Pith. sign in

REVIEW 2 major objections 5 minor 43 references

$d+1$ Measurement Bases are Sufficient for Determining $d$-Dimensional Quantum States: Theory and Experiment

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

Pith's one-line read d+1 projective measurement bases determine an arbitrary d-dimensional quantum state, with a proof for all d and a d=6 chip demonstration.

desk verdict A correct and clean explicit construction of d+1 tomographic bases, but the sufficiency claim is not new and the phase-existence gap is easy to fix. read the letter →

arxiv 2507.11204 v2 pith:AQ4WKVCG submitted 2025-07-15 quant-ph

classification quant-ph
keywords quantumstatetomographyminimalmeasurementbasesmutuallyunbiaseddensitymatrixreconstructionVandermondesiliconphotonicchipsix-dimensionalprojectivemeasurements
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

Full quantum state tomography normally requires on the order of $d^2$ measurements, and the best prior schemes that work for every dimension use $2d-1$ projective bases. This paper claims that $d+1$ projective measurement bases are sufficient to reconstruct an arbitrary $d$-dimensional quantum state, exactly matching the lower bound set by parameter counting. The construction uses the computational basis plus the Fourier basis and its $d-1$ phase-shifted variants, with phases chosen so that the reconstruction matrix is invertible. The authors demonstrate the scheme for $d=6$ on a silicon photonic chip, a dimension where complete mutually unbiased bases are unavailable, recovering two pure states and one mixed state with fidelities above $0.96$. If the construction holds, this settles the minimal number of projective measurement bases needed for complete tomography in every dimension.

What carries the argument

The central mechanism is the coefficient matrix $T$ defined elementwise by $T_{(d-1)j+k,\,d(c-1)+a}=\omega^{kc}e^{-i(\theta_a^{(j)}-\theta_{(a+c)\bmod d}^{(j)})}$, which maps the $d(d-1)$ measured quantities $q$ to the off-diagonal density-matrix elements $g$. The proof shows that after fixed full-rank row transformations, $T$ becomes block-diagonal with blocks $N^{(c)}$; under the phase choice $\theta_m^{(j)}=j\theta_m^{(1)}$ each block is a Vandermonde matrix (a matrix whose columns are powers of distinct numbers, invertible exactly when those numbers are distinct). Its determinant is a product of differences $e^{-i(\theta_t^{(1)}-\theta_{(t+c)\bmod d}^{(1)})}-e^{-i(\theta_{t'}^{(1)}-\theta_{(t'+c)\bmod d}^{(1)})}$, so the distinct-difference condition in Eq. (11) is exactly the condition that no Vandermonde determinant vanishes, making $T$ invertible and the reconstruction $g=T^{-1}q$ possible.

What would settle it

A direct falsifier would be to find a dimension $d$ for which no choice of $\theta^{(1)}_0,\ldots,\theta^{(1)}_{d-1}$ makes the differences $\theta^{(1)}_t-\theta^{(1)}_{(t+c)\bmod d}$ pairwise distinct modulo $2\pi$; then some Vandermonde block $N^{(c)}$ in the Supplemental proof has a zero determinant, $T$ is singular, and the $d+1$ bases fail. Checking the determinant for candidate phases at $d=7,8,\ldots$ is a concrete search.

