pith. sign in

arxiv: 2604.15029 · v1 · submitted 2026-04-16 · 🪐 quant-ph

Entanglement quantification with randomized measurements is maximally difficult

Julian Eisfeld , Nikolai Wyderka This is my paper

Pith reviewed 2026-05-10 11:47 UTC · model grok-4.3

classification 🪐 quant-ph
keywords randomized measurementstwo-qubit invariantsentanglement certificationlocal invariantsquantum certificationmeasurement settingsKempe invariant
0
0 comments X

The pith

Entanglement certification requires the largest number of randomized measurement settings among all two-qubit invariants.

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

The paper calculates the smallest number of randomized local measurement settings needed to determine every local invariant of a two-qubit state. It shows that the specific invariants required to certify entanglement sit at the top of this list and therefore demand strictly more settings than any other invariant. A reader would care because randomized measurements are designed to work without shared reference frames in quantum networks, yet the practical cost turns out to be highest precisely for the property most relevant to communication and distributed computing. The authors also apply the same counting method to the Kempe invariant of three-qubit states and obtain an improved protocol.

Core claim

We determine the minimal number of measurement settings required to access all two-qubit invariants. We further demonstrate that entanglement certification necessarily involves the most demanding invariants, establishing it as a maximally difficult task. The same optimization procedure improves known bounds for the Kempe invariant in three-qubit systems.

What carries the argument

The hierarchy of minimal randomized measurement settings required for each two-qubit local invariant, obtained by optimizing the choice of random local bases.

If this is right

  • Every two-qubit invariant becomes accessible with a finite, invariant-specific number of randomized settings.
  • Entanglement certification always requires the single largest number in this collection.
  • The same counting method yields a tighter protocol for the Kempe invariant of three qubits.
  • Experimental feasibility of quantum certification tasks can be ranked by consulting the position of the needed invariants in the hierarchy.

Where Pith is reading between the lines

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

  • Hybrid protocols could first measure the easier invariants to obtain partial information that reduces the number of settings ultimately needed for the harder entanglement-related ones.
  • Analogous difficulty orderings may appear for other resources such as coherence or non-stabilizerness when the same randomized-measurement counting is applied.
  • Including realistic noise and finite-shot statistics will likely raise the effective setting counts but may preserve or alter the relative ordering of the invariants.

Load-bearing premise

The minimal number of settings for each invariant can be optimized independently without regard to finite sample statistics, experimental noise, or the practical cost of realizing the random choices.

What would settle it

An experiment that accurately extracts an entanglement witness or monotone from a two-qubit state using fewer randomized settings than the paper's calculated maximum for the required invariants would contradict the maximality result.

Figures

Figures reproduced from arXiv: 2604.15029 by Julian Eisfeld, Nikolai Wyderka.

Figure 1
Figure 1. Figure 1: FIG. 1. Visualization of the randomized type [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
read the original abstract

The certification of quantum systems is essential for emerging quantum technologies, particularly in quantum communication, networks, and distributed computing, where maintaining a common reference frame across distant nodes poses significant challenges. Reference frame independent approaches, such as randomized measurement schemes, offer a promising route by reducing experimental demands while granting access to basis-independent quantities, including entanglement. However, the efficiency of such schemes in measuring such local invariants has remained unclear. In this work, we determine the minimal number of measurement settings required to access all two-qubit invariants. We further demonstrate that entanglement certification necessarily involves the most demanding invariants, establishing it as a maximally difficult task. Our results reveal a fundamental hierarchy among invariants, with direct implications for experimental feasibility and theoretical understanding of quantum certification. Finally, we extend our analysis beyond bipartite systems by applying it to the Kempe invariant in three-qubit systems, improving known measurement protocols and providing a first step toward uncovering similar hierarchies in higher dimensions.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The manuscript determines the minimal number of randomized measurement settings required to access all two-qubit invariants. It demonstrates that entanglement certification necessarily involves the most demanding invariants, establishing it as a maximally difficult task. The analysis is extended to the Kempe invariant in three-qubit systems, where it improves upon known measurement protocols.

