{"id":"467c055f-fc87-43e3-8d91-5f603751d49e","arxiv_id":"2411.13114","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"A scan of two phase angles in quantum PageRank reveals consistent cluster phases in rankings, fidelity, variance, coherence, entanglement, and power-law exponent on a 32-node scale-free graph.","lead":"This paper numerically explores how two phase angles in a quantum version of Google's PageRank algorithm change how nodes in a small network are ranked. It finds that the rankings group into distinct 'cluster phases' that also line up with quantum coherence and entanglement patterns, and it proposes a variant with four phase angles.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed cluster phases and cross-quantity correlations rest entirely on an unspecified 'KNN' clustering step and visual inspection of colormaps; no quantitative cluster-agreement or stability analysis is provided.","rationale":"The paper's intended contribution is an empirical map of how APR phase angles (theta1, theta2) affect quantum PageRank, leading to the claimed discovery of correlated cluster phases in ranking distributions, fidelity, variance, coherence, entanglement, and the power-law exponent beta. For that central claim to hold, the clusters must be genuine features of the PageRank dynamics rather than products of the clustering algorithm and its hyperparameters. The weakest point is exactly the clustering analysis: the method is misidentified as KNN, no distance metric or number of neighbors/clusters is stated, and the consistency across different observables is asserted from visual comparison of color maps. No quantitative measure of cluster agreement is provided, and no stability analysis with respect to k, graph instance, or time-averaging window is reported. This is not a matter of disagreeing with a consensus view; it is an internal reproducibility and validation gap. The paper does build on a known algorithm (Ref. [31]) and extends three special APR cases to the full parameter plane, and the proposed alternate PageRank is a reasonable variant to explore; those elements are useful. However, the main discovery is currently an unreproducible observation, not a demonstrated result. The reader's verdict of REJECT is therefore appropriate. I would not change the verdict, though the paper could become conditionally acceptable if the authors supplied code, data, exact clustering parameters, and a quantitative cluster-correlation and robustness analysis. Secondary issues such as the \"same 10 nodes ... these 9 nodes\" inconsistency in Section III reinforce the need for a careful revision but are not the primary load-bearing concern.","tokens_in":11937,"tokens_out":5679,"duration_ms":61274,"concrete_test":"Request the code and data, then rerun the analysis with a fully specified pipeline: k-means clustering (k=7, Euclidean distance on the 32-dimensional PageRank distributions, 100 random restarts) over the same theta grid; then compute the Adjusted Rand Index between the PageRank cluster labels and labels obtained by discretizing each of the variance, fidelity, coherence, entanglement, and beta maps into 7 quantile-based clusters. Repeat for k=3,...,12 and for at least five independently generated 32-node scale-free graphs. The cluster-phase claim is supported only if the ARI is substantially above chance (e.g., >0.7) and stable across k and graph instances; otherwise the reported phases are artifacts of the arbitrary clustering and single graph.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central empirical claim—that APR phases organize quantum PageRank into a small number of correlated clusters across ranking distributions, fidelity, variance, coherence, entanglement, and beta—requires that the cluster labels be intrinsic and stable. The paper does not establish this. In the section \"Cluster Phases in Quantum PageRank with APR,\" the authors say only that \"using the most traditional data clustering techniques, i.e. KNN\" they obtain seven clusters (Fig. 1a). KNN is k-nearest neighbors, a supervised classification rule, not an unsupervised clustering algorithm; if k-means or another method was actually used, the algorithm, distance metric, k value, initialization, and convergence criterion are never stated. The number 7 appears only as a visual color count. The subsequent claim that the PageRank-distribution clusters \"are consistent with\" the variance, fidelity, coherence, entanglement, and beta maps (Figs. 1c, 3a-d) is supported only by eye: no Adjusted Rand Index, mutual information, or boundary-overlap statistic is computed. Because every quantity in Figs. 3 is plotted on the same (theta1, theta2) grid, smooth gradients can create the appearance of matching regions without true cluster alignment. The time-averaged PageRank in Eq. (10) is also taken over the final Delta t=500 steps at T=5000 with no convergence diagnostic, so the maps could depend on the arbitrary cutoff. The paper's own conclusion that correlation signs change for 16- and 20-node graphs further undercuts the generic \"versatility\" claim. The cluster-phase discovery is therefore currently an unreproducible, unvalidated observation.