🔬 physics

Tomography-free quantum diagnostics of dissipative quantum chaos

This paper introduces a tomography-free diagnostic framework based on the propagator-norm form factor $\Phi(t;\theta)$ , which distinguishes chaotic from integrable dissipative quantum dynamics by analyzing profile shapes and excess magnitudes beyond simple spectral signatures, while linking these theoretical findings to practical measurement protocols via Choi-state purity and block-encoded propagators.

Original authors: Hikaru Wakaura, Taiki Tanimae

Published 2026-10-28

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Hikaru Wakaura, Taiki Tanimae

Original paper licensed under CC BY 4.0 (https://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

In the quantum world, the difference between order and chaos is not just a matter of curiosity; it determines how a system behaves, how it stores information, and whether it can be controlled. For systems that are completely isolated, scientists have long known how to spot chaos by looking at the spacing between their energy levels, much like listening for a specific rhythm in a drumbeat. However, the real world is rarely isolated. Most quantum systems interact with their environment, losing energy and information in a process called dissipation. When this happens, the mathematical tools used to describe the system change, becoming "non-Hermitian," a term that simply means the rules of symmetry that usually keep things tidy no longer apply. In these messy, open environments, chaos still exists, but the old ways of detecting it often fail. The challenge has been to find a new way to tell if a dissipative system is truly chaotic or just orderly, without needing to map out every single detail of its internal state, a task that is currently impossible for large systems.

A team of researchers has now developed a new method to diagnose this kind of chaos without needing to perform a full reconstruction of the system's state. Instead of trying to measure every tiny component, they focus on a specific quantity that tracks how the system evolves over time. Imagine watching a drop of ink spread through water. In a chaotic system, the ink spreads in a complex, unpredictable way that is deeply tied to how the different parts of the system interact. The researchers created a mathematical tool that measures the "size" of this spreading process. They found that for chaotic systems, this size grows in a very specific, predictable pattern over time, eventually settling into a steady, linear increase. In contrast, systems that are orderly or only partially chaotic show a different growth pattern, one that curves and flattens out differently. This difference in the shape of the growth curve acts as a fingerprint, allowing scientists to distinguish between chaos and order.

The researchers tested this idea using computer simulations of various quantum models, including some that are known to be chaotic and others that are known to be orderly. They found that the method works remarkably well when the data is sufficiently precise. In a test on 45 fresh instances, the method made 38 correct assignments and abstained from guessing on 7 cases, with no wrong-class assignments in that specific set. However, the method is not infallible; its accuracy depends heavily on the amount of data collected. For example, with a lower number of samples (10^5), the method produced a 19% wrong-class rate for certain orderly systems, whereas increasing the samples to 10^6 reduced this error to 2%, with the rest of the cases resulting in abstentions rather than mistakes. The team also discovered that simply measuring how large the effect is at a single moment is not enough to tell the difference; a system can be large but still orderly. It is the shape of the curve over time that carries the true signature of chaos. This distinction is crucial because it means scientists can identify chaos even in systems that look complex and messy on the surface but are actually following a hidden, orderly script.

One of the most significant aspects of this work is that it avoids the need for "tomography," a process that would require measuring millions of parameters to reconstruct the system's full state. Instead, the new method relies on a simpler measurement that can be performed with fewer resources. The researchers showed that this measurement is linked to a property called "purity," which describes how mixed or clean a quantum state is. By measuring this purity in a specific way, they can infer the chaotic nature of the system. In their tests, the method correctly identified chaotic systems in nearly all cases where a decision was made, and when it was unsure, it simply abstained from guessing rather than making a mistake. This reliability suggests the method is robust, even when the data is slightly noisy, provided enough samples are taken to minimize errors.

However, the researchers are careful to note that this is not a magic bullet for every situation. They tested the method on systems that are structured in a specific way, like chains of particles with local interactions, and found that while the method could sometimes tell the difference, it struggled in cases where the system was very large or had a specific type of disorder. In these difficult cases, the growth curves of chaotic and orderly systems looked too similar to distinguish. This limitation highlights that while the method is powerful, it is not a universal solution for all types of quantum systems. The researchers also emphasized that their results are based on computer simulations and theoretical calculations; they have not yet been tested on actual quantum hardware.

The study also clarifies a common misconception: that a system being "non-normal" (meaning it lacks certain symmetries) automatically means it is chaotic. The researchers showed that this is not true. A system can be non-normal and still be orderly, and the new method is designed to catch this distinction. By focusing on the shape of the growth curve rather than just the size of the effect, the method separates true chaos from mere complexity. This insight helps refine our understanding of how quantum systems behave in the real world, where they are constantly interacting with their surroundings.

Looking ahead, the researchers see this work as a stepping stone toward more advanced diagnostics. They plan to refine the method to handle larger and more complex systems, and they aim to test it on actual quantum computers to see how it performs in a real experimental setting. The ultimate goal is to have a reliable, efficient way to monitor quantum systems as they evolve, ensuring that they remain under control or behave as expected. For now, the discovery offers a clear, new lens through which to view the hidden dynamics of the quantum world, separating the true chaos from the noise.

