REVIEW 2 major objections 5 minor 27 references
Tightness of and counterexamples to several quantum estimates
T0 review · 2 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash
Pith's one-line read This paper proves exponential norm gaps and refutes the quantum Entropy–Influence conjecture.
desk verdict The quaternionic product-state gap and the Hermitian-dilation QFEI counterexample are real and check out, but the draft has a false inequality in Theorem 1's display and a gap in an auxiliary proof that need fixing before publication. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The argument rides on three interlocking objects. The first is the quaternion product $R_d=\prod_{r=1}^d(X_r i+Y_r j+Z_r k)$, whose scalar component $H_{2m}^{(0)}$ is Hermitian and homogeneous; the scalar-word count $A_\ell=(3^\ell+3(-1)^\ell)/4$ and the dimer overlap identity $\langle D_0,P_aD_1\rangle=2^{1-m}c_a$ pin down its exact Pauli coefficients and matrix elements. The second is the flat-spectrum assembly: take pairwise anti-commuting Pauli strings $P_\ell$ of weight $r$, attach commuting suffix operators $B_\ell$ built from a Hadamard matrix, and use the identity $\sum_\ell B_\ell^2=L^{k+1}I$ to make $H_d^2=N_d I$; this is what transfers a norm estimate into a statement about $\ell^p$ coefficients. The third is the Hermitian dilation $O_f$, the block operator with off-diagonal blocks $F_f$ and $F_f^*$, which converts a unimodular complex function $f$ into a quantum Boolean function while preserving influence exactly and entropy up to one bit.
What would settle it
Enumerate the Section 4.1 recursion for $r=4,5,6$: form the claimed $3^r$ strings on $(3^r-1)/2$ qubits and check every pair for anti-commutation and every row for distinctness; a single violation invalidates $H_d^2=N_d I$ and the derived lower bounds. Independently, for $m=1,2,3$ compute the exact operator norm of $H_{2m}$ by optimizing over tensor-product states and compare it with the lower bound $(9^m+2\cdot6^m+3)/(2^{m+1}+4)$.
Extended reading notes
Core claim
On the paper's own terms, the central discovery is a collection of exact formulas that force exponential constants. In Theorem 1, the quaternion recursion $R_d=\prod_{r=1}^d(X_r i+Y_r j+Z_r k)$ defines $H_{2m}=H^{(0)}_{2m}$, whose Pauli coefficients are the scalar parts of quaternion words; the scalar-word count $N_m=(9^m+3)/4$ and the dimer matrix elements give $\langle D_0,H_{2m}D_0\rangle=3^m$ and $\langle D_0,H_{2m}D_1\rangle=2^{1-m}N_m$, so the vector $|D_0\rangle+|D_1\rangle$ witnesses the lower bound $\|H_{2m}\|_{\mathrm{op}}\ge(9^m+2\cdot6^m+3)/(2^{m+1}+4)$. In Section 4, the Hadamard-suffix identity $\sum_{\ell=1}^L B_\ell^2=L^{k+1}I$ turns $L=2^k$ anti-commuting strings $P_\ell$ of weight $r$ into a quantum Boolean function $O_d$ with $O_d^2=I$ and $N_d=L^{k+1}$ equal Pauli coefficients; the ratio $\|\widehat{O_d}\|_{\ell^{2d/(d+1)}}/\|O_d\|_{\mathrm{op}}\ge 3^{d/8}$ forces the non-commutative Bohnenblust–Hille constant to be exponential. In Section 6, the Hermitian dilation of the scalar weighted Rudin–Shapiro phase family with $a_j=1/\sqrt n$ gives $\mathrm{Inf}[O_n]<2$ and $\mathrm{Ent}[O_n]=\log_2 n+O(1)$, a direct counterexample to the quantum Entropy–Influence conjecture. Finally, Theorem 4's counting bound $w\ge 3(N-1)/(2L)$ implies that exponentially many anti-commuting strings must concentrate qubit load, which yields $\max\mathrm{Inf}[P]\ge 3/(10r)$ and hence $\mathrm{Var}[P]\le 10r\,\max\mathrm{Inf}[P]$, proving the quantum Aaronson–Ambainis inequality for this class.
