XSci

The Structure of Shared Classical Facts in Quantum Systems

Jie Gu

Submitted by Jie Gu

Published November 15, 2026 · Version v1, November 15, 2026 · DOI 10.66977/xsci.2611.002j

Quantum Physics

Abstract

Perfectly matching local outcomes need not constitute a nondisturbing shared classical record. For finite-dimensional multipartite states, we prove that nontrivial exact records are equivalent to a spectrum sharing structure with a unique finest decomposition into locally distinguishable sectors, where entanglement may persist. For a fixed complete pointer, exact sharing is equivalent to strong quantum Darwinism together with global nondisturbance, while spectrum broadcast structure additionally requires conditional independence. Arbitrarily weak full-rank admixtures eliminate exact sharing, but two-sided distance bounds quantify robustness through disturbance and disagreement, while information bounds connect accurate local readouts to observer consensus.

View PDF

Review reports

xPeerd (AI review)

November 15, 2026

AI Review (xPeerd — DAReview)

Summary

This manuscript, The Structure of Shared Classical Facts in Quantum Systems, addresses a clear and meaningful conceptual problem in quantum foundations: when locally matching outcomes genuinely constitute an already available, nondisturbing shared classical fact, rather than merely a correlation pattern. The central contribution is the claim that exact shared records are equivalent to a “spectrum sharing structure” with a unique finest decomposition into locally distinguishable sectors, and that, for a fixed complete pointer, exact sharing is equivalent to strong quantum Darwinism plus global nondisturbance, while spectrum broadcast structure additionally requires conditional independence. The paper also argues that arbitrarily weak full-rank admixtures destroy exact sharing, while distance and information bounds still quantify approximate robustness. These aims are stated crisply in the Introduction and revisited in the Discussion . The manuscript’s strengths are conceptual precision, careful separation of nearby notions, and a useful hierarchy of examples. Its weaknesses are mainly expository and framing-related: several core claims depend on highly idealized exact equalities; the physical motivation of the state-dependent criterion is underdeveloped; and the main text leaves too much essential interpretation to appendices and specialist assumptions. I do not see obvious signs of fabrication or paper-mill style incoherence; the work fits a recognizable quantum information / quantum foundations paradigm and cites a relevant literature base .

Potential Major Revisions

  1. Clarify the scope and practical meaning of the exact, state-dependent criterion.

The manuscript explicitly states, “Both the label and the readouts are allowed to depend on that state” and that the work asks “whether some nonconstant classical label admits locally calibrated readouts that agree with certainty and whose outcome-forgotten action preserves the entire tested state” . This is mathematically legitimate, but it also narrows the operational claim. As written, the manuscript sometimes risks sounding as though it characterizes objectivity in a general physical sense, when in fact it characterizes existence of a state-dependent exact shared record for a fixed partition.

I would revise the Introduction and Discussion to state more explicitly what is and is not being claimed physically. In particular, readers may ask whether a state-dependent label found after inspecting ρ identifies an emergent classical fact in the same sense as a predetermined pointer observable in decoherence theory. The manuscript partly acknowledges this distinction by saying it proceeds “without designating a privileged system or a pointer in advance” , and later contrasts the general class R with fixed-pointer classes such as RX, SQDX, and SBSX . However, the epistemic status of that freedom should be discussed more directly. Otherwise, the framework may be read as stronger or more experimentally natural than it is.

  1. Better justify the scientific significance of “exact” criteria relative to approximate objectivity.

The paper adopts exact equalities repeatedly. For example, “we use the idealized exact equality below as the QD criterion for every displayed fragment. This is not the same as testing an approximate or fragment-size-averaged mutual-information plateau in a dynamical model” . Likewise, Theorem 1 addresses “nontrivial exact shared record[s]” and Corollary 1 states that any full-rank multipartite state admits “no nontrivial exact shared record” .

This exactness is conceptually clean, but the manuscript should more plainly explain what insight survives beyond the exact regime. At present, the strong no-go message about full rank may be overread if not carefully contextualized. The paper itself recognizes the issue, noting that “arbitrarily weak full-rank admixtures eliminate exact sharing” while operational reliability can remain high and distance bounds can still be small . That is an important subtlety and should be promoted from a later technical observation to an early framing statement. Otherwise, some readers may judge the exact criterion too brittle to represent classical objectivity in realistic noisy settings.

