Pith. sign in

REVIEW 3 major objections 4 minor 51 references

The paper proves that the fermionic entropy, defined from the squared Frobenius norm of the correlation matrix, is a strong pure-state monotone for fermionic non-Gaussianity and is estimable with O(ε⁻²) two-copy measurements, independent of

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-03 02:09 UTC pith:MCINS3D2

load-bearing objection A solid resource-theory paper whose central theorem relies on a black-box transfer from a prior same-author result; the proof is fixable but not self-contained. the 3 major comments →

arxiv 2607.29670 v1 pith:MCINS3D2 submitted 2026-07-31 quant-ph

Fermionic entropy: an efficiently measurable strong monotone for non-Gaussianity

Lorenzo Leone , Lennart Bittel This is my paper
classification quant-ph MSC 81P4081P4581P68 PACS 03.67.-a03.67.Mn
keywords fermionic non-Gaussianityfermionic entropycorrelation matrixstrong Gaussian monotoneresource theorymatchgate circuitsunitary designstolerant testing
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper establishes that a single number computed from a pure state's correlation matrix — the fermionic entropy Mf = n(1 - ||Γ||²/2n) — is a strong monotone: no Gaussian protocol can increase it, even on average across the branches of a probabilistic protocol. The same closed form makes it directly measurable: its fermionic purity can be unbiasedly estimated to error ε with O(ε⁻²) two-copy Bell measurements, with no dependence on the number of qubits. This combination of rigorous monotonicity and practical measurability gives the measure operational teeth: it upper-bounds the asymptotic rate of non-Gaussianity distillation. As an application, the paper proves that t-doped Matchgate circuits need a linear number of non-Gaussian gates before they can even form an approximate state 2-design, closing the gap between quadratic and linear lower bounds. If these claims hold, the fermionic entropy is one of the simplest resource measures that is both provably monotone and experimentally accessible.

Core claim

The central claim is that Mf(ψ) = n(1 - ||Γ(ψ)||₂²/(2n)) is a strong pure-state Gaussian monotone: for any pure state and any Gaussian protocol producing an ensemble {(pi, |φi⟩)}, Mf(ψ) ≥ Σ pi Mf(φi), with equality exactly on fermionic Gaussian states. The proof reduces any pure-state Gaussian protocol to an elementary two-branch decomposition and proves the inequality for that decomposition via a norm identity involving the branch correlation matrices. A new moment bound for the operator Λ = Σᵢ γᵢ⊗γᵢ — that tr(Λ²ʳρ⊗²) = O(nʳ) rather than the naive O(n²ʳ) — yields an unbiased estimator of the fermionic purity using O(ε⁻²) two-copy measurements, independent of n. From asymptotic continuity th

What carries the argument

The fermionic entropy Mf and its associated fermionic purity Pf = ||Γ(ψ)||₂²/(2n), where Γ is the antisymmetric correlation matrix of Majorana expectation values. The operator Λ = Σᵢ γᵢ⊗γᵢ satisfies tr(Λ²ρ⊗²) = 2Mf(ρ), and the crucial estimate is the moment bound tr(Λ²ʳρ⊗²) = O(nʳ), which replaces the naive O(n²ʳ) operator-norm bound and enables the system-size-independent sample complexity. The other load-bearing piece is the two-branch reduction lemma that rewrites the monotonicity claim for arbitrary Gaussian protocols into the inequality Mf(ψ) ≥ pMf(φ0) + (1-p)Mf(φ1).

Load-bearing premise

The proof of Theorem 1 treats as a black box a structural theorem asserting that every pure-state Gaussian protocol can be reduced to elementary two-branch decompositions of the form √p |0⟩|φ0⟩ + √(1-p)|1⟩|φ1⟩; the appendix says 'the remaining argument is identical' without spelling out the reduction for fermionic protocols, so Theorem 1 stands or falls with that reduction.