Load-bearing premise
The load-bearing premise is that the recursive family of $3^r$ pairwise anti-commuting, weight-$r$ Pauli strings supplied by [16] exists for every $r$ with no hidden collisions; if that construction fails, the flat-spectrum identity $H_d^2=N_d I$, the exponential Bohnenblust–Hille lower bound, and the Section 6 counterexample built on it all collapse.
Editorial extensions
If this is right
- For every even degree $d=2m$, no dimension-free constant smaller than $(3/\sqrt2)^d$ can bound the operator norm of a $d$-homogeneous Hamiltonian by its product-state expectation; product-state ansätze for ground-state estimation therefore lose exponentially in degree.
- The non-commutative Bohnenblust–Hille constant must grow exponentially in $d$, so the subexponential constants of the classical Hamming-cube inequality do not transfer to Pauli Hamiltonians.
- The quantum Entropy–Influence conjecture, as formulated with a universal constant $C$, is false for qubit quantum Boolean functions; entropy can grow logarithmically in the number of qubits while total influence is bounded.
- For Hamiltonians on pairwise anti-commuting Pauli strings with weight at most $10r$, the quantum Aaronson–Ambainis inequality holds with explicit constants $C=10$, $K=1$.
- The Bohnenblust–Hille exponent $2d/(d+1)$ is sharp for the Hamming cube, and a non-random $Z$-only construction exhibits the obstruction.
Reading between the lines
- The Hermitian-dilation transfer formulas imply a general principle the paper does not state as such: any sequence of unimodular complex functions with $\mathrm{Ent}[f]/\mathrm{Inf}[f]\to\infty$ yields a quantum Boolean counterexample with the same divergence, so the quantum Entropy–Influence conjecture fails for the entire family of phase functions, not just this example.
- A natural testable extension is to run the same weighted Rudin–Shapiro construction with other weight profiles, for example $a_j=j^{-\alpha}$; the paper computes only the equal-weight case $a_j=1/\sqrt n$, and the transfer formulas predict the entropy–influence ratio whenever $\sum_j p_j<2$ and $\sum_j h_2(p_j)\to\infty$.
- The quaternion algebra in Section 3 could be replaced by other Clifford-type products; the paper does not explore whether bases larger than $3/\sqrt2$ are attainable, so the true supremum of the homogeneous comparison constant remains an open question suggested by its method.
Signed reviews
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies dimension-free constants for several quantum inequalities. It constructs homogeneous qubit Hamiltonians from a quaternion-valued recursion and proves an exponential gap between operator norm and product-state norm; it reproduces and extends a construction of exponentially many pairwise anticommuting Pauli strings to show that the noncommutative Bohnenblust-Hille constant must grow exponentially in degree; it builds quantum Boolean functions by Hermitian dilation of a weighted scalar Rudin-Shapiro phase family, giving an explicit counterexample to the quantum Fourier Entropy-Influence conjecture; and it discusses the quantum Aaronson-Ambainis conjecture for Hamiltonians supported on anticommuting Pauli strings. The main tools are exact counts of scalar quaternion words, dimer overlap identities, a Hadamard suffix identity, and exact entropy/influence transfer formulas for the Hermitian dilation.
Significance. If the main results stand, the paper is significant: the quaternion construction gives a constructive lower bound with exponential base larger than 2 for the homogeneous product-norm comparison problem; the flat-spectrum construction gives an exponential lower bound for the noncommutative Bohnenblust-Hille constant; and the Hermitian-dilation counterexample is a clean, explicit obstruction to a dimension-free quantum Entropy-Influence inequality. The central constructions are explicit and largely checkable: the quaternion recursion, the dimer overlap identity, the Hadamard suffix identity, and the dilation transfer formulas are all concrete and exact. The dependence on the external anticommuting Pauli-string construction of Jiang-Kalev-Mruczkiewicz-Neven is acknowledged, and the described recursion is consistent with the standard construction. The main weaknesses are localized but real: one proof in Section 8.2 has a missing collision term, and the counting in Section 8.3 is inconsistent; several display-level identities also contain quantitative errors.
