Pith. sign in

REVIEW 2 major objections 4 minor 39 references

Quantum codes do not fix isotropic errors

T0 review · 2 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read This paper proves that quantum error-correcting codes cannot reduce the variance of isotropic quantum errors, and that any detected error leaves the corrected logical qubits uniformly distributed.

desk verdict A clean uniform-on-detection result, but the headline no-go claim is false as stated: under the paper's own variance, isotropic errors peaked near the antipode are corrected. read the letter →

arxiv 2502.07075 v1 pith:QZ74ODPK submitted 2025-01-27 quant-ph

Jesús García López de Lacalle , Luis Miguel Pozo Coronado , André L. Fonseca de Oliveira This is my paper · ORCID
classification quant-ph MSC 81P7081P68 PACS 03.67.Pp
keywords quantumerror-correctingcodesisotropiccomputingerrorserrorvariancerandomvariablesonspheresyndromedetectionuniformdistributioncorrectionlimitation
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

The paper tries to establish a limitation of quantum error correction: for a broad class of noise called isotropic errors, applying the code's correction circuit never reduces the variance of the error, even on the idealized assumption that the correction circuit adds no new noise. It models an n-qubit state as a point on a real sphere and an error as a random variable whose density depends only on the distance from the encoded state. For every isotropic density that is non-increasing on the whole interval, supported only before the equator, or symmetric and decreasing around the equator, the proof shows that the variance after correction is at least the variance before correction. It also proves that whenever a nonzero error syndrome is detected, the corrected logical state is uniformly distributed, so all computing information is already lost. If the argument is correct, standard quantum error correction cannot be said to fix isotropic errors when success is measured by variance.

What carries the argument

The central object is the real-sphere representation of quantum states: an $n$-qubit state is a point on the unit sphere of dimension $2^{n+1}-1$, written in spherical coordinates so that the reference state $\Phi=|0\rangle$ sits at the pole and isotropy means the density $f(\theta_0)$ depends only on the polar angle $\theta_0$, equivalently only on $\|\Phi-\Psi\|$. The comparison of variances reduces to a series of integrals of $\cos(\theta_0)\sin^{2d+2k-2}(\theta_0)$, and Theorem 5 holds exactly when these integrals are nonnegative. The geometric engine of the proof is that an isotropic density is constant on the parallels of the sphere, so projecting onto a syndrome subspace and applying the inverse error operator produces a uniform distribution on the code subspace for every detected error.

What would settle it

Choose a specific code and an isotropic density that violates all three conditions of Theorem 5, for instance one with a sharp peak at $\theta_0 = 3\pi/4$, and numerically evaluate $V(\tilde{\Phi})-V(\Psi)$ from the formulas of Lemma 3; a negative value would refute the unconditional version of the claim. Alternatively, compute the phase-invariant quantum variance $V_q$ before and after correction for a normal isotropic error in a 5-qubit code: if $V_q$ decreases while $V$ does not, the no-go is a property of the chosen variance measure rather than of information recovery.

Watch

Extended reading notes

Core claim

The central claim is Theorem 5: for an isotropic error $\Psi$ with density $f(\theta_0)$ that (a) is non-increasing on $[0,\pi]$, (b) vanishes for $\theta_0 \geq \pi/2$, or (c) satisfies $f(\pi/2-\theta) \geq f(\pi/2+\theta)$, the corrected state $\tilde{\Phi}$ satisfies $V(\tilde{\Phi})-V(\Psi) \geq 0$. Because the correction circuit is assumed to introduce no new errors, the gap is an intrinsic property of the code and not a hardware artifact. The mechanism is Theorem 3: for any nonzero syndrome $s$, meaning the measurement has flagged a specific discrete error $E_s$, the post-correction state $E_s^{-1}\Pi_s \Psi$ is uniform over the code subspace, so the logical $m$-qubit carries no computational information. The paper's stated conclusion is that quantum error-correcting codes do not fix isotropic errors.

Load-bearing premise

The load-bearing premise is that an error is well described by an isotropic random variable on the state sphere and that its size is the plain variance $V = E[\|\Psi-\Phi\|^2]$, a measure that depends on the unphysical global phase rather than on a directly observable quantity like fidelity.

Editorial extensions

If this is right

  • Under the conditions of Theorem 5, error correction cannot lower the variance of an isotropic error; the best possible outcome is equality, and for normal isotropic errors the inequality is strict.
  • If a correction circuit detects a nonzero syndrome for an isotropic error, the corrected logical state is uniform, so the computation cannot be recovered by further correction or post-processing.
  • These bounds hold in the idealized case where the correction circuit adds no new errors, so they constrain the theoretical capability of any quantum error-correcting code, not a particular implementation.
  • A code that appears to correct isotropic noise in practice must be judged by a different measure than this variance, or by an error model that is not isotropic.