What would settle it

Compute the fermionic entropy before and after a concrete Gaussian protocol on a small number of qubits, for example a conditional rotation followed by postselection on two branches with non-orthogonal auxiliary states; if the average output entropy exceeds the input entropy for any such protocol, strong monotonicity fails. Alternatively, search numerically over t-doped Matchgate circuits with t < n/(8κ) for an ensemble whose second moment matches Haar within error 1/256; such a circuit would falsify Theorem 5.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Gaussian protocols cannot increase fermionic non-Gaussianity on average; the convex-roof extension is a strong monotone for arbitrary mixed states.
  • Fermionic purity can be estimated to additive error ε with O(ε⁻² log δ⁻¹) two-copy measurements, independent of n.
  • Asymptotic continuity yields the operational bound: any sequence of Gaussian protocols converting ψ⊗N to approximately φ⊗RN has limsup R ≤ Mf(ψ)/Mf(φ), so the entropy ratio caps the distillation rate.
  • Tolerant testing of fermionic Gaussian states is solvable with O(n ε_B⁻² log(n ε_B⁻²)) copies, a quadratic improvement over the previous O(n²) scaling.
  • t-doped Matchgate circuits with t < n/(8κ) cannot form an ε-approximate state 2-design for ε ≤ 2⁻⁸, so the optimal doping level is Θ(n) up to logarithmic factors.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same two-branch reduction strategy may extend to the whole family of fermionic antiflatness measures, which the paper leaves open; the norm-identity proof appears transferable.
  • The purity gap used for the 2-design lower bound relies on second moments; higher moments of the fermionic purity could yield analogous lower bounds for k-designs with k > 2 without substantially new ideas.
  • The O(ε⁻²) sample complexity for purity estimation hints that an O(1)-sample tolerant tester might be achievable if the moment bound can be sharpened or combined with a different observable.
  • If the imported structural reduction is valid for fermionic protocols, similar strong-monotonicity results should hold for other additive correlation-matrix measures, broadening the resource theory beyond the entropy.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. The paper introduces and studies the fermionic entropy M_f(ψ)=n(1-‖Γ(ψ)‖_2^2/(2n)) as a measure of fermionic non-Gaussianity. The main results are: (i) Theorem 1: M_f is a strong pure-state Gaussian monotone, with a convex-roof extension to mixed states; (ii) Theorem 2: the fermionic purity can be estimated with O(ε^{-2}) two-copy measurements, independent of the system size; (iii) Theorem 3: a Fannes-like continuity bound |M_f(ρ)-M_f(ρ')|≤2n‖ρ-ρ'‖_1; (iv) Corollary 1: the asymptotic rate of non-Gaussianity distillation is bounded by M_f(ψ)/M_f(φ); (v) Theorem 4: tolerant testing of fermionic Gaussian states with sample complexity O(n ε_B^{-2} log(n ε_B^{-2})); and (vi) Theorem 5: t<n/(8κ) doped matchgate circuits cannot form an ε-approximate state 2-design for ε≤2^{-8}, yielding an extensive lower bound on the doping level. The proof of Theorem 1 reduces the central inequality to Lemma 1 for a two-branch decomposition and delegates the general reduction to Theorem 3 of Ref. [21].

Significance. If correct, the paper provides the simplest correlation-matrix-based non-Gaussianity monotone that is both a strong Gaussian monotone and efficiently measurable, and it closes the open question on the optimal doping level for matchgate designs. The strengths are substantial: the measure is parameter-free and has a closed-form expression; the O(ε^{-2}) sample complexity is established through explicit moment bounds (Lemma 2); Theorem 5 gives a concrete and falsifiable quantitative prediction. The main weakness is that the proof of Theorem 1 is not self-contained and relies on an unproved transfer of a structural theorem from Ref. [21]. The Note added, which mentions an independent proof of Theorem 1 in v2 of Ref. [12], increases confidence in the result but does not replace the missing derivation in this manuscript.

