{"id":"527538dd-dda3-433c-bf8e-b3a83b8d0e4a","arxiv_id":"1908.03842","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Bisynchronous nonlocal correlations with n inputs and n outputs are exactly traces on the quantum permutation group, and their associated completely positive maps are factorizable via quantum permutations.","lead":"This paper introduces a new class of nonlocal games called bisynchronous games, where equal questions force equal answers and different questions force different answers. It proves that when input and output sets have the same size, bisynchronous correlations are exactly traces on the quantum permutation group, and the associated completely positive maps are factorizable.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: core trace/factorizability theorems are sound modulo cited synchronous characterizations.","rationale":"The reader's CONDITIONAL verdict is appropriate. The mathematical core is internally consistent: Theorem 2.2 derives quantum-permutation structure from the synchronous trace representation, and Theorem 4.1 correctly reduces factorizability to that representation. The external dependency on [19], [10], [14,15], and [2, Theorem 4.9] is real but is the normal state of mathematical papers; without a specific failure in those cited results, it is a dependency risk rather than a correctness risk. The concrete defect is the abstract's unqualified claim that every bisynchronous density gives a factorizable map, contradicted by Remark 3.3 and by the square-map scope of factorizability. Since the body states the correct hypotheses, the paper should be accepted after the abstract is corrected. This does not change the reader's verdict.","tokens_in":17283,"tokens_out":34044,"duration_ms":385158,"concrete_test":"Specialize Theorem 2.2 to n=k=2, where O(S_2^+) is commutative and generated by a single projection e via u=(e,1-e;1-e,e). Enumerate all bisynchronous densities p(a,b|x,y)=tau(u_{x,a}u_{y,b}) over tracial states tau (equivalently, over e in an abelian C*-algebra) and compare this family with the convex hull of the two deterministic permutation correlations. Agreement verifies the theorem in the minimal nontrivial case and sharpens the abstract's scope; disagreement would falsify the qc/local characterization.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The core proof chain of Theorem 2.2 is internally sound: from a bisynchronous density in C_qc one gets a trace on C*(F(n,n)) by [19], GNS makes the column-sum argument valid, and the resulting matrix is a magic permutation. The same reduction carries through for qa/q/loc via [10] and for vect via [14,15]. The weakest point is therefore external: these cited characterizations are imported without re-proof, and Theorem 4.3 additionally imports [2, Theorem 4.9]. If any of those results has a hidden hypothesis (for example, if the [19] trace theorem applies only to tensor-product synchronous correlations rather than to the full qc class), the corresponding bisynchronous statement would need a separate proof. This is a dependency risk, not an internal inconsistency. Separately, the abstract's sentence 'Each bisynchronous density ... factorizable' is false as written: Remark 3.3 supplies a bisynchronous nonsignalling density whose Phi_p is not positive, and factorizable maps in this paper are only defined for square maps. Theorem 4.1 states the correct hypotheses (t in {loc,q,qa,qc}, n=k), so the body is safe.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a new class of synchronous non-local games and correlations, called bisynchronous, which require equal outputs for equal inputs and different outputs for different inputs. For the case of equal numbers of inputs and outputs, the authors prove a characterization: a bisynchronous correlation in C_bs_t(n,n) for t in {qc,qa,q,loc,vect} is of the form p(a,b|x,y) = τ(u_{x,a}u_{y,b}) for a suitable trace and quantum permutation. They then associate to each square bisynchronous density a completely positive map Φ_p and show, for t in {loc,q,qa,qc}, that Φ_p is t-factorizable via a quantum permutation; conversely, every such factorizable map gives a bisynchronous correlation. The paper also connects these maps to the graph isomorphism game, shows that local bisynchronous maps are exactly the mixed permutation maps, and analyzes the fixed-point algebra of Φ_p in terms of the commutant of the quantum permutation.","tokens_in":17433,"tokens_out":24713,"duration_ms":235271,"significance":"The paper's main contribution is an explicit bridge between bisynchronous correlations and traces on the quantum permutation group, extending the existing theory of