REVIEW 4 major objections 3 minor 44 references
Learning shadows to predict quantum ground state correlations
T0 review · 4 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash
Pith's one-line read This paper proposes that a bag of randomized measurement snapshots, optimized against a Hamiltonian's energy, can represent a spin system's ground state.
desk verdict The abstract and body are two different papers—the claimed shadow-tomography method is entirely absent, so the submission cannot be reviewed as is. 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 central object is the snapshot bag: a collection of $N$ individual randomized measurement outcomes that acts as a classical proxy for the state $\rho$. Its power is quantified by the classical shadow bound $N \sim 3^k \log M/\epsilon^2$, which guarantees that only logarithmically many snapshots in $M$ are needed to estimate $M$ $k$-local observables when bases are chosen locally at random. The second ingredient is the variational loop, in which the snapshots carry tunable parameters, the implied energy is minimized, and positivity of reduced density matrices is imposed to keep the predicted correlations consistent with a genuine Hilbert space; the claim that learning the measurement-outcome distribution captures correlations beyond those fixed by reduced-matrix positivity is the mechanism meant to distinguish the method from quantum-chemistry-style reduced density matrix approaches. The full text available with this record does not actually describe this machinery.
What would settle it
Two concrete checks settle it. Open the full text: the manuscript supplied here is a holography paper about volume and indices of operator algebras and contains no variational shadow scheme, no positivity-constraint algorithm, and no numerics, so the abstract's central claim is not substantiated in this record at all. And for a specific local spin Hamiltonian, such as the transverse-field Ising chain, run the proposed optimization and compare the predicted energy and spin–spin correlation functions with exact diagonalization or tensor-network results; disagreement with those benchmarks would falsify the claim that the optimized snapshot bag captures ground-state correlations.
Extended reading notes
Core claim
On its own terms, the paper's central discovery is a variational principle for classical data about a quantum state: rather than preparing copies of the ground state and measuring them, one starts from a bag of $N$ parametrized snapshots — individual outcomes of measurements in independently chosen random local bases — and optimizes that bag directly against the energy of the local spin Hamiltonian. Classical shadow tomography supplies the guarantee that such a bag can reconstruct the expectation values of $M$ $k$-local observables to accuracy $\epsilon$ using $N \sim 3^k \log M/\epsilon^2$ snapshots, which is what makes the bag a plausible efficient proxy for the state when only low-weight observables such as Hamiltonian terms matter. The algorithm adds positivity constraints on reduced density matrices, borrowed from quantum chemistry, so that the predicted correlations are compatible with the existence of an underlying Hilbert space; the distinctive claim is that learning the full distribution of measurement outcomes goes beyond those constraints and captures correlations that constrained density-matrix methods would miss. The abstract asserts that numerical results confirm the method is parallelizable, efficiently simulable, and gives a more complete description of the ground state. The manuscript body attached to this record is a different preprint, on bulk volume in holography and the index of an algebra inclusion, and presents none of this.
Load-bearing premise
The load-bearing premise is that energy minimization over the parametrized snapshot bag converges to the true ground-state correlations rather than to a local minimum, and that the learned distribution of measurement outcomes carries correlations beyond those already fixed by reduced density matrix positivity; neither step is demonstrated in the text supplied with this record.
Editorial extensions
If this is right
- Ground-state correlations of a local spin Hamiltonian could be predicted by optimizing a classical bag of measurement snapshots rather than by solving for the quantum state itself.
- The snapshot count scales as $3^k \log M/\epsilon^2$, so the cost is dominated by the locality $k$ of the observables of interest, matching the terms that appear in a local Hamiltonian.
- Positivity constraints on reduced density matrices carry over from quantum chemistry and can be imposed to keep predicted correlations physically consistent.
- The abstract's claims of parallelizability and efficient classical simulability would make the approach scalable to regimes where direct state tomography is impractical.
- If the learned measurement distribution indeed captures correlations beyond reduced-matrix positivity, the method would give a more complete description of the ground state than constrained density-matrix methods alone.