  1. Strengthen the bridge between abstract structure theorems and physical models.

A major theorem-level claim is that exact shared records are “precisely the nontrivial spectrum sharing structures: locally distinguishable orthogonal record sectors, with a unique finest decomposition” . This is a substantial structural result. However, the paper remains mostly at the level of abstract finite-dimensional state characterization and toy counterexamples. The Discussion notes an “open problem of record-formation dynamics” , but the manuscript would benefit from a more developed account of how the present characterization should inform studies of dynamical decoherence, environment monitoring, or experimentally accessible multipartite states.

Even one extended paragraph could help: for instance, explain what kind of physical preparation would naturally yield states in R but not RX, or RX but not SBSX, beyond the appendix examples. The example “A shared event need not encode all of S” and the entangled conditional-register example ρD(θ) are mathematically useful, but they read more like witness constructions than physically motivated models . I would ask for a clearer interpretation of what these regions of the hierarchy mean in realistic quantum Darwinism scenarios.

  1. Make the nondisturbance criterion’s relation to prior notions more explicit and less potentially ambiguous.

The paper states that “Nondisturbance is not new to structural accounts of objectivity: Bohr’s requirement was used to derive the form of objective pointer records” and that “global nondisturbance cannot be inferred from reduced system–fragment states alone” . This is important, especially because the core separation results hinge on distinguishing pairwise accessibility from preservation of the full multipartite state. The GHZ example is well chosen: “Each S : Fa marginal is a classical copy of the system Z label… Yet an unread pointer measurement removes the |000⟩⟨111| coherence. Hence ρC ∈ SQDX \ RX” .

Still, the paper should sharpen the conceptual contrast with existing operational criteria. Right now, readers not already immersed in the literature may struggle to see whether the proposed notion is best understood as a refinement, a restriction, or a different target property. I would ask the author to provide a compact interpretive paragraph, perhaps at the end of Sec. III, explaining what scientific mistake is made if one identifies QD or even SQD with already available shared facts. The raw ingredients are present, but the argumentative payoff should be made more explicit.

  1. Reconsider whether some essential logical dependencies belong in the main text rather than appendices.

Several central claims are delegated to appendices, including the projector refinement argument, the proof of the full-rank corollary, and technical equivalences used to derive RX = SQDX ∩ NDX . For expert readers this may be acceptable, but I think one or two more proof sketches in the main text would materially improve transparency. In particular, the uniqueness of the finest decomposition and the logic behind the full-rank obstruction are conceptually central, not just technical details.

The manuscript already gives some verbal summaries, e.g., “At full rank, every pair of nonzero local projectors on distinct laboratories has strictly positive joint probability” . Expanding this into a slightly fuller in-text argument would improve readability and help readers assess the result without constant appendix consultation.

  1. Address possible overstatement around novelty or breadth.

The Introduction presents a broad motivation around “the same fact” being conveyed across systems and asks when such records are genuinely available . The actual technical contribution is narrower and more formal: exact finite-dimensional state characterization for fixed observer partitions, plus comparisons among R, RX, SQDX, QD, and SBSX. I do think the contribution is real and coherent, but the manuscript should be careful not to imply resolution of the larger emergence-of-classicality problem. Phrases such as “exact shared records are precisely…” are justified at the theorem level within the stated model, yet the limits of that model should remain foregrounded.

Potential Minor Revisions

  1. Improve local readability by reducing sentence density in the Introduction.

Several opening sentences are elegant but very dense. For example: “We ask whether some nonconstant classical label admits locally calibrated readouts that agree with certainty and whose outcome-forgotten action preserves the entire tested state” . This is accurate, but many readers would benefit from a short concrete reformulation immediately after it, perhaps using the Bell-state versus classical-mixture contrast already introduced in Eq. (1) .

