Testing Genuine Multipartite Nonlocality via an Inflated Network with Multi-copy Entangled States

This paper proposes and experimentally verifies a noise-robust method using an inflated network with multi-copy entangled states to test genuine multipartite nonlocality, thereby extending Gisin's Theorem to arbitrary parties and establishing the equivalence between genuine multipartite nonlocality, steering, and entanglement for all pure states.

The Gist

Imagine you have a group of friends who are playing a game where they are supposed to coordinate their answers without talking to each other. In the world of quantum physics, these "friends" are particles, and their "coordination" is called entanglement.

For a long time, scientists knew that if two particles were entangled, they could do things that seemed impossible under normal rules of physics. But what happens when you have three or more particles? Sometimes, they look like they are working together, but they might actually just be two pairs of friends secretly whispering to each other, while the third person is left out. This is the difference between genuine teamwork (where everyone is truly connected) and fake teamwork (where only some are connected).

This paper introduces a clever new way to prove that a group of quantum particles is doing "genuine" teamwork, even when the particles are a bit noisy or imperfect.

The Problem: The "Fake Team" Trick

Usually, to prove a group of particles is truly connected, scientists use a specific test (like the Svetlichny inequality). Think of this test as a strict referee.

• The Issue: Some very special quantum teams (like certain "GHZ" and "W" states) are actually genuine, but they are so subtle that the referee's standard test fails to see them. It's like trying to hear a whisper in a noisy room; the referee thinks the team is just faking it, even though they are actually connected.
• The Old Solution: Scientists previously tried to solve this by looking at many copies of the same team at once. But the old methods were fragile; if there was even a tiny bit of noise (static), the test would break.

The New Idea: The "Inflated Network"

The authors propose a new strategy called an "inflated network."

Imagine you have a single, delicate origami crane (the quantum state). You want to prove it's a real, complex crane and not just a folded piece of paper.

1. The Setup: Instead of looking at just one crane, you make two identical copies of it.
2. The Swap: You take a piece from the first crane and a piece from the second crane and "swap" them, linking the two copies together in a specific way.
3. The Test: Now, you look at the remaining pieces. Because you linked the copies, the "noise" that usually hides the connection gets filtered out. The genuine teamwork becomes loud and clear, like turning up the volume on a radio.

The paper calls this "entanglement swapping." It's like taking two separate conversations, linking them in the middle, and suddenly hearing a clear, unified message that proves everyone was talking to everyone else all along.

What They Did in the Lab

The researchers built a physical machine using photons (particles of light).

• The Ingredients: They used light with two different properties: its color (polarization) and its path (which fiber optic cable it travels through). This allowed them to create two copies of complex quantum states simultaneously.
• The Test: They tested two famous types of quantum teams:
  1. GHZ States: Think of these as a team where everyone is perfectly synchronized.
  2. W States: Think of these as a team where the connection is more distributed and resilient.
• The Result: They successfully proved that these states were genuinely connected, even in situations where the old "referee" tests failed. They also showed that their method works even when the lab is a bit "noisy" (like having a little static in the room), which is a huge improvement over previous methods.

The Big Takeaway

The paper proves a fundamental rule of quantum physics: If a group of particles is genuinely entangled, they are also genuinely non-local (they can coordinate in ways impossible for normal objects).

Previously, this was only proven for simple cases. This paper extends that rule to any number of particles, provided you can use the "inflated network" trick with multiple copies.

In short: They found a way to use two copies of a quantum state to amplify the signal of "genuine teamwork," allowing them to prove that even the most stubborn, noisy quantum groups are truly connected, something that was impossible to prove with a single copy before.

Technical Summary

Technical Summary: Testing Genuine Multipartite Nonlocality via an Inflated Network with Multi-copy Entangled States

Problem Statement
While Bell's theorem and Gisin's theorem establish that pure entangled states violate local realism, the verification of Genuine Multipartite Nonlocality (GMN) in multipartite systems remains challenging. Standard Bell-type inequalities, such as the Svetlichny or Mermin-Ardehali-Belinskii-Klyshko (MABK) inequalities, often fail to detect GMN in certain pure entangled states (e.g., generalized Greenberger-Horne-Zeilinger (GHZ) states or W states) depending on measurement settings. Furthermore, existing methods utilizing multi-copy systems to prove GMN (e.g., joint systems of $n-1$ copies) typically verify the nonlocality of the joint system rather than the single-copy state and lack robustness against noise, rendering them unsuitable for experimental verification of noisy quantum states. There is a need for a noise-robust method that can verify the GMN inherent in a single copy of a genuinely multipartite entangled (GME) pure state.