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies quantum PageRank with arbitrary phase rotations (APR) on a 32-node scale-free graph. By scanning the two rotation angles (θ1, θ2), it reports a \"cluster phase\" structure in the PageRank distributions, obtained through an unsupervised procedure that the text calls \"KNN,\" and claims that the resulting clusters coincide with maps of the fidelity to classical PageRank, the variance of the ranking distribution, the coherence and entanglement of the PageRank state, and the power-law exponent β. It then introduces an alternate quantum PageRank operator with four phase parameters, studies three two-parameter cases, and analyzes the trackback graph of the original scale-free graph. The central claims are qualitative: that distinct APR phases organize quantum PageRank into a small number of correlated regimes, and that the associated quantum and ranking quantities are mutually correlated in a way that depends on the scheme and graph.","tokens_in":12157,"tokens_out":5043,"duration_ms":53921,"significance":"If the cluster-phase picture were established, it would provide a useful coarse-grained organization of the APR parameter space and could inform the design of quantum PageRank algorithms. The definitional part of the paper, especially Eqs. (1)-(9), is standard, and the alternate operator in Eq. (18) is clearly stated and easy to implement. However, the central claims rest on one graph, one clustering run, and visual comparisons of colormaps; no quantitative cluster-agreement or stability analysis is provided. The paper's own statement that the 16- and 20-node graphs show different correlation signs further limits the claimed universality. In its current form, the paper is better viewed as a preliminary numerical observation than as a demonstrated result.","major_comments":[{"comment":"The clustering procedure is not specified. \"KNN\" is k-nearest neighbors, a supervised classification rule, not an unsupervised clustering algorithm; the reference cited for it (Ref. [32]) is also about classification. The text never states the actual clustering algorithm, the distance metric, the number of clusters (the value seven is inferred only from the color count in Fig. 1(a)), the initialization, or any stability criterion. Since the cluster labels are the primary object on which every subsequent consistency claim depends, this omission is load-bearing and must be corrected with a complete, reproducible description.","section":"Cluster Phases in Quantum PageRank with APR, Fig. 1(a)"},{"comment":"The claimed consistency between the PageRank-distribution clusters and the maps of fidelity, variance, coherence, entanglement, and β is supported only by visual inspection. No quantitative agreement measure (e.g., adjusted Rand index, normalized mutual information, or boundary-overlap statistic) is computed. Because all the plotted quantities are smooth functions on the same (θ1, θ2) grid, visual resemblance of regions can arise from smooth gradients rather than from genuine cluster boundaries. The central claim that the cluster phases are correlated across these quantities needs a numerical test of cluster alignment.","section":"Cluster Phases in Quantum PageRank with APR, Figs. 1(a,c) and 3(a-d)"},{"comment":"The time-averaged PageRank is computed over the final Δt = 500 steps out of T = 5000, with no convergence diagnostic. The instantaneous PageRank is known to fluctuate, and the text asserts that the oscillations are \"stable\" without showing evidence. The results of Figs. 1, 3, 5, 8, 10, and 13 could depend on the arbitrary choices of T and Δt. A convergence study, or at least a demonstration that the maps are insensitive to the averaging window, is necessary for the quantitative claims to be reliable.","section":"Eq. (10) and all subsequent maps"},{"comment":"The power-law exponent β is obtained by \"linear data fitting\" of log I_i versus log i, but the fitting range, error bars, and goodness-of-fit are not reported. The logarithmic plots shown in Fig. 4 are visibly not linear over the full range of node indices, and for the trackback graph the paper itself states that no obvious linear behavior is present. Under these conditions, the extracted β values are not uniquely defined, and all correlation claims involving β are ambiguous. The authors should specify the fitting procedure and demonstrate its validity for each graph.","section":"Eqs. (16)-(17), Figs. 3(d) and 4"},{"comment":"The manuscript states that on 16- and 20-node scale-free graphs the signs of the correlations \"vary as the graph is different.