Reading between the lines
- A direct test of the 'more complete description' claim would be to run the optimization on a concrete chain (for instance the transverse-field Ising or Heisenberg model) and compare the predicted energy and spin–spin correlation functions against exact diagonalization or tensor-network benchmarks; the paper's abstract does not report such comparisons.
- Because shadow estimation yields many $k$-local expectation values from the same bag, a snapshot-bag variational method could double as a reusable resource for computing several observables of the same ground state at once; the abstract leaves this reuse implicit.
- Conceptually, the scheme is a classical analogue of a variational quantum eigensolver in which randomized measurement outcomes, rather than a parametrized quantum circuit, carry the variational degrees of freedom; drawing that connection could help transfer error-mitigation ideas from one setting to the other.
- If the attached full text is the document meant to accompany this abstract, then the record as assembled is not a single paper: the abstract announces a quantum-information method while the body advances an unrelated holographic proposal linking bulk volume to the index of an operator-algebra inclusion.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The abstract of arXiv:2508.00052 announces a variational scheme based on classical shadow tomography for predicting ground-state correlations of local spin Hamiltonians, with snapshot complexity N ~ 3^k log M / eps^2, positivity constraints on reduced density matrices, and numerical demonstrations. The full text supplied is not that paper: it is "Volume as an index of a subalgebra" (arXiv:2508.00056), a holography paper proposing that the exponential of the volume of a maximal slice equals an index of inclusion for boundary von Neumann algebras. The body contains no classical shadow formalism, no spin Hamiltonian, no parametrized snapshot ansatz, no optimization or positivity constraint, and no numerical results. The claims of the abstract are therefore not supported by the submitted document.
Significance. If the abstract's variational scheme existed as claimed, it would be significant: a sample-efficient classical representation of ground states for low-weight observables, with parallelization and efficient simulation, would be a useful tool. However, none of that content is present for evaluation. The holography body is a separate, internally coherent manuscript with concrete computations (e.g., Sec. III and Appendix B), but it does not bear on the abstract's claims. No credit for machine-checked proofs, derivations, or numerical data can be assigned to the claimed shadow scheme, as none are supplied.
major comments (4)
- [Abstract/full-text mismatch] The central claim of the abstract—that a bag of N parametrized snapshots measured in locally random bases can be variationally optimized to represent ground states with N ~ 3^k log M / eps^2 and positivity constraints—does not appear anywhere in the body. The body is entirely a holography manuscript on the volume-index relation for von Neumann algebras. No classical shadow formalism, no Hamiltonian, no snapshot ansatz, no energy functional, and no geometry of the variational problem is defined. The abstract's assertions are therefore unsupported by the submitted document.
- [Abstract] The scaling N ~ 3^k log M / eps^2 is quoted as if it transferred directly from classical shadow tomography to the variational setting. In standard shadow tomography that bound governs estimation of M expectation values from independent copies of a known state; the abstract supplies no argument that the same complexity controls a variational search over a parametrized bag of snapshots representing an unknown ground state. This is a nontrivial step that would need to be proved even in a correctly assembled manuscript.
- [Abstract] The sentence "learning the underlying distribution of measurement outcomes allows one to further correlations beyond those in the constrained density matrix" is stated without a mechanism. The manuscript does not define the learned distribution, explain how the snapshot bag relates to the reduced density matrices, or specify which correlations are captured beyond the positivity constraints. Without these definitions, the claim is untestable.
- [Abstract/body] The abstract promises "numerical results" demonstrating parallelizability, efficient simulability, and a more complete description of ground states. The body contains no figure, table, data, or even a description of a studied Hamiltonian or system size. This is a direct omission of the promised supporting evidence.
minor comments (3)
- [Title and metadata] The title, author list, preprint number, and body all identify the full text as arXiv:2508.00056, while the abstract is for arXiv:2508.00052; the submission metadata should be reconciled before any further review.
- [Throughout body] The typeset mathematics contains many garbled symbols and missing characters, making verification of the holography calculations unnecessarily difficult; a clean compile is needed.
- [Sec. VI] The paper states that equation (1.4) may be viewed as a new index defined by holography and that it does not generally agree with the Kosaki index; if the holography manuscript is resubmitted on its own, this status should be flagged as a conjecture in the introduction rather than only in the body.