Reading between the lines

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

  • The proof establishes nonnegativity for three large but not exhaustive families of isotropic densities; the fully unconditional statement that no code fixes isotropic errors is stronger than what Theorem 5 demonstrates, since the sign of the integrals is what carries the conclusion.
  • Because the plain variance $V$ depends on the unphysical global phase, the no-go result does not by itself imply that information cannot be recovered; a code might still increase fidelity or reduce the phase-invariant variance $V_q$ for the same isotropic error.
  • The uniformization mechanism is purely geometric, so a similar no-go may extend to any noise ensemble that is invariant under enough rotations of the state sphere to make the syndrome-subspace projections uniform, not only strictly isotropic errors.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 4 minor

Summary. The paper models isotropic quantum computing errors as random variables on the real sphere associated with an n-qubit state, defines the variance V(X)=E[||X−Φ||²] with respect to the no-error state Φ=|0>, and compares V before and after application of a quantum error correcting code's correction circuit, under the assumption that the correction circuit introduces no new errors. It proves (Theorem 3) that conditional on a nonzero syndrome, the corrected state is uniform over the code subspace, and (Theorem 5) that under certain monotonicity, support, or symmetry conditions on the isotropic density, V(Φ̃)−V(Ψ)≥0, concluding that codes do not fix isotropic errors. The abstract and conclusions state the no-fix result without the conditions of Theorem 5.

Significance. If the unconditional claim were true, it would be a significant negative result for quantum error correction. The paper does contain a clear geometric derivation of the conditional variance comparison, and the uniformization statement for detected isotropic errors is interesting and appears to be a genuine structural observation. However, the central claim as stated is false: Theorem 5 is conditional, and a simple isotropic density concentrated near θ0=3π/4 violates the theorem's conditions and makes the code reduce the paper's own variance. In addition, the paper's chosen measure V is not phase-invariant and hence not a physical error measure, and the no-go does not transfer to the paper's own Vq or to fidelity. The correct conditional statement is much weaker than the title and abstract claim.

major comments (2)
  1. [Section 4, Theorem 5; Abstract] The abstract states without qualification that quantum error correcting codes do not fix isotropic errors, citing Theorem 5, but Theorem 5 only proves V(Φ̃)−V(Ψ)≥0 for densities satisfying one of (a) non-increasing, (b) support in [0,π/2], or (c) f(π/2−θ)≥f(π/2+θ). The proof reduces the difference to a positive constant times Σ_{k≥1} c_k E[cosθ0 sin^{2d+2k−2}θ0], so the conclusion depends exactly on these expectations being nonnegative. For an isotropic density concentrated near θ0=3π/4, every such expectation is negative because cosθ0<0 and sinθ0>0 on the support, and all three conditions fail; the code then strictly reduces the authors' own variance V. Thus the paper's central claim is false for the class of errors it defines, and the abstract's and Section 5's statements that isotropic errors are uncontrollable are not supported.
  2. [Section 1, definitions of V and Vq] The paper defines 'fix' as reduction of V(X)=E[||Ψ−Φ||²], where Ψ and Φ are represented as points on the real sphere with coordinates including an unphysical global phase. States differing only by a global phase are physically identical, and the paper itself defines Vq(X)=E[min_φ ||Ψ−e^{iφ}Φ||²] and notes that Vq is equivalent to fidelity. Since Vq(X)≤V(X) pointwise, the theorem's conclusion V(Φ̃)≥V(Ψ) does not imply Vq(Φ̃)≥Vq(Ψ); the code could reduce the physically meaningful quantum variance or increase fidelity while increasing V. Hence the title's claim about quantum codes is not established by the theorem. A no-go for Vq or fidelity would be needed to support the physical conclusion.
minor comments (4)
  1. [Section 4, proof of Theorem 3] In the proof of Theorem 3, the sentence 'this random variable is also constant' should read 'this random variable has constant density'; as written it suggests the corrected state is deterministic, contradicting the uniform distribution established in the same paragraph.
  2. [Section 4, proof of Theorem 5] The variance difference is described as 'positive' in the proof of Theorem 5, but the theorem states nonnegativity; equality can occur, for example, for a uniform isotropic density, for which every E[cosθ0 sin^{2d+2k−2}θ0] vanishes.
  3. [Section 4, Lemma 3(d)] The derivation of Lemma 3(d) is summarized as 'concluded in a simple way' after an infinite series expansion; because this lemma carries the sign information used in Theorem 5, the intermediate algebra should be shown in full.
  4. [General] The manuscript refers to 'Formula (8)' and 'Figure 4' without equation or figure numbers; numbering and callouts should be added throughout.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the central no-go theorem is derived from explicit definitions and self-contained calculations, with only a minor non-load-bearing self-citation for the isotropic-error framework.