\" This directly qualifies the universality of the cluster-phase correlations claimed in the main text, yet these results are placed only in a supplementary file and are not accompanied by any details or figures in the main text. The main text should either present these results and explain how they limit or refine the main claims, or the claims of universality should be narrowed accordingly.","section":"Conclusions, supplementary 16- and 20-node results"}],"minor_comments":[{"comment":"The text first lists ten nodes (1,2,3,5,7,9,10,14,19,26) as the \"main hub\" set and then refers to \"these 9 nodes\"; the count does not match.","section":"Page 4, Fig. 1(b) discussion"},{"comment":"There are several typographical issues: \"Rf. [31]\" should be \"Ref. [31]\"; \"makers\" should be \"markers\"; \"eigen-spetrum\" should be \"eigenspectrum\"; and \"superpower data mining\" in the Alternate Quantum PageRank section is informal.","section":"General typography"},{"comment":"The captions for Figs. 13(a-d) and 14(a-b) refer to quantities \"in alternate fixing PageRank,\" but these figures concern the trackback graph; the model label should be corrected.","section":"Figs. 13 and 14 captions"},{"comment":"The scale-free graph in Fig. 2(a) is not described in terms of the generation model or parameters. For reproducibility, the authors should state how the 32-node graph was generated (e.g., preferential attachment parameters) and whether the results are robust to different instances.","section":"Section 2, graph generation"},{"comment":"The entanglement is defined through the spectrum of the reduced density matrix, but the text does not specify whether the entropy is the von Neumann entropy of the normalized reduced state or the linear entropy; Eq. (13) as written (with eigenvalues λ_i) suggests the former, which should be stated explicitly.","section":"Eq. (13), entanglement measure"}],"recommendation":"major_revision","confidential_remarks":"The paper is a numerical study that could become publishable after substantial revision, but the current version lacks the methodological detail needed to verify its central claims. The most important fix is to replace the unspecified 'KNN' clustering and visual consistency judgments with a complete clustering specification and quantitative cluster-agreement statistics. The varying correlation signs on the 16- and 20-node graphs should also be addressed in the main text rather than hidden in a supplementary file. I see no indication of deliberate misrepresentation, but the manuscript as written is not sufficiently reproducible for a journal publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper maps the full (θ1, θ2) plane of the APR quantum PageRank from Ref. [31] and introduces an alternate four-phase operator with three two-parameter subcases. That is the real advance: prior work studied three special cases; here you get the whole plane and a new evolution operator, plus a trackback-graph study. The definitions are standard and the numerical exploration is systematic. If the cluster picture holds, it gives practitioners a way to tune phase angles and emphasize different node sets, which is genuinely useful for quantum PageRank on complex networks.\n\nThe soft spots are in the empirical core, not the math. The central claim—that PageRank distributions, fidelity, variance, coherence, entanglement, and β all show consistent clusters—rests on an under-specified clustering step. The text calls it \"KNN,\" but KNN is a supervised classifier, not a clustering method. If the authors meant k-means, they never say k, distance metric, initialization, or convergence criterion. The number seven appears only as a color count. Then the \"consistency\" across quantities is assessed by eye. No Adjusted Rand Index, no mutual information, no boundary-overlap statistic. On a shared grid, smooth gradients can easily create the appearance of matching regions. This is the load-bearing weakness.\n\nThe rest of the concerns are proportionate but real: a single 32-node graph as the main demonstration; time averaging over Δt=500 at T=5000 with no convergence diagnostic; no error bars; an internal inconsistency where \"same 10 nodes\" becomes \"these 9 nodes\"; and no code or data, just \"available upon request.\" The authors do test 16- and 20-node graphs, but they put those in a supplementary file and admit that the correlation signs change with the graph. That directly undercuts the word \"versatility\" in the title: if the correlations flip sign, the cluster phases are not a stable feature of the algorithm.\n\nAlso worth saying: the correlations between variance, coherence, entanglement, and β are not surprising discoveries. All are computed from the same PageRank states, and they are mathematically related through the density matrix. The paper even gives an interpretive list of those relations. So the conceptual novelty is modest.