Contractivity of the Hilbert--Schmidt Speed in Unital Quantum Channels: Foundation for Witnessing Non-Markovianity and Discriminating Unital from Non-Unital Markovian Dynamics
A new theoretical result by Hossein Rangani Jahromi establishes precisely when a popular geometric diagnostic for quantum memory effects is trustworthy — and when it is not. The findings draw a sharp …
What Is the Hilbert–Schmidt Speed and Why Do Researchers Use It?
The Hilbert–Schmidt speed (HSS) is a geometric quantity that measures how quickly a quantum state changes as some physical parameter — an external field strength, a coupling constant — is varied. It is computed from the Hilbert–Schmidt norm of the tangent vector along a family of quantum states parameterized by that quantity. Because it requires no full quantum tomography and is computationally accessible, the HSS has attracted attention as a practical probe of quantum dynamics.
One of its headline applications has been detecting non-Markovianity — the presence of memory effects in open quantum systems. When a quantum system interacts with an environment, information can flow irreversibly outward (Markovian behavior) or backflow into the system (non-Markovian behavior). Researchers have used a rise in the HSS over time as a signal that such backflow is occurring, making it a candidate witness for quantum memory. The theoretical basis for exactly when this signal is reliable, however, had remained incomplete.
What Did Jahromi Actually Prove?
Working within the framework of finite-dimensional, parameter-independent completely positive trace-preserving (CPTP) maps — the standard mathematical description of physical quantum channels — Jahromi proved that the HSS can never increase under any unital CPTP map. A unital map is one that preserves the maximally mixed state, the quantum analogue of a perfectly unbiased noise process. This contractivity result is exact and general.
From this single theorem, two further results follow. First, for any CP-divisible evolution — the formal definition of Markovian dynamics — whose intermediate propagators are all unital, the HSS is guaranteed to decrease monotonically in time. Second, Jahromi derived the generator-level version of this statement for dynamics governed by the Gorini–Kossakowski–Sudarshan–Lindblad (GKSL) master equation, the workhorse equation of open quantum systems, when its Lindblad operators are Hermitian. In that setting, he produced an explicit, provably non-positive expression for the time derivative of the squared HSS, cementing the monotonic decay at the level of the differential equation itself.
Why Does Unitality Turn Out to Be the Crucial Ingredient?
Unitality is the structural property that makes HSS-based diagnostics rigorous. Think of it like a one-way valve: in a unital Markovian channel, geometric distinguishability between quantum states encoded in the parameter can only erode over time, never rebuild. Any observed increase in the HSS therefore signals a genuine violation of that erosion — a memory effect leaking information back into the system.
The result means that experimentalists and theorists working with physical noise models known to be unital — depolarizing channels and dephasing channels are common examples — now have a mathematically airtight justification for interpreting HSS backflow as evidence of non-Markovianity. The witness goes from being a well-motivated heuristic to a provably sufficient diagnostic within that class of dynamics.
What Is the Counterexample, and Why Does It Matter?
Jahromi constructed an explicit qutrit (three-level quantum system) example in which the HSS increases over time even though the dynamics is perfectly Markovian — that is, completely free of memory effects. The key feature of this counterexample is that the channel is non-unital: it does not preserve the maximally mixed state. This single example is enough to prove that HSS non-monotonicity is not a faithful indicator of non-Markovianity in general.
This is a significant clarification rather than a negative result. Prior applications of HSS as a non-Markovianity witness had implicitly assumed conditions that were never formally established. The qutrit counterexample shows that without confirmed unitality, an observed HSS increase could be an artifact of the non-unital character of the noise rather than genuine quantum memory. Misidentifying Markovian non-unital dynamics as non-Markovian would distort both fundamental studies and practical protocols that rely on accurate noise characterization.
What Are the Limitations and Open Questions?
The results apply to finite-dimensional systems with parameter-independent CPTP evolution, where the parameter enters only through the initial state. Extending the contractivity theorem to infinite-dimensional systems, or to parameter-dependent channels where the channel itself depends on the probed quantity, remains an open problem. Additionally, while Hermitian Lindblad operators cover an important class of physical models, GKSL equations with non-Hermitian Lindblad operators — which describe amplitude damping, for instance — are not covered by the generator-level theorem derived here.
A deeper open question concerns whether a similarly sharp structural characterization exists for other geometric speed indicators beyond the HSS, such as those built from the quantum Fisher information or the Bures metric. Unitality may play an equally decisive role there, but that connection has yet to be systematically explored.
What Comes Next for HSS-Based Quantum Diagnostics?
With a rigorous foundation now established, the immediate priority for the field is applying this framework to experimental platforms — superconducting qubits, trapped ions, photonic systems — where the noise model can be independently verified to be unital, enabling HSS measurements to serve as certified, resource-efficient witnesses of non-Markovian behavior without full process tomography.
As quantum devices grow in scale and complexity, having diagnostics whose validity conditions are precisely understood will be essential for distinguishing true environmental memory effects from artifacts of imperfect noise modeling.