Circularity Check
No circularity can be established: the submitted full text is an unrelated holography preprint, so the abstract's shadow-based variational claim has no derivation chain to reduce.
full rationale
The abstract of arXiv:2508.00052 promises a variational classical-shadow scheme for quantum ground-state correlations, with a sample complexity N ~ 3^k log M / eps^2, positivity-constrained reduced density matrices, and numerical results. However, the full text supplied is arXiv:2508.00056, 'Volume as an index of a subalgebra' by Leutheusser and Liu, a holography paper containing no shadow tomography formalism, spin Hamiltonian, snapshot ansatz, energy functional, or numerics. There is therefore no derivation chain in the submitted document that could exhibit an equation reduced to its own input or a fitted parameter renamed as a prediction. The abstract's sample-complexity statement is the standard classical-shadow tomography bound, not derived here, and the 'further correlations beyond those in the constrained density matrix' claim is stated without mechanism. Under the hard rule that circularity must be demonstrated by a specific quoted reduction, no circular step can be identified; the score is 0. This is a non-finding due to absence of the claimed content rather than a positive assessment of self-containment.
Assumptions & free parameters
assumptions (1)
- domain assumption The classical shadow tomography sample complexity bound N ~ 3^k log M / eps^2, quoted in the abstract, is valid and applies to the proposed snapshot representation.
Cite this review
Pith. "Pith review of Learning shadows to predict quantum ground state correlations." pith.science (2026). https://pith.science/paper/MUWFHY4L
@misc{pith2026250800052,
author = {Pith},
title = {Pith review of: Learning shadows to predict quantum ground state correlations},
year = {2026},
howpublished = {\url{https://pith.science/paper/MUWFHY4L}},
note = {Machine review of arXiv:2508.00052}
}
abstract
We introduce a variational scheme inspired by classical shadow tomography to compute ground state correlations of quantum spin Hamiltonians. Shadow tomography allows for efficient reconstruction of expectation values of arbitrary observables from a bag of repeated, randomized measurements, called snapshots, on copies of the state $\rho$. The prescription allows one to infer expectation values of $M$ $k-$local observables to accuracy $\epsilon$ using just $N \sim 3^k \text{log}M /\epsilon^2$ snapshots when measurements are performed in locally random bases. Turning this around, a bag of snapshots can be considered an efficient representation of the state $\rho$, particularly for estimating low-weight observables, such as terms in a local Hamiltonian needed to estimate the energy. Inspired by this, we consider a variational scheme wherein a bag of $N$ parametrized snapshots is used to represent the putative ground state of a desired local spin Hamiltonian and optimized to lower the energy with respect to it. Additional constraints in the form of positivity of reduced density matrices, motivated by work in quantum chemistry, are employed to ensure compatibility of the predicted correlations with the underlying Hilbert space. Unlike reduced density matrix approaches, learning the underlying distribution of measurement outcomes allows one to further correlations beyond those in the constrained density matrix. We show, with numerical results, that the proposed variational method can be parallelized, is efficiently simulable, and yields a more complete description of the ground state.
Reference graph
Works this paper leans on
-
[1]
Two disjoint intervals in vacuum 20
-
[2]
Single interval at finite temperature 20
-
[3]
Quantifying superadditivity of boundary subalgebras 21
Strip in � ≥ 3 20 C. Quantifying superadditivity of boundary subalgebras 21
-
[4]
Two intersecting intervals for vacuum CFT � 21
-
[5]
Possible finite � interpretation of the index 22 A
Two disjoint intervals for vacuum CFT � 22 IV. Possible finite � interpretation of the index 22 A. Further elaboration on the index formula 22 B. Algebras at finite � and generalized gravitational entropy 24 V. Algebraic probe of de Sitter volume growth 26 A. Kruskal and Static Coordinates on dS ��� 26 B. Algebras of North/South Pole Observables 27 C. App...
-
[6]
Kosaki’s index of inclusion 32 2 a. The Jones index 35 b. Longo’s index 36 B. Useful results on half-sided modular translations 37
-
[7]
Half-sided modular translations in the large dimension limit 37
-
[8]
AdS � calculations 39 a. Regulated AdS � ‘wormhole’ length 39 b. Evolution with HSMTs 41 References 43 �� ������������ Since the discovery of Bekenstein-Hawking entropy [1, 2], the area of spacetime regions in gravitational systems has been a subject of long-standing fascination. This relationship between area and entropy has played a fundamental role in ...
