Single Qubit Unlocks Full Analysis of Complex Quantum Processes

Abhinash Kumar Roy and colleagues at Macquarie University present a resource-efficient method for multi-time process tomography, a technique key for analysing noise and improving control in quantum systems. Their findings show that using a single qubit as ancillary memory is sufficient to reconstruct arbitrary multi-time dynamics, removing the need for complex mid-circuit measurements and reset procedures common in current quantum devices. The advancement offers a pathway towards more practical and scalable quantum process characterisation.

Single qubit ancilla enables complete multi-time quantum state reconstruction

Previously, complete informational sets on a qubit system required sixteen independent operations. This work demonstrates that a single qubit ancilla now suffices for full reconstruction of arbitrary multi-time dynamics. This breakthrough crosses a critical threshold, as prior methods demanded increasing ancilla resources with each additional lab intervention, making scalable multi-time process tomography impractical. By utilising sequential interactions with a single ancilla, researchers have eliminated the need for mid-circuit measurements and resets, dramatically reducing the complexity of quantum process characterisation.

The team demonstrated that sequential interactions with a single qubit ancilla can generate an informationally complete family of correlated probes for processes of arbitrary length, without mid-circuit measurements or reset. This provides a resource-efficient route for complete multi-time process tomography and establishes that one qubit of coherent ancillary memory suffices for full reconstruction of arbitrary multi-time dynamics. Correlated probes, generated from a single ancilla maintained in a coherent state, represent the entire range of multi-time operator spaces, establishing a fundamental limit on the ancillary resources required.

Earlier methods often required increasing ancilla resources with each experimental step, hindering scalability. This new technique avoids those limitations by removing the need for mid-circuit measurements and resets, simplifying quantum process characterisation. The ancilla’s initial state and final measurement outcome were fixed at ‘0’ for the constructive proof involving two intermediate labs, allowing for a specific family of probes to be generated.

Multi-time process matrices characterise non-Markovian dynamics in quantum systems

Scientists are currently investigating temporally correlated noise as a significant impediment to the reliable operation of quantum devices. Unlike Markovian noise, non-Markovian dynamics retain memory of earlier system-environment interactions, preventing their capture by a sequence of independent channels. Such memory effects invalidate the assumptions of standard error correction, gate characterisation, and control optimisation protocols, posing a fundamental challenge to scalable quantum computing.

The multi-time process matrix formalism provides a complete operational description of these dynamics, generalising quantum states and channels to sequences of interventions at different times. In principle, it enables the reconstruction of all experimentally accessible temporal correlations. Access to the full process matrix is important not only for diagnosing memory effects, but also for developing memory-aware control and verification strategies for noisy quantum hardware.

However, complete multi-time process tomography remains experimentally demanding for two main reasons. Firstly, the number of required experimental configurations grows rapidly with the number of probing times. Secondly, informational completeness requires non-deterministic intervention operations, such as measurement-and-repreparation maps, typically implemented using mid-circuit measurement, feed-forward, and reset on current platforms. Although such capabilities have enabled recent demonstrations of full process tomography, they remain slow, noisy, and technologically restrictive.

Consequently, most present-day experiments access only restricted process matrices constructed from deterministic operations alone, leaving a practical route to complete multi-time characterisation on near-term hardware still lacking. In this work, researchers study the minimal ancillary resources required for the complete characterisation of a fixed-size multi-time process, specified by the system dimension and the number of intervention times. They demonstrate that a single coherent qubit ancilla is sufficient to realise an informationally complete set of probes for arbitrary multi-time processes, without requiring either mid-circuit measurements or ancilla resets.

Sequential interactions between the system and a single qubit ancilla generate a family of correlated operations, constrained to have a matrix-product structure with fixed bond dimension. Despite this constraint, their linear span coincides with the full operator space needed for complete process reconstruction. This construction has implications beyond tomography, as the accessible probes span the full multi-time operator space. Therefore, arbitrary linear functionals of the process, including memory witnesses and other high-rank observables, can be evaluated from experimentally friendly probe statistics.

The resulting protocol offers a resource-efficient framework not only for complete multi-time tomography, but more broadly for the characterisation, verification, and control of temporally correlated quantum dynamics. A quantum system may evolve in time, possibly interacting with an environment. To probe this evolution, operations may be performed on the system at different times. The process matrix formalism provides a general operational framework to describe such multi-time experiments.

