{"id":"73725a76-6736-448a-b7f5-d13cd6667009","arxiv_id":"2506.19026","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"In star-shaped quantum networks, local filters applied before measurement can reveal hidden non n-locality, and a non-separable central filter can do so even when separable filters cannot.","lead":"This paper describes a protocol in which parties in a star-shaped quantum network apply local filters before measuring, and shows that this can reveal hidden network nonlocality that the usual inequality misses. It also reports that a joint, non-separable filter at the central node can activate such nonlocality in cases where separable filters provably cannot.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The non-separable filter advantage is computed with Eq. 18, whose product-structure derivation does not apply after a joint central filter; the reported violations need a direct check.","rationale":"I read the central claim as twofold: (i) separable filters in the preparation phase yield hidden non n-locality characterized by Eq. 18; (ii) non-separable central filters reveal violations even with a product-state source, where Theorem 3 forbids separable filters. Claim (ii) is the advertised topology-specific advantage, and its numerical support in Table III is the least secure link. The reader's weakest assumption is exactly right: Eq. 18 is derived (Appendix A, Eq. 34) from the factorized filter (Eq. 16), and F_NS breaks that factorization, so the post-filter correlations are not governed by pairwise filtered singular values. The missing edge settings and the odd B_3-star = 1.31468 entry in Table III make the non-separable numbers unreproducible. I verified parts of the separable framework: recomputing B_3-star = 1.9082 for the Section IV.A.1 example from Eq. 13 and the Bloch singular values of the stated states agrees with the paper, which supports the reader's decision to treat the separable claims as defensible. I also see no fundamental impossibility in the non-separable claim — with the product-source party outputting deterministically, the trilocality expression reduces to a bilocality-type test of the remaining two entangled sources — so the issue is verification, not falsity. The concrete test settles it: direct numerical evaluation of Eq. 11 on the post-filter state. The verdict CONDITIONAL matches the reader's conditional acceptance; it should not be hardened to REJECT on current evidence.","tokens_in":16603,"tokens_out":32564,"duration_ms":292796,"concrete_test":"Compute the left-hand side of Eq. (11) directly for the first example of Table III: take |ψ1(β1=0.785)⟩|ψ2(β2=−0.144)⟩|++⟩, apply F_NS of Eq. (30) with (α1, α2) = (0.26, 0.173) to the central qubits, normalize by the success probability, measure P1 in the GHZ basis and A2, A3, A4 in two binary settings each, and maximize Σ_{i=1}^{4}|J_i|^{1/3} numerically over all edge settings (the paper states none). If the maximum is ≤ 2, the claimed non-separable advantage is refuted; if it exceeds 2, the advantage is real but the paper still owes a valid upper bound and the concrete settings. A secondary check: test whether Eq. 18 with any per-source reduced filtered states reproduces the tabled 2.00716, which would indicate Eq. 18 was misapplied.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The paper's headline claim — that non-separable multi-qubit filters at the central node reveal non n-locality where separable filters cannot (Section VII vs Theorem 3) — rests entirely on the values B^(seq)_3-star = 2.00716, 2.0408, 2.0056 in Table III. These are labeled with the symbol defined in Theorem 1, but Theorem 1's derivation (Appendix A, Eqs. 34–36) requires the post-filter state to factor as ⊗_i ρ^(f)_{1,i}, which holds only for the factorized separable filter of Eq. 16. In Section VII the central filter F_NS (Eqs. 30, 32, 33) is explicitly non-separable, so ρ^(f) (Eq. 27) is entangled across sources and Eq. 18 has no justification. The paper never states whether Table III came from Eq. 18 or from a direct evaluation of Eq. 11: if the former, the computation is invalid; if the latter, the edge measurement directions are missing, so the numbers cannot be checked. Supporting inconsistency: for the first row of Table III, Eq. 13 applied to the stated states (β1 = 0.785, β2 = −0.144, product source) gives a maximum B_3-star of 2, not the reported 1.31468, so the table entries are not tied to the paper's own formulas. Section VIII candidly notes only discrete filter classes were studied, but that admission does not supply the missing bound or settings.