Significance. If the minimal-setting derivations and resulting hierarchy hold, the work clarifies the resource costs of reference-frame-independent certification methods and identifies which invariants drive experimental overhead. This has direct implications for designing efficient protocols in quantum networks and communication. The three-qubit extension provides a concrete step toward higher-dimensional generalizations.

major comments (2)
  1. The definition of difficulty as the cardinality of the minimal setting set (in the exact, infinite-shot limit) does not incorporate the statistical variance of finite-sample estimates of the correlators. Because entanglement functionals are nonlinear, this omission is load-bearing for the claim that entanglement certification is 'maximally difficult' in any experimentally relevant regime.
  2. The optimization treats the minimal setting count for each invariant independently. It is unclear whether a joint randomized-measurement scheme could simultaneously access multiple invariants (including those needed for entanglement) with a smaller total number of settings than the sum of the individual minima.
minor comments (2)
  1. The abstract states the hierarchy result but does not report the concrete minimal numbers obtained for the individual invariants; including these values would strengthen the summary.
  2. Notation for the two-qubit invariants and the randomized measurement operators should be introduced with explicit definitions before the main theorems are stated.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments on our manuscript. These points help delineate the theoretical scope of our results from practical experimental considerations. We address each major comment below.

read point-by-point responses

  1. Referee: The definition of difficulty as the cardinality of the minimal setting set (in the exact, infinite-shot limit) does not incorporate the statistical variance of finite-sample estimates of the correlators. Because entanglement functionals are nonlinear, this omission is load-bearing for the claim that entanglement certification is 'maximally difficult' in any experimentally relevant regime.

    Authors: We appreciate the referee's observation on the distinction between the exact infinite-shot limit and finite-sample regimes. Our work derives the minimal number of randomized measurement settings required for exact access to each two-qubit invariant, thereby establishing a fundamental lower bound and the resulting hierarchy. We agree that nonlinear entanglement functionals introduce additional statistical variance in finite-shot estimates, which could modify the effective resource cost in laboratory settings. Our analysis intentionally focuses on the ideal case to identify this bound; a full finite-sample treatment would require a separate statistical framework. In the revised manuscript we will add a dedicated paragraph in the discussion section acknowledging this limitation and identifying finite-shot analysis as an important direction for future work. revision: partial

  2. Referee: The optimization treats the minimal setting count for each invariant independently. It is unclear whether a joint randomized-measurement scheme could simultaneously access multiple invariants (including those needed for entanglement) with a smaller total number of settings than the sum of the individual minima.

    Authors: Our independent optimization for each invariant is used to reveal the hierarchy of minimal setting counts. While a joint scheme could in principle access several invariants together with a total setting count lower than the sum of the separate minima, this does not affect our central claim. Because entanglement certification requires the single most demanding invariant, any scheme—joint or separate—must employ at least the maximum number of settings identified for that invariant. Consequently the conclusion that entanglement is maximally difficult remains intact. We will insert a short clarifying paragraph explaining this point and noting that joint optimization, while potentially beneficial for simultaneous access to multiple invariants, cannot reduce the setting count below the maximum required by the hardest invariant. revision: partial

Circularity Check

0 steps flagged

No circularity: minimal setting counts derived from independent optimization over randomized schemes

full rationale

The paper computes the minimal number of randomized measurement settings needed for each two-qubit invariant by optimizing over possible schemes, then observes that the invariants required for entanglement certification are among those with the largest cardinalities. No equation or step reduces the target quantity to a fitted parameter or to a prior self-citation that itself assumes the result. The hierarchy follows directly from the explicit enumeration of invariants and the setting-count optimization; the derivation remains self-contained against external benchmarks and does not rename known results or smuggle ansatzes via citation.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The work rests on standard quantum-information assumptions about the existence and accessibility of local invariants via randomized measurements; no new free parameters, invented entities, or ad-hoc axioms are introduced in the abstract.

axioms (1)
  • domain assumption All two-qubit invariants are accessible through suitable randomized measurement schemes
    This is the foundational premise that allows the authors to count minimal settings for each invariant.

pith-pipeline@v0.9.0 · 5452 in / 1155 out tokens · 27534 ms · 2026-05-10T11:47:42.011585+00:00 · methodology

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

47 extracted references · 47 canonical work pages

  1. [1]

    Choose a random rotationU=U 1 ⊗...⊗U n

    work page

  2. [2]