full rationale

The paper's derivation chain is self-contained once its definitions are granted. Isotropic errors (Definition 1), the plain variance V (Definition 2), and the code-correction model (Section 2) are all stated explicitly in the paper. Theorem 5 computes V(Φ̃) − V(Ψ) from Theorems 1 and 4 via Lemmas 2 and 3; the difference is an explicit sum of terms proportional to E[cos(θ0) sin^{2d+2k−2}(θ0)], and conditions (a)–(c) are exactly the conditions that make each of those expectations nonnegative. No parameter is fitted to data, no external benchmark is used, and no previously published theorem is invoked to force the conclusion. The only self-citation is [13], which supplies the random-variable representation of quantum computing errors and the variance definition; these are definitions and prior formalism that the present paper restates, not unverified evidence for the no-go result itself. Therefore there is no circular reduction of the central claim. A separate concern, not a circularity, is that the abstract's unqualified wording 'quantum error correcting codes do not fix isotropic errors' exceeds what Theorem 5 actually proves, since the theorem covers only non-increasing densities, densities supported in [0, π/2], or densities symmetric-decreasing about π/2; that is a correctness or overclaim issue, not a self-reference or definitional equivalence.

Assumptions & free parameters 0 free parameters · 5 assumptions · 0 invented entities

The central proof rests on a nonstandard model of quantum errors as random pure states on the unit sphere, introduced in the authors' earlier paper [13]. Within that model, the derivations are self-contained and use standard spherical geometry and integral formulas from the appendix. No new physical entities are introduced, and no parameters are fitted to data. The use of the phase-sensitive variance V is the principal modeling axiom.

assumptions (5)
  • domain assumption Quantum computing errors are modeled as random variables on the unit sphere, with a density function f(x) over pure states.
    Introduced in Section 1 and used throughout. This is a nonstandard model, carried over from the authors' prior work [13].
  • domain assumption The correction circuit of the quantum code does not introduce new errors.
    Stated in Section 1: 'we will assume that the corrector circuit of C does not introduce new errors.' This gives an ideal upper bound on code performance.
  • domain assumption The code is non-degenerate and the syndrome subspaces S_s = E_s(C) form an orthogonal direct sum decomposition of the full space.
    Section 2, Formula (3): H_n = S_0 ⊥ S_1 ... ⊥ S_{d''−1}. This excludes degenerate codes.
  • domain assumption The mean of every quantum error can be taken as Φ = |0⟩ without loss of generality.
    Section 1: 'Without loss of generality, we will assume that the mean value of every quantum computing error will always be Φ = |0⟩.' This is a normalization choice.
  • ad hoc to paper The plain variance V(X) = E[||X − μ||²] is the appropriate measure of error for comparing corrected and uncorrected states.
    Section 1 defines V and Vq, then chooses V as the measure that 'best reflects the distribution.' This choice is not physically compelled because V is not invariant under global phase, and the paper does not prove the main results for Vq or fidelity.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum codes do not fix isotropic errors." pith.science (2026). https://pith.science/paper/QZ74ODPK

@misc{pith2026250207075,
  author       = {Pith},
  title        = {Pith review of: Quantum codes do not fix isotropic errors},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QZ74ODPK}},
  note         = {Machine review of arXiv:2502.07075}
}
abstract

In this work we prove that quantum error correcting codes do not fix isotropic errors, even assuming that their correction circuits do not introduce new errors. We say that a quantum code does not fix a quantum computing error if its application does not reduce the variance of the error. We also prove for isotropic errors that, if the correction circuit of a quantum code detects an error, the corrected logical $m-$qubit has uniform distribution and as a result, it already loses all the computing information.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

39 extracted references · 31 canonical work pages

  1. [1]

    Gaitan, F., Quantum error correction and fault tolerant quantum computing, CRC Press, 2008

    work page 2008

  2. [2]

    Calderbank, A.R., Shor, P.W., Good quantum error-correcting codes exist, Phys. Rev. A 54, 1098–1105 (1996)

    work page 1996

  3. [3]

    Steane, A.M., Multiple particle inference and quantum error correction, Proc. Roy. Soc. A. 452, 2551 (1996)

    work page 1996

  4. [4]

    PhD thesis, California Institute of Technology, 1997

    Gottesman, D., Stabilizer Codes and Quantum Error Correction. PhD thesis, California Institute of Technology, 1997

    work page 1997

  5. [5]