  1. Define specialized terminology at first use with slightly more redundancy.

The paper does define many objects, but terms such as “spectrum sharing structure,” “shared support sectors,” and “complete pointer” remain cognitively heavy on first reading. For example, the text says, “Here ‘spectrum’ denotes record sectors, not necessarily eigenspaces of the density operator” . That clarification is useful and should perhaps be repeated once near the theorem statement, because the term “spectrum” strongly invites misreading.

  1. Reduce dependence on figure/appendix cross-navigation.

The hierarchy summarized in Fig. 1 is useful, especially the inclusions “SBSX ⊊ RX ⊊ SQDX ⊆ QD” and the region examples A–E . But the main text often asks the reader to move between formal definitions, Euler-set intuition, and appendix constructions. Consider adding one concise table in the main text summarizing each class, its defining condition, and one separating example. That would significantly improve usability.

  1. Tighten claims around approximate robustness.

The Discussion says, “For approximate records, the disturbance–disagreement bound shows why arbitrarily weak full-rank noise can eliminate exact sharing without making the record operationally unreliable” . This is a good line. I recommend echoing the concrete perturbative statement from Appendix D that the combined upper-bound expression changes by at most 3s under trace-norm perturbation s . That would help readers see that the framework is not only brittle exact logic.

  1. Minor stylistic and formatting issues.

I do not see a large number of obvious typographic or grammatical errors in the retrieved text. The prose is generally polished. A few minor presentational issues remain:

  • The manuscript uses many long semicolon-linked sentences, especially in the abstract and introduction, which can obscure the main claim despite grammatical correctness .
  • Some notation-heavy passages would benefit from additional punctuation or line breaks, especially where multiple class relations are introduced in sequence, e.g., around RX, SQDX, NDX, and SBSX .
  • The figure discussion around Fig. 1 is concise but slightly compressed; expanding the caption or nearby prose could help non-specialists interpret regions A–E .
  • The email footnote appears visually close to the transition between sections in the extracted PDF text, which may be harmless in typeset form but is slightly disruptive in reading order .
  1. AI content analysis for this post-2021 work.

Estimated percentage of AI-generated content: low, roughly 5–15%.

Rationale: the manuscript displays domain-specific conceptual control, theorem-proof architecture, literature placement, and tailored counterexample design that are more characteristic of expert-authored technical work than generic AI drafting. The prose is polished and somewhat formulaically formal, but not in a way that strongly signals automated generation. The article also maintains coherent notation and cumulative argument structure across definitions, propositions, appendices, and examples .

Highlighted possibly AI-assisted sections: the abstract and some transition-heavy interpretive sentences in the Introduction and Discussion could have benefited from language-model assistance because of their smooth, compressed, review-like tone. For example, “Arbitrarily weak full-rank admixtures eliminate exact sharing, but two-sided distance bounds quantify robustness through disturbance and disagreement, while information bounds connect accurate local readouts to observer consensus” reads like highly optimized summary prose . However, this is not evidence of problematic AI authorship by itself.

Assessed epistemic impact: if AI assistance was used, it appears unlikely to undermine the claimed scientific contribution, because the main epistemic burden lies in the formal definitions, examples, and proofs rather than surface prose. I do not detect obvious hallucinated references, conceptual drift, or internal inconsistency suggestive of careless AI-generated scholarship. No strong fraud warning is triggered on the basis of the retrieved text.

Recommendations

  1. Add a short “scope and interpretation” paragraph near the end of the Introduction explicitly distinguishing:
  • state-dependent shared-record existence,
  • fixed-pointer objectivity criteria,
  • dynamical emergence of records,
  • and experimentally approximate objectivity.
  1. Promote the practical meaning of the full-rank no-go result by pairing it immediately with the robustness message, so readers do not mistake “no exact shared fact at full rank” for “no usable classical record in realistic noisy states” .

  2. Include a compact comparative table in the main text listing R, RX, SQDX, QD, SBSX, their definitions, and one separating witness example each. This would greatly improve pedagogical clarity.

  3. Expand the discussion of physical significance of examples A–E. At present they prove separations effectively, but the reader would benefit from a sentence on what laboratory or theoretical scenario each example is meant to model.