Methodology
The authors propose a novel network-based approach utilizing an inflated quantum network constructed from multiple identical copies of a given multipartite state. The core strategy involves performing a Bell test on a specific network topology that enables the verification of single-copy nonlocality through post-selection and entanglement swapping.

1. Theoretical Framework:

   • Assumption 1: In each round of the experiment, two identical copies of a multipartite state are distributed to independent nodes.
   • Network Topology (Tripartite Case): The authors construct a 5-partite network involving two sources ($S$ and $S'$), each distributing a tripartite state to observers $\{A, B, C\}$ and $\{A', B', C'\}$, respectively. Observers $B$ and $A'$ perform a joint measurement (entanglement swapping), while $C$ and $B'$ perform local projections. Observers $A$ and $C'$ perform a standard Bell test on the post-selected bipartite state.
   • Inequality Derivation: Under the Svetlichny biseparable non-signalling (NS) model, the authors derive a generalized Bell inequality (Eq. 3). They prove that any network realized with biseparable NS sources satisfies $L_3 \leq R_3$, where $L_3$ is a specific correlator sum and $R_3$ is a bound dependent on the correlation of the intermediate parties.
   • Generalization: The method is extended to $n$-partite systems under Assumption 2 (using $n-1$ copies) or via a modified two-copy approach requiring more measurement settings per party.

2. Experimental Implementation:

   • A hybrid photonic quantum network was constructed where qubits are encoded in both polarization and path degrees of freedom (DoFs).
   • Two copies of generalized GHZ and W states were generated using a Sagnac interferometer with a periodically poled MgO-doped lithium niobate (PPLN) crystal.
   • The setup implemented the specific joint measurements and local projections required by the theoretical network, allowing for the calculation of the inequality parameters $L_3$ and $R_3$.

Key Contributions

• Extension of Gisin's Theorem: The paper establishes that for all multipartite pure states, Genuine Multipartite Entanglement (GME) is equivalent to Genuine Multipartite Nonlocality (GMN) and Genuine Multipartite Steering (GMS) under the multi-copy assumption. This extends Gisin's theorem to an arbitrary number of parties.
• Noise-Robust Verification: Unlike previous multi-copy approaches, the proposed method is noise-robust and capable of verifying the GMN of a single-copy state.
• Unified Method: It provides a unified framework to explore multipartite quantum correlations, bridging the gap between entanglement, steering, and nonlocality in the context of network-distributed multi-copy states.
• Experimental Demonstration: The authors experimentally verified the genuine tripartite nonlocality (GTN) of generalized GHZ and W states in parameter regimes where standard Svetlichny tests fail.

Results

• GHZ and W States: The experiment successfully verified GTN for generalized GHZ states ($\cos \theta|000\rangle + \sin \theta|111\rangle$) with $\theta \in \{\pi/12, \pi/8, \pi/4\}$ and generalized W states.
• Comparison with Standard Tests: For generalized GHZ states with $\theta = \pi/8$ and $\pi/12$, the standard Svetlichny inequality yielded values $\leq 1$ (failing to detect nonlocality), whereas the proposed inflated network inequality was violated, confirming GTN.
• Robustness: The method demonstrated consistent robustness against white noise and decoherence noise (introduced via fused silica plates). The relative fidelity between the two prepared copies exceeded 0.975.
• Parameter Regimes: The experiment accessed previously inaccessible parameter regimes for verifying multipartite nonlocality, significantly extending the scope of experimental verification.

Significance and Claims
The paper claims to offer a significant advancement in quantum foundations by introducing a new avenue for exploring GMN through network-distributed multi-copy states. The authors emphasize that while their method relies on the multi-copy assumption, it uniquely certifies the nonlocality inherent in single copies of the state.

The work deepens the understanding of quantum correlations by demonstrating the equivalence of GME, GMS, and GMN for isolated systems under the multi-copy source assumption. The authors note that their experimental verification is subject to known loopholes (locality, freedom-of-choice, and detection loopholes) due to the encoding of two qubits in a single photon and the nature of the detection setup. Nevertheless, the study establishes a feasible and advantageous method for GMN verification within multi-copy inflated networks, opening new possibilities for studying genuine multipartite nonlocality across different network topologies.