Technical Summary: Tomography-free Quantum Diagnostics of Dissipative Quantum Chaos

Problem Statement
The distinction between chaotic and integrable dynamics in open (dissipative) quantum systems is a central challenge in quantum many-body physics. While isolated systems are diagnosed via spectral signatures like level repulsion and the spectral form factor (SFF), open systems are governed by non-Hermitian generators (e.g., Liouvillians). Existing diagnostics for dissipative chaos rely on complex spacing ratios (CSR) and Chalker–Mehlig (CM) overlaps, which quantify eigenvector non-orthogonality. However, experimentally accessing these signatures typically requires full process tomography, scaling exponentially with system size ( $O(16^n)$ parameters), or classical diagonalization of reconstructed maps. Furthermore, recent findings indicate that singular-value statistics of non-Hermitian generators (e.g., in Sachdev–Ye–Kitaev models) fail to distinguish chaotic from integrable dynamics, creating a need for new observables that capture the interplay of spectrum and eigenvectors without requiring full state reconstruction.

Methodology
The authors introduce a "tomography-free" diagnostic based on the propagator-norm form factor, defined as:
$\Phi(t; \theta) = \frac{1}{D} \| e^{t e^{i\theta} L} \|_F^2$
where $L$ is the non-Hermitian generator, $D$ is the dimension of the Liouville space, and $\theta$ rotates the complex plane. The core observable is the excess $X(t; \theta) = \Phi(t; \theta) - \Phi_0(t; \theta)$ , where $\Phi_0$ is a baseline determined solely by the eigenvalues of $L$ .

The paper establishes three analytic frameworks to interpret this quantity:

Short-time behavior: A trace expansion relates the initial quadratic growth of the excess to Henrici's departure from normality (the squared Frobenius norm difference between the operator and its eigenvalues).
Random Matrix Theory (RMT) limits: The authors derive exact curves for the complex Ginibre ensemble and the elliptic Ginibre ensemble. In the large-dimension limit, the excess over the baseline exhibits an asymptotically linear ramp.
Finite-size approximation: A planar-order approximation connects these RMT results to finite-size non-Hermitian SYK models, validating the theoretical curves against numerical data.

Measurement Protocols
The paper identifies two measurement routes for $\Phi$ that avoid full tomography:

Block Encoding: If a unitary block-encodes the propagator $M = e^{t e^{i\theta} L}$ , the acceptance probability on a maximally mixed input is proportional to $\Phi$ .
Choi Purity: For physical channels ( $\theta=0$ ), $\Phi$ corresponds to the normalized purity of the Choi state, accessible via a two-copy SWAP test.
The authors emphasize that while $\Phi$ is measurable, the subtraction of the baseline $\Phi_0$ requires accurate spectral information, which remains a separate computational challenge.

Key Results

Analytic Curves: The paper derives the "Ginibre curve" $X/\Phi_0 = t I_0(2t)/I_1(2t) - 1$ , showing that chaotic dynamics in the Ginibre class follow an asymptotically linear ramp. The elliptic ensemble solution provides a continuous interpolation between the Ginibre ramp and the zero-excess normal limit.
Classification Performance: Using a decision rule based on three features—magnitude ( $\chi$ $χ$ ), late-time log-log slope ( $s^*$ $s^{*}$ ), and angular anisotropy ( $A$ $A$ )—the authors tested 45 fresh instances across nine ensembles.
- The rule correctly classified 38 instances and abstained on 7 (instances falling in a "gray zone" between chaotic and integrable thresholds).
- Crucially, there were zero wrong-class assignments within the tested scope.
- The shape of the profile (specifically the slope $s^*$ ) proved more discriminative than the magnitude $\chi$ , which failed to separate integrable non-normal systems (like $q=2$ SYK) from chaotic ones.
Limitations: The method faces a "resolution barrier." While it accesses eigenvector signatures at $O(1)$ scales, it cannot resolve fine-grained spectral statistics (like CSR) that scale as $D^{-1/2}$ . Additionally, in structured locally dissipative chains (e.g., Hatano–Nelson or dephasing chains), spectral-spacing signatures may separate while profile slopes do not, indicating the method is not a universal classifier for all structured systems.

Significance and Claims
The paper claims to provide a framework that connects non-normal dynamics, random matrix theory, and quantum measurement without equating non-normality directly with chaos. Its primary contribution is demonstrating that the profile shape of the propagator-norm form factor complements magnitude in distinguishing chaotic from integrable ensembles.

The authors are explicit about the scope of their claims:

They distinguish their analytic results and classical sampling tests from unimplemented gate-level synthesis and unresolved large-system baseline reconstruction.
They do not claim to have built a universal chaos classifier; rather, they offer a diagnostic tool that works for specific ensembles (random Lindbladians, SYK models) but has known limitations in structured, quasi-1D systems.
The work establishes that while $\Phi$ is accessible without tomography, an efficient end-to-end classifier requires solving the separate problem of accurate baseline reconstruction ( $\Phi_0$ ), which remains an open challenge for large systems.

In summary, the paper proposes a measurable, tomography-free observable that captures the "eigenvector signature" of dissipative chaos, validated by analytic theory and numerical benchmarks, while carefully delineating the boundaries of its applicability and the remaining hurdles for experimental implementation.

Technical Summary: Tomography-free Quantum Diagnostics of Dissipative Quantum Chaos

More like this