    Calderbank, A.R., Rains, E.M., Shor, P.W., Sloane, N.J.A., Fellow, Quantum error correction via codes over GF(4), IEEE Trans. Inf. Theory44(4), 1369–1387 (1998)

    work page 1998

  6. [6]

    Symposium on the Foundations of Computer Science, Los Alamitos, CA, (1996)

    Shor, P.W., Fault-tolerant quantum computation. Symposium on the Foundations of Computer Science, Los Alamitos, CA, (1996)

    work page 1996

  7. [7]

    Preskill, J., Reliable quantum computers, Proc. Roy. Soc. Lond. A454, 385–410 (1998)

    work page 1998

  8. [8]

    Steane, A.M., Active stabilization, quantum computation and quantum state synthesis, Phys. Rev. Lett.78, 2252 (1997)

    work page 1997

Show all 39 references
  1. [9]

    Gottesman, D., Theory of fault-tolerant quantum computation, Phys. Rev. A57, 127– 137 (1998)

    1998

  2. [10]

    https://doi.org/10.1007/s11128-008-0087-4

    Gilbert, G., Hamrick, M., Weinstein, Y.S., End-to-end fault tolerance, Quantum Inf Process 7, 263 (2008). https://doi.org/10.1007/s11128-008-0087-4

    2008 doi

  3. [12]

    Nielsen, M.A., Chuang, I.L., Quantum Computation and Quantum Information, Cam- bridge University Press, 2010

    2010

  4. [13]

    Lacalle, J., Pozo Coronado, L.M., Variance of the sum of independent quantum com- puting errors, Quantum Information & Computation19 (15-16), 1294–1312 (2019)

    2019

  5. [14]

    https://doi.org/10.1007/s11128-005-0002-1

    Scott, A.J., Probabilities of Failure for Quantum Error Correction, Quantum Inf Process 4, 399–431 (2005). https://doi.org/10.1007/s11128-005-0002-1

    2005 doi

  6. [15]

    Quantum codes do not fix isotropic errors 19

    Preskill, J., Sufficient condition on noise correlations for scalable quantum computing, Quantum Information & Computation13(3-4), 181–194 (2013). Quantum codes do not fix isotropic errors 19

    2013

  7. [16]

    Gottesman, D., Fault-tolerant quantum computation with constant overhead, Quantum Information & Computation14(15), 1338–1372 (2014)

    2014

  8. [17]

    Cross, A.W., Divincenzo, D.P., Terhal, B.M., A comparative code study for quantum fault tolerance, Quantum Information & Computation9(7), 541–572 (2009)

    2009

  9. [18]

    Hill, C.D., Fowler, A.G., Wang, D.S., Hollenberg, L.C.L., Fault-tolerant quantum er- ror correction code conversion, Quantum Information & Computation13(5), 439–451 (2013)

    2013

  10. [19]

    Duclos-Cianci, G., Poulin, D., Fault-tolerant renormalization group decoder for abelian topological codes, Quantum Information & Computation14(9), 721–740 (2014)

    2014

  11. [20]

    https://doi.org/10.1007/s11128-016-1406-9

    Hocker, D., Zheng, Y., Kosut, R., Brun, T., Rabitz, H., Survey of control perfor- mance in quantum information processing, Quantum Inf Process15, 4361–4390 (2016). https://doi.org/10.1007/s11128-016-1406-9

    2016 doi

  12. [21]

    https://doi.org/10.1007/s11128-016-1337-5

    Hocker, D., Kosut, R., Rabitz, H., PEET: a Matlab tool for estimating physical gate errors in quantum information processing systems, Quantum Inf Process15, 3489–3518 (2016). https://doi.org/10.1007/s11128-016-1337-5

    2016 doi

  13. [22]

    Aliferis, P., Gottesman, D., Preskill, J., Quantum accuracy threshold for concatenated distance-3 codes, Quantum Information & Computation6, 97–165 (2006)

    2006

  14. [23]

    Wang, D.S., Fowler, A.G., Stephens, A.M., Hollenberg, and L.C.L., Threshold error rates for the toric and planar codes, Quantum Information & Computation10(5), 456–469 (2010)

    2010

  15. [24]

    https://doi.org/10.1007/s11128-010-0181-2

    Aggarwal, V., Calderbank, A.R., Gilbert, G., Weinstein, Y.S., Volume thresh- olds for quantum fault tolerance, Quantum Inf Process 9, 541–549 (2010). https://doi.org/10.1007/s11128-010-0181-2

    2010 doi

  16. [25]

