⚛️ quantum physics

Probing two-spin entanglement at quantum criticality on a quantum processor

This paper proposes and experimentally demonstrates an efficient, scalable method using the Positive Partial Transpose (PPT) criterion and overlapping state tomography to detect and map two-spin entanglement in quantum critical states on noisy quantum processors with up to 20 qubits.

Original authors: Anshumitra Baul, Xiao Xiao, Phillip C. Lotshaw

Published 2026-07-13

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Anshumitra Baul, Xiao Xiao, Phillip C. Lotshaw

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 a massive, invisible dance floor where thousands of tiny dancers (quantum spins) move in perfect, spooky synchronization. Sometimes, they dance in a simple, predictable line; other times, they get into a chaotic, hyper-connected frenzy where every move one dancer makes instantly influences a partner far away. This chaotic state is called a quantum phase transition, and it's where the most interesting "quantum magic" happens.

But here's the problem: our current quantum computers are like dancers with shaky hands. They are noisy and prone to mistakes. When scientists try to measure how connected these dancers are, the noise often hides the magic, making it look like the dancers are just randomly bumping into each other rather than performing a complex, entangled routine.

In this paper, a team of researchers set out to solve this mystery. They didn't just look at the whole dance floor; they zoomed in to check if two specific dancers were holding hands (entangled) or just standing near each other.

The Detective's Tool: The "Spooky Mirror"

To figure out if two dancers are truly connected, the team used a clever trick called the Positive Partial Transpose (PPT) criterion. Think of it like a "spooky mirror."

If you take a snapshot of two dancers and look at their reflection in this special mirror, a normal, unconnected pair will look perfectly normal. But if they are truly entangled, the reflection will look "broken" or impossible—like a mirror image that defies the laws of physics. If the mirror shows a "negative" value (a mathematical glitch that shouldn't exist in a normal world), the scientists know for sure: these two are entangled.

This method is special because it works even when the dance floor is messy and noisy (mixed states), which is exactly what happens on real quantum computers today.

The Experiment: Dancing on a Shaky Stage

The researchers tested this idea on two famous "dance routines" (quantum models):

The Transverse Field Ising Model (TFIM): A simple line of spins that can flip between being aligned and being chaotic.
The XXZ Model: A slightly more complex routine where spins interact in different directions.

They programmed a real quantum computer (the 156-qubit IBM Quantum ibm_boston device) to perform these dances. They used a special "brick-wall" circuit to prepare the states, trying to get the dancers into the most entangled positions possible.

The Results:

The Noise Problem: When they first looked at the raw data from the computer, the "mirror" was foggy. The noise made the results look like the dancers were connected even when they weren't, or it hid the connections that were there. The computer was "overestimating" the connections, making the dance floor look more chaotic than it really was.
The Fix: The team used a technique called Zero-Noise Extrapolation (ZNE). Imagine taking a blurry photo and then taking several more photos where you intentionally shake the camera even more. By comparing the blurry photo with the super-shakey ones, a computer algorithm can mathematically "un-shake" the image to reveal the clear picture underneath. They also used a "matrix-free" method to clean up the reading errors.
The Success: After cleaning up the noise, the results matched the perfect, theoretical simulations almost perfectly.
- For the TFIM model near the critical point (where the phase transition happens), the computer calculated an energy of -25.37, which is 99.53% accurate compared to the perfect theoretical value of -25.49.
- For the XXZ model, they got an energy of -12.58, matching the theoretical -12.70 with 99.05% accuracy.

What They Found (and What They Didn't)

The big discovery was that the "spooky mirror" (PPT) successfully spotted the quantum phase transitions.

The Peak of Connection: As the system approached the critical point (the moment the dance style changed), the "broken mirror" signal got strongest. This confirmed that the quantum computer could generate and sustain the highly entangled states needed for these transitions.
The Limits of Connection: Here is a crucial detail the paper explicitly rules out: Entanglement doesn't stretch forever.
- In the TFIM model, the mirror only showed entanglement between nearest neighbors (dancers standing right next to each other). Even though the system had long-range correlations, the two-spin entanglement didn't reach further than the immediate neighbor.
- In the XXZ model, the mirror showed entanglement between nearest neighbors and next-nearest neighbors (the dancers skipping one spot), but it stopped there.
- The Takeaway: The paper suggests that the long-range connections seen in these systems are likely "classical" correlations (like a crowd moving together) rather than deep, two-spin quantum entanglement. The true, deep quantum magic seems to be hidden in groups of three or more spins, which this specific two-spin test couldn't see.

Why This Matters

This work proves that we can use noisy, imperfect quantum computers to study complex materials if we use the right tools. By combining the PPT criterion (the spooky mirror) with error correction (the un-shaking), the team showed that these machines can act as reliable "entanglement witnesses."

They didn't just simulate a theory; they measured it on real hardware. They showed that while the noise is real, it can be tamed. The paper suggests that this approach is a scalable way to benchmark future quantum computers, ensuring they are ready to tackle even harder problems, like simulating high-temperature superconductors or exotic quantum spin liquids, by checking if they can truly hold hands across the quantum dance floor.

