{"id":"307bfa8e-4f36-48f3-bee0-6cf4fb423fb0","arxiv_id":"2412.20753","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":1,"one_line_summary":"Modifying one arc weight per vertex gives pretty good state transfer between antipodal vertices on every hypercube.","lead":"This math paper constructs simple weighted quantum coins that make pretty good state transfer work between opposite corners of any hypercube, a case previously settled only for prime dimensions. It is a concise result in algebraic graph theory with implications for quantum walk algorithms and state transfer.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central Theorem 4.7 is sound modulo the external Lemma 4.6; the decisive unproved input is the distinct-squarefree-part claim for odd primes.","rationale":"The reader's weakest assumption correctly identifies Lemma 4.6 as the load-bearing external input. I checked the internal hypercube proof in detail: Lemma 4.3 gives integer eigenvalues, Lemma 4.4 supplies strong cospectrality and the parity-preserving or parity-flipping symmetry between λ and −λ, and the applications of Lemma 4.5 and Kronecker's theorem are consistent. The only point where the proof could genuinely collapse is the number-theoretic distinctness of squarefree parts of r(p−r) for r = 1, …, (p−1)/2, which is imported verbatim from [5]. Since Lemma 4.6 is published but not re-proved here, and since the same author is involved in [5], a conditional acceptance is appropriate rather than outright rejection. The paper's Theorem 5.1 is also under-derived, and the reader's rationale already flags it; I agree that it is a real weakness, but it is not the single load-bearing assumption for the hypercube theorem itself. Therefore the reader's verdict should remain unchanged.","tokens_in":8517,"tokens_out":27331,"duration_ms":257833,"concrete_test":"Independently verify Lemma 4.6: (1) computationally check all odd primes p up to 10^6, computing the squarefree parts of j(p−j) for 1 ≤ j ≤ (p−1)/2 and confirming pairwise distinctness; (2) obtain or produce a complete proof, for example by showing that sf(j(p−j)) = sf(k(p−k)) forces j = k via the factorization structure of j(p−j) relative to the prime p. A counterexample would refute Theorem 4.7; a full proof would discharge the only external load-bearing step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 4.7 rests on the claim that the angles {π} ∪ {arccos(λ/p) : λ ∈ Λ_a, 0 < λ < p} are linearly independent over Q. This is obtained by applying Lemma 4.5 to the eigenvalues λ = p − 2r, using Lemma 4.6 to guarantee that the squarefree parts of r(p−r), hence of p² − λ² = 4r(p−r), are pairwise distinct for r = 1, …, (p−1)/2. Lemma 4.6 is cited from [5] and is not proved in this paper. If it failed for some odd prime p, the parity argument in both the even-d and odd-d cases would not force ℓ_λ = ℓ_{−λ} (or the required evenness), and condition (ii) of Theorem 3.6 would not follow. The remaining hypercube-specific steps—strong cospectrality, the λ ↔ −λ pairing via Lemma 4.4, and the Kronecker approximation step—are internal and check out. Thus the central claim is exactly as secure as Lemma 4.6, a nontrivial number-theoretic fact whose proof is external to this manuscript. The under-derived Theorem 5.1 does not affect Theorem 4.7 itself, but it is a separate concern for the paper's generalization claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that every hypercube Q_d admits pretty good state transfer (PGST) between antipodal vertices in a discrete-time coined quantum walk, using a specially chosen real weight matrix W_m. For prime d the known case m=2 (Grover coin) is invoked; for composite d, m is the smallest positive integer such that p = 2d - 2 + m is prime. The proof uses the spectral relation between the walk and the Hermitian adjacency matrix H_m, derives the eigenvalues explicitly (Lemma 4.3), proves strong cospectrality and the ± pairing of symmetric eigenvalues (Lemma 4.4), and reduces the parity condition of Theorem 3.6 to a rational-independence statement about the angles arccos((p-2r)/p), established via two number-theoretic lemmas