⚛️ quantum physics

Enhanced measurements on quantum computers via the simultaneous probing of non-commuting Pauli operators

This paper presents a first experimental implementation of a two-copy measurement scheme using Bayesian statistics and adaptive shot allocation to simultaneously probe non-commuting Pauli operators, demonstrating its ability to outperform state-of-the-art protocols in minimizing total measurement shots for high-precision estimation despite inherent variance overheads.

Original authors: Rick P. A. Simon, Zheng Shi, Charlie Nation, Andrew Jena, Luca Dellantonio

Published 2026-09-23

📖 4 min read🧠 Deep dive

CC BY 4.0

Original authors: Rick P. A. Simon, Zheng Shi, Charlie Nation, Andrew Jena, Luca Dellantonio

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

Quantum computers hold the promise of solving problems that are impossible for today's machines, from designing new medicines to modeling complex materials. However, these machines are notoriously fragile and difficult to read. To understand what a quantum computer is doing, scientists must measure its state, but this process is fraught with difficulty. The information they seek is often encoded in a collection of tiny, interacting parts that do not play nicely together; measuring one part can disturb the others, and many of the properties they need to check cannot be observed at the same time. This creates a bottleneck: to get a clear picture of the system, researchers must repeat their measurements thousands or millions of times, a process that consumes precious time and resources. The central challenge is how to gather the most accurate information possible while using the fewest number of these expensive measurements.

A team of researchers has developed a new strategy to tackle this problem, one that changes how they look at the data they collect. Instead of trying to measure the quantum state just once and hoping for the best, their method involves preparing two identical copies of the state at the same time. By probing these twin copies simultaneously, the researchers can extract information about the system's properties in a way that was previously impossible. They found that this "double" approach allows them to determine the average value of a property and, crucially, to calculate the uncertainty of that measurement as it happens. This real-time knowledge of the error allows their system to be smart about where it spends its effort. It can instantly recognize which parts of the system are still uncertain and focus the next measurement there, rather than wasting resources on parts that are already well understood.

The researchers tested this idea using a computer simulation of a quantum algorithm, pitting their new method against the best existing techniques. They focused on two types of systems: a model of nuclear spins arranged in a grid and a simulation of a hydrogen molecule. In these simulations, the quantum properties they needed to measure were often in conflict with one another, meaning they could not be measured together in the traditional way. The team's new protocol, which they call the double scheme, proved to be significantly more efficient in these difficult scenarios. When the properties were highly non-commuting—meaning they resisted being measured together—the new method required far fewer total measurements to reach the same level of precision as the older methods. In some cases, it reduced the number of measurements needed by a factor of two or more, a substantial saving in the world of quantum computing where every measurement counts.

The success of this approach relies on a clever feedback loop. The algorithm constantly updates its understanding of the system based on the results of each measurement. It uses a statistical method to predict which measurement will reduce the overall uncertainty the most. If the system is unsure about the "sign" of a property (whether it is positive or negative), the algorithm knows to measure that property alone. If it is unsure about the magnitude (how strong the property is), it can use the double scheme to get a clearer picture of the size without needing to know the sign yet. This adaptive strategy means the computer does not just blindly repeat measurements; it learns as it goes, prioritizing the most uncertain parts of the puzzle. The researchers confirmed that their estimates of the error were statistically sound, meaning the uncertainty they calculated matched the actual variation in the results, giving them a reliable gauge of their progress.

While the double scheme is a powerful tool, the researchers noted that it is not a magic bullet for every situation. The method requires more physical resources, specifically twice as many qubits to hold the two copies of the state, and it is most effective when the properties being measured are difficult to group together. In cases where the properties are already easy to measure simultaneously, the extra cost of the double scheme might not be worth it. However, for the complex problems that quantum computers are designed to solve, where many properties are intertwined and non-commuting, this new method offers a clear path forward. It suggests that by using multiple copies of a state and adapting the measurement strategy on the fly, scientists can extract more information with less effort, bringing the practical utility of quantum computers one step closer to reality.

Technical Summary: Enhanced Measurements on Quantum Computers via Simultaneous Probing of Non-Commuting Pauli Operators

Problem Statement
Accurately estimating the expectation value of an observable $\hat{O}$ on quantum computers is a fundamental bottleneck for algorithms, particularly variational quantum algorithms (VQAs). Observables are typically expressed as weighted sums of Pauli strings (PS), $\hat{O} = \sum c_i \hat{P}_i$ . Standard measurement protocols require grouping commuting PS to measure them simultaneously. However, when an observable contains many non-commuting PS, the number of required measurement groups increases, necessitating an efficient strategy to allocate a finite measurement budget ( $M$ ) to minimize estimation error. Existing methods, such as classical shadows or grouping strategies, often struggle to provide precise single-run estimates of both the mean and the statistical error, or they fail to optimally allocate shots when non-commuting terms dominate.

Methodology
The authors propose a "double" measurement scheme that combines multi-copy quantum state preparation with a Bayesian statistical framework.