synchronous correlations. The factorizability result for the associated completely positive maps is a strong structural property with potential applications in quantum information and operator algebras. The central proofs in Theorems 2.2 and 4.1 are coherent: the GNS faithfulness argument, the column-sum computation, and the factorization formula are all checkable and appear correct. The paper also honestly discusses limitations, for example in Remark 4.4. The main reservations are that the abstract overstates the hypotheses of the main theorem and that one step in the graph-isomorphism characterization (Theorem 4.3) is too terse and needs a missing argument.","major_comments":[{"comment":"The statement 'Each bisynchronous density gives rise to a completely positive map and we prove that these maps are factorizable maps' is false as written. Remark 3.3 provides a bisynchronous nonsignalling density on 3 inputs and 3 outputs for which Φ_p is not even positive, and factorizable maps are only defined in the paper for square maps (n=k). Theorem 4.1 correctly restricts to t in {loc,q,qa,qc} and n=k. Please revise the abstract and the introductory sentence to state the precise hypotheses under which the factorizability result holds.","section":"Abstract and Section 1, first paragraph"},{"comment":"After deriving the vanishing conditions from Φ(A_G)=A_H and Φ*(A_H)=A_G, the proof immediately concludes 'Hence 1 follows.' It does not explicitly show that these conditions imply the intertwining relation (A_G⊗1)u = u(A_H⊗1) from part (2), nor does it explain how the full graph-isomorphism-game conditions, including inputs from V(H) and cross-graph pairs, are recovered. The missing step is non-trivial: one must pass from τ(e_{x,a}e_{y,b})=0 on edge/non-edge mismatches to the operator equality using the row and column sums of the quantum permutation. Please expand this part of the proof.","section":"Theorem 4.3, proof of (3 ⇒ 1)"}],"minor_comments":[{"comment":"The symbol F(n,n) for the free product of n copies of the cyclic group of order n is used but never defined. Please add a definition or a reference.","section":"Proof of Theorem 2.2"},{"comment":"In the display computing Φ(1), the expression '∑_{x∼y} ∑_{a≠b}' appears to be a typo: the first sum should be over x, not over x∼y.","section":"Proof of Theorem 4.3"},{"comment":"The notation 'p ∈ C_b_qc(n,n)' should be 'p ∈ C_bs_qc(n,n)' for consistency with the rest of the paper.","section":"Corollary 4.7"},{"comment":"The notation is confusing because 'e_{x,y}' is used both for the matrix entries of the quantum permutation and for matrix units. Please disambiguate these uses.","section":"Proof of Theorem 4.5"},{"comment":"There are several typographical errors, including 'Anatharaman-Delaroche' for 'Anantharaman-Delaroche' and 'Neuamnn' in the proof of Theorem 4.1. A careful proofreading pass is recommended.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Chris,\n\nYou asked about the Paulsen–Rahaman bisynchronous games paper. Here is my take.\n\nThe genuinely new piece is the bisynchronous condition. Requiring same inputs to force same outputs and different inputs to force different outputs is a natural restriction on synchronous correlations, and it pays off: when n=k, the game algebra becomes a quantum permutation group, and Theorem 2.2 gives trace representations for qc, qa, q, loc, and vect bisynchronous correlations. The qc proof is clean: invoke the [19] synchronous trace theorem, use the bisynchronous condition to annihilate off-diagonal column products, and the GNS representation gives a magic unitary. The vectorial case has a nice column-sum argument. Theorem 4.1 showing that these correlations are t-factorizable via quantum permutations, and conversely, is the other main contribution. The flip game, the orbital algebra section, and the open problems are thoughtful extras.\n\nNow the soft spots. First, the abstract overclaims: it states that every bisynchronous density gives a completely positive map and that these maps are factorizable. That is not true as written. Remark 3.3 produces a bisynchronous nonsignalling density whose Phi_p is not positive, and factorizable maps in this paper are defined only for square maps. The body is careful—Theorem 4.1 restricts to t in {loc,q,qa,qc} and n=k—so it is a writing bug, but a serious one for citation.\n\nSecond, the proof architecture leans heavily on prior synchronous characterizations: [19] for qc, [10] for qa/q/loc, [14,15] for vect, and [2, Theorem 4.9] for the graph isomorphism equivalence. Many of those results come from the same community and are not reproved. That is a dependency risk, not circularity. A referee should verify that the cited theorems have no hidden hypotheses. I do not see evidence of a flaw, but the paper is not self-contained.