pith. sign in

arxiv: 2605.01438 · v1 · submitted 2026-05-02 · 🪐 quant-ph

Spectral Minimax Direct Fidelity Estimation for Generic Target States

Hyunho Cha , Jungwoo Lee This is my paper

Pith reviewed 2026-05-09 14:34 UTC · model grok-4.3

classification 🪐 quant-ph
keywords direct fidelity estimationminimax optimizationsemidefinite programmingunbiased linear estimatorsPauli measurementsquantum state verificationvariance reduction
0
0 comments X

The pith

A single operator identity plus semidefinite programming yields the exact minimax sampling law for direct fidelity estimation of any quantum target state.

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

The paper shows how to replace approximate linear programming with an exact spectral characterization for optimizing measurements in fidelity estimation. Unbiased estimators are pinned down by one operator equation, which lets the worst-case variance over all possible states be turned into a solvable semidefinite program. This produces an offline sampling design and an online estimator that uses only non-adaptive single-copy measurements. A reader cares because lower variance means fewer measurements are needed to certify a quantum state to a given precision. Numerical tests under noise confirm the new design beats the earlier surrogate method.

Core claim

Unbiased linear estimators for fidelity are fully characterized by the identity that the expected reconstruction operator equals the target projector. For any fixed target and reconstruction coefficients, the optimal sampling probabilities over measurement outcomes are found by solving a state-wise problem; the overall minimax problem is then cast as a semidefinite program whose solution gives the best possible worst-case variance.

What carries the argument

The single operator identity that enforces unbiasedness for linear estimators, which converts the minimax variance minimization into a semidefinite program.

If this is right

  • The resulting sampling probabilities can be precomputed offline for any chosen target state.
  • Local Pauli measurements suffice to implement the estimator with the optimized probabilities.
  • The achieved estimation variance is strictly lower than that of the OASIS surrogate under the same measurement budget.
  • Depolarizing noise simulations confirm the variance reduction for generic targets.

Where Pith is reading between the lines

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

  • If the SDP can be solved efficiently for large systems, the method scales to many-qubit states.
  • Similar spectral replacements might improve other shadow-based estimation tasks beyond fidelity.
  • Real-device tests could reveal whether the theoretical variance gain survives calibration errors and readout noise.

Load-bearing premise

That the single operator identity captures every possible unbiased linear estimator and that the semidefinite program solves the exact minimax problem without further relaxation or approximation.

What would settle it

A concrete counter-example: for some target state and measurement set, compute the SDP-derived variance and show that an alternative linear estimator (not derived from the identity) achieves lower worst-case variance.

read the original abstract

Direct fidelity estimation benefits from tailoring measurements to a fixed target, but the operator-aware shadow importance sampling (OASIS) method optimizes an outcome-wise linear-program surrogate rather than the exact worst-case variance over physical states. We propose an exact spectral replacement for arbitrary target states under the same non-adaptive single-copy measurement model. Specifically, we characterize unbiased linear estimators by a single operator identity, determine the state-wise optimal sampling law for fixed reconstruction coefficients, and convert the exact minimax problem into a semidefinite program. The resulting offline design and online estimator are presented as an algorithm and implemented with local Pauli measurements. Numerical simulations under depolarizing noise demonstrate that our exact spectral optimization outperforms the OASIS surrogate in terms of estimation variance.

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

3 major / 2 minor

Summary. The manuscript proposes an exact spectral minimax method for direct fidelity estimation of arbitrary target states under non-adaptive single-copy measurements. It characterizes all unbiased linear estimators via a single operator identity, derives the state-wise optimal sampling distribution for fixed reconstruction coefficients, and reformulates the resulting minimax variance problem as a semidefinite program. The approach is implemented with local Pauli measurements, presented as an algorithm, and shown via numerical simulations under depolarizing noise to yield lower estimation variance than the OASIS surrogate.