Multi-Copy Principle: Instead of measuring a single copy of the state $|\phi\rangle$ , the protocol prepares two identical copies, $|\psi\rangle = |\phi\rangle \otimes |\phi\rangle$ . This allows for the simultaneous measurement of "double" Pauli strings $\hat{P}_i \otimes \hat{P}_i$ . Crucially, while $\hat{P}_i$ and $\hat{P}_j$ may not commute, their doubled versions $\hat{P}_i \otimes \hat{P}_i$ and $\hat{P}_j \otimes \hat{P}_j$ always commute, enabling the simultaneous assessment of magnitudes for all Pauli terms regardless of their commutation relations.
Bayesian Estimation: The scheme employs Bayesian inference to estimate the probability $\theta_i$ $θ_{i}$ of obtaining $+1$ $+ 1$ for a single Pauli measurement and $\phi_i$ $ϕ_{i}$ for the double measurement.
- The relationship between single and double outcomes is derived as $\phi_i = \theta_i^2 + (1-\theta_i)^2$ .
- The protocol constructs posterior distributions $P_1$ (for single/double outcomes of one PS) and $P_2$ (for correlated outcomes of commuting PS pairs) to estimate the mean $\hat{O}$ and, critically, the variance of the estimator $(\Delta \hat{O})^2$ .
- Unlike previous works that calculate variance from covariances of PS, this method calculates the variance of the estimate itself, allowing for real-time error monitoring.
Adaptive Shot Allocation: An algorithm dynamically allocates measurement shots between the "double" scheme (measuring all $\hat{P}_i \otimes \hat{P}_i$ $\hat{P}_{i} \otimes \hat{P}_{i}$ ) and the "single" scheme (measuring groups of commuting $\hat{P}_i$ $\hat{P}_{i}$ ).
- The algorithm uses "virtual measurements" to predict the reduction in $(\Delta \hat{O})^2$ for each possible choice.
- It accounts for the resource overhead of the double scheme (requiring twice the qubits and potentially more circuit depth) by defining an effective shot count $M_{eff} = M_{single} + 2\alpha M_{double}$ .
- The algorithm prioritizes sampling the most uncertain Pauli terms, specifically targeting the resolution of sign ambiguities that arise from the symmetric nature of double measurements when $\langle \hat{P}_i \rangle \approx 0$ .

Key Contributions

First Implementation: This work presents the first implementation of a two-copy scheme within a measurement protocol that integrates Bayesian statistics.
Enhanced Error Monitoring: The method provides an accurate, single-run estimate of the statistical error $(\Delta \hat{O})^2$ alongside the mean. While AEQuO is noted as a state-of-the-art approach capable of single-run error estimation, this work extends the capability by integrating it with a multi-copy framework that offers superior error reduction in specific regimes.
Adaptive Algorithm: The authors develop an algorithm that intelligently switches between single and double measurements based on the current state of knowledge, optimizing the trade-off between resource cost and error reduction.
Theoretical Confirmation: The work numerically confirms theoretical predictions regarding the overhead of multi-copy schemes, specifically the square-root relationship between variance and shot count for terms with zero expectation values.

Results
Numerical benchmarks were performed on Hydrogen molecule ( $H_2$ ) Hamiltonians and quantum Ising models with varying lattice sizes.

Performance vs. State-of-the-Art: The double scheme significantly outperforms the state-of-the-art AEQuO protocol and fixed-allocation strategies in regimes where the observable contains many non-commuting Pauli terms (e.g., molecular Hamiltonians). In these cases, the double scheme achieves lower estimation errors for a given effective measurement budget. However, the authors note that in specific scenarios (such as certain Ising model instances), AEQuO can sometimes yield better errors, likely due to differences in shot allocation strategies based on instantaneous knowledge.
Regime Dependence: The advantage is highly dependent on the observable's properties. For systems with many non-commuting terms (where grouping is difficult), the double scheme is superior. However, for terms with expectation values near zero, the double scheme's efficiency degrades (variance scales as $1/\sqrt{M}$ rather than $1/M$ ), and the algorithm adaptively reduces the allocation to double measurements in these specific regimes.
Statistical Validity: The estimated error bars $(\Delta \hat{O})^2$ were shown to be statistically meaningful, correctly capturing the uncertainty of the estimator without requiring repeated experimental runs.
Resource Scaling: The algorithm's time complexity is bounded by $O(p^2 M^3)$ but can be optimized to $O(M)$ with parallelization and batch allocation strategies.

Significance and Claims
The paper claims that the proposed "double" scheme is a practical, implementation-ready advancement for quantum measurement. Its primary significance lies in its ability to minimize the total number of shots required to reach a target precision, particularly for complex observables dominated by non-commuting terms, such as those found in quantum chemistry simulations. By providing a robust, single-run error estimate, the method enables adaptive measurement strategies that are often more efficient than current best practices, though the relative performance depends on the specific commutation structure of the observable. The authors note that while the scheme incurs a resource overhead (factor of 2 in shots/qubits), this cost is often outweighed by the reduction in total shots needed for convergence in non-trivial systems. The work concludes that this approach is particularly well-suited for near-term quantum simulations where measurement budgets are strictly limited.

Technical Summary: Enhanced Measurements on Quantum Computers via Simultaneous Probing of Non-Commuting Pauli Operators

More like this