Stabilizer entropy is trustworthy for mixed states

Gianluca Esposito, Michele Viscardi, Alioscia Hamma · 2026

Read Stabilizer entropy is trustworthy for mixed states by Gianluca Esposito, Michele Viscardi, Alioscia Hamma (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2606-29443.txt.

Introduction

Esposito, Viscardi, and Hamma examine stabilizer entropy, a measure of how far a quantum state sits from the easy to simulate stabilizer family. That distance, called non-stabilizerness or magic, is a resource for quantum computing. For mixed states, states that blend several possibilities, computing a strict measure is provably intractable. Any exact monotone, a measure that never grows under free operations, needs more than exponential time.

The authors defend a simpler linear version of stabilizer entropy. They show it behaves as a proper monotone with overwhelming probability for flat mixed stabilizer states under non-adaptive Clifford channels. The chance of a violation shrinks exponentially as the system grows. Tests cover Haar-random states, Clifford orbits, and random matrix product states. They also prove safety for certain many-body systems under partial measurements. Their verdict: the linear measure is trustworthy in practice.

Catalog record

CREATORGianluca Esposito, Michele Viscardi, Alioscia Hamma
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE29 KB
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2606.29443

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Stabilizer entropy is trustworthy for mixed states Gianluca Esposito, Michele Viscardi, Alioscia Hamma Abstract Quantifying non-stabilizerness in mixed states is provably intractable, as any strict monotone requires superexponential time. We propose a linear Stabilizer Entropy that acts as a proper non-stabilizerness monotone with overwhelming probability when restricted to non-adaptive Clifford channels acting on flat mixed stabilizer states. Analytical and numerical results for Haar-random states, Clifford orbits, and random matrix product states show that monotonicity violation probabilities decay as exp − η ​ N \exp-\eta N . We also prove the validity of Stabilizer Entropy in specific many-body systems undergoing partial measurements, where the amount of resource never increases for each measurement outcome as well as when averaged over outcome probabilities. Given the hardness of strict alternatives, Stabilizer Entropy emerges as a practical and theoretically justified resource measure. Introduction — In recent years, non-stabilizerness has established itself as a necessary ingredient, together with entanglement, to achieve quantum advantage and to uncover the true nature of quantum systems, from exposing the breaking of efficient classical simulation algorithms Gottesman 1998 ; Aaronson and Gottesman 2004 ; Gu et al. 2024 to signaling onset of chaotic behavior both in quantum circuits Leone et al. 2021a ; Magni and Turkeshi 2025 and in many-body systems Odavić et al. 2022 ; Jasser et al. 2025 ; Odavić et al. 2025 ; Tirrito et al. 2025 ; Turkeshi et al. 2025 . Many non-stabilizerness monotones have emerged in the literature Veitch et al. 2014 ; Leone et al. 2022 ; Howard and Campbell 2017 ; Haug and Kim 2023 , with the aim of tracking the behavior of resource during a quantum computation or a physical process, and to obtain bounds on the extractable resource from a given quantum state. However, a recent work proves computational hardness of quantifying non-stabilizerness in the realm of mixed states Leone et al. 2026 , stating that any strict non-stabilizerness monotone for mixed states lies in the complexity class \rm QP^{2} , meaning that its cost scales superexponentially in the number of qubits. In this scenario, a practical non-stabilizerness measure seems unfeasible if applied to realistic physical applications, both experimental and theoretical, where many physical processes are modeled through maps that do not preserve state purity. This situation is akin as the quantum separability problem for entanglement in mixed states, which is also computationally hard Gharibian 2010 ; Gurvits 2003 : in that field the hurdle has been partially overcome by the creation of entanglement witnesses Peres 1996 ; Eisert and Plenio 1999 ; Vidal and Werner 2002 ; Plenio 2005 ; Życzkowski et al. 1998 , which sacrifice faithfulness to obtain efficient computability, and more generic frameworks for resource witnesses are also being pursued Tang et al. 2026 ; Bermejo et al. 2025 . A different, more general question that can be asked for mixed states is whether they admit a purification that is a stabilizer state, in other