⚛️ quantum physics

Measurement Geometry as a Resource for Certifying Network Nonlocality

This paper establishes measurement geometry as an independent resource for certifying network nonlocality by developing an ancilla-assisted framework that successfully demonstrates bilocal nonlocality and near-threshold fully network nonlocal correlations on a 156-qubit superconducting processor, revealing that specific joint measurement settings are critical for maximizing these violations.

Original authors: Leon Adachi, Le Bin Ho

Published 2026-07-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Leon Adachi, Le Bin Ho

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

The Big Picture: A Quantum "Game of Telephone"

Imagine a game of "Telephone" played by three people: Alice, Bob, and Charlie. Alice and Charlie are on opposite ends of a room, and they cannot talk to each other directly. They can only talk to Bob, who stands in the middle.

In a normal, classical world, if Alice and Charlie want to coordinate their answers, they would need a secret plan (a "hidden variable") agreed upon beforehand. But in the quantum world, they share something special called entanglement. It's like they are holding two halves of a magic coin that always land on the same side, no matter how far apart they are.

The scientists in this paper wanted to prove that Alice and Charlie are truly using this "quantum magic" and not just following a secret classical plan. They did this by testing two different levels of "quantum-ness":

Level 1 (Bilocal): Proving the two sources of magic coins are independent and not cheating.
Level 2 (Full Network Nonlocality - FNN): Proving that neither source is classical. This is a much harder, stricter test.

The Problem: The Middleman's Job

In this setup, Bob has a very specific job. He receives a particle from Alice and a particle from Charlie. He has to perform a joint measurement on both of them at the same time. Think of Bob as a referee who has to look at two balls simultaneously and decide how they interact.

The paper's main discovery is that how Bob looks at the balls matters just as much as the balls themselves.

If Bob looks at the balls from the "wrong angle," even if the balls are perfectly entangled (the best quantum magic possible), the test will fail. It's like trying to open a door with a key: even if you have the perfect key, if you hold it at the wrong angle, the door won't open.

The Experiment: A Digital Test Drive

The researchers built a computer simulation and then ran the experiment on a real, massive quantum computer (IBM's "Kingston" processor with 156 qubits).

1. The Simulation (The Perfect World):
In their computer simulation, they showed that their method works perfectly. They could successfully prove both Level 1 and Level 2 quantum magic.

2. The Real Hardware (The Noisy World):
When they ran it on the real quantum computer, things got a bit messy because real machines make mistakes (noise).

Level 1 (Bilocal): They passed! Even with the noise, they proved the quantum connection existed. The score was 1.067, which is higher than the classical limit of 1.0.
Level 2 (FNN): They came very close, but didn't quite cross the finish line. The scores were 99% and 96% of what was needed to pass. This shows that proving the "stronger" version of quantum magic is much harder and requires a much cleaner machine.

The Big Discovery: Measurement Geometry is a Resource

The most exciting part of the paper is what they found when they changed how Bob measured the particles.

They treated Bob's measurement angle like a dial they could turn. They found that:

Different Angles for Different Goals: The angle that makes the "Level 1" test pass is different from the angle that makes the "Level 2" test pass.
The "Disappearing Act": They found settings where the quantum connection was so strong (maximally entangled), but because Bob measured at the "wrong" angle, the test results looked completely classical. The quantum magic effectively "disappeared" from the test results.

The Analogy:
Imagine you have a perfect radio signal (entanglement).

If you tune your radio to Station A, you hear music (Level 1 passes).
If you tune to Station B, you hear a different, clearer broadcast (Level 2 passes).
If you tune to Station C, you hear static, even though the signal is still strong.

The paper concludes that the way you tune the radio (the measurement geometry) is just as important as the strength of the signal (entanglement). It is a separate "resource" that you need to manage to prove quantum nonlocality.

Summary of Results

Success: They successfully proved "Level 1" quantum nonlocality on a real quantum computer.
Challenge: "Level 2" (the stronger proof) was almost achieved but fell just short due to current hardware noise.
Key Insight: You cannot just rely on having good entangled particles; you must also choose the exact right measurement angle for the specific type of proof you want to make. If you get the angle wrong, the proof fails, even with perfect particles.

This work provides a practical "user manual" for how to set up these experiments on future quantum computers to ensure they actually work.