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies a sequential star-shaped n-local network in which parties apply local filtering operations before performing measurements, with the goal of characterizing 'hidden non n-locality'. The authors derive an upper bound, Eq. (18), on the n-local inequality for the case where the central party's filter is a tensor product of single-qubit filters, and prove two no-go results: Theorem 2, that hidden non n-locality cannot be detected if every source state remains Bell-CHSH local under all local filtering operations, and Theorem 3, that no violation is possible in the separable-filter protocol when at least one source distributes a two-qubit product state. They then provide numerical examples of hidden non-trilocality, including cases with one or two hidden nonlocal sources, and discuss locally accessible hidden non-trilocality and enhanced noise robustness. In Section VII, they claim that non-separable multi-qubit filters at the central node reveal hidden non-trilocality even when a product state is distributed, and report violations in Table III. The central claim of the paper is that joint non-separable central filtering is a new resource in star networks that has no analogue in the bipartite Bell scenario.","tokens_in":17015,"tokens_out":6667,"duration_ms":73839,"significance":"If the non-separable filtering advantage were established, this would be a conceptually interesting result: it would show that the star topology gives the central party a joint-access resource beyond separable SLOCC operations, leading to network nonlocality that cannot be reproduced in a bipartite Bell setup. The separable-filter upper bound Eq. (18) and the no-go theorems Theorems 2 and 3 are valuable and appear to follow from the stated arguments. The explicit numerical examples, if correct, would also demonstrate an improved noise robustness of the sequential protocol. However, the central novelty of the paper, namely the non-separable filter advantage in Section VII, is currently unsupported because the derivations and numerical tables suffer from a missing justification and internal inconsistencies. The paper's significance is therefore conditional on a successful direct check of the non-separable filter examples.","major_comments":[{"comment":"The values B^(seq)_3-star in Table III are labeled with the symbol defined in Theorem 1, but Theorem 1 and its proof in Appendix A rely essentially on the post-filter state factoring as a tensor product, Eq. (34). This factorization holds for the separable central filter of Eq. (16), but not for the non-separable filter F_NS used in Eq. (27). After applying F_NS, the central party's qubits are entangled across the different sources, so Eq. (18) has no justification in this case. The paper provides no replacement bound and no statement of how the numbers in Table III were obtained. The authors should either provide a direct evaluation of the n-local inequality Eq. (11) for explicit measurement directions, or prove a new theorem that extends Eq. (18) to non-separable central filters.","section":"Section VII / Eq. (18) / Appendix A"},{"comment":"The reported value B_3-star = 1.31468 for the first row is inconsistent with the paper's own formula Eq. (13). For the stated states, beta_1 = 0.785 and beta_2 = -0.144 are pure states of the form Eq. (28), whose correlation tensors have singular values (1, |sin 2 beta_i|, |sin 2 beta_i|), and S_3 distributes the product state Eq. (29), whose correlation tensor has only one nonzero singular value. Eq. (13) then gives B_3-star = 2, not 1.31468. This discrepancy indicates that the table entries are not reliably generated from the formulas in the paper, and it undermines confidence in the non-separable filter claim.","section":"Table III, first row"},{"comment":"Even if the non-separable filter numbers were intended to come from a direct computation of the n-local inequality rather than from Eq. (18), the manuscript does not specify the edge measurement directions or any other measurement settings used to obtain B^(seq)_3-star = 2.00716, 2.0408, and 2.0056. Without these settings, the reported violations cannot be checked or reproduced. This is a load-bearing omission because the non-separable advantage is the main new claim of the paper.","section":"Section VII.B / Table III"}],"minor_comments":[{"comment":"The text states w11 = 0.01 and success probability approximately 37%, while the Figure 3 caption states w11 = 0.05 and approximately 32% probability of success; these specifications should be reconciled.","section":"Section IV.A.1 and Figure 3"},{"comment":"The sentence 'none of P2, P2, P3 perform any operation' appears to contain a typo; it should list P2, P3, and P4.","section":"Section V.A"},{"comment":"The normalization is written ambiguously: the factor 1/Π_i C_i appears inside the tensor product in Eq. (34) and again in the definition of ρ^(f)_{1,i} in Eq. (35). This should be rewritten so that each normalized two-qubit state has unit trace.","section":"Appendix A, Eqs. (34)-(35)"},{"comment":"The