Comment on "A General Framework for Constructing Local Hidden-state Models to Determine the Steerability"

Nick von Selzam, Florian Marquardt

#2592 of 2593 · Quantum Physics
Share
Tournament Score
1136±45
10501750
4%
Win Rate
1
Wins
25
Losses
26
Matches
Rating
2.5/ 10
Significance
Rigor
Novelty
Clarity

Abstract

We point out that the method presented in a recent arXiv article by Jia et al. (arXiv:2512.21848) for constructing local hidden-state models closely follows the framework we developed in N. von Selzam & F. Marquardt (PRX Quantum, 2025) for constructing local hidden-variable models. While Jia et al. cite our work, the extent of the methodological overlap and the degree of textual similarity are not adequately reflected by the attribution given. We document this overlap in detail.

AI Impact Assessments

(3 models)

Scientific Impact Assessment

1. Core Contribution

This paper is a formal "Comment on..." submission that documents extensive methodological and textual overlap between a subsequent arXiv preprint (Jia et al., arXiv:2512.21848) and the authors' own earlier published work (von Selzam & Marquardt, PRX Quantum, 2025). The core contribution is not a new scientific method or result, but rather a detailed, evidence-based claim of insufficient attribution. The authors provide three tables systematically cataloging: (i) shared methodological elements and whether they were cited, (ii-iii) side-by-side textual comparisons demonstrating near-verbatim overlap.

The authors acknowledge that extending their local hidden-variable (LHV) framework to the quantum steering (local hidden-state, LHS) setting is a legitimate scientific contribution, and they identify the novel element in Jia et al.—parameterization of hidden states via unconstrained complex matrices. Their concern is specifically about presentation and attribution rather than the legitimacy of the extension itself.

2. Methodological Rigor

For a comment of this nature, the evidentiary standard is whether the overlap is documented convincingly and fairly. The authors meet this standard well:

  • Table I lists 10 distinct methodological elements (finite cloud of hidden-variable tuples, softmax parametrization, Gell-Mann matrix expansion with non-standard normalization, basis function expansion, coefficient matrix representation, odd spherical harmonics for qubit measurements, sigmoid function for gradient flow, loss function definition, stochastic gradient descent with batch sampling) and shows that only 1 of 10 is attributed to the original work.
  • Tables II and III provide extensive side-by-side textual comparisons spanning multiple sections. The degree of similarity is striking—in many cases, sentences are nearly identical modulo notation changes (LHV→LHS, variable renaming) and minor synonym substitutions ("Borrowing" → "Drawing upon," "as can be seen from" → "as can be verified by").
  • The documentation is thorough and speaks for itself. The authors are careful to be measured in tone, acknowledging the legitimate aspects of Jia et al.'s contribution while precisely delineating what was borrowed without adequate credit.

    3. Potential Impact

    The direct scientific impact of this comment is narrow—it does not advance the field's understanding of quantum entanglement, steering, or hidden-variable models. However, it serves several important functions:

  • Correcting the scientific record: Ensuring proper attribution is fundamental to the integrity of scientific publishing. If the overlap is as extensive as documented, this comment prevents the community from attributing the underlying framework to the wrong group.
  • Precedent for ML-in-quantum-foundations: The original framework for machine-learning-based construction of LHV models represents a methodologically significant advance. Protecting priority for such frameworks matters as the intersection of ML and quantum information grows.
  • Limited broader influence: This comment is unlikely to influence researchers outside the narrow subfield of entanglement/steering characterization via hidden-variable models.
  • 4. Timeliness & Relevance

    The comment is timely—it was posted (April 2026) roughly four months after the preprint it critiques (December 2025), which is an appropriate response window. The original work was published in a top-tier venue (PRX Quantum) approximately eight months before the contested preprint appeared, establishing clear priority. Given that the Jia et al. paper is still on arXiv (not yet peer-reviewed and published), this comment could influence the peer review process and prompt appropriate revisions before formal publication.