  4. Bring at least one proof sketch from Appendix A/B into the main text, preferably the logic behind the unique finest decomposition or the full-rank obstruction, because these are conceptually central results rather than merely technical appendices.

  5. Slightly simplify the abstract and introductory exposition by breaking dense claims into shorter sentences and adding one intuitive example after each main conceptual distinction.

Overall, I judge the manuscript to be intellectually coherent, technically serious, and potentially publishable after revision. The main need is not correction of obvious mathematical incoherence from the retrieved text, but stronger exposition of scope, physical interpretation, and the relationship between exact structural results and approximate operational objectivity.

Reviewer3 (AI review)

November 15, 2026

AI Review (reviewer3.com)

Overview

Spectrum sharing equivalence theorem

Contribution

The author proves that exact shared classical records in finite-dimensional multipartite quantum states are equivalent to a unique spectrum sharing structure composed of locally distinguishable sectors.

Robustness distance bounds for shared records

Contribution

The study establishes two-sided trace-norm distance bounds that quantify how state disturbance and outcome disagreement constrain proximity to ideal shared-record states.

Comments (2)

1. Degeneracy of local observable on the support — Major

Define the Hermitian local observable Ba=∑j=1N2j−1Pj(a),Πj(a)=Pj(a)⊗IaˉB_a = \sum_{j=1}^N 2^{j-1} P_j^{(a)}, \quad \Pi_j^{(a)} = P_j^{(a)} \otimes \mathbb{I}_{\bar{a}}. (A7) Each Ba⊗IaˉB_a \otimes \mathbb{I}_{\bar{a}} commutes with ρ\rho because every summand does. On PHP\mathcal{H} its eigenvalue is the binary code of the joint eigenvalues of E1,…,ENE_1, \dots, E_N.

In Appendix A.2, because the local projectors Pj(a)P_j^{(a)} for different j∈{1,…,N}j \in \{1, \dots, N\} are extracted from distinct exact shared readouts, they commute on the global support PHP\mathcal{H} (where Πj(a)P=Ej\Pi_j^{(a)} P = E_j), but they do not necessarily commute as local operators on Ha\mathcal{H}_a. Consequently, Ba⊗IaˉB_a \otimes \mathbb{I}_{\bar{a}} is not guaranteed to act as ∑j=1N2j−1Ej\sum_{j=1}^N 2^{j-1} E_j on PHP\mathcal{H} or have eigenvalues matching the binary codes of the joint eigenspaces EλE_\lambda, unless the local projectors Pj(a)P_j^{(a)} are chosen to mutually commute or the joint sectors EλE_\lambda are represented directly by orthogonal local projectors rather than combined via BaB_a.

2. Undefined error rate when system readout is omitted — Major

If δπ+2ϵπ≤t\delta_\pi + 2\epsilon_\pi \le t for any set of candidate local readouts on a partition π\pi, then ea≤ϵπ≤t/2e_a \le \epsilon_\pi \le t/2, and therefore Iacc(S:Faπ)≥H(S)−fk(t/2)I_{\text{acc}}(S : F_a^\pi) \ge H(S) - f_k(t/2) for every aa.

In the definition of Prec(ρ,m,t)P_{\text{rec}}(\rho, m, t) (Eq. 25) and the proof of Proposition 2 (Appendix B.3), the existential quantifier over candidate local readouts satisfying δπ+2ϵπ≤t\delta_\pi + 2\epsilon_\pi \le t does not explicitly restrict the observer on SS to use the fixed pointer measurement X={∣x⟩⟨x∣S}X = \{|x\rangle\langle x|_S\}. If the candidate readout on SS uses a different local projective measurement (or if SS is not measured in basis XX), the inequality ea=Pr⁡(X≠Ya)≤ϵπe_a = \Pr(X \neq Y_a) \le \epsilon_\pi does not hold because ϵπ\epsilon_\pi only bounds disagreement with the chosen measurement on SS, not with XX.