notation for the sequential network with non-separable filters (referred to as N(seq)_n-star) is too easily confused with the separable-filter network N (seq)_n-star; distinct symbols would improve readability.","section":"Section VII"},{"comment":"The term '0.2089∥11⟩⟨00|' contains a typo: the first ket should be '|11⟩' rather than '∥11⟩'.","section":"Section VII.B, Eq. (32)"}],"recommendation":"major_revision","confidential_remarks":"The main novel claim, the non-separable central filter advantage, is currently not supported by a valid derivation or reproducible computation. The inconsistency in Table III's first row is particularly concerning because it suggests the numerical tables may have been produced by a method different from the paper's stated formulas. I would recommend that the authors provide a direct, fully specified evaluation of Eq. (11) for the non-separable filter examples, including measurement directions, and that they re-check all numerical entries. If the direct evaluation does not reproduce the claimed violations, the non-separable advantage claim would need to be retracted or substantially revised."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis paper is a mixed bag. The genuinely useful part is the separable-filter analysis for hidden non n-locality in star networks: Theorem 1 gives a closed form for the post-filtered n-local bound when each party applies single-qubit filters, and the no-go results (Theorems 2–4) are clean consequences of known facts. The examples with one or two hidden-nonlocal sources are internally coherent, and the locally accessible variants in Section V are a reasonable addition. Credit where due: transferring the linear-network framework to star topology, and the explicit noise-robustness comparison in Section VI, are worth having.\n\nBut the headline claim about non-separable multi-qubit filters at the central node does not hold up. The values in Table III are labeled as B^(seq)_3-star, the quantity defined in Theorem 1. That theorem's derivation (Appendix A) requires the post-filtered state to factor as a tensor product across sources (Eq. 34). After a joint non-separable filter like Eq. (30) or (32), the filtered state (Eq. 27) does not factor, so Eq. (18) is not a valid bound. The paper never provides a replacement derivation or the edge measurement directions used to obtain the numbers. The stress test is right that the first row of Table III is also inconsistent with the paper's own Eq. (13): for the stated β1, β2, and product source, B_3-star should be 2, not 1.31468. There are minor numeric mismatches elsewhere (32% vs 37% success, w11=0.01 vs 0.05 in text vs captions), and no code or data is provided. The citation pattern looks fine; prior work on linear networks and the Kundu et al. bound are properly acknowledged.\n\nThe non-separable section is the reason the paper is interesting, and that section is currently unsupported. The separable-filter results survive the critique and are a legitimate, if incremental, contribution. I would send this to a serious referee, with the recommendation that the non-separable claim either be re-derived with a direct evaluation of the n-local expression for the joint-filtered state (with stated settings) or removed. Researchers working on sequential network nonlocality will find the separable part useful, but they should not rely on Table III until it is fixed.","headline":"The separable-filter extension to star networks is legitimate and worth having, but the headline non-separable-filter advantage is computed with an inapplicable formula and inconsistent numbers, so that claim needs a direct derivation before it can be believed.","tokens_in":17532,"tokens_out":5948,"would_cite":false,"duration_ms":58708,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81P40","81P15"],"pacs":["03.65.Ud"],"model":"deepseek-v4-flash","headline":"Filtering first, then measuring, can reveal non n-locality in star-shaped quantum networks that ordinary measurements miss, and a non-separable filter at the central node works even when one source sends a product state.","keywords":["hidden non-locality","star network","n-locality","local filtering","non-separable filter","Bell-CHSH locality","trilocality","entanglement detection"],"falsifier":"Take the first numerical row of Table III, explicitly optimize the edge-party measurement directions, and evaluate the left-hand side of the n-local inequality directly on the post-filter state produced by the stated F_NS; if the maximal value does not exceed 2, then Eq. (18) was not validly applied after the non-separable filter and the claimed non-separable advantage would fail. The missing measurement directions used for the Table III values are exactly what such a check would