Watch

Extended reading notes

Core claim

The paper's central claim is that the $d+1$ projective bases consisting of the computational basis, the Fourier basis $\{|\psi_k^{(0)}\rangle\}_{k=0}^{d-1}$, and the $d-1$ bases obtained by applying diagonal unitaries $R^{(j)}$ with phases $\theta_m^{(j)}$ to the Fourier basis completely determine any $d$-dimensional density matrix $\rho$. The computational basis gives the diagonal elements directly; the remaining $d$ bases produce the linear system $q=T g$, where $g$ collects the $d(d-1)$ off-diagonal elements $g_{a,b}$ and $q$ collects the measured quantities $q_k^{(j)}=d\,p_k^{(j)}-1$. The paper proves that $T$ is invertible whenever $\theta_m^{(j)}=j\theta_m^{(1)}$ and the differences $\theta_t^{(1)}-\theta_{(t+c)\bmod d}^{(1)}$ are all distinct modulo $2\pi$, so the off-diagonal elements are uniquely recovered as $g=T^{-1}q$. It reports a $d=6$ experiment on a silicon photonic chip that reconstructs two pure states and one mixed state with fidelities above $0.96$, in a dimension where no complete set of mutually unbiased bases is known.

Load-bearing premise

The load-bearing premise is that for every dimension $d$ a choice of phases satisfying Eq. (11) exists—the paper gives a numerical example only for $d=6$—and that, experimentally, the chip implements those bases with negligible calibration error.

Editorial extensions

If this is right

  • If the construction is correct, full tomography of any $d$-dimensional state needs only $d+1$ projective bases, meeting the parameter-counting lower bound and improving on the previous universal $2d-1$ scheme.
  • The scheme applies to dimensions where complete mutually unbiased bases are not known, such as $d=6$, so it removes a long-standing obstacle to minimal-basis tomography.
  • The reconstruction is explicit: diagonal elements come from computational-basis probabilities and off-diagonal elements from $g=T^{-1}q$, so no iterative search is required in the ideal case.
  • The $d=6$ experiment shows the bases can be realized on a programmable silicon photonic chip, with fidelities above $0.96$.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A generic choice of phases $\theta_m^{(1)}$ satisfies the distinct-difference condition with probability one, so the $d+1$-basis scheme does not depend on fine-tuned values; almost any phases should work.
  • The remaining freedom in the phases can be used to control noise: minimizing the condition number of $T$ (extending the paper's one-parameter optimization for $d=6$) would make the reconstruction more stable for larger dimensions.
  • For states known to be low-rank, the same measurement design might allow reconstruction with fewer than $d+1$ bases, though this is an extension rather than a claim of the paper.
  • When comparing tomography schemes, reporting the condition number of $T$ alongside fidelity would sharpen the practical assessment, since measurement noise is amplified by $T^{-1}$.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 5 minor

Summary. The manuscript proposes a quantum state tomography protocol using d+1 projective measurement bases: the computational basis, the Fourier basis, and d-1 variants obtained by applying diagonal unitary operators R^(j) with phases θ_m^(j)=jθ_m^(1). The authors derive a linear system q = T g that connects the measured basis probabilities to the off-diagonal density-matrix elements and show, in the Supplemental Material, that T decomposes into Vandermonde blocks and is invertible when the first-order phase differences in Eq. (S14) are distinct modulo 2π. The state is then reconstructed as g = T^{-1}q. The paper also reports a silicon-photonic-chip experiment for d=6, reconstructing three states with fidelities above 0.96.

Significance. If the missing existence argument is supplied, the scheme gives an explicit family of d+1 bases saturating the parameter-counting lower bound in every dimension, including composite dimensions such as d=6 where a complete set of MUBs is not known. The algebraic derivation is transparent: the Supplemental Vandermonde reduction is sound, and the conditional invertibility criterion is clearly stated. The experiment is a useful proof-of-principle, though its quantitative claims are weakened by the absence of calibration and systematic-uncertainty analysis. The main theoretical gap is that the paper never proves that phases satisfying Eq. (S14) exist for arbitrary d; this is readily fixable by an explicit construction.

