⚛️ quantum physics

Scalable Test of Genuine Multipartite Entanglement via Partially Randomized Measurements

This paper introduces a scalable criterion for certifying genuine multipartite entanglement using partially randomized measurements that avoids exponential scaling with system size, and experimentally validates its effectiveness by demonstrating genuine five-partite entanglement on an ion-trap quantum computer.

Original authors: Jan Wojcik, Pawel Chrabkowski, Wieslaw Laskowski

Published 2026-08-17

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Jan Wojcik, Pawel Chrabkowski, Wieslaw Laskowski

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 world where information isn't just written in zeros and ones, but in a mysterious, ghostly dance of particles that can be in two places at once. This is the realm of quantum physics, a field that promises to revolutionize how we communicate, compute, and measure the universe. At the heart of this revolution is a strange phenomenon called "entanglement." You can think of entangled particles as a pair of magic dice: no matter how far apart you roll them, if one lands on a six, the other instantly shows a six too. They aren't just linked; they are sharing a single, unified existence.

But here is the tricky part: sometimes, this magic isn't just between two dice. It can happen among a whole crowd of them all at once. This is called "genuine multipartite entanglement." It's like a secret handshake shared by an entire group of friends, where no one can be left out or paired off separately. Scientists care deeply about this because it's the ultimate fuel for powerful quantum computers and unbreakable codes. However, proving that a group of particles is truly sharing this deep, all-in-one connection is incredibly hard. Traditionally, checking if a group of $N$ particles is entangled required taking a number of measurements that grew exponentially with the size of the group. It was like trying to find a specific grain of sand on a beach by counting every single grain one by one; as the beach got bigger, the task became impossible.

This is where a new study by Jan Wójcik, Paweł Chrabkowski, and Wiesław Łaskowski from the University of Gdańsk steps in with a clever shortcut. They have developed a new way to test for this "all-in-one" entanglement that doesn't require counting every single grain of sand. Instead of trying to map out every possible connection, they use a method called "partially randomized measurements." Imagine trying to guess the shape of a hidden object in a dark room. The old way was to feel every single inch of the object with your hands, which took forever. The new way is to spin around randomly and tap the object from different angles, then use the pattern of those taps to figure out what it is.

The researchers showed that by measuring particles in random directions within specific flat "planes" (like spinning a coin on a table rather than looking at it from every angle in 3D space), they could calculate a specific number called the "average squared correlation." If this number is high enough, it proves the particles are genuinely entangled. They derived a mathematical rule (a threshold) that says: if your random tapping yields a result greater than a certain value (specifically $2^{-k}$ , where $k$ relates to how the group might be split), the group cannot be separated into smaller, independent chunks. In simpler terms, if the random taps show a strong enough "group vibe," the particles are definitely sharing a genuine, multipartite secret.

The team didn't just do the math; they tested it on a real quantum computer made of trapped ions (charged atoms held in place by electric fields). They created a state of five entangled particles (a five-qubit GHZ state) and ran their new test. The results were clear: both their new randomized method and the old, heavy-duty method confirmed that the five particles were genuinely entangled. The exciting part is that while the old method required measuring 16 specific combinations to get this result, the new randomized method achieved the same certainty with far fewer resources and without needing to know the exact settings in advance.

The paper demonstrates that this new criterion works for various types of entangled states, including famous families like GHZ states and Dicke states. They found that for some states, the test is very sensitive, while for others, it might miss the entanglement if the state sits exactly on the edge of the detection limit. However, the key takeaway is that this method scales beautifully. As the number of particles grows, the number of measurements needed doesn't explode exponentially like it used to; instead, it stays manageable. This means that as quantum computers get bigger and more complex, we will finally have a practical, scalable way to verify that they are actually doing the magic they are supposed to do. The authors confirm that this approach is not just a theory but a working tool that can certify genuine entanglement in real, noisy quantum machines, paving the way for larger and more reliable quantum technologies.

Technical Summary: Scalable Test of Genuine Multipartite Entanglement via Partially Randomized Measurements

Problem Statement
Certifying genuine multipartite entanglement (GME) in quantum systems is a fundamental challenge in quantum information science. While various methods exist—including entanglement witnesses, density-matrix criteria, and Bell-type inequalities—many suffer from significant limitations. Specifically, traditional approaches often require prior knowledge of the target state, carefully optimized measurement settings, or well-aligned local reference frames. Furthermore, direct evaluation of correlation tensors typically incurs a measurement cost that grows exponentially with the system size ( $N$ ), rendering these methods impractical for large-scale multipartite systems.