major comments (3)
  1. [Appendix S.1 / Theorem 1] The proof of strong monotonicity is not self-contained. Lemma 1 only establishes the inequality (S1) for the specific two-branch decomposition |ψ⟩=√p|0⟩|φ0⟩+√(1-p)|1⟩|φ1⟩. The step from this inequality to arbitrary pure-state Gaussian protocols is delegated to Theorem 3 of Ref. [21]. Appendix S.1 asserts that "the lemma reduces any pure-state Gaussian protocol to the same elementary decomposition" but no such reduction is exhibited for fermionic Gaussian protocols. Since Gaussian protocols include matchgate unitaries, measurements, conditioned operations, and partial traces, it is not automatic that every branch ensemble can be brought to the single-qubit computational-basis form of Lemma 1. Without this reduction, Lemma 1 does not establish strong monotonicity, and Corollary 1 loses its foundation. Please either prove the reduction explicitly or state Theorem 3 of Ref. [21] in a form wh
  2. [Appendix S.1, Eq. (S30)] The displayed inequality in Eq. (S30) is invalid as typeset. From Eqs. (S28)-(S29), Γ12(φ'_0)-Γ12(φ'_1)= -2(x-y) and Γ21 = -Γ12, so ‖Γ'_0-Γ'_1‖_F^2 ≥ (Γ12 diff)^2 + (Γ21 diff)^2 = 8(x-y)^2. The expression (1/8)(|Γ12 diff|+|Γ21 diff|)^2 equals 2(x-y)^2, not (x-y)^2; if the right-hand side 1/4[2(x-y)]^2 is intended, the preceding factor should be 1/8(|Γ12 diff|+|Γ21 diff|)^2 with the sum equal to 4|x-y|. The desired bound (x-y)^2 follows directly from the two squared blocks, so the proof is locally repairable, but Eq. (S30) must be corrected.
  3. [Appendix S.4] The proof of Theorem 4 begins with the identity "1/n tr(Λ²ρ⊗2)=1-Pf(ρ)". This is inconsistent with Eq. (5), which gives tr(Λ²ρ⊗2)=2M_f(ρ)=2n(1-Pf(ρ)); the correct identity is tr(Λ²ρ⊗2)/(2n)=1-Pf(ρ), or equivalently (1/n)tr(Λ²ρ⊗2)=2(1-Pf(ρ)). The subsequent use of X=Λ²/(2n) suggests the intended estimator is the former, so the constants and the promise-gap condition are off by factors of two as written. The asymptotic scaling is unaffected, but the proof should be rewritten consistently.
minor comments (4)
  1. [Definition 2 and Theorems 2-4] M_f is defined only for pure states, but the paper repeatedly uses M_f(ρ) and P_f(ρ) for arbitrary mixed states. Please state explicitly that the same formula is extended to all density matrices.
  2. [Theorem 1] The mixed-state convex-roof extension fM_f is stated as a strong monotone, but no proof is given beyond the pure-state sketch. Even if this follows by standard convex-roof arguments, the claim should be justified or marked as an open point.
  3. [Eq. (5)] The notation Λ := Σ_i γ_i^{⊗2} is ambiguous; it should be written as Σ_i γ_i ⊗ γ_i, which is what the surrounding text uses.
  4. [Appendix S.5, Eqs. (S88)-(S89)] The passage from (S88) to (S89) drops the square without comment. The final lower bound is plausible, but the inequality should be written out explicitly.

Circularity Check

1 steps flagged

Theorem 1's strong monotonicity is delegated to a same-author structural theorem (Ref. [21]) whose fermionic reduction is asserted but never exhibited.

specific steps
  1. self citation load bearing [Theorem 1 proof sketch (main text) and Appendix S.1, after Lemma 1]
    "The claim then follows from Theorem 3 of Ref. [21]. ... The lemma reduces any pure-state Gaussian protocol to the same elementary decomposition considered in the proof of Theorem 3 of Ref. [21]. Once this reduction is established, the remaining argument is identical: one applies the monotonicity inequality to each branch and averages over the corresponding probabilities."

    Lemma 1 proves the binary-split inequality only for |ψ> = √p|0>|ϕ0> + √1−p|1>|ϕ1>. The step from that inequality to strong monotonicity under arbitrary Gaussian protocols (matchgate unitaries, partial trace, measurements, conditioned operations) is not derived in this manuscript; it is asserted to be 'the same' as Theorem 3 of Ref. [21], a same-author structural theorem from the stabilizer-entropy setting. The fermionic reduction is never exhibited, so the central theorem's generality rests on an unverified self-citation rather than on the appendix's Lemma 1. This is load-bearing but not definitional: the cited theorem is a published result, and the note added reports that v2 of Ref. [12] independently contains a proof of Theorem 1, so the claim has independent support outside this manuscr

full rationale

The fermionic entropy Mf is defined directly from the correlation matrix, and no parameter is fitted from the quantities being predicted. Theorems 2, 3, 4, and 5 are self-contained: the sample-complexity proof follows from Lemma 2 and a Chernoff bound, the continuity bound follows from a Fannes-type argument, the tolerant tester follows from the continuity bound and a truncation argument, and the design lower bound follows from Haar purity versus doped-Matchgate purity. None of these reduce to their inputs by construction. The only load-bearing reliance on prior same-author work is in the proof of Theorem 1: the reduction of arbitrary Gaussian protocols to the two-branch decomposition of Lemma 1 is delegated to Theorem 3 of Ref. [21] and stated as 'identical' without being shown. That is a self-citation carrying the main theorem's generality, which justifies a moderate score. It is not full circularity because Lemma 1 is an original nontrivial inequality, the cited theorem is an external published result, and the note added reports an independent proof of Theorem 1 in v2 of Ref. [12]; the gap is one of unshown transfer, not equivalence by definition.

