⚛️ quantum physics

Theory-independent monitoring of the decoherence of a superconducting qubit with generalized contextuality

This paper presents a theory-independent method to monitor a superconducting qubit, demonstrating that it transitions from a nonclassical, contextuality-rich state to a classical, noncontextual one over time while undergoing non-Markovian evolution, all without assuming quantum theory or trusting the devices.

Original authors: Albert Aloy, Matteo Fadel, Thomas D. Galley, Caroline L. Jones, Markus P. Mueller

Published 2026-03-18

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Albert Aloy, Matteo Fadel, Thomas D. Galley, Caroline L. Jones, Markus P. Mueller

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 you have a mysterious black box. You put things in, wait a moment, and take things out. You want to know: Is this box truly magical (quantum), or is it just a very clever trick (classical)?

Usually, to answer this, scientists say, "Okay, let's assume the laws of quantum physics are true, and then check if the box behaves that way." But what if we don't want to assume anything? What if we want to prove the box is magical without relying on the rulebook of quantum mechanics?

This paper does exactly that. The researchers took a superconducting qubit (a tiny, artificial atom made of metal that acts like a quantum bit) and watched it "die" (decohere) over time. They did this using a method called Theory-Independent Tomography.

Here is the story of what they found, explained with simple analogies.

1. The "Blind" Detective (Theory-Independent Analysis)

Imagine you are a detective trying to figure out what a new video game console looks like inside, but you aren't allowed to open the box or read the manual. You can only press buttons (preparations) and watch the screen (measurements).

Most scientists would say, "I bet it's a Nintendo Switch," and then check if the buttons match. These researchers said, "No, we won't guess. We will just map out every possible button press and screen result to build a 3D map of what the machine actually does."

They call this a Generalized Probabilistic Theory (GPT). Think of it as a shape-shifting clay model.

If the machine is a Quantum Bit (Qubit), the clay shape is a perfect ball (the famous Bloch sphere).
If the machine is a Classical Bit (like a light switch), the clay shape is a simple line.
If it's something weird, the clay could be a cube, a pyramid, or a blob.

By pressing their buttons 100 times in different combinations, they sculpted this clay model based only on the data they saw. They didn't assume it was a ball; they let the data shape the ball for them.

2. The "Magic" of Contextuality (The Hidden Trap)

The researchers were looking for a specific kind of "magic" called Contextuality.

The Analogy: Imagine you have a deck of cards.

Classical World: If you shuffle the deck and deal a card, the card has a fixed identity. It's an Ace of Spades whether you look at it with your left eye or your right eye. The "context" (how you look at it) doesn't change the card.
Quantum World (Contextual): Imagine the card changes its identity depending on which other cards you are looking at alongside it. If you look at it with a King, it acts like a Queen. If you look at it with a Jack, it acts like a 2. The card doesn't have a single, hidden "true self" that exists independently of how you measure it.

The researchers showed that at the very beginning (Time = 0), their superconducting qubit was Contextual. It was behaving like a true quantum system where the "card" changed based on the context. It could not be explained by a simple, hidden list of instructions (a "hidden variable model").

3. The Slow Death (Decoherence)

Quantum systems are fragile. They lose their "magic" when they interact with the noisy environment. This is called decoherence.

The researchers watched their clay model (the state space) over time:

At the start: The model was a big, bumpy ball. It was full of "quantum magic" (contextuality).
As time passed: The ball started to shrink and flatten. It was losing its volume.
The Result: Eventually, the ball shrank so much that it lost its "magic." It became Non-Contextual. It turned into a boring, classical object that could be explained by a simple list of hidden instructions.

The Big Discovery: They proved that the system literally lost its quantum nature and became classical, and they proved this without ever assuming quantum physics was true. They just watched the shape of the data change.

4. The Ghost in the Machine (Non-Markovianity)

Here is the twist. Usually, when something decays, it just gets smaller and smaller, like a melting ice cube. It never gets bigger again.

But the researchers noticed something weird between 20 and 30 microseconds. The clay model expanded slightly. It got bigger for a moment before shrinking again.