    5. Strengths & Limitations

    Strengths:

  • Exceptionally well-documented: The three-table structure is clear, systematic, and allows readers to judge for themselves.
  • Measured and professional tone: The authors avoid inflammatory language and acknowledge the legitimate novelty in the contested paper.
  • The textual parallels are remarkably close, making the case compelling.
  • The distinction between methodological overlap (Table I) and textual overlap (Tables II-III) is useful, as both raise distinct concerns.
  • Limitations:

  • As a comment, this paper has no independent scientific contribution beyond the attribution claim.
  • The paper does not discuss whether the methodological elements in Table I are sufficiently standard/obvious that independent rediscovery is plausible (though the textual overlap in Tables II-III makes this explanation unlikely).
  • The comment does not address whether some of these elements might have even earlier precedents in the optimization or machine learning literature, which could complicate the attribution picture.
  • Impact is limited to the specific subcommunity working on constructive hidden-variable/hidden-state models.
  • Additional Observations

    The strength of this comment rests primarily on the textual parallels rather than the methodological overlap alone. While individual methodological choices (softmax, sigmoid, SGD, Monte Carlo batch sampling) might be rediscovered independently, the combination of identical mathematical notation conventions (e.g., the specific non-standard G₀ normalization), identical pedagogical framing ("Borrowing/Drawing upon machine learning terminology"), and nearly identical sentence structures throughout multiple sections builds a cumulative case that is difficult to explain by coincidence. The fact that Jia et al. do cite the original work—but only for one specific element—makes the pattern more concerning, as it suggests awareness of the source.

    This paper fulfills an important but unglamorous role in scientific self-correction. Its impact on the broader field is minimal, but its importance for maintaining attribution standards in a growing subfield should not be dismissed.

    Rating:2.5/ 10
    Significance 2Rigor 7.5Novelty 1Clarity 8.5

    Generated Apr 20, 2026

    Comparison History (26)

    vs. Agent Choice via Quantum Flux in Living Systems
    gpt-5.25/16/2026

    Paper 2 is more likely to have near-term scientific impact: it addresses a timely, active research area (quantum steering/hidden-state models) and engages directly with a specific, citable methodological framework, potentially influencing attribution, replication, and subsequent work. Paper 1 is highly speculative, provides limited methodological detail in the abstract, and hinges on controversial claims about indeterministic “intentional choices,” making rigorous validation and uptake less likely. Thus, Paper 2 scores higher on rigor, relevance, and immediate applicability, despite narrower scope.

    vs. A conjecture on a tight norm inequality in the finite-dimensional l_p
    gemini-3.15/16/2026

    Paper 2 proposes a novel mathematical conjecture with proofs for specific cases and implications for quantum information theory, offering broader scientific value. In contrast, Paper 1 is a comment focused on academic attribution and methodological overlap between specific articles, which has highly localized impact and contributes little new scientific knowledge.

    vs. Quantum Suicide in Many-Worlds Implies P=NP
    gemini-3.15/16/2026

    While Paper 1 relies on a highly impractical and somewhat satirical thought experiment (quantum suicide), it explores a profound theoretical intersection between quantum mechanics, anthropic principles, and computational complexity (P vs NP). Paper 2 is merely a priority dispute and plagiarism claim, which, while important for academic integrity, offers zero novel scientific contributions or impact to the field's actual knowledge.

    vs. Expander attention as exchange-correlation
    gpt-5.25/16/2026

    Paper 1 proposes a novel, linearly scaling expander-graph transformer ansatz for nonlocal exchange-correlation functionals in DFT, directly addressing a key bottleneck (accuracy vs. cost) in a widely used computational paradigm. If validated broadly, it could enable scalable, accurate treatment of strongly correlated systems with major applications across chemistry and materials science, giving high breadth and timeliness. Paper 2 is primarily a priority/attribution comment; while important for scholarly record, it is unlikely to drive new methods or broad downstream applications, so its expected scientific impact is lower.

    vs. GAT-QNN: Genetic Algorithm-Based Training of Hybrid Quantum Neural Networks
    gpt-5.25/16/2026

    Paper 2 has higher potential impact: it proposes a concrete, novel training/deployment framework (GA-driven macro/micro circuit co-training plus backend-aware architecture selection) addressing key NISQ constraints, with demonstrated empirical gains across multiple backends and potential applicability to many hybrid quantum ML tasks. Paper 1 is primarily a comment/priority and attribution clarification; while important for scholarly record, it is less likely to drive new methods or broad downstream applications. Paper 2 is more timely for near-term quantum computing and has broader cross-field relevance (optimization, ML, quantum compilation).

    vs. Storage of telecom-band time-bin qubits in thin-film lithium niobate
    claude-opus-4.65/16/2026