Axiom & Free-Parameter Ledger

0 free parameters · 6 axioms · 0 invented entities

No free parameters, no data fitting, and no invented entities. The paper is analytic; all constants (e.g., beta in Lemma 5) are optimization variables chosen for the proof, not fitted to data. The central claim rests on prior results: Ref [21] (same-author structural theorem), Wick's theorem, the normal form of t-doped states, and Bell sampling.

axioms (6)
  • standard math Theorem 3 of Ref. [21]: general reduction of strong monotonicity to a two-branch inequality for pure states
    Invoked in Appendix S.1 to lift Lemma 1 to arbitrary Gaussian protocols; not proved here and authored by the same authors.
  • standard math Wick's theorem / correlation-matrix characterization of fermionic Gaussian states
    Used throughout to define Mf via Gamma(rho) and to compute Gaussian-state purity (Refs [4,9]).
  • domain assumption Normal form of t-doped Matchgate states: there exists a Matchgate U with U|psi_t> = |0>^{otimes n-kappa t} otimes |phi>_{kappa t}
    Used in Theorem 5 lower bound (Appendix S.5, citing Ref [39]) to bound Pf(psi_t) >= (n-kappa t)/n.
  • domain assumption The shallow relative-error unitary design construction of Ref. [32] can be compiled into Matchgates plus kappa=3,4 Majorana rotations with O(1) overhead per gate
    Basis for the upper bound t=O(n log n); asserted in the main text but not proved or cited in detail.
  • standard math Standard resource-theory bookkeeping: strong monotonicity + additivity + asymptotic continuity imply an upper bound on asymptotic distillation rate
    Used in Corollary 1, following Ref [26].
  • domain assumption Bell sampling can estimate tr(Lambda^2 rho^{otimes 2})
    Assumed from Refs [11,12] for Theorem 2; Lambda is diagonal in the Bell basis.

pith-pipeline@v1.3.0-daily-deepseek · 17122 in / 21788 out tokens · 214808 ms · 2026-08-03T02:09:21.084076+00:00 · methodology

0 comments
read the original abstract

Fermionic Gaussian states form a central class of classically tractable quantum states, while fermionic non-Gaussianity provides the resource required to go beyond free-fermion dynamics. A key challenge is to quantify this resource through monotones that are both mathematically rigorous and experimentally accessible. Here, we show that the fermionic entropy, defined through the squared Frobenius norm of the correlation matrix, is a strong pure-state Gaussian monotone. Its simple closed-form expression also makes it directly measurable: we show that the associated fermionic purity can be unbiasedly estimated up to additive error $\varepsilon$ using $O(\varepsilon^{-2})$ two-copy measurements, independently of the system size. Moreover, we prove that the fermionic entropy obeys asymptotic continuity and, as a direct consequence, establish its operational meaning as the upper bound to the asymptotic rate of non-Gaussianity distillation. We further derive a linear sample complexity bound for tolerant testing of fermionic Gaussian states, providing a quadratic improvement over the state of the art. As a further application of our results, we study unitary designs generated by Matchgate circuits supplemented with Majorana-local non-Gaussian gates. We prove that a linear number of such gates is necessary even to achieve an approximate state $2$-design with error below $0.4\%$. Combined with known nearly linear upper bounds for relative-error designs, this determines the optimal doping level, up to logarithmic factors, across all relevant design notions and reveals the extensive non-Gaussianity cost required to generate Haar-like quantum dynamics in this architecture.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

51 extracted references · 2 canonical work pages

  1. [1]

    Mind the gaps: The fraught road to quantum advantage,

    J. Eisert and J. Preskill, “Mind the gaps: The fraught road to quantum advantage,” (2026), arXiv:2510.19928 [quant-ph]

    Pith/arXiv arXiv 2026

  2. [2]