The Analogy: Imagine you are walking away from a campfire. You expect to get colder and colder. But suddenly, for a few seconds, you feel a warm breeze hit you from behind, making you warmer than you were a moment ago, before you get cold again.

In physics, this is called Non-Markovianity. It means information (heat) flowed back from the environment into the system. The system "remembered" its past and briefly regained some of its lost complexity.

The researchers detected this "breathing" of the system purely by looking at the shape of their data map, again without needing to assume the laws of quantum mechanics.

Why Does This Matter?

This is a huge step forward for science.

No Crutches: Usually, to prove something is quantum, you need to assume quantum theory is correct. This is like proving a car is fast by assuming the speedometer works. These researchers proved the car is fast by measuring the distance and time directly, without trusting the speedometer.
Future-Proof: If we discover a new theory of physics tomorrow that replaces quantum mechanics, these results will still be true. They didn't rely on the old rules; they relied on the raw data.
Certification: This gives us a new way to check if quantum computers are actually working as quantum machines, even if we don't fully understand the theory behind them yet.

In a nutshell: The team built a "blind" map of a quantum computer, watched it slowly turn into a classical object, and spotted a moment where it briefly "inhaled" information from the environment—all without ever opening the instruction manual of the universe.

1. Problem Statement

Characterizing the nonclassicality of quantum systems typically relies on assumptions inherent to quantum theory (e.g., that measurements are noiseless projective measurements, outcomes depend deterministically on hidden variables, or that the system is strictly a qubit). These assumptions limit the ability to experimentally certify nonclassicality in a "device-independent" or "theory-independent" manner.

The authors address the challenge of monitoring the decoherence of a physical system and its transition from a nonclassical to a classical state without assuming the validity of quantum theory. Specifically, they aim to:

Determine the dimensionality and geometry of the system's state space directly from data.
Monitor the loss of generalized contextuality (a robust form of nonclassicality) over time.
Detect non-Markovian evolution (information backflow) without relying on quantum dynamical maps.

2. Methodology: Theory-Agnostic Tomography

The authors employ Generalized Probabilistic Theories (GPTs) as a framework to describe the system. In this framework, a system is defined by a convex set of states ( $S$ ) and a convex set of effects ( $E$ ), without assuming they correspond to density matrices and POVMs.

Experimental Setup:

System: A transmon superconducting qubit ( $\omega_q \approx 5.05$ GHz) coupled to a 3D readout cavity.
Procedure: A "prepare-transform-measure" scenario.
1. Preparation: $m=100$ distinct preparation procedures ( $P_i$ ) are generated using Rabi pulses with amplitudes/phases chosen via a Fibonacci distribution to uniformly cover the Bloch sphere surface.
2. Evolution: The system evolves for a variable waiting time $\tau \in \{0, 5, ..., 50\} \, \mu\text{s}$ .
3. Measurement: $n=100$ measurement bases ( $M_j$ ) are applied, followed by a standard circuit-QED readout in the $Z$ -basis.
4. Data: For each $(P_i, M_j)$ pair, the experiment is repeated 2000 times to estimate frequencies $F_{ij}$ (approximating probabilities $p(0|P_i, M_j)$ ).

Data Analysis (Theory-Agnostic Tomography):

Rank Determination: The authors construct a frequency matrix $F$ $F$ and seek the best-fit GPT model of rank $k$ $k$ . They minimize a weighted $\chi^2$ $χ^{2}$ statistic:
$\min_{D} \sum_{ij} \frac{(F_{ij} - D_{ij})^2}{(\Delta F_{ij})^2} \quad \text{subject to } \text{rank}(D)=k, \, 0 \le D_{ij} \le 1$
where $D = SE$ (states $\times$ $\times$ effects).
- They use a training/test split (10 frequency tables) to detect overfitting. The optimal rank is identified where the test error is minimized.
State/Effect Reconstruction: Once the rank $k$ is determined, they factorize the data to reconstruct the polytope of realized states ( $S$ ) and effects ( $E$ ). They also compute the consistent spaces ( $S_{consistent}, E_{consistent}$ ), which represent the maximal sets of states/effects compatible with the observed statistics (duals of the realized sets).
Contextuality Test: They use a linear programming approach to test for simplex embeddability. If the GPT state space can be embedded into a classical probability simplex, the system is noncontextual. If not, it is contextual. They quantify this via a robustness coefficient $r$ (amount of depolarizing noise needed to make the system noncontextual).
Non-Markovianity Test: They calculate the volume of the realized state space $Vol(S_\tau)$ over time. In a Markovian process, the state space volume must be non-increasing (volume contraction). An increase in volume indicates non-Markovian evolution (information backflow).

