{"id":"611982ab-92b5-4168-803d-adec0eac9e4e","arxiv_id":"2611.17216","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Explicit AME(9,2q), AME(11,2q), AME(13,2q), AME(15,2q), AME(17,2q) families, a general odd-N bound, and the classification AME(9,d) exists iff d≥3.","lead":"Researchers construct explicit families of multipartite quantum states where every group of at most half the particles is maximally entangled, in local dimensions twice an odd prime power. The result also settles, for every integer dimension, whether a nine-party state of this kind exists.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek/deepseek-v4-flash","headline":"The nine-party classification is gated by a single external state, Kitamura's AME(9,6), whose matrix is absent from the companion repository and whose check the paper reports as not run rather than passed (App. D2); an exact verification of that input for d=6 would close the gap.","rationale":"Reading in good faith: the paper's core construction is the coupling of Proposition II.1 plus three fully specified ingredient families. The coupling proof is clean and parameter-free. The qubit mixtures for N = 13, 15, 17 are stabilizer projectors whose uniformity criterion (Appendix B1) is a standard weight argument, and the claimed weights, generator counts, and branch counts are backed by exact certificates in the companion repository; the Atlas and Kitamura mixtures for N = 9, 11 are independently checked in-repo through the average-sign test (Appendix B3). The Cauchy parents (Lemma C.1) are proven in-paper for every odd prime power q ≥ N, with a correct determinant argument for the extended all-ones column. The small-field parents are either explicitly constructed with stated check counts (e.g., the F_121 matrix with all 12,869 square submatrices, the F_49 codes, the extended RS codes) or cited to published sources (Bevins–Bidav, Hontarenko–Życzkowski) whose relevant equations are named. The rank bounds of Proposition II.7 are Farkas certificates, with the N = 9 and N = 13 identities displayed so they can be checked by hand. I found no internal inconsistency in the coupling or the family proofs. The single load-bearing concern is the classification's reliance on the external AME(9,6) state. A useful refinement to the reader's statement: the dependency is narrower than 'every pure power of 3' — Theorem II.3 already gives AME(9,2·3^m) for m ≥ 2, so the genuinely load-bearing instance is d = 6. That instance is exactly the one the paper's rank bound excludes from its own coupling, and it is the case the artifact does not verify. The existence of AME(9,6) has independent support (Kitamura's Lean proof, HZ amplitude description), so this is not an attack on the cited result; it is a reproducibility gap the paper itself flags, sitting at the entry gate of the headline classification. The error probability is low, but the verification is cheap and the claim is strong, so a CONDITIONAL verdict with a request to close the d = 6 check is proportionate. No verdict change from the reader is needed.","tokens_in":26680,"tokens_out":31667,"duration_ms":186705,"concrete_test":"Obtain Kitamura's AME(9,6) state from the cited Lean witness (ame-9-6-lean, commit 500ee827dffb, lean/AME96/Data.lean) or from the Hontarenko–Życzkowski amplitude description (arXiv:2610.17170v1, Secs. 7.2–7.4 and Apps. D–E), and run the exact check the repository skips: in exact arithmetic, verify that for each of the 126 four-party subsets A of {1,...,9} the reduction ρ_A = Tr_{A^c}|ψ⟩⟨ψ| equals I_6^4/6^4, and that grouping by four binary digits yields sixteen branches each equal to an AME(9,3) graph state as claimed in Appendix D2. If the check passes, the classification's only external dependency is closed and the artifact can be updated to report 'passed'.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Corollary II.4 claims AME(9,d) exists if and only if d ≥ 3, and the abstract presents the completed classification as a headline. The proof routes every d = 2s with 3|s through Kitamura's AME(9,6) state [29], an input the paper's own machinery cannot produce: Proposition II.7 proves every 4-uniform nine-qubit mixture has rank at least 4, so with the capacity bound K ≤ q this excludes q = 3 (Sec. IIE). Appendix D2 states that the companion repository does not include Kitamura's original matrix and that 'without it the runner reports the AME(9,6) check as not run rather than passed'; the same appendix earlier says 'We checked both in exact arithmetic,' a reporting tension. The Lean proof at commit 500ee827dffb was read but not rebuilt. A refinement matters: the external state is truly load-bearing only for d = 6, because d = 2·3^m with m ≥ 2 is already covered by Theorem II.3 with q = 3^m ≥ 9, and odd factors with a prime divisor p ≥ 5 are covered by AME(9,2p); the written proof, however, uses Kitamura for all 3|s cases. If AME(9,6) did not exist, the if-and-only-if claim would fail