    L. G. Valiant, in Proceedings of the thirty-third annual ACM symposium on Theory of computing (2001) pp. 114–123

    2001

  3. [3]

    Fermionic linear optics and matchgates,

    E. Knill, “Fermionic linear optics and matchgates,” (2001), arXiv:quant-ph/0108033 [quant-ph]

    Pith/arXiv arXiv 2001

  4. [4]

    S. B. Bravyi and A. Y. Kitaev, Annals of Physics 298, 210 (2002)

    2002

  5. [5]

    Jozsa and A

    R. Jozsa and A. Miyake, Proceedings of the Royal Soci- ety A: Mathematical, Physical and Engineering Sciences 464, 3089–3106 (2008)

    2008

  6. [6]

    B. M. Terhal and D. P. DiVincenzo, Physical Review A 65, 032325 (2002)

    2002

  7. [7]

    Hebenstreit, R

    M. Hebenstreit, R. Jozsa, B. Kraus, and S. Strelchuk, Phys. Rev. A 102, 052604 (2020)

    2020

  8. [8]

    Matchgate circuit represen- tation of fermionic gaussian states: optimal prepara- tion, approximation, and classical simulation,

    M. Langer, R. Morral-Yepes, A. Gammon-Smith, F. Poll- mann, and B. Kraus, “Matchgate circuit represen- tation of fermionic gaussian states: optimal prepara- tion, approximation, and classical simulation,” (2026), arXiv:2603.05675 [quant-ph]

    Pith/arXiv arXiv 2026

  9. [9]

    Surace and L

    J. Surace and L. Tagliacozzo, SciPost Physics Lecture Notes (2022), 10.21468/scipostphyslectnotes.54

    doi:10.21468/scipostphyslectnotes.54 2022

  10. [10]

    Hebenstreit, R

    M. Hebenstreit, R. Jozsa, B. Kraus, S. Strelchuk, and M. Yoganathan, Phys. Rev. Lett. 123, 080503 (2019)

    2019

  11. [11]

    Sierant, P

    P. Sierant, P. Stornati, and X. Turkeshi, PRX Quantum 7 (2026), 10.1103/3yx4-1j27

    doi:10.1103/3yx4-1j27 2026

  12. [12]

    Computable measures of fermionic non-gaussianity from the covari- ance matrix,

    P. S. Tarabunga, B. Jobst, R. Morral-Yepes, M. Langer, B. Kraus, F. Pollmann, and S.-H. Lin, “Computable measures of fermionic non-gaussianity from the covari- ance matrix,” (2026), arXiv:2607.02242 [quant-ph]

    Pith/arXiv arXiv 2026

  13. [13]

    Practical tests and witnesses of fermionic non-gaussianity,

    T. Haug, X. Turkeshi, and P. Sierant, “Practical tests and witnesses of fermionic non-gaussianity,” (2026), arXiv:2605.26218 [quant-ph]

    Pith/arXiv arXiv 2026

  14. [14]

    A. D. Gottlieb and N. J. Mauser, Physical Review Letters 95 (2005), 10.1103/physrevlett.95.123003

    doi:10.1103/physrevlett.95.123003 2005

  15. [15]

    Properties of non- freeness: an entropy measure of electron correlation,

    A. D. Gottlieb and N. J. Mauser, “Properties of non- freeness: an entropy measure of electron correlation,” (2006), arXiv:quant-ph/0608171 [quant-ph]

    Pith/arXiv arXiv 2006

  16. [16]

    Gaussian decomposition of magic states for matchgate computations,

    J. Cudby and S. Strelchuk, “Gaussian decomposition of magic states for matchgate computations,” (2025), arXiv:2307.12654 [quant-ph]

    Pith/arXiv arXiv 2025

  17. [17]

    Dias and R

    B. Dias and R. Koenig, Quantum 8, 1350 (2024)

    2024

  18. [18]

    Reardon-Smith, M

    O. Reardon-Smith, M. Oszmaniec, and K. Korzekwa, Quantum 8, 1549 (2024)

    2024

  19. [19]

    The un- bearable hardness of deciding about magic,

    L. Leone, J. Eisert, and S. F. E. Oliviero, “The un- bearable hardness of deciding about magic,” (2026), arXiv:2602.22330 [quant-ph]

    arXiv 2026

  20. [20]

    Leone, S

    L. Leone, S. F. E. Oliviero, and A. Hamma, Physical Review Letters 128, 050402 (2022)

    2022

  21. [21]

