and Development Program of China (Grant No. 2021YFA1400400), Shenzhen Fundamental Research Program (Grant No.

Improvements for AI systems

textbfProcessing of 1-form Symmetries in PEPS State Space: Improved Optimization Convergence Stability (Page 5). This improvement allows AI systems to perform gradient-based optimization on PEPS states using symmetry matrices derived from Eq. (15) and a 2 × 2 identity matrix, which significantly improved optimization efficiency and successfully converged to the eigenstate of the 1-form symmetry operator. This enables the AI to reliably find ground states for complex quantum many-body systems where initial nonsymmetric gradient-based optimization fails to converge to a satisfactory variational state.
textbfCharacterization of Topological Order via Braiding Phases (Page 3, 4). The system can now distinguish between different topological phases by analyzing the braiding phase, as the braiding phase ωU can be expressed as ωU = −u, where u is an integer, which is a critical indicator of the 1-form symmetry anomaly. This allows for precise classification of states based on whether the solution involves bosonic anyon related to the 1-form symmetric operator or a non-trivial (fermionic) braiding phase for the anyons.
textbfPredictive Modeling of Excitation Spectra (Page 1). The framework enables characterizing the 1-form symmetry structure in excited states, providing insights into their physical implications. Specifically, the system can predict how these symmetries influence the ground state and tangent space in PEPS, offering new methods to analyze excited state properties beyond standard entanglement measures.