Significance. If the operator identity fully parametrizes the unbiased linear estimators and the SDP reformulation is exact (no hidden relaxations), the work supplies a rigorous, parameter-free improvement to measurement design and estimation for fidelity, which is a core primitive in quantum state verification and benchmarking. The explicit algorithm and local-Pauli implementation make the result immediately usable.

major comments (3)
  1. [§3] §3 (characterization of unbiased estimators): the claim that a single operator identity fully parametrizes all unbiased linear estimators for generic mixed targets is load-bearing for the subsequent minimization and SDP step. An explicit proof is required showing that no other linear unbiased estimators exist outside this parametrization, particularly when the target is not pure.
  2. [§4] §4 (SDP reformulation of the minimax problem): the conversion is asserted to be exact, yet it is unclear whether the program retains all original constraints (e.g., complete positivity and normalization of the POVM elements, or any implicit rank restrictions). If a relaxation is used, the claimed “exact” minimax solution would only be a lower bound.
  3. [Numerical simulations] Numerical section (simulations under depolarizing noise): the reported variance improvement is demonstrated only for depolarizing channels. Because the minimax objective is defined over the worst-case physical state, additional benchmarks on other noise models or on states that saturate the minimax bound are needed to substantiate the outperformance claim.
minor comments (2)
  1. [Abstract] The abstract introduces the phrase “exact spectral replacement” without definition; a one-sentence clarification of what “spectral” refers to would aid readers who encounter the paper via the abstract alone.
  2. [Algorithm presentation] The algorithm box would benefit from explicit pseudocode for the offline SDP solve and the online estimator, including the solver library used and any numerical tolerances.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for their careful reading and constructive comments on our manuscript. We address each major comment point by point below and indicate the revisions planned for the next version.

read point-by-point responses

  1. Referee: [§3] §3 (characterization of unbiased estimators): the claim that a single operator identity fully parametrizes all unbiased linear estimators for generic mixed targets is load-bearing for the subsequent minimization and SDP step. An explicit proof is required showing that no other linear unbiased estimators exist outside this parametrization, particularly when the target is not pure.

    Authors: We agree that an explicit proof is necessary to fully substantiate the parametrization. In the revised manuscript we will add a self-contained proof (in Section 3 or an appendix) showing that any linear estimator unbiased for all states must satisfy the stated operator identity, and conversely that any coefficients obeying the identity yield an unbiased estimator. The argument uses linearity of the expectation together with the requirement that the estimator reproduce the fidelity for every density operator, which applies directly to mixed targets without additional assumptions. revision: yes

  2. Referee: [§4] §4 (SDP reformulation of the minimax problem): the conversion is asserted to be exact, yet it is unclear whether the program retains all original constraints (e.g., complete positivity and normalization of the POVM elements, or any implicit rank restrictions). If a relaxation is used, the claimed “exact” minimax solution would only be a lower bound.

    Authors: The SDP is an exact reformulation; no relaxation is introduced. The decision variables are the sampling distribution and the reconstruction operator, and the semidefinite constraints together with the trace-normalization condition enforce complete positivity and normalization of the underlying POVM elements exactly as in the original problem. No extraneous rank restrictions are added. We will insert a short clarifying paragraph in Section 4 that explicitly maps each original constraint to its SDP counterpart to remove any ambiguity. revision: partial

  3. Referee: Numerical section (simulations under depolarizing noise): the reported variance improvement is demonstrated only for depolarizing channels. Because the minimax objective is defined over the worst-case physical state, additional benchmarks on other noise models or on states that saturate the minimax bound are needed to substantiate the outperformance claim.

    Authors: The referee is correct that the present numerical results focus on depolarizing noise. To strengthen the empirical support we will augment the numerical section with simulations under additional noise models (amplitude damping and dephasing) and for states that attain the minimax bound, thereby demonstrating that the observed advantage is not limited to the depolarizing case. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation uses standard operator characterization and exact SDP reformulation

full rationale

The paper derives the unbiasedness condition for linear estimators of Tr(ρ_target ρ) from the definition of unbiasedness, yielding a single operator identity that constrains the reconstruction operator. It then optimizes sampling probabilities for fixed coefficients and converts the resulting state-wise minimax variance problem into an SDP without introducing fitted parameters that are later renamed as predictions. No self-citations are load-bearing, no ansatz is smuggled via prior work, and the numerical comparison to OASIS under depolarizing noise serves as external validation rather than tautological confirmation. The central claims therefore remain independent of their inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 0 invented entities

Based on abstract only; no explicit free parameters, new entities, or non-standard axioms are stated. Relies on standard quantum mechanics and convex optimization.