major comments (2)
  1. [Supplemental Material, Eqs. (S11) and (S14); main text Eq. (11)] The invertibility theorem is conditional: T is invertible only when phases satisfying Eqs. (S11) and (S14) are chosen, but the manuscript never proves that such phases exist for every dimension d. The main text supplies only the numerical d=6 choice θ_m^(1)=0.5671 m^2 in footnote [42]. Since the claim that the protocol 'applies to arbitrary dimension d' rests on this existence, the theorem is incomplete as stated. The gap is easily repaired: take θ_m^(1)=πα m^2 with α irrational; then, for each c, the difference θ_t^(1)-θ_(t+c mod d)^(1) = -πα(2ct+c^2), and equality of two such differences modulo 2π would force α c(t-t') to be an integer, impossible for α irrational and t≠t'. Please add this existence argument explicitly and phrase the theorem as 'for every d there exist phases satisfying Eq. (11)'.
  2. [Experimental section, Fig. 2 and text near Eq. (14)] The experimental verification does not include a calibration or systematic-uncertainty analysis for the seven implemented unitaries. The reconstruction assumes that the chip realizes exactly the computational basis, the Fourier basis, and the diagonal-phase-varied Fourier bases; residual phase and transmission errors in the MZI network directly bias the inferred density matrices, and this bias is not reflected in the quoted statistical uncertainties. Please characterize the implemented bases (e.g., through interferometric calibration or tomographic characterization of the unitaries) and propagate systematic errors, or explicitly limit the experimental claim to a proof-of-principle demonstration.
minor comments (5)
  1. [Abstract and Introduction] The abstract states that in d=6 'a complete set of mutually unbiased bases does not exist,' but the introduction correctly says that 'only three MUBs have been found to date' and that the construction remains an open problem. The nonexistence of a complete set of seven MUBs in dimension six is not an established fact; please rephrase the abstract (e.g., 'for which a complete set is not known').
  2. [Eq. (11) and Eq. (S14)] The condition written as 'θ_t^(1)-θ_(t+c mod d)^(1) ≠ θ_t'^(1)-θ_(t'+c mod d)^(1)+2πn' is clearer as a congruence statement: the two differences should be non-congruent modulo 2π. Please adjust the notation.
  3. [Experimental section, paragraph before Fig. 2] The maximum-likelihood estimation step is invoked but not described; please specify the likelihood model, the parameterization of the density matrix, and the constraints (e.g., positivity and trace) used to obtain the reconstructed matrices in Fig. 2, so that the reader can separate the effect of the linear inversion from that of the estimator.
  4. [Fig. 1 caption] The caption defines VOA, FPC, and SNSPD but not DAC; please define the acronym and state the role of the 'Classical Processing' block (e.g., whether it performs only counting or also unitary calibration).
  5. [Footnote [42]] For the numerical optimization that yielded φ=0.5671, please state the domain of the search and the tolerance, and clarify that exact optimality of the mutual-unbiasedness deviation is not required for invertibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the d+1-basis reconstruction is a direct linear inversion with basis phases chosen independently of the target states.

full rationale

The derivation is self-contained and does not reduce any prediction to a fitted input. The diagonal elements are taken directly from computational-basis probabilities in Eq. (6), and the off-diagonal elements are obtained by solving the linear system q = T g in Eq. (10), with T defined explicitly from the chosen measurement phases. Invertibility of T is proved in the Supplemental Material by reducing T to block-diagonal Vandermonde matrices N(c); the required phase conditions (S11) and (S14) are state-independent conditions on the measurement design, not conditions on the reconstructed density matrix. The experimental phase phi = 0.5671 for d = 6 is selected by minimizing a state-independent mutual-unbiasedness deviation f (footnote [42]), not by fitting to the prepared states or to the measured fidelities, so no fitted parameter is renamed as a prediction. The measured fidelities are computed against independently chosen known states (Eq. 13), providing an external check rather than an assumed result. Citations to prior tomography and MUB work are contextual or comparative and are not load-bearing for the invertibility proof, which is contained in the paper itself. The only notable caveat is that the paper does not prove the existence of phases satisfying Eq. (S14) for arbitrary dimension d, giving a numerical instance only for d = 6; this is a mathematical completeness or correctness gap, not circularity, since even a generic-phase existence argument would not make the derivation assume the target quantum-state reconstruction.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The central inversion is a linear algebra result over the d(d-1) off-diagonal density matrix amplitudes. No new physical entities are introduced. The only adjustable inputs are the phase parameters, and the main unstated input is the implicit existence of non-degenerate phases in every dimension.