from [5]. The paper also states a general sufficient condition (Theorem 5.1) for PGST on other graphs.","tokens_in":8828,"tokens_out":21739,"duration_ms":183448,"significance":"Theorem 4.7 is a notable extension of the prime-dimensional result of Chan and Zhan and gives an explicit, real, one-arc-per-vertex coin construction for every hypercube. The internal algebra is carefully presented: the eigenvalue formula, the strong cospectrality argument, and the even/odd parity analysis all check out. The proof is self-contained modulo Lemma 4.6 from [5], an external number-theoretic fact about distinct squarefree parts of j(p-j); this dependency is legitimate but should be kept in view. The advertised generalization in Theorem 5.1 is not proved to the same standard and requires additional hypotheses or a complete proof; this is the main gap in the manuscript as it stands.","major_comments":[{"comment":"The proof asserts the Q-linear independence of the angles {π} ∪ {arccos(λ/p) : λ ∈ Λ_a, 0 < λ < p} without deriving it from the hypotheses; in Theorem 4.7 this independence is obtained from Lemmas 4.5 and 4.6, which require an odd prime p and the distinct-squarefree-part property, but Theorem 5.1 states neither of these conditions and does not cite the lemmas. Consequently, the parity condition (ii) of Theorem 3.6 is not established, and the theorem as stated is unsupported.","section":"Section 5, Theorem 5.1"},{"comment":"The case −p ∈ Λ^-_ab is dismissed by saying that X is bipartite and that 'a similar argument to Theorem 4.7' applies, but the required linear-independence argument is not supplied for this case and the assertion of bipartiteness is not proved.","section":"Section 5, proof of Theorem 5.1"},{"comment":"The definition W = H^{∘1/2} does not make H the Hermitian adjacency matrix of the resulting quantum walk unless H has constant row sum; in general the associated Hermitian adjacency matrix is D^{1/2} H D^{1/2} with D = diag(row sums of H), so the spectral data of H cannot be used directly in Theorem 3.6. The theorem should either impose regularity or specify the correct normalization.","section":"Section 5, Theorem 5.1"}],"minor_comments":[{"comment":"The interval should be (q/2, q), not (d/2, q).","section":"Lemma 4.5(i)"},{"comment":"Delete the duplicated phrase 'for any for any set set'.","section":"Proof of Theorem 3.6"},{"comment":"The symbol Λ_ab is used in the displayed congruences but is not defined; it should be Λ_a throughout.","section":"Proof of Theorem 4.7"},{"comment":"The claim about Q_4 (periodicity and maximum transfer probability 9/16) should be accompanied by a reference to [5] or a brief computation.","section":"Introduction"},{"comment":"The step concluding μ_λ = ±γ from the simultaneously small errors is correct but terse; an extra sentence explaining that γ/μ_λ must be real for all λ would improve readability.","section":"Corollary 3.4"},{"comment":"Reference [17] contains a corrupted author name ('Pawe/suppress l Kurzy´ nski') and should be fixed.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper's main theorem (4.7) is sound and significant. The primary weakness is Theorem 5.1, which is advertised in the abstract but is under-proved. The author is also a co-author of the heavily used reference [5]; this is a normal scientific dependency, but the editor may wish to ensure that the external Lemma 4.6 is indeed established in [5]."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline here is that the paper delivers a real, modest result: pretty good state transfer between antipodal vertices on every hypercube Q_d, using weighted Grover coins that only change one arc weight per vertex. The composite-dimension case was genuinely open, and the construction is simple enough to be believable and useful.\n\nWhat is new is the explicit choice of m making p = 2d-2+m prime, and the parity argument that handles even and odd d separately. I checked the chain: Lemma 4.3 gives the eigenvalues of H_m, Lemma 4.4 gives strong cospectrality and the lambda <-> -lambda pairing, and the Kronecker approximation step in Theorem 3.6 is applied correctly. The algebra is clean; the typos I spotted ('for any for any set set', 'd/2' in Lemma 4.5 for 'q/2') are trivial.