\n\nWho should read this? People working on quantum PageRank and quantum walk search algorithms will want to know that a full phase scan exists and that an alternate operator has been proposed. But the current presentation is not reproducible enough to cite as evidence for cluster phases. My recommendation: send it to peer review with the expectation of major revision. A serious referee should demand the code, exact clustering parameters, and a quantitative measure of cluster agreement. Without those, this is an interesting observation, not a demonstrated result.","headline":"A systematic phase-plane scan of a known quantum PageRank variant, with a plausible but unproven cluster claim that needs code, clustering details, and quantitative validation before it can be trusted.","tokens_in":12795,"tokens_out":1700,"would_cite":false,"duration_ms":19438,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P68","68Q12"],"pacs":["03.67.Lx","03.67.Mn"],"model":"deepseek-v4-flash","headline":"This paper claims that scanning the two rotation phases of a Szegedy-walk quantum PageRank on a scale-free graph splits rankings into distinct, correlated cluster phases.","keywords":["quantum PageRank","Szegedy quantum walks","arbitrary phase rotations","cluster phases","scale-free networks","quantum coherence","entanglement","power-law ranking"],"falsifier":"Compute the same phase sweep on several other scale-free graphs with different sizes and degree exponents, and also vary the KNN number of clusters k and the distance metric; if the cluster boundaries and the sign of the fidelity-coherence correlation do not survive these changes, the claimed generic cluster phases would be refuted.","tokens_in":11671,"feed_emoji":"🌐","tokens_out":6468,"duration_ms":59966,"temperature":0.7,"pith_summary":"Quantum PageRank, a proposed ranking algorithm for a future quantum internet, lets the underlying Szegedy walk carry two adjustable phase rotations. This paper sweeps both phases over their full circle and asks what the algorithm does for each choice. It finds that the resulting rankings do not change smoothly: they fall into a small number of cluster phases, and the same cluster boundaries show up in the fidelity of the quantum ranking to the classical PageRank, in the variance of the ranking distribution, in the coherence and entanglement of the quantum state, and in the exponent of the ranking's power-law tail. The paper also proposes an alternate four-parameter PageRank and shows that, in three two-parameter specializations, it produces differently shaped cluster families that emphasize the same top hubs with different relative weights. If the cluster phases are generic, the practical point is that tuning the phase angles selects a whole profile of ranking behavior, giving quantum PageRank a versatility classical PageRank lacks.","feed_headline":"Two phase knobs sort quantum PageRank into distinct ranking regimes","feed_subtitle":"Sweeping the rotation angles of a Szegedy-walk PageRank reveals seven correlated clusters spanning rankings, coherence, and entanglement.","key_machinery":"The machinery is the two-phase evolution operator $W(\\theta_1, \\theta_2) = U(\\theta_2)U(\\theta_1)$ of the Szegedy quantum walk, with $U(\\theta) = S([1 - e^{i\\theta}]\\Pi - I)$ acting on the duplicated directed graph. PageRank probabilities come from projecting the evolved state onto one copy, time-averaged over a window to smooth instantaneous fluctuations. The analysis then measures, for every phase pair, the ranking vector and five derived quantities: fidelity to classical PageRank, distribution variance, $l_1$-norm coherence, entanglement entropy of the reduced state, and the power-law exponent $\\beta$ fitted to the ranked node probabilities. KNN clustering of the ranking vectors is what turns the continuous phase plane into labelled cluster regions.","core_discovery":"The central discovery is that the two rotation angles $\\theta_1$ and $\\theta_2$ organize quantum PageRank on a scale-free graph into distinct cluster phases. Within the parameter plane, a time-averaged quantum PageRank distribution takes only a few qualitatively different forms; the paper identifies seven clusters by KNN clustering of the distributions. The same clustered geography appears, with matching boundaries, in the fidelity between the quantum and classical PageRank vectors, in the variance of the ranking distribution, in the $l_1$-norm coherence and entanglement of the reduced PageRank state, and in the fitted power-law exponent $\\beta$. Across the standard model the quantum PageRank fidelity is positively correlated with ranking variance and negatively correlated with coherence, entanglement, and $\\beta$; in the alternate phase models the correlation signs can change, yielding different shapes of clusters. On the trackback graph (the reversed edges of the same graph) the cluster structure survives but the set of emphasized nodes changes, and the paper offers this as a view on network traffic tracking.","pith_inferences":["The sharp boundaries between clusters in the $(\\theta_1, \\theta_2)$ plane may correspond to spectral transitions of the walk operator; checking whether the boundaries align with avoided crossings or gap closings of $W(\\theta_1, \\theta_2)$ would give an analytic explanation of the phases.","A practical tuner could use the phase map as a control surface: pick a phase pair to hit a desired trade-off between classical fidelity and quantum resource content, which the paper does not explicitly propose.","The general four-parameter alternate operator was explored only in three