Technical Summary: Probing Two-Spin Entanglement at Quantum Criticality on a Quantum Processor

Problem Statement
Characterizing quantum correlations in many-body systems is essential for understanding quantum materials and benchmarking quantum hardware. However, traditional measures like entanglement entropy (EE) are difficult to interpret for the mixed states produced by noisy intermediate-scale quantum (NISQ) devices and require complex circuits (e.g., controlled-SWAP operations) that are highly sensitive to noise. Furthermore, classical simulation methods like Density Matrix Renormalization Group (DMRG) face exponential scaling bottlenecks when dealing with the "volume law" entanglement found near quantum critical points. There is an urgent need for scalable, model-independent witnesses that can effectively detect pairwise entanglement in noisy, mixed states to validate quantum processor performance.

Methodology
The authors propose a scalable framework utilizing the Positive Partial Transpose (PPT) criterion coupled with overlapping state tomography (QOT) to witness two-spin entanglement. The approach involves:

Quantum Critical State Preparation: The team prepares ground states of one-dimensional spin models—the Transverse-Field Ising Model (TFIM) and the XXZ model—using variational quantum circuits. These circuits employ a "brick-wall" ansatz structure inspired by Matrix Product States (MPS) to encode the ground states of up to 20 qubits.
Overlapping State Tomography (QOT): Instead of full state tomography, the authors use QOT to efficiently reconstruct the reduced density matrices ( $\rho_{AB}$ ) for all spin pairs. This method requires $O(\log N)$ measurement settings to determine two-point correlations for all pairs.
PPT Entanglement Witness: For each reconstructed $\rho_{AB}$ , the partial transpose with respect to one subsystem is computed. According to the Peres-Horodecki criterion, if the minimum eigenvalue ( $\lambda_{min}$ ) of the partially transposed matrix is negative, the state is entangled. This metric is applicable to both pure and mixed states.
Error Mitigation: To address hardware noise on the 156-qubit IBM Quantum ibm_boston processor, the authors implement:
- Matrix-Free Readout Mitigation (M3): To correct measurement assignment errors without the exponential overhead of full calibration.
- Zero-Noise Extrapolation (ZNE): Specifically using "partial folding" to generate noise-scaled circuits with fractional noise factors (e.g., 1.5x, 2.5x), allowing for more accurate extrapolation to the zero-noise limit compared to standard integer folding.
- Symmetry Constraints: Utilizing the global parity (TFIM) and U(1) rotational symmetry (XXZ) to eliminate vanishing correlation terms, reducing the measurement burden.

Key Results
The study was executed on the IBMQ Boston backend using 20,000 shots per circuit for systems with $N=12$ and $N=20$ qubits.

State Preparation Fidelity: The variational circuits achieved high fidelity with DMRG-calculated ground states (99.8% for TFIM, 98.2% for XXZ at $N=20$ ).
Entanglement Detection:
- TFIM: Near the critical point ( $h=1.0$ ), the PPT criterion successfully identified nearest-neighbor entanglement ( $\lambda_{min} < 0$ ). While unmitigated hardware data showed noise-induced shifts that obscured these signatures, error-mitigated results aligned closely with ideal MPS benchmarks. The analysis revealed that while long-range correlations exist, the two-spin entanglement is restricted to nearest neighbors; longer-range correlations appear classical in nature.
- XXZ Model: In the gapless phase ( $-1 < \Delta \le 1$ ), the mitigated hardware results detected entanglement not only between nearest neighbors but also between next-nearest neighbors. The PPT witness correctly identified the transition boundaries at $\Delta = -1$ and $\Delta = 1$ .
Quantitative Agreement: After applying error mitigation, the experimental energy estimates and correlation functions matched numerical MPS results with high accuracy (e.g., 99.53% accuracy for TFIM energy, 99.05% for XXZ energy).
Critical Signatures: The minimum eigenvalue $\lambda_{min}$ exhibited a distinct dip (becoming most negative) near the quantum critical points, serving as a robust signature of the phase transition and peak pairwise entanglement.

Significance and Claims
The paper claims to establish a practical, model-independent workflow for benchmarking quantum hardware using many-body physics problems. By demonstrating that the PPT criterion can effectively distinguish quantum from classical correlations in noisy, mixed states, the authors provide a universal metric for hardware reliability. The work validates that current quantum processors, when combined with robust error mitigation, can faithfully simulate and characterize the entanglement structure of quantum critical systems. The authors position this approach as a scalable tool for the near-term era, capable of probing condensed matter systems and paving the way for future extensions to multipartite entanglement and finite-temperature studies. They explicitly note that while the current analysis is limited to two-spin entanglement, the framework is designed to be extensible to higher-order correlations and diverse quantum architectures.

The Detective's Tool: The "Spooky Mirror"

The Experiment: Dancing on a Shaky Stage

What They Found (and What They Didn't)

Why This Matters

Technical Summary: Probing Two-Spin Entanglement at Quantum Criticality on a Quantum Processor

More like this