⚛️ quantum physics

Toward Quantum Utility in Correlated Topological Matter: Variational Preparation of Fractional Quantum Hall Manifolds

This paper demonstrates that variational quantum algorithms, specifically VQE and VQD, can successfully prepare and characterize the topological ground-state manifold of the $\nu=1/3$ fractional quantum Hall state on both Haldane sphere and torus geometries, thereby establishing a pathway toward simulating strongly correlated topological phases on near-term quantum hardware.

Original authors: Sergio F. Expósito, Unai Aseginolaza, Raúl Guerrero-Avilés, Joaquim Jornet-Somoza, Francisco Guinea, Juan Borge

Published 2026-07-14

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Sergio F. Expósito, Unai Aseginolaza, Raúl Guerrero-Avilés, Joaquim Jornet-Somoza, Francisco Guinea, Juan Borge

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are trying to build a tiny, perfect crystal out of invisible, dancing electrons. But these aren't ordinary electrons; they are part of a special club called the "Fractional Quantum Hall" team, where they move in a synchronized, topological dance that creates a state of matter so complex that even the most powerful supercomputers struggle to figure out their steps.

Scientists have long wanted to use quantum computers to simulate this dance, but the machines we have today are still a bit "noisy"—like trying to hear a whisper in a rock concert. This paper is a report on a team of researchers who tried to teach a real, noisy quantum computer to learn the steps of this electron dance, specifically for a famous pattern called the $\nu = 1/3$ Laughlin state.

The Two Dance Floors: A Sphere and a Donut

To test if their quantum computer could really learn the dance, the researchers set up two different "dance floors" (geometries) to see how the electrons behaved.

The Sphere (The Solo Act): Imagine the electrons dancing on a perfect ball. Here, the rules are strict: there is only one special, zero-energy way for them to arrange themselves to be perfectly happy. It's like a solo performance where there is only one correct pose. The researchers used a method called VQE (Variational Quantum Eigensolver) to try to find this single perfect pose.
The Torus (The Group Hug): Now, imagine the dance floor is a donut (a torus). This is trickier. Because of the donut's shape, the electrons can arrange themselves in three different, equally perfect ways at the same energy level (zero energy). It's like a group hug where three different people can all be the "leader" at the same time. This is called a "threefold topological ground-state degeneracy." To find all three, the researchers had to use a more advanced tool called VQD (Variational Quantum Deflation), which helps the computer find the first solution, then "deflate" it to find the next one, and so on.

The Challenge: Noise vs. The Perfect Dance

The researchers ran their experiments on a real quantum processor (an IBM machine). Because the machine is noisy, the energy readings it gave were often too high and shaky—like a shaky camera trying to film a fast dancer.

However, the team didn't just look at the shaky camera footage. They took the "moves" (parameters) the noisy computer learned and replayed them on a perfect, noiseless simulator.

The Result: Even though the real machine was noisy, the moves it learned were actually very good! When replayed perfectly, the computer successfully recreated the three distinct states on the donut and the single state on the sphere.
The Fix: To make the noisy readings look more like the perfect ones, they used "error mitigation" tricks (called ZNE and TREX). Think of this like using a photo filter to smooth out the graininess of a shaky video. These tricks didn't make the video perfect, but they brought the energy numbers much closer to the true values. For example, on the donut, the first excited state (the next level of the dance) was measured at 0.710 after mitigation, which is very close to the exact theoretical value of 0.703.

How Do We Know They Got It Right?

Just getting the energy numbers close isn't enough. The researchers had to prove the computer actually learned the right dance, not just a lucky guess.

On the Sphere: They checked the "total spin" (angular momentum). The perfect dance should have a value of 0. The noisy computer's raw data was way off (5.51), but after cleaning up the noise, it got much closer (2.22). The perfect simulation of the learned moves showed it was very close to the target (1.306).
On the Donut: They checked the "magnetic momentum" (a specific direction the electrons are facing). The three perfect states should be in specific sectors labeled 1, 3, and 5. The computer's learned moves put about 88% to 89% of the electrons in the correct sectors. This proves the computer didn't just find a low-energy state; it found the right topological states.

What This Means (and What It Doesn't)

The paper shows that hybrid quantum algorithms (mixing a quantum computer with a classical one) can approximately reconstruct the low-energy structure of these tiny, complex systems. They successfully found the "topological ground-state manifold" (the group of three perfect states) on the donut and the unique state on the sphere.

However, the authors are careful not to claim they have "solved" the problem for all materials. They explicitly state that their results are for small systems (like 2 electrons on the donut and 3 electrons on the sphere) where they could check the answers against exact math. They also note that the energy differences they are measuring are tiny, and the "noise" in the machine is actually much larger than the energy gaps they are trying to see. The fact that they could get close using simple error-mitigation tricks suggests that with better tools, we might be able to tackle even bigger, more realistic materials in the future.

In short, this isn't a finished masterpiece yet; it's a very promising sketch. The researchers have shown that noisy quantum computers can learn the steps of the most complex electron dances, provided we give them the right geometry (like a donut) and the right tools to clean up the noise. This opens a door to simulating "Fractional Chern insulators" and other strange topological materials that we can't currently calculate with classical computers.