    Criger, B., Terhal, B., Noise thresholds for the [4,2,2]-concatenated toric code, Quantum Information & Computation16(15) 1261–1281 (2016)

    2016

  17. [26]

    Ozen, M., Guzeltepe, M., Quantum codes from codes over Gaussian integers with respect to the Mannheim metric, Quantum Information & Computation12(9), 813–819 (2012)

    2012

  18. [27]

    Li, R., Zou, F., Liu, Y., Xu, Z., Hermitian dual containing BCH codes and Construction of new quantum codes, Quantum Information & Computation13(1), 21–35 (2013)

    2013

  19. [28]

    https://doi.org/10.1007/s11128-018-2156-7

    Chen, J., Chen, Y., Huang, Y., Feng, C., New optimal asymmetric quantum codes and quantum convolutional codes derived from constacyclic codes, Quantum Inf Process18, 40 (2019). https://doi.org/10.1007/s11128-018-2156-7

    2019 doi

  20. [29]

    https://doi.org/10.1007/s11128-011- 0269-3

    La Guardia, G.G., Asymmetric quantum Reed-Solomon and generalized Reed-Solomon codes, Quantum Inf Process 11, 591–604 (2012). https://doi.org/10.1007/s11128-011- 0269-3

    2012 doi

  21. [30]

    https://doi.org/10.1023/A:1019623208633

    Boulant, N., Pravia, M.A., Fortunato, E.M., Havel, T.F., Cory, D.G., Experimental Concatenation of Quantum Error Correction with Decoupling, Quantum Inf Processing 1, 135–144 (2002). https://doi.org/10.1023/A:1019623208633

    2002 doi

  22. [31]

    https://doi.org/10.1007/s11128- 011-0312-4

    Evans, Z.W.E., Stephens, A.M., Optimal correction of concatenated fault-tolerant quan- tum codes, Quantum Inf Process11, 1511–1521 (2012). https://doi.org/10.1007/s11128- 011-0312-4

    2012 doi

  23. [32]

    Dias de Albuquerque, C., Palazzo Jr., R., Brandani da Sil, E., Families of codes of topological quantum codes from tessellations tessellations {4i+2,2i+1}, {4i,4i}, {8i-4,4} and {12i-6,3}, Quantum Information & Computation14(15), 1424–1440 (2014)

    2014

  24. [33]

    https://doi.org/10.1007/s11128-015-1115-9

    Naghipour, A., Jafarizadeh, M.A., Shahmorad, S., Topological quantum codes from self-complementary self-dual graphs, Quantum Inf Process 14, 4057–4066 (2015). https://doi.org/10.1007/s11128-015-1115-9

    2015 doi

  25. [34]

    015007 (2018)

    Greenbaum, D., Dutton, Z., Modeling coherent errors in quantum error correction, Quantum Science and Technology3(1), pp. 015007 (2018)

    2018

  26. [35]

    55 (2018)

    Bravyi, S., Englbrecht, M., Konig, R., Peard N., Correcting coherent errors with surface codes, npj Quantum Information4, art. 55 (2018)

    2018

  27. [36]

    4679 (2014)

    Piltz, C., Sriarunothai, T., Var´ on, A.F., Wunderlich, C., A trapped-ion-based quantum byte with 10−5 next-neighbour cross-talk, Nature Communications5, art. 4679 (2014)

    2014

  28. [37]

    Buterakos, D., Throckmorton, R.E., Das Sarma, S., Crosstalk error correction through dynamical decoupling of single-qubit gates in capacitively coupled singlet-triplet semi- conductor spin qubits, Phys. Rev. B97(4), pp. 045431 (2018)

    2018

  29. [38]

    In: Duplantier B., Raimond JM., Rivasseau V

    Zurek W.H., Decoherence and the Transition from Quantum to Classical - Revisited. In: Duplantier B., Raimond JM., Rivasseau V. (eds) Quantum Decoherence. Progress in Mathematical Physics, vol 48. Birkh ¨ ı¿½user Basel (2006). 20 J. Lacalle et al

    2006

  30. [39]

    Bennet, C.H., DiVincenzo, D.P., Smolin, J.A., Wootters, W.K., Mixed state entangle- ment and quantum error correction, Los Alamos Physics Preprint Archivehttp://xxx.- lanl.gov/abs/quant-ph/9909058, (1999)

    1999 arXiv

  31. [40]

    Laflamme, R., Miquel, C., Paz, J.-P., Zurek, W.H., Perfect quantum error correction codes, Phys. Rev. Lett.77, 198 (1996). Los Alamos Physics Preprint Archive http://- xxx.lanl.gov/abs/quant-ph/9602019

    1996 arXiv

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.