    (b) In each subsystemk, measure the observableU † k Ok Uk

    Estimate the quantity⟨U †OU⟩ ϱ by repeating suitably of- ten: (a) Prepare a copy ofϱ. (b) In each subsystemk, measure the observableU † k Ok Uk

    work page

  3. [3]

    If, however,Ois not a product observable, we de- compose it in terms ofr≥2product observables, O= rP j=1 O(j) 1 ⊗...⊗ O (j) n ≡ rP j=1 O(j)

    EstimateR (t) O by repeating steps 1 and 2 suitably often. If, however,Ois not a product observable, we de- compose it in terms ofr≥2product observables, O= rP j=1 O(j) 1 ⊗...⊗ O (j) n ≡ rP j=1 O(j) . Then, above mea- surement protocol is modified as follows: Protocol 2Typermeasurement

    work page

  4. [4]

    Choose a random rotationU=U 1 ⊗...⊗U n. 2.1. Estimate the quantity⟨U †O(1)U⟩ ϱ by repeating suitably often: FIG. 1. Visualization of the randomized typermeasurement proto- col to measure the momentsR (t) O (ϱ)[Eq. (1)] for a given non-local observableOwith tensor rankr. The prepared bipartite stateϱun- dergoes an unknown local unitary rotation, visualized...

    work page

  5. [5]

    The measurement setup is visualized in Fig

    EstimateR (t) O by repeating steps 1 and 2 suitably often. The measurement setup is visualized in Fig. 1 for the exam- ple ofn= 2. Experimentally, the difficulty lies in obtaining ⟨U †OU⟩ ϱ for fixedU=U 1 ⊗...⊗U n. This involves keeping a fairly constant reference frame for a lot of measurements. The number of measurements with a constant reference frame ...

    work page

  6. [6]

    A local invariant of degreedeg(I)is called type 1 if it can be measured through a series of type 1 measure- ments ofR (tj ) Oj (ϱ)witht j ≤deg(I)

    work page

  7. [7]

    A local invariant of degreedeg(I)is called type≤r if it can be measured through a series of typer j ≤r measurements ofR (tj ) Oj (ϱ)witht j ≤deg(I)

    work page

  8. [8]

    Bipartite invariants of typer max = min(d 2 1 ,d 2 2 )are called maximally difficult

    An invariant that is of type≤rand not of type≤(r−1) is called type r. Bipartite invariants of typer max = min(d 2 1 ,d 2 2 )are called maximally difficult. 4 Note that any function of type≤rinvariants is again a type ≤rinvariant. This induces some additional structure, imply- ing that the sets of all≤rinvariants are non-decreasingly large subrings of the ...

    work page

  9. [9]

    It is non-zero if and only ifM 1M T 2 is invertible

    Let M1 := √s1 ˜P(A 1), ..., √sr ˜P(A r) ∈R 3×r ,(3) M2 := √s1 ˜P(B 1), ..., √sr ˜P(B r) ∈R 3×r .(4) Then the prefactor ofdet(T)in the affine expression of R(3) O (ϱ)is given bydet M1M T 2 /8. It is non-zero if and only ifM 1M T 2 is invertible. This is proven as Corollary B.4 in Appendix B. Under the additional restriction of a symmetric operator Schmidt ...

    work page

  10. [10]

    There is no product observableO=A⊗B⊗Csuch that the invarianttr T AB T BC T CA is uniquely re- coverable from randomized measurement moments of order up to 3

    work page

  11. [11]

    (7) can be obtained from at= 3randomized measurement moment

    There exists a product observable such that 3||W|| 2 2 + 6 X jkℓ Wjkℓ T AB jk γℓ +α j T BC kℓ +T CA ℓj βk + 6 tr T AB T BC T CA . (7) can be obtained from at= 3randomized measurement moment. 6 This is proven as Proposition F.3 in the Appendix. Due to the large amount of symmetry in the explicit cal- culations forr= 1, the five different invariants that ap...

    work page

  12. [12]

    Ac ´ın, I

    A. Ac ´ın, I. Bloch, H. Buhrman, T. Calarco, C. Eichler, J. Eisert, D. Esteve, N. Gisin, S. J. Glaser, F. Jelezko,et al., New J. Phys. 20, 080201 (2018)

    work page 2018

  13. [13]