at d = 6. This is a verification gap in an otherwise self-contained artifact, not an internal inconsistency; it is the single most load-bearing premise of the central claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek/deepseek-v4-flash","summary":"The paper develops a coupling construction for absolutely maximally entangled states in local dimension 2q: it superposes K qubit branches with K projected branches from an AME(N+1,q) parent state, so that only the average of the qubit branches needs uniform reductions. The central Proposition II.1 is short and clean. Using known and newly constructed k-uniform qubit mixtures and Cauchy/finite-field parents, the paper proves explicit thresholds: AME(9,2q) and AME(11,2q) for all odd prime powers q≥5, AME(13,2q) for q≥17, AME(15,2q) for q≥9, and AME(17,2q) for q≥67, plus a general bound for every odd N. It also proves rank lower bounds for k-uniform qubit mixtures, an exponential lower bound for this coupling, and a fifteen-party family with compensated cyclic symmetry. In Section II.C the nine-party families are combined with prime-power AME(9,q) states, the product rule, and Kitamura's external AME(9,6) state to claim the classification AME(9,d) exists if and only if d≥3.","tokens_in":27077,"tokens_out":10469,"duration_ms":69832,"significance":"If the claims are correct, the paper gives valuable explicit infinite families in dimensions that the product construction cannot reach, and it completes the nine-party local-dimension classification. The coupling proof in Proposition II.1 is simple and convincing, and the explicit qubit groups, parent graphs, and linear-programming/Farkas certificates are a real strength: the manuscript ships exact arithmetic certificates and reconstruction code for the finite ingredients. The paper is also unusually transparent about which checks were run and which remain open. The main caveat is that the headline classification depends on Kitamura's external AME(9,6) state, and the paper's own appendix reports that this input was not re-verified in the companion repository. This affects the classification's status as a self-contained artifact, not the internal validity of Theorem II.3.","major_comments":[{"comment":"Corollary II.4 is stated as an unconditional if-and-only-if classification, but for d=2s with 3|s and no other odd prime divisor the proof depends entirely on Kitamura's AME(9,6) state [29]. Appendix D2 first says \"We checked both in exact arithmetic\" and then states that the companion repository does not include Kitamura's original matrix and that without it the runner reports the AME(9,6) check as \"not run rather than passed\". Because Proposition II.7 together with the capacity bound K≤q excludes q=3 inside this paper's own coupling, this external input is load-bearing for the headline classification. I am not asking that the cited Lean proof be reproduced; I am asking that the manuscript either include the matrix and a passed exact certificate, or state explicitly in Corollary II.4 and the abstract that the classification is conditional on the external Lean proof, and remove the conflicting \"checked both in exact arithmetic\" wording.","section":"Appendix D2 / Corollary II.4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The written proof applies Kitamura's state to every d=2s with 3|s, although d=2·3^m with m≥2 is already covered by Theorem II.3 with q=3^m≥9, and cases with a prime factor p≥5 are covered by AME(9,2p). The external state is therefore indispensable only at d=6. Please either reorganize the case split so that this is visible, or add a remark stating that [29] is used only for d=6. This would not change the mathematical conclusion but would make the verification target precise.","section":"Section II.C","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The notation q_N for the sufficient threshold is easy to confuse with the local dimension q; a symbol such as q_min(N) or q_thr(N) would be clearer.","section":"Theorem II.3","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The phrase \"the AME(9,6) check as not run rather than passed\" should be reconciled with the earlier sentence \"We checked both in exact arithmetic\"; as written, a reader cannot tell which of the two external states passed an exact check in the companion repository.","section":"Appendix D2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The final paragraph repeats the abstract's classification claim without the caveat about the external AME(9,6) input; after the Appendix D2 issue is resolved, the conclusion should mirror the precise dependency stated there.","section":"Conclusions","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"major_revision"},"author_rebuttal":null,"desk_editor":{"model":"deepseek/deepseek-v4-flash","letter":"Bottom line: this is a solid constructive paper and deserves a serious referee. The headline claim—AME(9,d) exists for every integer d≥3—looks correct, and the one real soft spot is narrow and honestly disclosed: the d=6 case rests on Kitamura's external AME(9,6) state, which the companion artifact does not re-verify.