axioms (2)
  • domain assumption Unbiased linear estimators for fidelity are characterized by a single operator identity.
    Invoked to reduce the estimator design space before optimization.
  • domain assumption The minimax variance problem over physical states can be exactly converted to an SDP.
    Central step enabling the offline design.

pith-pipeline@v0.9.0 · 5409 in / 1283 out tokens · 22454 ms · 2026-05-09T14:34:01.534104+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

13 extracted references · 13 canonical work pages

  1. [1]

    Certifying almost all quantum states with few single-qubit measurements,

    H.-Y . Huang, J. Preskill, and M. Soleimanifar, “Certifying almost all quantum states with few single-qubit measurements,”Nature Physics, vol. 21, no. 11, pp. 1834–1841, 2025

    work page 2025

  2. [2]

    Direct fidelity estimation from few Pauli measurements,

    S. T. Flammia and Y .-K. Liu, “Direct fidelity estimation from few Pauli measurements,”Physical Review Letters, vol. 106, no. 23, p. 230501, 2011

    work page 2011

  3. [3]

    Predicting many properties of a quantum system from very few measurements,

    H.-Y . Huang, R. Kueng, and J. Preskill, “Predicting many properties of a quantum system from very few measurements,”Nature Physics, vol. 16, no. 10, pp. 1050–1057, 2020

    work page 2020

  4. [4]

    Learning quantum states from their classical shadows,

    H.-Y . Huang, “Learning quantum states from their classical shadows,” Nature Reviews Physics, vol. 4, no. 2, pp. 81–81, 2022

    work page 2022

  5. [5]

    Sampling groups of Pauli operators to enhance direct fidelity estimation,

    J. Barber `a-Rodr´ıguez, M. Navarro, and L. Zambrano, “Sampling groups of Pauli operators to enhance direct fidelity estimation,”Quantum, vol. 9, p. 1784, 2025

    work page 2025

  6. [6]

    Efficient sampling for Pauli-measurement-based shadow tomography in direct fidelity estimation,

    H. Cha and J. Lee, “Efficient sampling for Pauli-measurement-based shadow tomography in direct fidelity estimation,”Physical Review A, vol. 112, no. 3, p. 032427, 2025

    work page 2025

  7. [7]

    Operator-aware shadow importance sampling for accurate fidelity estimation,

    H. Cha, S. Hong, and J. Lee, “Operator-aware shadow importance sampling for accurate fidelity estimation,”Physical Review A, vol. 113, no. 4, p. 042602, 2026

    work page 2026

  8. [8]

    Shadow tomography based on informationally complete positive operator-valued measure,

    A. Acharya, S. Saha, and A. M. Sengupta, “Shadow tomography based on informationally complete positive operator-valued measure,” Physical Review A, vol. 104, no. 5, p. 052418, 2021

    work page 2021

  9. [9]

    Shadow tomography on general measurement frames,

    L. Innocenti, S. Lorenzo, I. Palmisano, F. Albarelli, A. Ferraro, M. Pater- nostro, and G. M. Palma, “Shadow tomography on general measurement frames,”PRX Quantum, vol. 4, no. 4, p. 040328, 2023

    work page 2023

  10. [10]

    Optimizing quantum tomography via shadow inversion,

    A. Caprotti, J. Morris, and B. Daki ´c, “Optimizing quantum tomography via shadow inversion,”Physical Review Research, vol. 6, no. 3, p. 033301, 2024

    work page 2024

  11. [11]

    Optimal data processing for quantum measurements,

    G. M. D’Ariano and P. Perinotti, “Optimal data processing for quantum measurements,”Physical Review Letters, vol. 98, no. 2, p. 020403, 2007

    work page 2007

  12. [12]

    On general minimax theorems,

    M. Sion, “On general minimax theorems,”Pacific Journal of Mathe- matics, vol. 8, no. 1, pp. 171–176, 1958

    work page 1958

  13. [13]

    Der Massbegriff in der Theorie der kontinuierlichen Grup- pen,

    A. Haar, “Der Massbegriff in der Theorie der kontinuierlichen Grup- pen,”Annals of Mathematics, vol. 34, no. 1, pp. 147–169, 1933

    work page 1933