3. Key Contributions

Theory-Independent Process Tomography: The first extension of theory-agnostic tomography to time-evolving systems, allowing the monitoring of decoherence and dynamical properties without assuming quantum mechanics.
Direct Observation of Contextuality Loss: Demonstrating that a superconducting qubit transitions from a contextual (nonclassical) regime to a noncontextual (classical) regime solely due to decoherence, verified without quantum assumptions.
Model-Free Non-Markovianity Detection: Providing a theory-independent witness for non-Markovian dynamics based on the expansion of the state space volume.

4. Key Results

System Dimensionality: The data analysis reveals that the optimal rank for the GPT model is $k=4$ . This corresponds to a 3-dimensional normalized state space (since the first component is normalization), which matches the geometry of a qubit's Bloch ball. The reconstructed state space is a "bumpy" polytope close to a sphere.
Decoherence and State Space Contraction:
- For $\tau = 0$ , the state space is large and contextual.
- As $\tau$ increases, the state space contracts towards a single point (the ground state), indicating irreversible decoherence.
- The system loses its ability to be described by a noncontextual hidden-variable model.
Loss of Generalized Contextuality:
- At $\tau = 0, 5, 10 \, \mu\text{s}$ , the system is contextual ( $r > 0$ ).
- Between $\tau = 10$ and $15 \, \mu\text{s}$ , the robustness coefficient drops to $r=0$ .
- For $\tau \ge 15 \, \mu\text{s}$ , the system is noncontextual (effectively classical). This marks the precise time window where quantum nonclassicality is lost due to environmental noise.
Non-Markovian Evolution:
- The volume of the state space generally decreases (Markovian).
- However, between $\tau = 20 \, \mu\text{s}$ and $\tau = 30 \, \mu\text{s}$ , the relative volume of the state space increases.
- This expansion is a signature of non-Markovianity, attributed to residual off-resonant coupling with higher energy levels or other degrees of freedom in the cavity.
Purity Bounds: Assuming quantum theory holds for interpretation, the analysis implies that the initial prepared states have a maximum eigenvalue $\lambda_{max} \ge 0.846$ , confirming they are close to pure states.

5. Significance

Foundational Robustness: The results hold regardless of whether quantum theory is the ultimate description of nature. Even if quantum theory were replaced by a new theory, the observation that the system's state space contracts and loses contextuality would remain valid, provided the assumption of tomographic completeness (that the chosen preparations and measurements are sufficient to characterize the subsystem) holds.
Practical Certification: This method offers a rigorous way to certify the "quantumness" of a device and monitor its degradation (decoherence) without trusting the internal workings of the device or assuming a specific theoretical model.
Dynamical Insights: It demonstrates that non-Markovian effects (often considered noise) can be detected and quantified in a model-independent way, offering new tools for characterizing open quantum systems.
Bridge to Classicality: The experiment provides a clear, data-driven timeline of how a quantum system "becomes classical" (loses contextuality) as it interacts with its environment, bridging the gap between quantum foundations and practical quantum engineering.

In summary, the paper successfully applies generalized probabilistic theories to a superconducting qubit to demonstrate that decoherence leads to the loss of generalized contextuality and that non-Markovian dynamics can be detected purely through statistical analysis of preparation and measurement outcomes, independent of quantum mechanical assumptions.