\n\nWhat's actually new: explicit AME(N,2q) families at N=13,15,17 with thresholds 17,9,67; the extension of the N=9,11 families beyond the known q=5 members; the if-and-only-if classification at nine parties; the cyclic N=15 family; the general odd-N bound; and the rank lower bounds of Prop. II.7. The coupling principle itself is explicitly credited to Atlas, Kitamura, and Shi et al., which is the right call. No overclaiming about novelty.\n\nWhere the paper earns its keep: Prop. II.1 is short and clean; the qubit mixtures are specified by explicit Pauli groups with a weight criterion; the finite-field parents come with exact checks (M \\bar M^T = −I, every square submatrix invertible, every balanced cut verified); the rank bounds are Farkas certificates with displayed nonnegative combinations. The companion repo ships the finite data and a checker. This is reproducible work in the sense that matters.\n\nThe soft spot is the one the stress test identifies. Appendix D2 first says \"We checked both in exact arithmetic\" and later says the runner reports the AME(9,6) check as not run because Kitamura's original matrix is not in the repository; the Lean proof was read but not rebuilt. That is a reporting tension and a genuine verification gap for the load-bearing external input. But the gap is narrower than the written proof makes it look: Theorem II.3 with q=9,27,... already covers d=18,54,..., and odd factors with a prime divisor p≥5 are covered by AME(9,2p). Only d=6 actually needs Kitamura. A one-line refinement of the proof shrinks the dependency to that single case, so the fix is cheap: include the matrix and run the exact check, or rebuild the Lean proof. Everything else feeding the classification is internally certified.\n\nMinor: the (15,7) and (17,22) rank-bound certificates live in the repo rather than on the page; the nine- and thirteen-party identities are written out, the other two are replayed by the checker. Not a problem, just a place where a referee has to run code or trust the repo.\n\nAudience: anyone tracking AME existence, multipartite entanglement, or quantum MDS codes. I would accept this for peer review and ask the authors to close the AME(9,6) gap and resolve the Appendix D2 wording before acceptance. The thresholds are explicitly sufficient rather than optimal, so there is no issue there.","headline":"Solid, reproducible construction paper that completes the AME(9,d) classification; the only real soft spot is the single external AME(9,6) input behind d=6, which the artifact does not re-verify and which a one-line proof change would isolate.","tokens_in":27630,"tokens_out":4680,"duration_ms":30270},"pith_extraction":{"msc":["81P45","81P70","94B25"],"pacs":["03.67.Mn","03.67.Pp","03.67.-a"],"model":"deepseek/deepseek-v4-flash","headline":"Explicit AME states exist for 9 to 17 parties in dimension twice a prime power","keywords":["absolutely maximally entangled states","AME states","multipartite entanglement","k-uniform states","quantum error correction","qubit mixtures","cauchy matrices","nine-party classification"],"falsifier":"Construct an explicit AME(9,6) state from the data promised in the cited source [29] and verify all four-party reductions are maximally mixed. If that check fails or the source contains no constructible state, then the nine-party classification is not fully established for dimensions of the form 2·3^k.","tokens_in":30306,"feed_emoji":"🔀","tokens_out":2577,"duration_ms":67520,"temperature":0.7,"pith_summary":"The paper constructs explicit infinite families of absolutely maximally entangled (AME) states for 9, 11, 13, 15, and 17 parties, in local dimension twice an odd prime power. The key trick is to couple a mixture of qubit states, which need only be uniform on average, with a set of $q$-level states that cancel interference exactly, obtained by projecting one party of a larger AME state. This bypasses the known impossibility of a pure qubit AME state for these party numbers. The families are explicit enough to specify each amplitude as a sum of at most 64 terms, and the paper completes the nine-party classification: an AME(9,d) state exists for every integer d≥3, and none exists at d=2. If the construction is correct, these states are available as concrete multipartite entangled resources for quantum error correction and secret sharing, with required local dimensions that grow at most exponentially in the number of parties.","feed_headline":"Explicit AME states for 9 to 17 parties in dimension twice a prime power","feed_subtitle":"A qubit-plus-parent coupling completes the nine-party dimension classification and yields infinite families.","key_machinery":"The coupling of Proposition II.1: a state |Ψ⟩ = ∑_a √λ_a |u_a⟩⊗|v_a⟩, where the q-level states |v_a⟩ satisfy