\n\nThe real soft spot is the dependence on Lemma 4.6, cited from the author's own [5], which asserts that the squarefree parts of j(p-j) are pairwise distinct for j up to (p-1)/2. That lemma is the load-bearing input for the linear independence of the angles, and if it failed for some odd prime, the parity argument collapses. It is a published result, presumably proved in [5], but since it is decisive and external to this manuscript, any referee should verify that proof or ask for it to be included. This is not a fatal flaw, but it is the right thing to check.\n\nThe second soft spot is Theorem 5.1, the purported sufficient condition for other graphs. The proof is a sketch. It jumps from linear independence to conditions (a)-(c) without deriving them, and the 'otherwise' case for -p is left vague. As written, that theorem is more of a research note than a proof. Since Theorem 4.7 does not depend on it, this does not damage the main result, but it should be either fully proved or explicitly flagged as a conjecture/direction.\n\nOverall: this is a solid contribution to the quantum walk state transfer literature. It does not reorganize the field, but it closes an open case in a natural way, and the main theorem deserves referee time. I would recommend accepting it after a moderate revision: flesh out or weaken Theorem 5.1, and either prove Lemma 4.6 in an appendix or provide a precise pointer to where it is proved. The central result is secure.","headline":"A clean, modest extension of the prime-case result to every hypercube via a one-weight coin change; the main proof checks out, with the main caveat being a heavy but legitimate reliance on a cited number-theoretic lemma.","tokens_in":9303,"tokens_out":3163,"would_cite":true,"duration_ms":30099,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05C50","05C25"],"pacs":[],"model":"deepseek-v4-flash","headline":"A slight modification of the Grover coin, changing one arc weight per vertex, gives pretty good state transfer between antipodal vertices on every hypercube $Q_d$.","keywords":["pretty good state transfer","discrete-time quantum walk","hypercube","Grover coin","weighted coin","Cayley graph","Kronecker approximation","spectral graph theory"],"falsifier":"For the first composite dimension, $d=4$, the construction sets $m=1$ and $p=7$; a direct check of Theorem 3.6(ii) would settle the parity step: list the plus/minus sets for the eigenvalues $7, 5, 3, 1, -1, -3, -5, -7$ of $H_1$ and test whether every integer relation among the angles $\\arccos(\\lambda/7)$ has even total weight on the minus set. A simpler arithmetic falsifier also exists: compute, for all odd primes $p$ up to any bound, the squarefree parts of $j(p-j)$ for $1 \\le j \\le (p-1)/2$; the first repeated squarefree part would disprove the external lemma on which the angle independence rests.","tokens_in":8344,"feed_emoji":"⚛️","tokens_out":12690,"duration_ms":119074,"temperature":0.7,"pith_summary":"Pretty good state transfer means a quantum walk can be brought arbitrarily close to a target state by choosing the right time. The paper proves that every hypercube $Q_d$, for every dimension $d \\ge 2$, admits such transfer between antipodal vertices, using a deliberately simple coin: a weighted Grover coin that changes the weight of only one arc per vertex. For prime $d$, the usual unweighted Grover coin already works; the new content is composite $d$, where the standard walk fails (on $Q_4$ it is periodic with maximum transfer probability $9/16$). The construction chooses the smallest positive $m$ making $2d-2+m$ a prime, and the proof reduces the transfer condition to a parity statement about integer relations among $\\arccos$ values, which is forced by a number-theoretic lemma about squarefree parts. The same recipe yields a general sufficient condition for other graphs.","feed_headline":"Pretty good state transfer now works on every hypercube","feed_subtitle":"Changing one arc weight per vertex suffices even where the standard Grover coin fails.","key_machinery":"The