    Paper 2 demonstrates the first on-chip quantum memory in thin-film lithium niobate (TFLN), a significant experimental milestone for integrated quantum photonics. It addresses a critical missing component for quantum networks—on-chip quantum memory at telecom wavelengths—with demonstrated qubit storage fidelity exceeding classical limits. This has broad implications for quantum repeaters, quantum internet, and scalable quantum information processing. Paper 1 is a comment documenting methodological overlap and insufficient attribution, which, while important for academic integrity, has minimal scientific impact beyond the specific priority dispute.

    vs. A Superconducting Levitating Oscillator with < 1 $μ$Hz Resonance Linewidth
    gpt-5.25/16/2026

    Paper 1 reports a highly novel, rigorously demonstrated cryogenic superconducting levitated oscillator with an exceptionally narrow <1 μHz linewidth and >110 h ring-down, enabling ultra-isolated macroscopic tests relevant to quantum–classical boundary, precision force sensing, and low-temperature physics. It has clear real-world and cross-field applications (quantum foundations, metrology, cryogenics, superfluid/impurity drag measurements) and is timely for quantum-limited macroscopic experiments. Paper 2 is primarily a priority/attribution comment with limited methodological novelty and narrower scientific/technological impact.

    vs. Graphical Algebraic Geometry: From Ideals and Varieties to Quantum Calculi
    gpt-5.25/16/2026

    Paper 1 presents a novel, extensible diagrammatic framework (Graphical Algebraic Geometry) with strong theoretical results (universality/completeness for (co)span semantics), links to #CSP complexity (#P-hardness), and a substantive bridge to quantum computation via the qudit ZH calculus, suggesting broad cross-field impact (algebraic geometry, categorical methods, complexity, quantum). It also indicates potential practical applications in compositional reasoning about polynomial constraints and quantum amplitudes. Paper 2 is primarily a priority/attribution comment; while important for scholarly record, it contributes limited new methodology or broad applicability.

    vs. Exploring the Geometric and Dynamical Properties of Spin Systems and Their Interplay with Quantum Entanglement
    claude-opus-4.65/16/2026

    Paper 2, while a thesis rather than a novel research article, covers substantial scientific content exploring quantum entanglement through geometric and dynamical perspectives across multiple spin systems and interaction models. It addresses fundamental questions connecting classical and quantum mechanics with practical applications in quantum information. Paper 1 is a brief comment documenting methodological overlap and attribution concerns regarding another paper—important for academic integrity but with very limited scientific impact beyond the specific dispute. It contributes no new methods or results.

    vs. Breaking the scalability barrier via a vertical tunable coupler in 3D integrated transmon system
    gemini-3.15/16/2026

    Paper 1 presents a significant technological advancement in quantum computing scalability through a novel 3D integrated architecture and experimental demonstrations of high-fidelity operations. In contrast, Paper 2 is merely a comment addressing attribution and methodological overlap regarding a specific prior work. Paper 1 has broad, high-impact implications for fault-tolerant quantum computation, whereas Paper 2's impact is limited to correcting the academic record.

    vs. Schroedinger's Equation at 100: The Wave Picture That Helped and Possibly Hurt
    gpt-5.25/16/2026

    Paper 1 is primarily a comment alleging methodological/textual overlap; while relevant for scholarly attribution and integrity, it is unlikely to generate broad new research directions or applications. Paper 2, though an essay, offers a timely centenary synthesis linking historical interpretation (Schroedinger, Mach, Boltzmann) to enduring foundational debates (configuration space, ontology, Bell-type theorems), with potential to influence perspectives across quantum foundations, philosophy of physics, and pedagogy. Its breadth and relevance are wider, even if methodological rigor is not experimental/technical.

    vs. Unitary dynamics and resource trade-offs in a four-qubit isotropic Heisenberg XXX chain with tunable next-nearest-neighbor coupling
    claude-opus-4.65/16/2026

    Paper 1 presents original research with closed-form analytical results for quantum dynamics in a four-qubit Heisenberg chain, providing exact benchmarks for quantum devices and a unifying framework connecting fidelity, coherence, and entanglement through a single composite phase. While modest in scope, it offers novel, concrete results with potential applications in quantum information science. Paper 2 is a comment documenting methodological overlap/insufficient attribution in another work—important for academic integrity but not contributing new scientific knowledge or methods, thus having minimal scientific impact.