Technical Summary: Toward Quantum Utility in Correlated Topological Matter

Problem Statement
The paper addresses the challenge of simulating strongly correlated topological phases, specifically the fractional quantum Hall effect (FQHE), on noisy intermediate-scale quantum (NISQ) processors. While quantum computers offer a natural framework for many-body problems due to the exponential scaling of the Hilbert space, preparing and characterizing topological states remains difficult. Existing approaches often rely on quasi-one-dimensional geometries (cylinders or thin-torus limits) which simplify the problem but obscure the genuinely two-dimensional nature of the quantum Hall liquid. The authors aim to benchmark hybrid variational quantum algorithms on a fully two-dimensional periodic geometry (the torus) to reconstruct not only ground states but also degenerate topological manifolds and excited states, while comparing these results against a spherical geometry where the ground state is unique.

Methodology
The authors formulate the problem using the fermionic $\nu = 1/3$ Laughlin phase described by the $V_1$ Haldane pseudopotential, projected onto the lowest Landau level (LLL).

Hamiltonian Formulation: The problem is mapped to second quantization in two distinct geometries:
1. Haldane Sphere: A rotationally invariant system with a unique zero-energy Laughlin ground state ( $L=0$ ) and a flux-particle relation $N_\phi = 3(N_e - 1)$ .
2. Torus: A periodic system with no curvature, exhibiting a threefold topological ground-state degeneracy associated with magnetic-translation momentum sectors ( $K$ ), with a flux-particle relation $N_\phi = 3N_e$ .
Quantum Algorithms: The study employs a hybrid workflow combining:
- Particle-Number-Preserving Ansatz: A parametrized quantum circuit designed to maintain the fixed electron number sector ( $N_e$ ).
- Variational Quantum Eigensolver (VQE): Used to find the lowest-energy state.
- Variational Quantum Deflation (VQD): Used sequentially to reconstruct excited states and the degenerate ground-state manifold by penalizing overlap with previously optimized states.
Hardware and Diagnostics: Calculations were performed on the IBM Heron-R2 processor (IBM_BasqueCountry System Two). The workflow involves optimizing parameters on hardware and subsequently evaluating the resulting parameter vectors using noiseless statevector simulations.
- Error Mitigation: The study utilizes Zero-Noise Extrapolation (ZNE) and Twirled Readout Error eXtinction (TREX) to mitigate gate and readout errors.
- Metrics: Performance is benchmarked against exact diagonalization (ED) using energy estimates, subspace containment (projection of the variational state onto exact eigenspaces), and symmetry diagnostics (total angular momentum $\langle \hat{L}^2 \rangle$ for the sphere and magnetic momentum $K$ for the torus).

Key Results
The study successfully reconstructed the low-energy structure of small FQHE systems ( $N_e=2$ on the torus, $N_e=3$ on the sphere) using real quantum hardware.

Torus Geometry:
- The VQD algorithm successfully identified the threefold degenerate zero-energy ground-state manifold and the first excited state.
- While raw hardware energies were systematically shifted upward due to noise, error-mitigated estimates (ZNE + TREX) clustered close to the exact zero-energy manifold (best mitigated values: 0.056, 0.197, 0.129) and accurately reproduced the first excited state energy (0.710 vs. exact 0.703).
- Subspace Containment: The variational states showed high projection weights onto the exact ground-state manifold (0.970, 0.883, 0.901), confirming that the algorithm reconstructs the correct wave-function subspace, not just low energies.
- Symmetry: The prepared states concentrated probability in the correct magnetic-momentum sectors ( $K \in \{1, 3, 5\}$ for the ground state), with target weights exceeding 84%.
Sphere Geometry:
- VQE and VQD recovered the unique $L=0$ ground state and the first excited state.
- Mitigated energy estimates reproduced the qualitative spectral gap.
- Symmetry: The variational states approached the correct angular momentum sectors. For the ground state, the statevector $\langle \hat{L}^2 \rangle$ was 1.306 (close to 0), and for the first excited state, it was 12.30 (close to exact 12). Error mitigation significantly improved the $\langle \hat{L}^2 \rangle$ estimates obtained from the QPU.
Error Analysis: The authors note that the Hamiltonian expectation value is reconstructed from Pauli strings, leading to an error-amplification scale larger than the physical energy differences. Despite this, the lightweight mitigation strategy recovered the main spectral features.

Significance and Claims
The paper claims that this work establishes a controlled quantum-computing workflow for strongly correlated topological matter that goes beyond simple spin models or molecular benchmarks.

Geometry Resolution: By working directly on the torus, the authors demonstrate the ability to handle the genuine two-dimensional character of the quantum Hall liquid, including topological degeneracy and magnetic-translation symmetry, which are often lost in quasi-1D approximations.
Wave-Function Reconstruction: The results show that hybrid algorithms can approximately reconstruct the low-energy structure, including the topological ground-state manifold, on real hardware. The subspace-containment analysis proves that the hardware-optimized parameters define states with the correct wave-function character, not merely reasonable energy estimates.
Path to Utility: The authors position the fractional quantum Hall system as a canonical benchmark and a controlled starting point for developing workflows aimed at more general, realistic topological phases, such as fractional Chern insulators (FCI) and interacting topological materials in moiré systems. They argue that while the current work uses parent Hamiltonians with known solutions, the demonstrated methodology provides a route toward quantum simulations where exact solutions are unavailable and classical methods are intractable.

The Two Dance Floors: A Sphere and a Donut

The Challenge: Noise vs. The Perfect Dance

How Do We Know They Got It Right?

What This Means (and What It Doesn't)

More like this