major comments (2)
- [§8.2] The proof of (88) uses the assertion that the probability that two independently chosen strings have an anticommuting local overlap is 1. With independent sampling with replacement from the distribution |alpha_l|^2, the probability that the two choices are distinct is 1 - sum_l |alpha_l|^4, and for equal choices there is no local anticommutation. The displayed lower bound should therefore contain the factor 1 - sum_l |alpha_l|^4, and since the manuscript imposes no control on this collision term, the conclusion max_q Inf_q(P) >= 3/(10r) is not established as written. The stated theorem may be repairable by a case split on sum_l |alpha_l|^4 (if the collision probability is large, a single dominant string directly gives a large influence), but that argument is absent and is needed.
- [§8.3] The counting in this section is internally inconsistent. If the operator has L^{d-1} Pauli strings each of coefficient magnitude L, then its ell_p norm satisfies ||hat B||_{ell_p}^p = L^{d-1} L^p, so ||hat B||_{ell_p} = L^{(d+p-1)/p}, not L^{d/p} as stated. If a different normalization is intended, it is not specified. In addition, the bound ||B||_{op} <= L^{(d+1)/2} is asserted after 'the same doubling trick' without the promised derivation. Thus the claimed non-random proof of the sharp exponent p = 2d/(d+1) is not established as written, even though the underlying commutative result is classical.
minor comments (5)
- [§3.7, after Eq. (53)] The displayed inequality L_m > (9/2)^m is false for every m >= 1; for example, m = 1 gives L_1 = 3 < 9/2. The correct behavior is L_m = (1/2)(9/2)^m(1+o(1)), which still gives the intended exponential gap with base 3/sqrt(2) per degree. The false comparison should be corrected or removed.
- [§2.2] The displayed formula for P_n appears to have the wrong qubit count and normalization. With the recursion of (11), P_n acts on n+1 qubits and the exact form is P_n = i 2^n (|1^{n+1}><0^{n+1}| - |0^{n+1}><1^{n+1}|), giving operator norm 2^n, not 2^{n-1}. Section 8 also states ||P_n||_{op} = 2^n. The discrepancy does not affect the exponential-gap conclusion but should be fixed.
- [Title and abstract] The word 'estimates' is misspelled as 'estima tes' in the title and abstract; this should be corrected.
- [§2.1, Eq. (15)] The sentence says the Bloch coordinates are those 'of pure state e1 ⊗ e1'; this should presumably be the Bloch coordinates of e1 (or of the first tensor factor), and the wording should be cleaned up.
- [§8.3] The symbol B is reused for different operators in the same paragraph: B denotes the suffix operators from (61) and then a new operator S = B_1 + ... + B_L is also called B in (89). Please introduce separate notation.
Circularity Check
No load-bearing circularity identified; the central derivations are self-contained.
full rationale
The paper's central constructions are derived in full rather than fitted or imported by definition: the quaternion recursion defines H_{2m}, and Theorem 1 proves both the product-norm upper bound and the dimer-matrix-element lower bound from Lemmas 1-4, with no fitted quantity later renamed as a prediction. The quantum Entropy-Influence counterexample is likewise self-contained: Proposition 3 derives the exact Hermitian-dilation transfer formulas, and the evaluation on the weighted Rudin-Shapiro family gives Inf[O_n] < 2 and H[O_n] = log_2 n + O(1) by direct algebra. The noncommutative Bohnenblust-Hille exponential lower bound is reproduced in the text from the explicit anti-commuting Pauli-string construction of [16], with the proof of [23] written out rather than merely cited; although [5] and [23] are self-citations, the load-bearing arguments do not reduce to those citations. The external dependence on [16] is a standard, independently available construction, not a conclusion from the present paper. No displayed equation is equivalent to its own input by construction, so no circular step is present.
Assumptions & free parameters
free parameters (1)
- a_j =
a_j = 1/sqrt(n) for every j
assumptions (5)
- domain assumption Existence and pairwise anti-commutation of 3^r weight-r Pauli strings on (3^r - 1)/2 qubits, from Jiang-Kalev-Mruczkiewicz-Neven.
- standard math Hadamard orthogonality H^T H = L I and D_s^2 = I for diagonal Z-operator matrices.