Technical Summary: Measurement Geometry as a Resource for Certifying Network Nonlocality

Problem Statement
Quantum networks generate nonclassical correlations that cannot be explained by classical models with independent sources. While the role of entanglement in these networks is well-established, the impact of measurement design—specifically the geometry of joint measurements at intermediate nodes—remains largely unexplored. Existing studies have primarily focused on demonstrating witness violations for specific measurement configurations, leaving a gap in understanding how the choice of nonlocal measurements influences the certifiable level of network nonlocality. Furthermore, there is a need for an operational framework to evaluate the complex nonlocal observables required for certifying Full Network Nonlocality (FNN) on programmable quantum hardware.

Methodology
The authors develop an operational framework based on ancilla-assisted meter circuits to evaluate nonlocal observables in a bilocal Alice–Bob–Charlie network.

Network Configuration: The network consists of two independent sources distributing entangled pairs to Alice and Bob, and Bob and Charlie, respectively. Bob holds two subsystems ( $B_A$ and $B_C$ ) and performs joint measurements to mediate correlations between the outer parties.
Witnesses: The framework targets two levels of nonlocality:
1. Bilocal Hidden-Variable (BLHV) violation: Certifying correlations that violate the assumption of two independent classical sources ( $S_{BLHV} > 1$ ).
2. Full Network Nonlocality (FNN): Certifying correlations that exclude any model where at least one source remains classical. This requires the simultaneous violation of two hybrid witnesses, $R_{CNS}$ and $R_{NSC}$ , which bound classical–non-signaling (NS) and NS–classical models, respectively.
Implementation: The nonlocal observables, including multi-body correlators and partial Bell-state measurements on Bob's qubits, are extracted using ancilla-assisted circuits that preserve the network state during measurement.
Hardware: The protocol is implemented on a 156-qubit superconducting processor (IBM Kingston, Heron architecture) with readout-error mitigation. The study sweeps the entanglement parameter $\theta$ and, in a separate analysis, varies Bob's joint measurement angles $(\phi, \varphi)$ to map the "measurement geometry."

Key Contributions

Operational Framework: The paper establishes a method for evaluating BLHV and FNN witnesses on quantum hardware using ancilla-assisted circuits, successfully reproducing theoretical predictions in simulation.
Experimental Certification: The authors report the successful certification of bilocal nonlocality on hardware ( $S_{BLHV} = 1.067(6)$ ) and demonstrate that FNN witnesses reach 99% and 96% of their certification thresholds, highlighting the stringent experimental demands of FNN.
Measurement Geometry as a Resource: A central finding is that the accessible level of network nonlocality is determined not only by the entanglement strength but also by the geometry of the intermediate node's measurement. The study reveals that BLHV and FNN certifications are optimized by different measurement settings.

Results

BLHV Violation: On the IBM Kingston processor, the bilocal inequality was violated with $S_{BLHV} = 1.067 \pm 0.006$ at maximal entanglement ( $\theta = \pm \pi/2$ ), exceeding the classical bound by more than six standard deviations. The violation was sustained over a wide range of entanglement ( $|\theta| \in [0.4\pi, 0.6\pi]$ ).
FNN Certification: While the FNN witnesses ( $R_{CNS}$ and $R_{NSC}$ ) followed the theoretical $\theta$ dependence, hardware noise attenuated the values to $2.97$ and $2.87$ respectively at maximal entanglement. These values represent 99% and 96% of the classical bound (3), falling just short of certification. This gap underscores that FNN certification is significantly more fragile to noise than BLHV violation.
Measurement Dependence: By varying Bob's joint measurement operator $B_1(\phi, \varphi)$ $B_{1} (ϕ, φ)$ , the authors mapped the parameter space of measurement settings.
- The region satisfying BLHV violation covers approximately 53% of the parameter space, while the FNN certification region covers only ~14%.
- The optimal settings for BLHV and FNN are shifted relative to each other by $\pi/4$ in the $\phi$ parameter.
- Crucially, the study demonstrates that even with maximally entangled states, both violations can disappear if the measurement geometry is suboptimal.

Significance
The paper identifies measurement geometry as an independent resource for network nonlocality, distinct from the entanglement distributed by sources. It establishes that certifying stronger forms of nonlocality (FNN) requires not only high-fidelity entanglement but also precise alignment of joint measurement settings. The work provides a practical route for certifying network nonlocality on programmable quantum processors and illustrates the substantial experimental gap between verifying bilocality and achieving full network nonlocality certification. The findings suggest that future network protocols must treat measurement design as a critical optimization variable alongside source quality.