Methodology
The authors propose a scalable criterion based on correlation-tensor subsector lengths restricted to local measurement planes. The core methodology involves:

Partial Randomization: Instead of sampling measurement directions over the entire Bloch sphere, local measurement directions are sampled randomly but restricted to specific local planes ( $P_i$ ) spanned by pairs of orthogonal vectors $\{\vec{n}^1_i, \vec{n}^2_i\}$ .
Theoretical Derivation:
- The authors define the subsector length $\|T\|^2_{P_1...P_N}$ as the sum of squared correlation tensor elements within these restricted planes.
- They derive a universal upper bound for $k$ -separable states (states that can be written as a convex mixture of $k$ -product states). The bound is established as $\|T\|^2_{P_1...P_N} \leq 2^{N-k}$ .
- Consequently, a violation of this bound ( $\|T\|^2_{P_1...P_N} > 2^{N-k}$ ) certifies that the state is not $k$ -separable. For $k=2$ , this certifies GME.
Connection to Randomized Measurements: The subsector length is linked to the average squared correlation $R^{(2)}_{P_1...P_N}$ via the relation $\|T\|^2_{P_1...P_N} = 2^N R^{(2)}_{P_1...P_N}$ . This allows the criterion to be evaluated experimentally by estimating the average squared correlation over angles $\phi_i$ sampled uniformly from $[0, 2\pi)$ within the chosen planes.
Statistical Scaling Analysis: The paper compares the measurement resources required for the direct subsector-length approach versus the randomized approach.
- Direct Approach: Requires measuring all $2^N$ correlation terms, leading to a total measurement count scaling as $M \sim 4^N/\epsilon^2$ for a target uncertainty $\epsilon$ .
- Randomized Approach: Requires $m \sim 1/\epsilon^2$ random settings with a constant number of shots per setting. The total measurement count scales as $M \sim 1/\epsilon^2$ , showing no explicit exponential dependence on $N$ .

Key Contributions and Results

Scalable Criterion: The paper introduces a GME criterion that avoids exponential scaling in measurement resources by utilizing partially randomized measurements within local planes.
General Bounds: The authors provide rigorous bounds for $k$ -separable states, establishing a "ladder-like" structure of thresholds (e.g., $R^{(2)} > 1/4$ excludes biseparability, $R^{(2)} > 1/8$ excludes 3-separability).
Theoretical Validation: The criterion is tested on representative families of states:
- GHZ States: The criterion successfully certifies GME for all $N$ -qubit GHZ states using the $xy$-plane.
- Dicke States: The method detects GME and higher-order separability levels for various Dicke states $|D^{(e)}_N\rangle$ , with optimal detection in the $xz$-plane.
- Four-Qubit Family: For a specific family of four-qubit states $|\psi(\alpha)\rangle$ , the criterion certifies GME for all $\alpha \in [0, 1]$ except at specific points ( $\alpha=0, 2/3$ ) where the state lies exactly on the detection threshold. The authors note this illustrates the sufficient-but-not-necessary nature of entanglement witnesses.
Experimental Demonstration: The method was implemented on a 5-qubit trapped-ion quantum computer (PIAST-Q).
- The team prepared a 5-qubit GHZ state.
- Using the standard subsector-length approach, they measured $\|T\|^2_{xy} = 8.55 \pm 0.07$ , exceeding the bound of $2^{5-2}=8$ .
- Using the randomized approach, they obtained $R^{(2)}_{xy} = 0.27775 \pm 0.00998$ , exceeding the bound of $2^{-2}=0.25$ .
- Both methods successfully certified genuine five-partite entanglement.

Significance
The paper claims that its primary significance lies in the scalability of the proposed method. By shifting from explicit measurement of all correlation tensor elements to estimating average squared correlations via partially randomized measurements, the method eliminates the explicit exponential dependence on the number of qubits. This makes the certification of GME feasible for larger multipartite systems on realistic, noisy quantum processing platforms without the need for complex state-specific optimization or perfect reference frame alignment. The experimental demonstration on a trapped-ion processor confirms the practical applicability of the method in current quantum hardware environments.

Technical Summary: Scalable Test of Genuine Multipartite Entanglement via Partially Randomized Measurements

More like this