Tr_{A^c}|v_a⟩⟨v_b| = δ_{ab} I_A/q^k for every set A of k parties, and the qubit mixture ρ_bin = ∑_a λ_a |u_a⟩⟨u_a| satisfies Tr_{A^c}ρ_bin = I_A/2^k. The q-level states are obtained by projecting one party of an AME(N+1,q) parent (Lemma A.1), which guarantees the required orthogonality and vanishing cross blocks. The qubit mixture is built from a commuting Pauli group whose nonidentity words all have weight greater than k, so its normalized projector is k-uniform (Appendix B). The coupling separates the problem: the q-level side can supply at most q branches (capacity bound, Appendix F1), and the qubit side must have rank at most q (Theorem II.6). The paper gives explicit qubit mixtures of ranks 4, 2, 16, 8, and 64 for N=9, 11, 13, 15, and 17, together with parents for each required q, including Cauchy graph states (Lemma C.1) and finite-field codes (Appendix C3).","core_discovery":"The paper proves that for every odd number of parties N=9,11,13,15,17 and every odd prime power q above a stated threshold (5, 5, 17, 9, 67 respectively), an AME(N,2q) state exists. The construction uses a superposition of K branches, each branch pairing an N-qubit state with an N-party q-level state. The q-level states are obtained by projecting the reference party of an AME(N+1,q) parent; their cross-block reductions vanish, so they cancel all interference between distinct branches. The qubit mixture therefore needs only to have maximally mixed reductions on every set of k=⌊N/2⌋ parties — a strictly weaker condition than requiring a pure AME qubit state. The paper supplies explicit qubit mixtures, built from known codes or commuting Pauli groups, and explicit parents, built from Cauchy matrices, Reed–Solomon codes, or finite-field constructions, for each N and each q in the stated range. Combined with product constructions and an external AME(9,6) state, the nine-party case yields the complete integer-dimension classification: AME(9,d) exists if and only if d≥3.","pith_inferences":["The nine-party classification inherits a load-bearing external premise: the paper relies on an AME(9,6) state whose 4-party uniformity check it reports as not run rather than passed. If that state were invalid, the classification would still cover all d not divisible by 3, but the product argument would miss dimensions d=2·3^k (e.g., 18, 54).","The rank bounds in Proposition II.7 are proven by necessary linear inequalities on Pauli correlators, but the paper notes that satisfying those inequalities does not guarantee a valid density matrix. The true minimal ranks for the qubit mixtures at N=13,15,17 could be larger than the bounds, which would raise the thresholds of the families.","The construction treats q as an odd prime power to get finite-field parents. For composite odd q that are not prime powers, the method would need a different way to supply the parent, so the stated infinite families do not immediately extend to all odd integers q.","Because the rank packing bound shows exponential growth of the required local dimension within this coupling, large-party states from this method quickly become inefficient; alternate constructions that let the qubit and q-level factors share the interference cancellation may reach smaller dimensions for the same N."],"forward_implications":["For quantum error correction and secret sharing, AME(N,2q) states provide explicit pure ⌊N/2⌋-uniform states that can encode a secret such that any set of at most half the parties learns nothing.","The nine-party local-dimension question is settled: AME(9,d) exists for every integer d≥3 and fails only at d=2, removing a long-standing gap in the existence table.","The coupling method is general: for every odd N≥3, an AME(N,2q) state exists whenever q≥max{N,2^{N-2}}, and for q≥N feasibility inside this coupling is characterized by the minimum rank of a k-uniform qubit mixture.","The cyclic family for N=15 and primes p≡1 mod 15 produces AME(15,2p) states with a compensated cyclic permutation symmetry, adding a structural constraint that may be useful for symmetric quantum protocols."],"supporting_citations":[{"why":"Supplies the nine-qubit mixture and the AME(9,10) state used for the N=9 family and classification.","marker":"[21]"},{"why":"Supplies the eleven-qubit mixture and the AME(11,10) state used for the N=11 family.","marker":"[23]"},{"why":"Supplies the external AME(9,6) state that fills the q=3 gap in the nine-party classification.","marker":"[29]"},{"why":"Supplies the AME(12,5) parent used for N=11 at q=5.","marker":"[24]"},{"why":"Supplies the AME(16,9) and AME(16,13) parents used for the N=15 family at q=9 and q=13.","marker":"[25]"},{"why":"Introduces the Cauchy-matrix AME construction that the paper extends to Lemma C.1 for large q.","marker":"[7]"},{"why":"Provides the weight criterion for k-uniformity of normalized projectors from commuting Pauli groups, used for the qubit mixtures.","marker":"[4]"},{"why":"Provides the code-to-AME construction used