free parameters (2)
  • Diagonal phase scale phi = 0.5671
    Chosen for d=6 by numerically minimizing the mutual-unbiasedness deviation f in footnote [42]. It is not fitted to the reconstructed states and is not needed for the theorem's generic existence argument; it is a design parameter for the experimental basis set.
  • Diagonal phase set theta_m^(1) = phi m^2 for d=6; otherwise constrained by Eq. (11)
    The invertibility proof requires phases satisfying Eq. (11). The paper leaves the general existence implicit, so this is an adjustable input to the construction rather than a constant fixed by the physics.
assumptions (4)
  • domain assumption Each projective measurement basis yields d-1 independent probabilities through the Born rule.
    Used in the parameter count and in Eq. (9), where the k=d-1 probability is excluded because probabilities sum to one. This is standard for projective measurements.
  • standard math A Vandermonde matrix with distinct nodes is invertible.
    This is the key tool in Supplemental Eq. (S13), used to show each block N(c) is nonsingular when the phase differences are distinct modulo 2 pi.
  • domain assumption The density operator is Hermitian with trace one, so diagonal elements are probabilities and off-diagonal elements carry the remaining information.
    This structure is assumed in Eqs. (1) and (6). It is standard in quantum state tomography.
  • ad hoc to paper A phase choice satisfying Eq. (11) exists in every dimension d.
    The paper states the sufficient condition and gives a numerical phi for d=6, but it does not explicitly prove that such phases exist for all d. Generic irrational phases would satisfy the open condition, so the gap is minor but real.

how reviews work

0 comments
Cite this review

Pith. "Pith review of $d+1$ Measurement Bases are Sufficient for Determining $d$-Dimensional Quantum States: Theory and Experiment." pith.science (2026). https://pith.science/paper/AQ4WKVCG

@misc{pith2026250711204,
  author       = {Pith},
  title        = {Pith review of: $d+1$ Measurement Bases are Sufficient for Determining $d$-Dimensional Quantum States: Theory and Experiment},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/AQ4WKVCG}},
  note         = {Machine review of arXiv:2507.11204}
}
abstract

A long-standing problem in quantum physics is to determine the minimal number of measurement bases required for the complete characterization of unknown quantum states, a question of particular relevance to high-dimensional quantum information processing. Here, we propose a quantum state tomography scheme that requires only $d+1$ projective measurement bases to fully reconstruct an arbitrary $d$-dimensional quantum state. As a proof-of-principle, we experimentally verified this scheme on a silicon photonic chip by reconstructing quantum states for $d=6$, in which a complete set of mutually unbiased bases does not exist. This approach offers new perspectives for quantum state characterization and measurement design, and holds promise for future applications in quantum information processing.

Figures

Figures reproduced from arXiv: 2507.11204 by the authors.

Figure 1
Figure 1. FIG. 1. Schematic of the silicon quantum photonic chip and external setup. The system consists of three functional modules: (1) Single [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2. Experimentally reconstructed density matrices. (a), (c), and (e) show the real parts, while (b), (d), and (f) show the imaginary parts of [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

43 extracted references · 40 canonical work pages

  1. [42]

    S. Xue, Y . Wang, J. Zhan, Y . Wang, R. Zeng, J. Ding, W. Shi, Y . Liu, Y . Liu, A. Huang, et al. , Physical Review Letters 129, 133601 (2022)

  2. [1]

    Paris and J

    M. Paris and J. Rehacek, Quantum state estimation , V ol. 649 (Springer Science & Business Media, 2004)

  3. [2]

    The fi- delities between the reconstructed density matrices ρ′ and the ideal statesρ are calculated as F = tr √ ρ1/2ρ′ρ1/2, (14) 4 yielding values of 0 .9654 ± 0.0067, 0.9698 ± 0.0042, and 0.9761 ± 0.0033, respectively. In summary, we have theoretically proposed and exper- imentally demonstrated a QST scheme that reconstructs d- dimensional quantum states ...