Show all 44 references
-
[9]
It applies to algebras that are defined only in the large � limit
-
[10]
Kruskal coordinates
It applies to algebras that can be intrinsically defined at finite � , and gives the leading order behavior of the indices of such algebras in the large � limit. This would be analogous to the black hole entropy formula. First consider interpretation 1 above, which appears nat...
-
[11]
J. D. Bekenstein, Lett. Nuovo Cim. �, 737-740 (1972) doi:10.1007/BF02757029
1972 doi
-
[12]
S. W. Hawking, Commun. Math. Phys. �� , 199-220 (1975) [erratum: Commun. Math. Phys. �� , 206 (1976)] doi:10.1007/BF02345020
1975 doi
-
[13]
’t Hooft, Conf
G. ’t Hooft, Conf. Proc. C ������ , 284-296 (1993) [arXiv:gr-qc/9310026 [gr-qc]]
1993 arXiv
-
[14]
Susskind, J
L. Susskind, J. Math. Phys. �� , 6377-6396 (1995) doi:10.1063/1.531249 [arXiv:hep-th/9409089 [hep-th]]
1995 arXiv
-
[15]
Ryu and T
S. Ryu and T. Takayanagi, Phys. Rev. Lett. �� , 181602 (2006) doi:10.1103/PhysRevLett.96.181602 [arXiv:hep-th/0603001 [hep-th]]
2006 arXiv
-
[16]
V. E. Hubeny, M. Rangamani and T. Takayanagi, JHEP �� , 062 (2007) doi:10.1088/1126- 6708/2007/07/062 [arXiv:0705.0016 [hep-th]]
2007 arXiv
-
[17]
Engelhardt and A
N. Engelhardt and A. C. Wall, JHEP �� , 073 (2015) doi:10.1007/JHEP01(2015)073 [arXiv:1408.3203 [hep-th]]
2015 arXiv
-
[18]
Susskind, Fortsch
L. Susskind, Fortsch. Phys. �� , 24-43 (2016) doi:10.1002/prop.201500092 [arXiv:1403.5695 [hep-th]]
2016 arXiv
-
[19]
Stanford and L
D. Stanford and L. Susskind, Phys. Rev. D �� , no.12, 126007 (2014) doi:10.1103/PhysRevD.90.126007 [arXiv:1406.2678 [hep-th]]. �� Note that this had to be the case since exp( i(P� + P� )s) implements an isometry in AdS �. 43
2014 arXiv
-
[20]
Miyaji, T
M. Miyaji, T. Numasawa, N. Shiba, T. Takayanagi and K. Watanabe, Phys. Rev. Lett. ��� , no.26, 261602 (2015) doi:10.1103/PhysRevLett.115.261602 [arXiv:1507.07555 [hep-th]]
2015 arXiv
-
[21]
Alishahiha, Phys
M. Alishahiha, Phys. Rev. D �� , no.12, 126009 (2015) doi:10.1103/PhysRevD.92.126009 [arXiv:1509.06614 [hep-th]]
2015 arXiv
-
[22]
Jefferson and R
R. Jefferson and R. C. Myers, JHEP �� , 107 (2017) doi:10.1007/JHEP10(2017)107 [arXiv:1707.08570 [hep-th]]
2017 arXiv
-
[23]
Chapman, M
S. Chapman, M. P. Heller, H. Marrochio and F. Pastawski, Phys. Rev. Lett.��� , no.12, 121602 (2018) doi:10.1103/PhysRevLett.120.121602 [arXiv:1707.08582 [hep-th]]
2018 arXiv
-
[24]
Belin, R
A. Belin, R. C. Myers, S. M. Ruan, G. S´ arosi and A. J. Speranza, Phys. Rev. Lett. ��� , no.8, 081602 (2022) doi:10.1103/PhysRevLett.128.081602 [arXiv:2111.02429 [hep-th]]