- standard math Pauli matrix identities, including sigma_a^T epsilon = -epsilon sigma_a with epsilon = iY, and the trace identity c_a = (-1)^m 2^{-1} Tr(sigma_{a1} ... sigma_{a2m}).
- domain assumption Every pure single-qubit state has Bloch coordinates satisfying (a_x)^2 + (a_y)^2 + (a_z)^2 = 1.
- domain assumption Quantum Boolean function definition and normalized Hilbert-Schmidt norm conventions from [19].
Cite this review
Pith. "Pith review of Tightness of and counterexamples to several quantum estimates." pith.science (2026). https://pith.science/paper/B4JDT27U
@misc{pith2026260804411,
author = {Pith},
title = {Pith review of: Tightness of and counterexamples to several quantum estimates},
year = {2026},
howpublished = {\url{https://pith.science/paper/B4JDT27U}},
note = {Machine review of arXiv:2608.04411}
}
abstract
We prove here several tightness results for such quantum inequalities as the comparison of operator norm and product norm of $d$-local hamiltonians, Bohnenblust--Hille inequality for $d$-local hamiltonians and for quantum Fourier entropy-influence conjecture, we also discuss the quantum Aaronson--Ambainis conjecture in a special case of anti-commuting Pauli strings.
Reference graph
Works this paper leans on
-
[23]
J. Slote, Dense Hamiltonians at the Parseval Limit, arX iv:2608.01424, pp.1–8
-
[5]
Approximating the operator norm of local Hamiltonians via few quantum states
L. Becker, J. Slote, A. Volberg, H. Zhang, Approximating the operator norm of local Hamil- tonians via few quantum states, arXiv:2509.11979v3, pp. 1– 34
-
[1]
S. Aaronson and A. Ambainis, The need for structure in quantum speedups , Theory Comput. 10 (2014), 133–166
work page 2014
-
[2]
Approximation algorithms for quantum many-body problems
Sergey Bravyi, David Gosset, Robert König, and Kristan T emme. Approximation algorithms for quantum many-body problems. J. Math. Phys. , 60(3):032203, 18, 2019
work page 2019
-
[3]
K. Bu, R. J. Garcia, A. Jaffe, D. E. Koh, and L. Li, Complexity of quantum circuits via sensitivity, magic, and coherence , Commun. Math. Phys. 405 (2024), article 161, doi:10.1007/s00220-024-05030-6
-
[4]
L. Ben Efraim, F. Lust-Piquard, Poincaré type inequalit ies on the discrete cube and in the CAR algebra. Probability Theory and Related Fields 141 (200 8), no. 3–4, 569–602. DOI: 10.1007/s00440-007-0094-x
-
[6]
H. F. Bohnenblust, E. Hille, On the Absolute Convergence of Dirichlet Series. ” Annals of Mathematics (Second Series) 32 (1931), no. 3, 600–622
work page 1931
-
[7]
Chen, S., Huang, H.-Y., Preskill, J. (2023). Learning to Predict Arbitrary Quantum Processes. PRX Quantum, 4, Article 040337. https://doi.org/10.1103/ prxquantum.4.040337
work page 2023
Show all 27 references
-
[8]
Defant, D
A. Defant, D. Garcia, M. Maestre, P. Sevilla-Peris Diric hlet Series And Holomorphic Func- tions In High Dimensions, Cambridge University Press, 2019 , ISBN 978-1-108-47671-3 Hard- back
2019
-
[9]
Defant, L
A. Defant, L. Frerick, J. Ortega-Cerdà, M. Ounaies, and K . Seip, The Bohnenblust–Hille inequality for homogeneous polynomials is hypercontracti ve. Ann. Math. (2), 174(1):485–497, 2011
2011
-
[10]
Defant, M
A. Defant, M. Mastylo, and A. Pérez, On the Fourier spect rum of functions on boolean cubes. Mathematische Annalen, 374(1–2):653–680, 2019. 27
2019
-
[11]