for prime-power nine-party inputs and several parents.","marker":"[5]"},{"why":"Supplies the quantum weight enumerator inequalities that the paper adapts to arbitrary-spectrum rank bounds.","marker":"[17]"}],"fun_headline_variants":["AME states for 9 to 17 parties in twice-a-prime-power dimensions","Qubit–q-level coupling yields explicit AME families for 9–17 parties","Twice a prime power: explicit AME states for up to 17 parties","Nine-party AME classification completed via qubit–parent coupling","Explicit AME(2q) states for 9, 11, 13, 15, 17 parties"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The nine-party classification depends on the existence of an AME(9,6) state that the paper does not independently verify; if that external state does not exist, the product construction would fail for dimensions whose odd factor is a pure power of three.","fun_headline_variants_meta":{"raw":{"variants":["AME states for 9 to 17 parties in twice-a-prime-power dimensions","Qubit–q-level coupling yields explicit AME families for 9–17 parties","Twice a prime power: explicit AME states for up to 17 parties","Nine-party AME classification completed via qubit–parent coupling","Explicit AME(2q) states for 9, 11, 13, 15, 17 parties"]},"model":"deepseek/deepseek-v4-flash","effort":"low","cost_usd":0.000105,"raw_usage":{"cost":0.0001051232,"is_byok":false,"cost_details":{"upstream_inference_cost":0.0001051232,"upstream_inference_prompt_cost":8.63772e-05,"upstream_inference_completions_cost":1.8746e-05},"total_tokens":1257,"prompt_tokens":1154,"completion_tokens":103,"prompt_tokens_details":{"audio_tokens":0,"video_tokens":0,"cached_tokens":256,"cache_write_tokens":0},"completion_tokens_details":{"audio_tokens":0,"image_tokens":0,"reasoning_tokens":0}},"tokens_in":1154,"tokens_out":103,"cost_source":"openrouter_usage","duration_ms":5093,"image_pages":[],"temperature":1.0,"billed_attempts":1,"reasoning_tokens":0,"transport_provider":"openrouter","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-11-13T09:16:29.031515+00:00","model_set":{"reader":"deepseek/deepseek-v4-flash"},"pipeline_version":"pith-pipeline@v1.4.1-daily-deepseek-or-medium","input_kind":null,"input_chars":null,"input_total_chars":null,"passes_missing":null,"published_standing":null,"published_confidence":null,"objection_audit":null,"run_record":null,"basis":{"kind":"not_recorded","source":null,"chars_given":null,"chars_total":null},"receipt":null,"falsifier":"Construct an explicit AME(9,6) state from the data promised in the cited source [29] and verify all four-party reductions are maximally mixed. If that check fails or the source contains no constructible state, then the nine-party classification is not fully established for dimensions of the form 2·3^k.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the nine-qubit mixture and the AME(9,10) state used for the N=9 family and classification."},{"cited_title":"Kitamura, A Lean proof of AME(11,10) (2026), ame-11-10-lean, commit 6b7d009a4397, README and lean/AMEProbe.lean","cited_arxiv_id":null,"evidence_quote":"Supplies the eleven-qubit mixture and the AME(11,10) state used for the N=11 family."},{"cited_title":"Kitamura, A Lean proof of an AME(9,6) state (2026), ame-9-6-lean, commit 500ee827dffb, witness in lean/AME96/Data.lean","cited_arxiv_id":null,"evidence_quote":"Supplies the external AME(9,6) state that fills the q=3 gap in the nine-party classification."},{"cited_title":"Symmetry-guided constructions of absolutely maximally entangled states in five open cases","cited_arxiv_id":"2608.05781","evidence_quote":"Supplies the AME(12,5) parent used for N=11 at q=5."},{"cited_title":"Absolutely Maximally Entangled States of $2q$ Parties in Every Odd Prime-Power Dimension $q$","cited_arxiv_id":"2609.11796","evidence_quote":"Supplies the AME(16,9) and AME(16,13) parents used for the N=15 family at q=9 and q=13."},{"cited_title":"Constructing optimal quantum error correcting codes from absolute maximally entangled states","cited_arxiv_id":"1701.03359","evidence_quote":"Introduces the Cauchy-matrix AME construction that the paper extends to Lemma C.1 for large q."},{"cited_title":"$k$-uniform mixed states","cited_arxiv_id":"1906.01311","evidence_quote":"Provides the weight criterion for k-uniformity of normalized projectors from commuting Pauli groups, used for the qubit mixtures."},{"cited_title":"Absolutely Maximally Entangled States: Existence and Applications","cited_arxiv_id":"1306.2536","evidence_quote":"Provides the code-to-AME construction used for prime-power nine-party inputs and several parents."},{"cited_title":"Quantum Weight Enumerators","cited_arxiv_id":"quant-ph/9612015","evidence_quote":"Supplies the quantum weight enumerator inequalities that the paper adapts to arbitrary-spectrum rank bounds."}],"lane_accuracy":{"status":"not yet measured","lane":{"model":"deepseek/deepseek-v4-flash","input_kind":null},"url":"https://pith.science/accuracy"},"review_version":1}