need.","tokens_in":16357,"feed_emoji":"🕸️","tokens_out":4268,"duration_ms":44958,"temperature":0.7,"pith_summary":"The paper claims that inserting local filtering operations before the measurement phase can turn n-local correlations in a star-shaped quantum network into detectable non n-local correlations. It derives a closed-form upper bound for the filtered n-local inequality in terms of the singular values of each filtered two-qubit state, and it shows that Bell-CHSH nonlocality of every distributed state is not necessary: at least one source must supply a state that is Bell-CHSH nonlocal or hidden-Bell-nonlocal under filtering. The paper also claims a new network-specific effect: a non-separable multi-qubit filter at the central node can reveal hidden non-trilocality even when one source distributes a two-qubit product state, a case where its Theorem 3 rules out all separable-filter protocols. If correct, this gives a concrete resource-theoretic distinction between separable and joint filtering in star networks and extends hidden-nonlocality ideas beyond the bipartite Bell scenario.","feed_headline":"Joint filters expose hidden non-locality in star networks","feed_subtitle":"In star-shaped quantum networks, one non-separable filter at the hub can beat the product-state restriction that blocks every…","key_machinery":"The central object is the star-network n-local inequality of Eq. (11) together with its upper bound B_n-star in Eq. (13), expressed through the two largest singular values of each source's two-qubit correlation tensor. The protocol adds a preparation phase in which every party applies a local filter before measurement; because separable filters keep the post-filter state a tensor product across sources, Theorem 1 reduces the task to computing the largest singular values of each individually filtered two-qubit state. The non-separable case replaces the central party's separable filter by a joint filter F_NS acting on qubits from different sources, which couples the formerly independent subsystems and is claimed to generate violations that separable filtering cannot.","core_discovery":"The central claim is that the n-local inequality for a star network can be violated after suitable local filtering even when the unfiltered correlations satisfy the inequality, and that the maximal violation after separable single-qubit filters is governed by the closed expression B_n-star^(seq)=2 $\\sqrt$( product-of-largest-singular-values-squared plus product-of-second-singular-values-squared ) for the filtered two-qubit states, given in Eq. (18). The paper proves that if every source distributes a two-qubit Bell-CHSH local state that remains local under any filtering, no violation is possible (Theorem 2), and that if at least one source distributes a two-qubit product state, no separable-filter protocol can violate the inequality (Theorem 3). It then presents numerical examples in which a non-separable filter at the central node produces violations in exactly the product-state setting where Theorem 3 forbids separable filters, along with instances where filtering by only the central party or only the edge parties suffices to reveal hidden non-trilocality.","pith_inferences":["The non-separable advantage, if it survives direct evaluation, suggests that joint filtering at a network hub is a topology-specific resource with no analogue in the bipartite Bell scenario, and it invites a systematic search over non-separable filters for other network geometries such as linear and bilocal networks.","A natural testable extension is to derive a genuine upper bound for the n-local expression after a joint non-separable filter, since the tensor-product factorization used to prove Eq. (18) fails there; without such a bound, the Table III violations are not yet anchored by a theorem.","The paper's examples also indicate a possible route to device-independent entanglement detection in networks whose sources are contaminated by product-state noise, which would be practically relevant for diagnosing untrusted quantum-network nodes."],"forward_implications":["Violation of the n-local inequality in the sequential protocol certifies that at least one source is entangled, while saying nothing about how many sources are entangled.","Filtering can enhance noise robustness: for amplitude-damped states there is a range of noise parameters in which the sequential network violates the trilocal inequality even though the usual network does not.","Hidden non-trilocality can be generated with only one source supplying a hidden-Bell-nonlocal state while the other two sources distribute Bell-CHSH local states.","Locally accessible hidden non-trilocality exists when only the central party filters, when only edge parties filter, or when a proper subset of parties filters.","If the non-separable examples are correct, a joint central filter is a genuine