    Kliesch and I

    M. Kliesch and I. Roth, PRX Quantum2, 010201 (2021)

    work page 2021

  14. [14]

    N ¨oller, N

    J. N ¨oller, N. Miklin, M. Kliesch, and M. Gachechiladze, Quan- tum9, 1825 (2025)

    work page 2025

  15. [15]

    Islam, L

    T. Islam, L. Magnin, B. Sorg, and S. Wehner, New J. Phys.16, 063040 (2014)

    work page 2014

  16. [16]

    Islam and S

    T. Islam and S. Wehner, New J. Phys.18, 033018 (2016)

    work page 2016

  17. [17]

    Cie ´sli´nski, S

    P. Cie ´sli´nski, S. Imai, J. Dziewior, O. G ¨uhne, L. Knips, W. Laskowski, J. Meinecke, T. Paterek, and T. V ´ertesi, Phys. Rep.1095, 1 (2024)

    work page 2024

  18. [18]

    J. G. Rarity, P. Tapster, P. Gorman, and P. Knight, New J. Phys. 4, 82 (2002)

    work page 2002

  19. [19]

    Kurtsiefer, P

    C. Kurtsiefer, P. Zarda, M. Halder, H. Weinfurter, P. Gorman, P. Tapster, and J. Rarity, Nature419, 450 (2002)

    work page 2002

  20. [20]

    Liao, W.-Q

    S.-K. Liao, W.-Q. Cai, W.-Y . Liu, L. Zhang, Y . Li, J.-G. Ren, J. Yin, Q. Shen, Y . Cao, Z.-P. Li,et al., Nature549, 43 (2017)

    work page 2017

  21. [21]

    Li, W.-Q

    Y . Li, W.-Q. Cai, J.-G. Ren, C.-Z. Wang, M. Yang, L. Zhang, H.-Y . Wu, L. Chang, J.-C. Wu, B. Jin,et al., Nature640, 47 (2025)

    work page 2025

  22. [22]

    Elben, R

    A. Elben, R. Kueng, H.-Y . Huang, R. van Bijnen, C. Kokail, M. Dalmonte, P. Calabrese, B. Kraus, J. Preskill, P. Zoller,et al., Phys. Rev. Lett.125, 200501 (2020). 7

    work page 2020

  23. [23]

    Ketterer, S

    A. Ketterer, S. Imai, N. Wyderka, and O. G ¨uhne, Phys. Rev. A 106, L010402 (2022)

    work page 2022

  24. [25]

    G ¨uhne and G

    O. G ¨uhne and G. T´oth, Phys. Rep.474, 1 (2009)

    work page 2009

  25. [26]

    Bartkiewicz, J

    K. Bartkiewicz, J. Beran, K. Lemr, M. Norek, and A. Mira- nowicz, Phys. Rev. A91, 022323 (2015)

    work page 2015

  26. [27]

    Knips, J

    L. Knips, J. Dziewior, W. Kłobus, W. Laskowski, T. Paterek, P. J. Shadbolt, H. Weinfurter, and J. D. Meinecke, npj Quantum Inf.6, 51 (2020)

    work page 2020

  27. [28]

    Knips, Quantum Views4, 47 (2020)

    L. Knips, Quantum Views4, 47 (2020)

    work page 2020

  28. [29]

    S. Imai, N. Wyderka, A. Ketterer, and O. G ¨uhne, Phys. Rev. Lett.126, 150501 (2021)

    work page 2021

  29. [30]

    Elben, S

    A. Elben, S. T. Flammia, H.-Y . Huang, R. Kueng, J. Preskill, B. Vermersch, and P. Zoller, Nat. Rev. Phys.5, 9 (2023)

    work page 2023

  30. [31]

    Kempe, Phys

    J. Kempe, Phys. Rev. A60, 910 (1999)

    work page 1999

  31. [32]

    Barnum and N

    H. Barnum and N. Linden, J. Phys. A: Math. Gen.34, 6787 (2001)

    work page 2001

  32. [33]

    Wyderka and A

    N. Wyderka and A. Ketterer, PRX Quantum4, 020325 (2023)

    work page 2023

  33. [34]

    Wyderka, A

    N. Wyderka, A. Ketterer, S. Imai, J. L. B ¨onsel, D. E. Jones, B. T. Kirby, X.-D. Yu, and O. G ¨uhne, Phys. Rev. Lett.131, 090201 (2023)

    work page 2023

  34. [35]