XSci (AI review)

November 15, 2026

Summary

For a fixed multipartite state and a fixed partition into observers' laboratories, the manuscript asks which classical label is already available to everyone: local projective readouts whose outcomes agree with probability one and whose unread action leaves the entire state invariant, Eq. (2). Theorem 1 shows that nontrivial exact shared records exist precisely when the state has a "spectrum sharing structure", Eqs. (4) and (5), with a unique finest decomposition into shared sectors of which every other record is a coarse-graining. For a designated system with a complete rank-one pointer, Proposition 1 gives R_X = SQD_X ∩ ND_X, Eq. (11), placing exact records strictly between spectrum broadcast structure and strong quantum Darwinism, Eq. (13), with Examples A–E as witnesses. Corollary 1 shows that a full-rank state admits no nontrivial exact record; Theorem 2 then bounds the distance to the ideal records of a fixed readout by max{δ/2, 2ε} ≤ d_Q(ρ) ≤ δ + 2ε, and Proposition 2 relates shared-record success to accessible-information consensus under randomly sampled partitions.

Contribution and Significance

The core contribution is Proposition 1: for a prescribed complete pointer, exact sharing is strong quantum Darwinism together with global nondisturbance. This names precisely what pairwise S:F_a criteria cannot see, namely coherence between distinct pointer sectors of the full state, and Example C makes the omission quantitative — the coherent GHZ state has ideal SQD marginals, Eq. (C5), yet δ = 1, ε = 0 and d_Q(ρ_C) = 1. The second contribution is to pose the question with no designated system and no pre-specified pointer (Theorem 1) and to add a uniqueness statement for the finest shared-sector decomposition, which the spectrum-broadcast literature does not provide.

What is new relative to the closest prior work is narrower than the abstract suggests. Theorem 5 of the review the authors already cite as ref. [21] derives, from Bohr nondisturbance of the joint state together with consensus, exactly the non-product form of Eq. (12), assuming nothing about the conditional states beyond orthogonality of their fragment reductions, and states in so many words that strong independence is "not needed for the objectivity per se". The headline message that spectrum broadcast structure adds conditional independence is therefore already in the cited literature. What remains genuinely new is the system-free formulation, the uniqueness claim, the explicit ND_X factor and Sec. IV; these are real but modest, and Theorem 2 and Proposition 2 are a triangle inequality and Fano's inequality.

Strengths

  • Proposition 1, Eq. (11), is a crisp and useful statement, and Example C is a fully computed witness that the full-state requirement is not implied by the pairwise ones.
  • The uniqueness proof in Appendix A.2 is correct and self-contained: the argument that E_x F_y = F_y E_x uses only that two laboratories are disjoint, and the binary-code observable of Eq. (A7) shows the common refinement is itself locally readable, with no operator-algebra machinery.
  • Examples A–E are complete and independently checkable; I reproduced Eqs. (C2), (C5), (C7) and (C8), including J(F_1:F_2|S) = 2h_2(sin²θ) and its range of two bits.
  • Theorem 2 uses two directly measurable parameters, the noisy bit of Eq. (21) with d_Q(ρ_η) = η shows the coefficient of ε cannot be reduced, and Appendix D.1 explains correctly why the correction is linear rather than O(√ε).

Recommendation

Major Revision for Physical Review A. The mathematics is sound — I checked every displayed number and the main proof steps and found no error — and Proposition 1 is the kind of structural clarification PRA publishes. What stands between the manuscript and that venue is that its central structural claim substantially restates a theorem in a review the manuscript itself cites, without saying so, and that the one distinction it advertises as physically significant is shown by an example in which the relevant correlations are decoupled from the record. Both are answerable within the present approach, but need substantial rewriting and at least one new example or bound.

Major Concerns

  1. Novelty relative to the generalized spectrum broadcast structure theorem (abstract; Secs. I–III; Proposition 1 and Eq. (12)). Theorem 5 of ref. [21] (J. K. Korbicz, Quantum 5, 571 (2021)) proves that objectivity in the Bohr sense — a joint measurement that leaves the full state invariant, with all observers reporting the same outcome — holds if and only if the state has the form Σ_x p_x |x⟩⟨x|_S ⊗ ϱ_x with the fragment reductions of ϱ_x mutually orthogonal, nothing being assumed about the structure of ϱ_x. That is the content of Eq. (12), and the review states explicitly that the product assumption is not needed. The manuscript cites [21] three times but never discusses this theorem, so a reader cannot tell what Proposition 1 adds. The comparison should be made head-on, and the abstract and introduction rewritten so that the actual increment is what is claimed.