words, what is the non-stabilizerness of purification of a mixed state. It is well-known that generic convex combinations of pure stabilizer states, referred to as the stabilizer polytope, do not admit a stabilizer purification despite having no distillable non-stabilizer resources. The only mixtures of stabilizer states that do admit a stabilizer purification are those whose Pauli tomography weights are flat, and we dub this set as {\rm STAB_{0}} with an explicit definition in the following section. The computational problem of quantifying non-stabilizerness of purification has not (to our knowledge) been fully studied, but it is likely to be hard, akin to it counterpart in entanglement theory, namely entanglement of purification Terhal et al. 2002 ; Nguyen et al. 2018 ; Leone et al. 2025 . The unresolved nature of this question led us to investigate whether it exists a measure of non-stabilizerness of purification that is efficiently computable. Stabilizer Entropy (SE) Leone et al. 2022 appeared as the perfect candidate: computable Haug and Piroli 2023a ; Huang et al. 2026 , faithful and experimentally accessible Oliviero et al. 2022 ; Ahmad et al. 2025 . For pure states, as expected, SE coincides with distillable non-stabilizerness, but its naive extension to mixed states, despite being faithful on the set of {\rm STAB_{0}} , leads to violations of monotonicity under stabilizer operations, as correctly pointed out in Haug and Piroli 2023b , and a proper mixed-state extension of SE, presented in Leone and Bittel 2024 , leads again to a computationally hard quantity. In this paper, we present computable extensions of Stabilizer Entropy to mixed states, and provide analytical and numerical guarantees on its monotonicity under free operations of non-stabilizerness resource theory, namely partial trace and non-post-selected measurement of the system in the computational basis. We tested the validity of our claims in three relevant state ensembles: Haar random states, states with fixed non-stabilizerness, and random Matrix Product States with fixed bond dimension \chi . Across the state ensembles, the message is clear: violation of monotonicity of linear Stabilizer Entropy are exponentially rare, i.e. linear SE behaves as a true mixed-state non-stabilizerness monotone with overwhelming probability. We introduce the formalism of the resource theory of non-adaptive stabilizer computation, and our definition of linear Stabilizer Entropy for mixed states. We also define the concept of \eta- resource proxy as a resource quantifier whose probability of monotonicity violations decays as \exp(-\eta N) . We then compute the monotonicity violation probability for SE in Haar random states, in the Clifford orbit of fixed-non-stabilizerness states, and in random MPS with a fixed bond dimension. Moreover, we numerically analyze the efficacy of logarithmic stabilizer entropy in the ground space of the XY model. We ascertain that, upon partial measurement of the system in the computational basis, SE averaged over outcome probabilities never increases with respect to its values before measurement. While this is to be expected for linear SE Leone and Bittel 2024 , it holds for its logarithmic mixed state Leone et al. 2022 counterpart as well across the XY model phase diagram. This work paves the way for the use of Stabilizer Entropy to reliably and efficiently quantify non-stabilizer resources in realistic physical settings, from many-body systems where access to probe non-stabilizerness properties of subsystems is needed Frau et al. 2024 ; Viscardi et al. 2026 ; Iannotti et al. 2025 ; Tirrito et al. 2023 ; Tarabunga and Haug 2025 ; Iannotti et al. 2026 , to noisy quantum information processes involving measurements. Resource proxies and linear stabilizer entropy —Quantum Resource Theories (QRT) are a useful conceptual tool for quantifying cost of quantum operations Chitambar and Gour 2019 ; Gour 2024 : the main idea consists in categorizing quantum states and operations as either ”free” or ”costly”, motivated on operational constraints or the laws of physics: the set of free states and operations uniquely characterize the QRT. Along with a resource theory, a resource monotone is also needed, namely a function f:\mathcal{D}(\mathcal{H})\rightarrow\mathbb{R}^{+} that quantifies the amount of the resource contained in a specific quantum state. Such a function needs to be i) faithful, namely f(\rho)=0\iff\rho is a free state, and ii) monotone under free operations, namely f(\mathcal{E}(\rho))\leq f(\rho) whenever \mathcal{E} is a free operation in the resource theory. In this work we relax the condition

Read the complete text