Govia and Colleagues Models Pauli Fidelity Symmetries for Quantum Noise Characterisation

A new method simplifies the challenging task of characterising errors in quantum circuits. Moein Malekakhlagh and colleagues at IBM T. J. Watson Research Centre demonstrate that the physical structure of noise imposes constraints on gate noise channels, relating the fidelity of a Pauli error and its conjugate. The findings reveal that coherent errors do not induce asymmetry, and only specific dissipative errors can break symmetry at first order, offering a means to fix the gauge using error type rather than magnitude. By using these symmetries, systematic identification of state preparation and measurement (SPAM) errors becomes possible, ultimately streamlining error characterisation and mitigation strategies, as validated on IBM Kingston.

Symmetry constraints resolve gauge ambiguities in single-qubit error characterisation

Scientists at IBM T. J. Watson Research Centre have reduced the number of undetermined single-qubit depolarizing parameters from ‘n’ to zero, a feat previously impossible due to inherent gauge ambiguities in quantum error characterisation. The physical structure of realistic noise imposes approximate symmetry constraints on Pauli fidelities, linking the accuracy of a quantum operation to its transformed counterpart under gate operations. By exploiting these symmetries, the team established a method to fix the gauge, the reference point for measuring quantum properties, using error type rather than error magnitude, streamlining the identification of state preparation and measurement (SPAM) errors.

Quantum computations are inherently susceptible to errors arising from various sources, including imperfections in qubit control, environmental noise, and limitations in measurement fidelity. Characterising these errors is crucial for developing effective error correction and mitigation techniques. However, a fundamental challenge arises from gauge degrees of freedom. This means that certain error parameters, such as the individual contributions of state preparation and measurement (SPAM) errors, are fundamentally unlearnable from any experiment performed within the native gate set of the quantum computer. This ambiguity stems from the fact that the absolute scale of error probabilities is often undefined, only relative differences being measurable. Traditional error characterisation methods often require imposing arbitrary gauge choices, introducing further complexity and potential inaccuracies.

Coherent errors, originating from the precise control of qubits, do not induce first-order asymmetry in Pauli fidelities. Only a limited range of dissipative errors, specifically those affecting transitions between qubit states, can break this symmetry at first order, and rules were derived to identify these problematic errors. Common sources of single-qubit noise, such as T1-relaxation and T2φ-pure-dephasing, cause asymmetry only at second order, meaning their impact is less direct.

The team’s finding simplified the identification of state preparation and measurement errors on the IBM Kingston quantum computer, and this approach could be extended to more complex systems. Establishing a reliable method for pinpointing the sources of error in quantum computers is vital as these machines scale towards practical applications. Fully untangling all error sources remains an immense challenge, but this provides an important simplification.

Predictable symmetries within the noise affecting quantum bits, or qubits, have been identified, allowing scientists to narrow down the possibilities when diagnosing problems and improving the reliability of quantum computations. IBM T. J. Watson Research Centre has established a new method for simplifying the characterisation of errors in quantum computers by exploiting inherent symmetries within noise. Realistic noise imposes constraints on how errors affect quantum operations, enabling the reference point for measurements, the gauge, to be fixed using the type of error rather than its precise strength. Consequently, identifying errors stemming from the preparation and measurement of quantum states becomes more systematic, paving the way for more efficient error mitigation strategies.

The core of this advancement lies in recognising that the physical mechanisms generating noise impose constraints on the structure of error channels. These channels describe how quantum states are transformed by noise during computation. The researchers focused on single-qubit errors, which are often dominant in early-stage quantum devices. They analysed how noise affects the Pauli operators, a set of fundamental quantum gates (I, X, Y, and Z), and their corresponding fidelities, which quantify the accuracy of the operations. The team demonstrated that coherent errors, which arise from imperfections in the control pulses applied to the qubits, preserve a symmetry between the Pauli fidelities. This means that the error induced on a Pauli operator is directly related to the error on its conjugate operator.

However, certain dissipative errors, those caused by interactions with the environment leading to energy loss or dephasing, can break this symmetry. The researchers identified that only errors affecting transitions between the qubit’s |0⟩ and |1⟩ states can induce first-order asymmetry in the Pauli fidelities. This finding is significant because it provides a way to distinguish between different types of errors based on their symmetry properties. By analysing the symmetry of the observed errors, the gauge can be fixed, effectively eliminating the ambiguity in determining SPAM errors. SPAM errors represent the inaccuracies in the initial state preparation and final state measurement, which can significantly degrade the performance of quantum algorithms.

The validation of this method on the IBM Kingston quantum computer, a superconducting transmon qubit processor, demonstrates its practical applicability. The ability to systematically identify and mitigate SPAM errors is crucial for improving the fidelity of quantum computations and enabling the development of more complex quantum algorithms. As quantum computers continue to scale in size and complexity, the challenges associated with error characterisation and mitigation will only intensify. This new approach offers a promising pathway towards building more robust and reliable quantum machines, bringing us closer to realising the full potential of quantum computing.

Furthermore, the principles established in this work are not limited to single-qubit errors. The researchers suggest that similar symmetry constraints may exist in multi-qubit systems, potentially leading to further simplifications in error characterisation for more complex quantum circuits. Future research will focus on extending these techniques to address the challenges posed by correlated noise and higher-order error terms, ultimately paving the way for fault-tolerant quantum computation.

The research demonstrated that the structure of noise in quantum circuits possesses symmetries which constrain how errors affect qubit behaviour. This matters because it allows scientists to resolve ambiguities when identifying errors in state preparation and measurement, known as SPAM errors, which degrade quantum algorithm performance. By leveraging these symmetries to fix the gauge, researchers were able to systematically identify SPAM errors on the IBM Kingston processor. The authors intend to extend these techniques to address more complex noise scenarios and multi-qubit systems.