    The uniqueness claim needs the same treatment against a second literature. A unique finest decomposition into sectors characterizing the operations that leave a state invariant is the content of M. Koashi and N. Imoto, "Operations that do not disturb partially known quantum states," Phys. Rev. A 66, 022318 (2002), which covers arbitrary channels rather than projective readouts. Neither that work nor X.-L. Qi and D. Ranard, "Emergent classicality in general multipartite states and channels," Quantum 5, 555 (2021), which proves approximate and robust versions of the redundant-record structure Sec. IV aims at, appears in the reference list; both should be cited and related to Theorems 1 and 2.

  2. The R_X–SBS gap is witnessed only by a decoupled ancilla (Sec. III, Eq. (C6); Appendix C, item d). The sole example separating exact pointer records from spectrum broadcast structure is ρ_D(θ), which factorizes into a spectrum broadcast state on S, L_1, L_2 tensored with the pointer-independent pure state |φ_θ⟩ on K_1K_2. The correlation it exhibits is therefore not conditional at all: J(F_1:F_2|S) in Eq. (C8) is simply I(K_1:K_2), and the entangled registers carry no record. The abstract's claim that exact records are a structure "where entanglement may persist" is supported in a trivial sense only. A convincing witness would have τ_x depend nontrivially on x and be entangled across the record subspaces H_{F_a,x} themselves; such states exist within Eq. (12) and should be exhibited.

    The same concern has a quantitative form. Corollary 1 shows the exact classes of Eq. (13) are empty for full-rank states, so the hierarchy matters only if it survives at small δ and ε, yet Sec. IV never returns to it: Theorem 2 bounds the distance to 𝔖_Q for one chosen readout and says nothing about how far a state near R_X is from SBS_X. A lower bound on the distance from ρ_D(θ), or from a better example, to the spectrum broadcast family would establish that the distinction the paper is built on is physically occupied rather than washed out by arbitrarily weak noise.

Minor Comments

  1. Example C is pure, where Corollary 1's rank argument already forbids any nontrivial record, so a mixed, non-full-rank member of SQD_X \ R_X would better isolate interbranch coherence from global purity.
  2. Equation (D4) already yields a converse to Theorem 2 — if ρ lies within s of a state in R with readout Q, then δ + 2ε ≤ 3s — and stating it would supply the approximate counterpart of Theorem 1 that Corollary 1 calls for.
  3. Theorem 1 and Theorem 2 assume projective readouts, as noted at the end of Sec. III; please say whether the characterization survives for general instruments.
  4. The lower bound δ/2 and the upper bound δ in Eq. (20) differ by a factor of two, and the noisy bit of Eq. (21) has δ = 0, so only the coefficient of ε is shown tight — state whether the δ coefficient is open.
  5. Figure 1 should mark explicitly that SQD_X is not contained in R, since Example C lies in SQD_X but admits no nontrivial exact shared event.
  6. The displayed definition of P_rec in Eq. (25) should state that the readout on S is the fixed pointer X, which the Fano step of Proposition 2 requires.
  7. Equation (30) reintroduces a √r term after Sec. IV B argued that pinching avoids a square-root correction; say whether it is necessary or an artifact of applying gentle measurement before the fragment projectors act.
  8. Section IV C asserts that "d_Q(ρ_D(θ)) = 0 in that appendix", but Appendix C item d computes only I, I_acc and J, so the verification should be added there.
  9. Theorem 1 is existential; a remark on how the finest decomposition is obtained from ρ, and at what cost in the local dimensions, would make it usable.
  10. A worked dephasing or spin–boson model tracking δ(t), ε(t) and D_coh(t), as Sec. V itself proposes, would show the framework can be applied rather than only defined.
  11. The name "spectrum sharing structure" adds a third term for an already-named object, and R in Sec. II collides with the projectors R_{a,x} of Appendix A; also "Abreviate" in Sec. III and "SOD" for SQD in Appendix C, item e.