resource in star networks because it can circumvent the product-state obstruction stated in Theorem 3."],"supporting_citations":[{"why":"Defines n-locality through source independence and the factorized n-local correlation decomposition, which the paper's inequality and definitions rest on.","marker":"[17]"},{"why":"Supplies the star-network n-local inequality, the GHZ measurement scheme, and the classically post-processed output bits used in Eq. (11).","marker":"[22]"},{"why":"Provides the closed upper bound B_n-star in Eq. (13) for the usual star network, which is the baseline the sequential protocol must beat.","marker":"[30]"},{"why":"Gives the necessary and sufficient hidden-Bell-CHSH-nonlocality criterion, used by the paper to classify which two-qubit states can feed hidden non n-locality.","marker":"[12]"},{"why":"Introduced hidden non n-locality in linear networks, the concept this paper extends to star topology.","marker":"[39]"},{"why":"Establishes that Bell-CHSH local states remain local under local filtering, a key input to the no-go Theorem 2.","marker":"[49]"},{"why":"Provides the family of two-qubit hidden-nonlocal states used in the trilocal numerical examples.","marker":"[50]"},{"why":"Underpins Theorem 4 by providing the result on network nonlocality without entanglement of all sources.","marker":"[52]"}],"fun_headline_variants":["Nonseparable filters reveal hidden non-locality in star networks","Star-network non-locality hidden until joint filtering at hub","Filtering exposes non-locality that n-local inequality misses","Nonseparable filter at central node beats separable in star nets","Hidden non-locality in star networks emerges from nonseparable filters"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the filtered-network bound of Eq. (18), proved for separable filters where the post-filter state factors across sources, remains a valid upper bound after a joint non-separable filter has coupled qubits from different sources; the paper gives no proof for that case.","fun_headline_variants_meta":{"raw":{"variants":["Nonseparable filters reveal hidden non-locality in star networks","Star-network non-locality hidden until joint filtering at hub","Filtering exposes non-locality that n-local inequality misses","Nonseparable filter at central node beats separable in star nets","Hidden non-locality in star networks emerges from nonseparable filters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00104,"raw_usage":{"total_tokens":4380,"prompt_tokens":957,"completion_tokens":3423,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":573,"completion_tokens_details":{"reasoning_tokens":3334}},"tokens_in":573,"tokens_out":3423,"duration_ms":25681,"temperature":1.0,"reasoning_tokens":3334,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:41:03.217221+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the first numerical row of Table III, explicitly optimize the edge-party measurement directions, and evaluate the left-hand side of the n-local inequality directly on the post-filter state produced by the stated F_NS; if the maximal value does not exceed 2, then Eq. (18) was not validly applied after the non-separable filter and the claimed non-separable advantage would fail. The missing measurement directions used for the Table III values are exactly what such a check would need.","supporting_citations":[{"cited_title":"Locally inaccessible hidden quantum correlations","cited_arxiv_id":null,"evidence_quote":"Defines n-locality through source independence and the factorized n-local correlation decomposition, which the paper's inequality and definitions rest on."},{"cited_title":"Branciard, N","cited_arxiv_id":null,"evidence_quote":"Supplies the star-network n-local inequality, the GHZ measurement scheme, and the classically post-processed output bits used in Eq. (11)."},{"cited_title":"Andreoli, et al., New.J.Phys","cited_arxiv_id":null,"evidence_quote":"Provides the closed upper bound B_n-star in Eq. (13) for the usual star network, which is the baseline the sequential protocol must beat."},{"cited_title":"Gisin, Phys","cited_arxiv_id":null,"evidence_quote":"Gives the necessary and sufficient hidden-Bell-CHSH-nonlocality criterion, used by the paper to classify which two-qubit states can feed hidden non n-locality."},{"cited_title":"Wolfe, A","cited_arxiv_id":null,"evidence_quote":"Introduced hidden non n-locality in linear networks, the concept this paper extends to star topology."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes that Bell-CHSH local states remain local under local filtering, a key input to the no-go Theorem 2."},{"cited_title":"Quantum information science","cited_arxiv_id":null,"evidence_quote":"Underpins Theorem 4 by providing the result on network nonlocality without entanglement of all sources."}],"review_version":2}