  4. [3]

    H ¨affner, W

    H. H ¨affner, W. H ¨ansel, C. Roos, J. Benhelm, D. Chek-al Kar, M. Chwalla, T. K ¨orber, U. Rapol, M. Riebe, P . Schmidt, et al., Nature 438, 643 (2005)

  5. [4]

    D. P . DiVincenzo, Science 270, 255 (1995)

  6. [5]

    J. L. O’brien, Science 318, 1567 (2007)

  7. [6]

    C. H. Bennett, G. Brassard, C. Cr ´epeau, R. Jozsa, A. Peres, and W. K. Wootters, Physical review letters 70, 1895 (1993)

  8. [7]

    Bouwmeester, J.-W

    D. Bouwmeester, J.-W. Pan, K. Mattle, M. Eibl, H. Weinfurter, and A. Zeilinger, Nature 390, 575 (1997)

Show all 43 references
  1. [8]

    Giovannetti, S

    V . Giovannetti, S. Lloyd, and L. Maccone, Physical review let- ters 96, 010401 (2006)

  2. [9]

    Giovannetti, S

    V . Giovannetti, S. Lloyd, and L. Maccone, Nature photonics 5, 222 (2011)

  3. [10]

    D. F. James, P . G. Kwiat, W. J. Munro, and A. G. White, Phys- ical Review A 64, 052312 (2001)

  4. [11]

    R. T. Thew, K. Nemoto, A. G. White, and W. J. Munro, Physi- cal Review A 66, 012303 (2002)

  5. [12]

    F. T. Hioe and J. H. Eberly, Physical Review Letters 47, 838 (1981)

  6. [13]

    R. A. Bertlmann and P . Krammer, Journal of Physics A: Math- ematical and Theoretical 41, 235303 (2008)

  7. [14]

    Kimura, Physics Letters A 314, 339 (2003)

    G. Kimura, Physics Letters A 314, 339 (2003)

  8. [15]

    I. P . Menda ˇs, Journal of Physics A: Mathematical and General 39, 11313 (2006)

  9. [16]

    R. G. Newton and B.-L. Y oung, Annals of Physics 49, 393 (1968)

  10. [17]

    M. A. Perlin, D. Barberena, and A. M. Rey, Physical Review A 104, 062413 (2021)

  11. [18]

    Y . Wang, H. Jiang, Y . Liu, and K. Li, arXiv preprint arXiv:2409.03435 (2024)

  12. [19]

    Ivonovic, Journal of Physics A: Mathematical and General 14, 3241 (1981)

    I. Ivonovic, Journal of Physics A: Mathematical and General 14, 3241 (1981)

  13. [20]

    W. K. Wootters and B. D. Fields, Annals of Physics 191, 363 (1989)

  14. [21]

    Zauner, International Journal of Quantum Information 9, 445 (2011)

    G. Zauner, International Journal of Quantum Information 9, 445 (2011)

  15. [22]

    Durt, B.-G

    T. Durt, B.-G. Englert, I. Bengtsson, and K. ˙Zyczkowski, In- ternational journal of quantum information 8, 535 (2010)

  16. [23]

    Horodecki, Ł

    P . Horodecki, Ł. Rudnicki, and K. ˙Zyczkowski, PRX Quantum 3, 010101 (2022)

  17. [24]

    McNulty and S

    D. McNulty and S. Weigert, arXiv preprint arXiv:2410.23997 (2024)

  18. [25]