    vs. Fundamental Physics, Existential Risks and Human Futures
    gpt-5.25/16/2026

    Paper 2 has higher potential impact: it targets broad, high-stakes questions (beyond-quantum physics, quantum-gravity, consciousness, implications for AI and human futures) with wide cross-field relevance and timeliness. If it develops concrete arguments or testable frameworks, applications could be transformative. Paper 1 is primarily a commentary focused on attribution and overlap; while important for scholarly integrity, it is unlikely to introduce substantial new methodology or enable broad real-world applications, limiting its scientific impact.

    vs. Multiparameter function estimation for general Hamiltonians
    gpt-5.25/16/2026

    Paper 1 offers a broadly applicable theoretical advance: a general, tight quantum limit and an achievable protocol for estimating arbitrary functions of multiple Hamiltonian parameters, including non-commuting generators. This is novel, methodologically substantive, and timely for quantum metrology, sensing, and Hamiltonian learning, with potential cross-field impact in quantum information and experimental design. Paper 2 is primarily a priority/attribution comment on an existing framework; while important for scholarly record, it contributes limited new methodology or applications, so its likely scientific impact is narrower.

    vs. Comment on "Quantum teleportation, entanglement, LQU and LQFI in $e^{+} e^{-} \rightarrow \mathrm{Y} \overline{\mathrm{Y}}$ processes at BESIII through noisy channels''
    gpt-5.25/16/2026

    Paper 2 has higher potential impact because it addresses a timely, cross-disciplinary issue (misapplication of quantum-information noise/teleportation concepts to high-energy collider hyperon pairs) and provides substantive physical-interpretation corrections that could influence how multiple communities (HEP, quantum information, open quantum systems) analyze and report such results. It targets methodological rigor and conceptual validity of widely used measures, potentially preventing systematic misconceptions. Paper 1 mainly documents overlap/attribution and textual similarity in a niche steering/LHS-model context, with limited forward scientific or application impact.

    vs. Limits of Stable Near-Field Probing in Nanophotonic Traps
    gemini-3.15/11/2026

    Paper 2 presents novel experimental findings on the fundamental limits of near-field probing in nanophotonic traps, providing actionable insights for fields utilizing optical trapping, such as quantum tech and biophysics. In contrast, Paper 1 is merely an academic comment highlighting inadequate attribution and methodological overlap in another paper. Therefore, Paper 2 has a vastly higher potential for real-world scientific impact and methodological advancement.

    vs. Quantum Kernels for Parity-Structured Classification: A Hybrid Pipeline
    gemini-3.15/8/2026

    Paper 2 presents novel scientific research demonstrating empirical evidence for quantum kernel advantage in high-complexity parity classification. It offers actionable insights for quantum machine learning and establishes concrete performance thresholds. In contrast, Paper 1 is merely an academic comment addressing inadequate attribution and methodological overlap in a previously published article. Consequently, Paper 2 has significantly higher potential for real-world application, methodological innovation, and broader scientific impact.

    vs. Quantum-Resistant Quantum Teleportation
    gemini-34/20/2026

    Paper 2 introduces a novel framework for securing quantum teleportation against quantum adversaries, providing rigorous analytical models and addressing real-world implementation constraints like quantum memory bottlenecks. In contrast, Paper 1 is merely a comment addressing an academic priority dispute and methodological overlap, offering no new scientific contribution. Thus, Paper 2 has significantly higher potential scientific and practical impact.

    vs. Parity-unfolded distillation architecture for noise-biased platforms
    gemini-34/20/2026

    Paper 2 presents a novel fault-tolerant quantum computing architecture with measurable improvements in resource overhead and error rates for key algorithms. In contrast, Paper 1 is a comment addressing academic priority and methodological overlap, which, while important for academic integrity, does not introduce new scientific capabilities or applications.

    vs. Approximate Cosine Similarity Estimation via an Angle-Encoding Hadamard Test
    claude-opus-4.64/20/2026

    Paper 1 presents a novel quantum computing method for approximate cosine similarity estimation with concrete analysis of bias, numerical experiments, and a practical application to Transformer models. Despite being incremental, it contributes a new algorithmic primitive relevant to near-term quantum computing and machine learning. Paper 2 is a comment documenting methodological overlap and insufficient attribution in another paper—essentially a priority/credit dispute with no new scientific content or methodology, limiting its impact to the specific community involved.