2022 arXiv
-
[25]
Belin, R
A. Belin, R. C. Myers, S. M. Ruan, G. S´ arosi and A. J. Speranza, JHEP �� , 154 (2023) doi:10.1007/JHEP01(2023)154 [arXiv:2210.09647 [hep-th]]
2023 arXiv
-
[26]
Jørstad, R
E. Jørstad, R. C. Myers and S. M. Ruan, JHEP �� , 223 (2023) doi:10.1007/JHEP07(2023)223 [arXiv:2304.05453 [hep-th]]
2023 arXiv
-
[27]
R. C. Myers and S. M. Ruan, [arXiv:2403.17475 [hep-th]]
-
[28]
Leutheusser and H
S. Leutheusser and H. Liu, Phys. Rev. D ��� , no.8, 086020 (2023) doi:10.1103/PhysRevD.108.086020 [arXiv:2112.12156 [hep-th]]
2023 arXiv
- [29]
-
[30]
Kosaki, J
H. Kosaki, J. Funct. Anal. �� , no.1, 335-348 (1986). doi:10.1016/0022-1236(86)90085-6
1986 doi
-
[31]
V. F. R. Jones, Invent. Math. �� , 1-25 (1983). doi:10.1007/BF01389127
1983 doi
- [32]
-
[33]
van der Heijden and E
J. van der Heijden and E. Verlinde, JHEP �� , 207 (2025) doi:10.1007/JHEP02(2025)207 [arXiv:2408.00071 [hep-th]]
2025 arXiv
-
[34]
A. R. Brown, D. A. Roberts, L. Susskind, B. Swingle and Y. Zhao, Phys. Rev. Lett. ��� , no.19, 191301 (2016) doi:10.1103/PhysRevLett.116.191301 [arXiv:1509.07876 [hep-th]]
2016 arXiv
-
[35]
Ben-Ami and D
O. Ben-Ami and D. Carmi, JHEP �� , 129 (2016) doi:10.1007/JHEP11(2016)129 [arXiv:1609.02514 [hep-th]]. 44
2016 arXiv
-
[36]
R. Abt, J. Erdmenger, H. Hinrichsen, C. M. Melby-Thompson, R. Meyer, C. Northe and I. A. Reyes, Fortsch. Phys. �� , no.6, 1800034 (2018) doi:10.1002/prop.201800034 [arXiv:1710.01327 [hep-th]]
2018 arXiv
-
[37]
V. E. Hubeny, H. Maxfield, M. Rangamani and E. Tonni, JHEP �� , 092 (2013) doi:10.1007/JHEP08(2013)092 [arXiv:1306.4004 [hep-th]]
2013 arXiv
-
[38]
V. E. Hubeny and M. Rangamani, JHEP �� , 114 (2012) doi:10.1007/JHEP06(2012)114 [arXiv:1204.1698 [hep-th]]
2012 arXiv
- [39]
-
[40]
Faulkner and A
T. Faulkner and A. J. Speranza, JHEP �� , 099 (2024) doi:10.1007/JHEP11(2024)099 [arXiv:2405.00847 [hep-th]]
2024 arXiv
-
[41]
Bousso, Z
R. Bousso, Z. Fisher, S. Leichenauer and A. C. Wall, Phys. Rev. D �� , no.6, 064044 (2016) doi:10.1103/PhysRevD.93.064044 [arXiv:1506.02669 [hep-th]]
2016 arXiv
-
[42]
G. W. Gibbons and S. W. Hawking, Phys. Rev. D �� , 2738-2751 (1977) doi:10.1103/PhysRevD.15.2738
1977 doi
-
[43]
Susskind, JHAP �, no.1, 1-22 (2021) doi:10.22128/jhap.2021.455.1005 [arXiv:2109.14104 [hep-th]]
L. Susskind, JHAP �, no.1, 1-22 (2021) doi:10.22128/jhap.2021.455.1005 [arXiv:2109.14104 [hep-th]]
2021 arXiv
-
[44]
Lashkari, K
N. Lashkari, K. L. Leung, M. Moosa and S. Ouseph, [arXiv:2412.19882 [hep-th]]. 45
Reviewed August 6, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.