    Leone and L

    L. Leone and L. Bittel, Phys. Rev. A 110, L040403 (2024)

    2024

  22. [22]

    Fermionic magic resources in disordered quantum spin chains,

    P. R. N. Falc˜ ao, J. Zakrzewski, and P. Sierant, “Fermionic magic resources in disordered quantum spin chains,” (2026), arXiv:2602.00245 [quant-ph]

    arXiv 2026

  23. [23]

    Fermionic gaussian testing and non-gaussian measures via convolution,

    X. Lyu and K. Bu, “Fermionic gaussian testing and non-gaussian measures via convolution,” (2024), arXiv:2409.08180 [quant-ph]

    Pith/arXiv arXiv 2024

  24. [24]

    Measuring non- gaussian magic in fermions: Convolution, entropy, and the violation of wick’s theorem and the matchgate iden- tity,

    L. Coffman, G. Smith, and X. Gao, “Measuring non- gaussian magic in fermions: Convolution, entropy, and the violation of wick’s theorem and the matchgate iden- tity,” (2025), arXiv:2501.06179 [quant-ph]

    Pith/arXiv arXiv 2025

  25. [25]

    Gigena and R

    N. Gigena and R. Rossignoli, Physical Review A 94 (2016), 10.1103/physreva.94.042315

    work page doi:10.1103/physreva.94.042315 2016

  26. [26]

    Chitambar and G

    E. Chitambar and G. Gour, Review of Modern Physics 91, 025001 (2019)

    2019

  27. [27]

    Popescu, A

    S. Popescu, A. J. Short, and A. Winter, Nature Physics 2, 754 (2006)

    2006

  28. [28]

    Sekino and L

    Y. Sekino and L. Susskind, Journal of High Energy Physics 2008, 065 (2008)

    2008

  29. [29]

    Munson, N

    A. Munson, N. B. T. Kothakonda, J. Haferkamp, N. Yunger Halpern, J. Eisert, and P. Faist, PRX Quan- tum 6, 010346 (2025)

    2025

  30. [30]

    Haferkamp, F

    J. Haferkamp, F. Montealegre-Mora, M. Heinrich, J. Eis- ert, D. Gross, and I. Roth, Commun. Math. Phys. 397, 995 (2023)

    2023

  31. [31]

    Bittel and L

    L. Bittel and L. Leone, Phys. Rev. Lett. 136, 210802 (2026)

    2026

  32. [32]

    Schuster, J

    T. Schuster, J. Haferkamp, and H.-Y. Huang, Science 389, 92 (2025), arXiv:2407.07754 [quant-ph]

    Pith/arXiv arXiv 2025

  33. [33]

    F. d. Melo, P. ´Cwikli´ nski, and B. M. Terhal, New Journal of Physics 15, 013015 (2013)

    2013

  34. [34]

    Vershynina, Physical Review A 90 (2014), 10.1103/physreva.90.062329

    A. Vershynina, Physical Review A 90 (2014), 10.1103/physreva.90.062329

    work page doi:10.1103/physreva.90.062329 2014

  35. [35]

    High- temperature fermionic gibbs states are mixtures of gaus- sian states,

    A. Ramkumar, Y. Cai, Y. Tong, and J. Jiang, “High- temperature fermionic gibbs states are mixtures of gaus- sian states,” (2026), arXiv:2505.09730 [quant-ph]

    arXiv 2026

  36. [36]

    Synak-Radtke and M

    B. Synak-Radtke and M. Horodecki, Journal of Physics A: Mathematical and General 39, L423–L437 (2006)

    2006

  37. [37]

    Unitary designs from doped matchgate circuits,

    F. B. Trigueros, Z.-H. Sun, X. Turkeshi, P. Sierant, and P. S. Tarabunga, “Unitary designs from doped matchgate circuits,” (2026), arXiv:2606.23800 [quant-ph]

    Pith/arXiv arXiv 2026

  38. [38]

    B. M. Terhal, M. Horodecki, D. W. Leung, and D. P. DiVincenzo, Journal of Mathematical Physics 43, 4286–4298 (2002)

    2002

  39. [39]

    A. A. Mele and Y. Herasymenko, PRX Quantum 6, 010319 (2025), arXiv:2402.18665 [quant-ph]

    Pith/arXiv arXiv 2025

  40. [40]