    Bandyopadhyay, P

    S. Bandyopadhyay, P . O. Boykin, V . Roychowdhury, and F. V atan, Algorithmica34, 512 (2002)

  19. [26]

    Klappenecker and M

    A. Klappenecker and M. R ¨otteler, in Finite Fields and Appli- cations: 7th International Conference, Fq7, Toulouse, France, May 5-9, 2003. Revised Papers (Springer, 2004) pp. 137–144

  20. [27]

    Archer, Journal of mathematical physics 46 (2005)

    C. Archer, Journal of mathematical physics 46 (2005)

  21. [28]

    Brierley, S

    S. Brierley, S. Weigert, and I. Bengtsson, arXiv preprint arXiv:0907.4097 (2009)

  22. [29]

    Adamson and A

    R. Adamson and A. M. Steinberg, Physical review letters 105, 030406 (2010)

  23. [30]

    G. Lima, L. Neves, R. Guzm ´an, E. S. G ´omez, W. Nogueira, A. Delgado, A. V argas, and C. Saavedra, Optics Express 19, 3542 (2011)

  24. [31]

    Giovannini, J

    D. Giovannini, J. Romero, J. Leach, A. Dudley, A. Forbes, and M. J. Padgett, Physical review letters 110, 143601 (2013)

  25. [32]

    Bengtsson, W

    I. Bengtsson, W. Bruzda, Å. Ericsson, J.-Å. Larsson, W. Tadej, and K. ˙Zyczkowski, Journal of mathematical physics 48 (2007)

  26. [33]

    Grassl, arXiv preprint quant-ph /0406175 (2004)

    M. Grassl, arXiv preprint quant-ph /0406175 (2004)

  27. [34]

    Jaming, M

    P . Jaming, M. Matolcsi, P . M ´ora, F. Sz ¨oll˝osi, and M. Weiner, Journal of Physics A: Mathematical and Theoretical 42, 245305 (2009)

  28. [35]

    Butterley and W

    P . Butterley and W. Hall, Physics Letters A 369, 5 (2007)

  29. [36]

    Brierley and S

    S. Brierley and S. Weigert, Physical Review A—Atomic, Molecular, and Optical Physics 79, 052316 (2009)

  30. [37]

    Raynal, X

    P . Raynal, X. L ¨u, and B.-G. Englert, Physical Review A—Atomic, Molecular, and Optical Physics 83, 062303 (2011)

  31. [38]

    Goyeneche, Journal of Physics A: Mathematical and Theo- retical 46, 105301 (2013)

    D. Goyeneche, Journal of Physics A: Mathematical and Theo- retical 46, 105301 (2013)

  32. [39]

    W. R. Clements, P . C. Humphreys, B. J. Metcalf, W. S. Kolthammer, and I. A. Walmsley, Optica 3, 1460 (2016)

  33. [40]

    Li, G.-F

    Z.-H. Li, G.-F. Y u, Y .-X. Wang, Z.-Y . Xing, L.-W. Kong, and X.-Q. Zhou, Science China Physics, Mechanics & Astronomy 66, 290311 (2023)

  34. [41]

    Peruzzo, J

    A. Peruzzo, J. McClean, P . Shadbolt, M.-H. Y ung, X.-Q. Zhou, P . J. Love, A. Aspuru-Guzik, and J. L. O’brien, Nature com- munications 5, 4213 (2014)

  35. [43]

    d + 1 Measurement Bases Are Su fficient for Determining d-Dimensional Quantum States: Theory and Experiment

    To select the diagonal phases, we require that θ( j) m satisfy Eq. ( 11), ensuring T is invertible. To fur- ther optimize the scheme performance [ 28], we min- imize the mutual-unbiasedness deviation function f = ∑ 0≤ j<u≤d−1 ∑d−1 k,v=0 (|⟨ψ( j) k |ψ(u) v ⟩| − d−1/2)2. Accordi...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.