In this formalism, a sequence of laboratories, A1, A2, , AN, are placed along the system’s time evolution. Each lab implements a quantum instrument on the system, associated with a linear map from an input Hilbert space Hi to an output Hilbert space Ho, and with a classical outcome AI. The instrument implemented in lab Ai is described by a completely positive map Mai: Hi →Ho, represented by its Choi operator, Mai = (I ⊗Mai)|I⟩⟨I|, which belongs to L(Hi ⊗Ho). Here I denote the identity map on L (Hi), and |I⟩= P n |n⟩⊗|n⟩is the unnormalised maximally entangled state on Hi ⊗Hi, equivalent to the trace operator. Therefore, any informationally complete set of operations for a single lab must contain at least d4 linearly independent elements.

If we restrict the intermediate probes to unitary operations acting on the system, the accessible set of lab operations is U = { |U⟩⟨U| : U†U = UU† = I }. In the Choi representation, the unitarity constraint is equivalent to requiring that tracing over either the input or the output space yields the identity operator. In an expansion in a generalised Pauli operator basis, this implies that all coefficients multiplying terms of the form I ⊗σi with i, 0 must vanish. The space spanned by deterministic operations therefore has dimension only d2(d2 −1) + 1. Since this is still smaller than the full dimension d4 required for informational completeness, deterministic operations alone are insufficient for complete reconstruction of a multi-time process.

Therefore, an informationally complete probe set must necessarily include non-deterministic operations. This is one of the main obstacles to experimental characterisation of multi-time processes. In practice, implementing such operations typically requires mid-circuit measurement and conditional re-preparation, which are substantially more demanding than gate-only interventions. As a result, most existing experiments have been limited to partial characterisation of multi-time processes.

Full multi-time tomography has only recently been demonstrated experimentally on superconducting devices, although the mid-circuit measurement and re-preparation steps contribute additional overhead and can introduce further errors through idling and measurement-induced crosstalk. It is therefore highly desirable to realise informationally complete probe sets using alternatives to mid-circuit measurement. A standard example of a non-deterministic operation is a measure-and-prepare map, described in the Choi representation by the operator |a⟩⟨a’T ⊗|ψ⟩⟨ψ|, where |a⟩is a state vector specifying the rank-one measurement effect on the input space, |ψ⟩is the prepared output state, and the transpose is taken with respect to the fixed computational basis.

Such operations are sufficient to form an informationally complete set, since their linear span covers the full operator space. The ability to fully map quantum processes is now within reach, promising advances in tackling the pervasive problem of noise that plagues quantum systems and hinders their development. However, this breakthrough isn’t without its caveats; the paper acknowledges that having enough information to describe a process doesn’t guarantee the most efficient way to gather it. Still, acknowledging that efficient data collection isn’t guaranteed simply because it’s possible doesn’t diminish this achievement.

Researchers have demonstrated a way to fully map complex quantum processes using only a single qubit of ancillary memory; a significant reduction in required resources. This avoids complicated mid-circuit measurements and resets, common limitations in current quantum devices, streamlining the process of ‘quantum process tomography’, which is essential for identifying and mitigating errors. Researchers have shown that fully mapping complex quantum processes requires fewer resources than previously thought.

This advance establishes a fundamental limit on the ancillary resources needed for detailed quantum process characterisation; previously, increasing the number of experimental steps demanded ever more supporting quantum bits. By demonstrating complete multi-time process tomography with just one coherently maintained qubit, researchers have bypassed a significant obstacle to scaling up quantum experiments, removing reliance on complex mid-circuit measurements and resets. This simplification unlocks new possibilities for analysing temporally correlated noise, a key impediment to building stable quantum technologies, and designing strategies to mitigate its effects.

Researchers demonstrated that complex quantum processes can be fully mapped using only one qubit of ancillary memory. This represents a reduction in the resources needed for ‘quantum process tomography’, a technique used to identify and mitigate errors in quantum systems. The study avoids the need for complicated mid-circuit measurements and resets, simplifying the process and enabling more efficient characterisation of multi-time dynamics. This advance establishes a fundamental limit on ancillary resources and facilitates analysis of noise that hinders the development of stable quantum technologies.