    Ma and H.-Y

    F. Ma and H.-Y . Huang, inProceedings of the 57th Annual ACM Symposium on Theory of Computing(2025) pp. 806–809

    work page 2025

  35. [36]

    Schmidt, Mathematische Annalen63, 433 (1907)

    E. Schmidt, Mathematische Annalen63, 433 (1907)

    work page 1907

  36. [37]

    Grassl, M

    M. Grassl, M. R ¨otteler, and T. Beth, Phys. Rev. A58, 1833 (1998)

    work page 1998

  37. [38]

    A. L. Onishchik and E. B. Vinberg,Lie groups and algebraic groups(Springer Science & Business Media, 2012)

    work page 2012

  38. [39]

    T. A. Springer,Invariant theory, V ol. 585 (Springer, 2006)

    work page 2006

  39. [40]

    Makhlin, Quantum Inf

    Y . Makhlin, Quantum Inf. Process.1, 243–252 (2002)

    work page 2002

  40. [41]

    Vidal and R

    G. Vidal and R. F. Werner, Phys. Rev. A65, 032314 (2002)

    work page 2002

  41. [42]

    R. F. Werner, Phys. Rev. A40, 4277 (1989)

    work page 1989

  42. [43]

    Gross, K

    D. Gross, K. Audenaert, and J. Eisert, J. Math. Phys.48, 052104 (2007)

    work page 2007

  43. [44]

    Weingarten Calculus, February 2021

    G. K ¨ostenberger, arXiv preprint arXiv:2101.00921 (2021)

    work page arXiv 2021

  44. [45]

    A naive look at Schur–Weyl duality,

    G. Aubrun, “A naive look at Schur–Weyl duality,” (2018),https://math.univ-lyon1.fr/ ˜aubrun/ recherche/schur-weyl.pdf[Accessed: 09.04.2026]. Appendix A: Notation and general calculation of integrals with respect to the Haar measure Definition A.1.The twirling operator (see [31, 32] ) is given by: Φ(t) :L(C d)⊗n → L(Cd)⊗n ⊗t O 7→ Z (U † 1 ⊗...⊗U † n)⊗t O⊗t...

    work page 2018

  45. [46]

    In this case the statement requires an easy modification of the proof of Proposition C.1, since onlyN ⊗2 3 ̸= 0and its prefactor is now s1s2s3√ 2 6

    SinceRis invariant underO →(U A ⊗U B)†O(UA ⊗U B), it is sufficient to considerO= 3P j=1 sj√ 2 2 σj ⊗σ j . In this case the statement requires an easy modification of the proof of Proposition C.1, since onlyN ⊗2 3 ̸= 0and its prefactor is now s1s2s3√ 2 6 . An explicit calculation allows the proof of the other direction in Proposition C.2, but only for symm...

    work page

  46. [47]

    +B 22(4λ1λ2 + 2λ2 2 −4λ 2 2) = 2 tr(B) h −tr(A) 2 + tr A2 −2λ 1λ2 i = 2 tr(B) −(λ1 +λ 2)2 +λ 2 1 +λ 2 2 −2λ 1λ2 = 0. Thus, it only remains to discuss the choices ofV (12),(12),(12) (and the other two terms of similar form),V (12),(23),(31) (and its five other permutations), as well asV (12),(12),(13) (and the 17 other terms of similar structure). They yie...

    work page

  47. [48]

    Using the same observable (but permuting on which subsystem the 1−σ 1 terms act) gives access to P jkℓ Wjkℓ T CA ℓj βk andP jkℓ Wjkℓ αj T BC kℓ via symmetry

    Furthermore, ˆc O3 =1 ⊗2 ⊗(1−σ 1) +σ ⊗2 3 ⊗(1−σ 1) =   8/9 16/9 0 0 0   ,(F6) additionally allowing measurement of P jkℓ Wjkℓ T AB jk γℓ. Using the same observable (but permuting on which subsystem the 1−σ 1 terms act) gives access to P jkℓ Wjkℓ T CA ℓj βk andP jkℓ Wjkℓ αj T BC kℓ via symmetry. Thus,tr T AB T BC T CA can now be recovered from th...

    work page