load-bearing objects are the walk's coin and its associated Hermitian adjacency matrix. For a weighted graph with arc weights coming from a matrix $W$, the coin is the reflection $2N_t^*N_t - I$, where $N_t$ is the weighted arc-tail incidence matrix; the associated Hermitian adjacency matrix is $H = (I \\circ WW^*)^{1/2}(W \\circ W^*)(I \\circ WW^*)^{1/2}$, which for real positive weights has entries equal to the squared arc weights. On the hypercube the paper uses $H_m = m A_0 + 2\\sum_{j \\ge 1} A_j$, diagonalized by the characters $\\psi_g(x) = (-1)^{\\langle g,x\\rangle}$ of $\\mathbb{Z}_2^d$, giving eigenvalues $\\lambda_g = 2d - 4\\,\\mathrm{wt}(g) + (-1)^{g_0}(m-2) = p - 2r$. The spectral characterization (Theorem 3.6) reduces pretty good state transfer to strong cospectrality plus a parity condition: every integer relation among the angles $\\arccos \\lambda$ must have even total weight on the 'minus' part of the eigenvalue support. That parity condition is enforced by Kronecker's approximation theorem together with the number-theoretic linear independence of the angles, which comes from the distinct squarefree parts of $j(p-j)$.","core_discovery":"The paper's central theorem is that pretty good state transfer occurs between antipodal vertices of $Q_d$ for every $d \\ge 2$, relative to a real weighted Grover coin. Concretely, set $m=2$ when $d$ is prime; when $d$ is composite, let $m$ be the smallest positive integer for which $p = 2d-2+m$ is prime. In the arc-reversal walk on $Q_d$ whose coin is the reflection built from the real weighted adjacency matrix $W_m$ — weight $\\sqrt{m}$ on arcs in one coordinate direction and $\\sqrt{2}$ on arcs in every other direction — the antipodal vertices enjoy pretty good state transfer. The eigenvalues of the associated Hermitian adjacency matrix $H_m$ are exactly $p - 2r$ for $r = 0, \\dots, p$, and the proof verifies the two conditions of the paper's spectral characterization of pretty good state transfer: strong cospectrality of the antipodal vertices, and an even-parity condition on integer relations among the angles $\\arccos((p-2r)/p)$. The parity condition is established through linear independence of those angles, which follows from a cited lemma on pairwise distinct squarefree parts of $j(p-j)$ for odd prime $p$. The paper also proves a general sufficient condition: any connected graph with a real nonnegative adjacency matrix $H$ whose spectral radius is a prime $p$, with strongly cospectral vertices whose eigenvalue support lies in $\\{p-2r : 0 \\le r \\le p\\}$ and whose plus/minus sets behave symmetrically under $\\lambda \\mapsto -\\lambda$, admits antipodal pretty good state transfer in the walk driven by the entrywise square root $W = H^{\\circ 1/2}$.","pith_inferences":["A natural extrapolation from Theorem 5.1 is that many Cayley or distance-regular graphs whose spectra lie in an arithmetic progression of the form $p - 2r$ might admit pretty good state transfer with the same one-parameter family of real coins; the main work would be verifying strong cospectrality, which the hypercube gets for free from its characters.","The construction suggests an explicit experimental recipe: for a given dimension $d$, take the smallest prime $p \\ge 2d-1$ and weight one coordinate direction by $\\sqrt{p - 2d + 2}$, leaving the other directions at $\\sqrt{2}$; this is a finite search that could be tested numerically on small hypercubes.","The proof splits by the parity of $d$: in even dimensions the minus set is symmetric under $\\lambda \\mapsto -\\lambda$, while in odd dimensions it is swapped. A testable extension would be to check whether this same even/odd dichotomy controls pretty good state transfer on other bipartite graphs."],"forward_implications":["Every hypercube, including $Q_4$ where the unmodified Grover walk is periodic with maximum antipodal transfer probability $9/16$, acquires a built-in coin that achieves pretty good state transfer.","The coins are real and modify the weight of only one arc per vertex, so the departure from the standard Grover coin is a minimal local change rather than a reweighting