two-parameter slices; a full four-dimensional scan is a direct extension that might reveal regimes inaccessible to these slices.","The count of seven clusters is tied to the KNN setting, so a stability check across cluster counts, $k$ values, and distance metrics would separate genuine phase structure from clustering artifacts."],"forward_implications":["Choosing a pair of phase angles selects one of a few discrete ranking regimes, so phase tuning can emphasize different top hubs without changing the graph or the algorithm.","Because variance, fidelity, coherence, entanglement, and $\\beta$ share the same cluster boundaries, any one of these observables can serve as a proxy for the regime of a given phase pair.","The alternate PageRank operator extends the accessible behavior to four phase parameters, and the three two-parameter cases already produce cluster patterns with different correlation signs, enabling multi-perspective network interpretation.","On a trackback graph the cluster structure persists but highlights a different node set, pointing toward applications in identifying important nodes for network traffic tracking and defense.","The consistent correlation pattern offers a compact classification: higher quantum resource content (coherence, entanglement) comes with lower fidelity to classical ranking and lower variance, effectively quantifying how quantum a chosen ranking regime is."],"supporting_citations":[{"why":"Defines the classical PageRank whose fixed-point vector is the baseline for quantum fidelity comparisons.","marker":"[12]"},{"why":"Introduces the first Szegedy-walk-based quantum PageRank scheme that the APR model generalizes.","marker":"[13]"},{"why":"Supplies the Szegedy quantum walk construction that underlies the unitary evolution operator used throughout.","marker":"[16]"},{"why":"Introduces the generalized quantum PageRank with arbitrary phase rotations and defines the equal, opposite, and alternate phase schemes the paper scans and extends.","marker":"[31]"},{"why":"Provides the KNN clustering algorithm used to identify cluster phases in the PageRank distributions.","marker":"[32]"},{"why":"Defines the l1-norm coherence measure used to quantify the coherence of PageRank states.","marker":"[33]"}],"fun_headline_variants":["Two rotation angles carve quantum PageRank into seven clusters","Quantum PageRank phases switch with two angle knobs","Cluster diversity in quantum PageRank from rotation phases","Seven ranking regimes emerge from quantum PageRank phases","Rotation angles steer quantum PageRank between ranking clusters"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The cluster phases and their correlation signs are generic properties of APR quantum PageRank, not features peculiar to the single 32-node scale-free graph and the chosen time-averaging and KNN settings.","fun_headline_variants_meta":{"raw":{"variants":["Two rotation angles carve quantum PageRank into seven clusters","Quantum PageRank phases switch with two angle knobs","Cluster diversity in quantum PageRank from rotation phases","Seven ranking regimes emerge from quantum PageRank phases","Rotation angles steer quantum PageRank between ranking clusters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000203,"raw_usage":{"total_tokens":1400,"prompt_tokens":977,"completion_tokens":423,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":593,"completion_tokens_details":{"reasoning_tokens":350}},"tokens_in":593,"tokens_out":423,"duration_ms":4806,"temperature":1.0,"reasoning_tokens":350,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:48:45.402001+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the same phase sweep on several other scale-free graphs with different sizes and degree exponents, and also vary the KNN number of clusters k and the distance metric; if the cluster boundaries and the sign of the fidelity-coherence correlation do not survive these changes, the claimed generic cluster phases would be refuted.","supporting_citations":[{"cited_title":"The pagerank citation ranking: Bringing or- der to the web","cited_arxiv_id":null,"evidence_quote":"Defines the classical PageRank whose fixed-point vector is the baseline for quantum fidelity comparisons."},{"cited_title":"Google in a quantum network","cited_arxiv_id":null,"evidence_quote":"Introduces the first Szegedy-walk-based quantum PageRank scheme that the APR model generalizes."},{"cited_title":"Ortega and Miguel A","cited_arxiv_id":null,"evidence_quote":"Introduces the generalized quantum PageRank with arbitrary phase rotations and defines the equal, opposite, and alternate phase schemes the paper scans and extends."},{"cited_title":"A brief review of nearest neighbor algo- rithm for learning and classification","cited_arxiv_id":null,"evidence_quote":"Provides the KNN clustering algorithm used to identify cluster phases in the PageRank distributions."},{"cited_title":"Quantifying coherence","cited_arxiv_id":null,"evidence_quote":"Defines the l1-norm coherence measure used to quantify the coherence of PageRank states."}],"review_version":1}