H. A. Dye, B. Russo, A note on unitary operators in C ∗-algebras, Duke Mathematical Journal, Vol. 33, No. 2, pp. 413–416 (1966)
1966
-
[12]
Eskenazis, P
A. Eskenazis, P. Ivanisvili, Learning Low-Degree Func tions from a Logarithmic Num- ber of Random Queries. Proceedings of the 54th Annual ACM SIG ACT Symposium on Theory of Computing4(STOC 2022), Rome, Italy, June 20–24, 2 022, pp. 203–207.5DOI: 10.1145/3519935.3519981
2022
-
[13]
Einstein, B
A. Einstein, B. Podolsky, and N. Rosen, Can Quantum-Mec hanical Description of Physical Reality Be Considered Complete? Phys. Rev. 47, 777 (1935)
1935
-
[14]
Bell’s Theorem, Quantum Theory and Conception of the Universe
Daniel M. Greenberger, Michael A. Horne, and Anton Zeil inger, in "Bell’s Theorem, Quantum Theory and Conception of the Universe", edited by M. Kafatos (Kluwer Academic, Dordrecht, 1989), p. 69
1989
-
[15]
Huang, R
H.-Y. Huang, R. Kueng, J. Preskill, (2020). Predicting many properties of a quantum system from very few measurements. Nature Physics, 16, 1050–1057. https://doi.org/10.1038/s41567- 020-0932-7
2020 doi
-
[16]
Jiang, A
Z. Jiang, A. Kalev, W. Mruczkiewicz, and H. Neven, Optimal fermion-to-qubit mapping via ternary trees with applications to reduced quantum states l earning, Quantum 4 (2020), 276, doi:10.22331/q-2020-06-04-276
2020 doi
-
[17]
Elliott H. Lieb. The classical limit of quantum spin sys tems. Comm. Math. Phys. , 31:327–340, 1973
1973
-
[18]
Mermin, Extreme Quantum Entanglement in a Superposi tion of Macroscopically Distinct States, Phys
D. Mermin, Extreme Quantum Entanglement in a Superposi tion of Macroscopically Distinct States, Phys. Review Letters, v. 65, no.15 (1990), 1838–184 0
1990
-
[19]
Montanaro, T
A. Montanaro, T. J. Osborne Quantum Boolean Functions. Chicago Journal of Theoretical Computer Science,42010, Article 1, pp. 1–45.5DOI: 10.4086 /cjtcs.2010.001
2010
-
[20]
O’Donnell, M
R. O’Donnell, M. Saks, O. Schramm, R.A. Servedio, Every decision tree has an influential variable. Proceedings of the 46th Annual IEEE Symposium on F oundations4of Computer Science (FOCS 2005), pp. 31–39
2005
-
[21]
Rouzé, M
C. Rouzé, M. Wirth, and H. Zhang, Quantum Talagrand, KKL and Friedgut’s theorems and the learnability of quantum Boolean functions , Commun. Math. Phys. 405 (2024), article 95, doi:10.1007/s00220-024-04981-0
2024 doi
-
[22]
Schechtman, Entropy versus influence for complex fun ctions of modulus one, arXiv:2009.12753 [math.CO], 2020
G. Schechtman, Entropy versus influence for complex fun ctions of modulus one, arXiv:2009.12753 [math.CO], 2020
2009 arXiv
-
[24]
Tomiyama, On the projection of norm one in C ∗-algebras
J. Tomiyama, On the projection of norm one in C ∗-algebras. Proceedings of the Japan Academy, Series A, Mathematical Sciences, 33(10), 608-612 (1957)
1957
-
[25]
Volberg, H
A. Volberg, H. Zhang, Noncommutative Bohnenblust–Hil le inequalities. Mathematische Annalen 389 (2024), no. 2, 1609–1652.4DOI: 10.1007/s00208 -023-02680-0
2024 doi
-
[26]
Volberg, H
A. Volberg, H. Zhang Two tensor Rudin–Shapiro construc tions, Preprint, 2026, pp. 1–8
2026
-
[27]
Zhang, The Boolean case of the Aaronson–Ambainis infl uence conjecture without OSSS
H. Zhang, The Boolean case of the Aaronson–Ambainis infl uence conjecture without OSSS. Preprint, 2026, pp. 1–5. 28
2026
Reviewed August 15, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.