of the whole graph.","Theorem 5.1 gives a general recipe: a graph with prime spectral radius, eigenvalue support inside $\\{p-2r\\}$, strong cospectrality, and either of two $\\lambda \\mapsto -\\lambda$ symmetries admits pretty good state transfer with real entrywise-square-root coins.","The proof ties transport properties of quantum walks to arithmetic: on hypercubes, the required parity condition holds because the squarefree parts of $j(p-j)$ are distinct for prime $p$."],"supporting_citations":[{"why":"Supplies the entire quantum-walk framework, the spectral characterization of pretty good state transfer used as Theorem 3.6, the prime-hypercube result, and the two number-theoretic lemmas (7.5 and 7.6) that give angle independence.","marker":"[5]"},{"why":"Provides the characters of $\\mathbb{Z}_2^d$ that jointly diagonalize the weighted adjacency matrices of Cayley graphs, yielding the explicit eigenvalue formula for $H_m$.","marker":"[8]"},{"why":"Supplies Kronecker's approximation theorem, which converts the existence of arbitrarily good hitting times into a parity condition on integer linear relations among the walk's angles.","marker":"[9]"}],"fun_headline_variants":["Weighted Grover coins fix state transfer on every hypercube","One-arc tweak enables state transfer on all hypercubes","Pretty good transfer on hypercubes, finally for every dimension","All hypercubes achieve pretty good state transfer","Single arc weight change makes every hypercube transferable"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The construction relies on a cited number-theoretic fact it does not prove: for every odd prime $p$, the numbers $j(p-j)$ for $j = 1, \\dots, (p-1)/2$ must all have different squarefree parts (the factor left after removing square factors); if that fact failed, the angle-independence argument that enforces the transfer condition would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Weighted Grover coins fix state transfer on every hypercube","One-arc tweak enables state transfer on all hypercubes","Pretty good transfer on hypercubes, finally for every dimension","All hypercubes achieve pretty good state transfer","Single arc weight change makes every hypercube transferable"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000359,"raw_usage":{"total_tokens":1965,"prompt_tokens":988,"completion_tokens":977,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":604,"completion_tokens_details":{"reasoning_tokens":895}},"tokens_in":604,"tokens_out":977,"duration_ms":6545,"temperature":1.0,"reasoning_tokens":895,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:13:25.265723+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For the first composite dimension, $d=4$, the construction sets $m=1$ and $p=7$; a direct check of Theorem 3.6(ii) would settle the parity step: list the plus/minus sets for the eigenvalues $7, 5, 3, 1, -1, -3, -5, -7$ of $H_1$ and test whether every integer relation among the angles $\\arccos(\\lambda/7)$ has even total weight on the minus set. A simpler arithmetic falsifier also exists: compute, for all odd primes $p$ up to any bound, the squarefree parts of $j(p-j)$ for $1 \\le j \\le (p-1)/2$; the first repeated squarefree part would disprove the external lemma on which the angle independence rests.","supporting_citations":[{"cited_title":"16, 165305, Publisher: IOP Publishing","cited_arxiv_id":null,"evidence_quote":"Supplies the entire quantum-walk framework, the spectral characterization of pretty good state transfer used as Theorem 3.6, the prime-hypercube result, and the two number-theoretic lemmas (7.5 and 7.6) that give angle independence."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the characters of $\\mathbb{Z}_2^d$ that jointly diagonalize the weighted adjacency matrices of Cayley graphs, yielding the explicit eigenvalue formula for $H_m$."},{"cited_title":"Gonek and Hugh L","cited_arxiv_id":null,"evidence_quote":"Supplies Kronecker's approximation theorem, which converts the existence of arbitrarily good hitting times into a parity condition on integer linear relations among the walk's angles."}],"review_version":1}