    Emergence of generic entanglement structure in doped matchgate circuits,

    A. Paviglianiti, L. Lumia, E. Tirrito, A. Silva, M. Col- lura, X. Turkeshi, and G. Lami, “Emergence of generic entanglement structure in doped matchgate circuits,” (2025), arXiv:2507.12526 [quant-ph]

    Pith/arXiv arXiv 2025

  41. [41]

    A. W. Harrow and R. A. Low, Comm. Math. Phys. 291, 302 (2009)

    2009

  42. [42]

    F. G. S. L. Brand˜ ao, A. W. Harrow, and M. Horodecki, Communications in Mathematical Physics 346, 397 (2016)

    2016

  43. [43]

    Leone, S

    L. Leone, S. F. Oliviero, A. Hamma, J. Eisert, and L. Bittel, PRX Quantum 7, 020321 (2026)

    2026

  44. [44]

    Theory of the matchgate commutant,

    P. Sierant, X. Turkeshi, and P. S. Tarabunga, “Theory of the matchgate commutant,” (2026), arXiv:2603.12392 [quant-ph]

    arXiv 2026

  45. [45]

    Fermionic non-gaussianity via bell sampling: Monotones and efficient quantum algorithms,

    P. S. Tarabunga, “Fermionic non-gaussianity via bell sampling: Monotones and efficient quantum algorithms,” (2026), arXiv:2606.05066 [quant-ph]

    Pith/arXiv arXiv 2026

  46. [46]

    Witness expansion: A unified framework for analytical and measurable mixed- 7 state resource detection,

    Y. Tang, C. Zhu, Y. Zhang, J. Eisert, Z.-W. Liu, I. Roth, O. G¨ uhne, X. Wang, and Z. Liu, “Witness expansion: A unified framework for analytical and measurable mixed- 7 state resource detection,” (2026), arXiv:2606.27105 [quant-ph]

    Pith/arXiv arXiv 2026

  47. [47]

    Bittel, A

    L. Bittel, A. A. Mele, J. Eisert, and L. Leone, PRX Quantum 6, 030341 (2025), arXiv:2409.17953 [quant-ph]. 8 Appendix CONTENTS S.1. Strong monotonicity of the fermionic entropy: proof of Theorem 1 8 S.2. System-size independent measurement scheme for the fermionic purity: proof of Theorem 2 10 S.3. Fannes-like inequality and asymptotic continuity of the f...

    arXiv 2025

  48. [48]

    We have Γ′ 12(ϕ0) = ⟨ϕ′ 0|Z1|ϕ′ 0⟩ = ∥v0∥2 2 − ∥v1∥2 2 = 1 − 2x (S28) Γ′ 12(ϕ1) = ⟨ϕ′ 1|Z1|ϕ′ 1⟩ = ∥w0∥2 2 − ∥w1∥2 2 = 1 − 2y (S29) Hence: 1 8 ∥Γ′(ϕ0) − Γ′(ϕ1)∥2 2 ≥ 1 8 (|Γ12(ϕ′

  49. [49]

    − Γ12(ϕ′ 1)| + |Γ21(ϕ′

  50. [50]

    2nY i=1 (I + i√zΓi) + 2nY i=1 (I − i√zΓi) # (S55) Indeed, we have Q2n i=1(I + i√zΓi) = P S⊆[2n](i√z)|S|ΓS and therefore 1 2

    − Γ21(ϕ′ 1)|) = 1 4 [2(x − y)]2 = (x − y)2 (S30) which proves the desired inequality and concludes the proof. The proof of Theorem 1 follows directly from Lemma 1. The lemma reduces any pure-state Gaussian protocol to the same elementary decomposition considered in the proof of Theorem 3 of Ref. [21]. Once this reduction is established, the remaining argu...

  51. [51]

    (S64) = Pf (ρ) − Pf (ρ′) (S65) Hence, we can bound |Pf (ρ) − Pf (ρ′)| ≤ ∥Kρ,ρ′∥∞∥ρ − ρ′∥1. To bound the infinity norm of Kρ,ρ′, notice that by linearity we can write Kρ,ρ′ = −2i n X i<j Γij ρ + ρ′ 2 γiγj (S66) Since Γ ρ+ρ′ 2 is a valid correlation matrix, we know that there exists a Gaussian unitaryO such that OΓ ρ+ρ′ 2 OT = Lm i=1 0 ci −ci 0 , where ci a...