\n\nThird, the converse direction in Theorem 4.1 depends on identifying the map from its matrix units and comparing with the trace formula; that works. Theorem 4.3 imports the quantum t-permutation characterization of graph isomorphism, so it inherits that theorem's standing.\n\nOverall, this is a solid and useful paper. The core math holds up under standard cited results. Fix the abstract, spell out the hypotheses in the body, and it is a good contribution. I would send it to a serious referee. If I worked on synchronous games or factorizable maps, I would cite it.\n\nBest.","headline":"Bisynchronous correlations give a clean trace representation on the quantum permutation group; the core theorems hold, but the abstract's blanket CP/factorizability claim is false as stated.","tokens_in":17961,"tokens_out":3185,"would_cite":true,"duration_ms":32454,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L10","46L30","05C60","81P45"],"pacs":[],"model":"deepseek-v4-flash","headline":"Bisynchronous correlations are trace-form densities on the quantum permutation group.","keywords":["bisynchronous games","bisynchronous correlations","synchronous games","quantum permutation group","magic unitary","factorizable maps","graph isomorphism game","tracial states"],"falsifier":"Find a bisynchronous density in $C^{bs}_{qc}(3,3)$ whose entries cannot be written as $\\tau(u_{x,a}u_{y,b})$ for any tracial state on $O(S_3^+)$; since the set of such trace evaluations is convex and semidefinite-representable, a concrete counterexample would be a finite matrix violating the feasibility problem. Alternatively, exhibit a local bisynchronous density whose associated map is not a mixed permutation map; Theorem 4.2 predicts no such density exists.","tokens_in":19,"feed_emoji":"🎲","tokens_out":8322,"duration_ms":141838,"temperature":0.7,"pith_summary":"Bisynchronous games are non-local games in which matching questions force matching answers and distinct questions force distinct answers. The paper proves that when the number of questions equals the number of answers, every bisynchronous correlation $p(a,b|x,y)$ is exactly the trace of a product of two entries of a quantum permutation: $p(a,b|x,y)=\\tau(u_{x,a}u_{y,b})$ for a tracial state on the quantum permutation group $O(S_n^+)$. It further shows that the completely positive map naturally attached to such a density is factorizable, meaning it can be implemented by a quantum-permutation unitary and an ancilla trace. A sympathetic reader should care because this identifies a clean algebraic object—traces on $O(S_n^+)$—as the carrier of an entire family of correlation densities, and because the same mechanism connects graph isomorphism games to factorizable channels.","feed_headline":"Bisynchronous correlations reduce to traces on a quantum group","feed_subtitle":"When questions and answers match in number, every bisynchronous density is a trace, and its channel is factorizable.","key_machinery":"The load-bearing object is the bisynchronous density, equivalently the relations $p(a,b|x,x)=0$ for $a\\neq b$ and $p(a,a|x,y)=0$ for $x\\neq y$ imposed on a conditional probability density, and the corresponding game-algebra relations $e_{x,a}e_{y,a}=0$ for $x\\neq y$. When $|I|=|O|=n$, these relations combine with the synchronous relations to make the projections $\\{e_{x,a}\\}$ satisfy $\\sum_a e_{x,a}=1$ and $\\sum_x e_{x,a}=1$, i.e. to form a magic permutation, also called a quantum permutation: a square matrix of projections whose rows and columns each sum to the identity. The quantum permutation group $O(S_n^+)$ is the universal $C^*$-algebra generated by such entries, and the paper's central identity is $p(a,b|x,y)=\\tau(u_{x,a}u_{y,b})$, which turns a trace on $O(S_n^+)$ into a bisynchronous correlation and, via $\\Phi_p$, into a factorizable map with ancilla $N$: $\\Phi_p(X)=\\mathrm{id}\\otimes\\tau_N(u^*(X\\otimes 1_N)u)$.","core_discovery":"The paper's central discovery is that bisynchronicity, in the square case $n=k$, forces the projections of a game's $*$-algebra to assemble into a magic permutation, and every trace on the quantum permutation group produced this way yields a bisynchronous density. Theorem 2.2 states the precise equivalence: $p\\in C^{bs}_{qc}(n,n)$ iff $p(a,b|x,y)=\\tau(u_{x,a}u_{y,b})$ for a tracial state on $O(S_n^+)$, with parallel characterizations for qa, q, loc, and vectorial correlations using representations of $O(S_n^+)$ into an ultrapower of the hyperfinite II$_1$-factor, a finite-dimensional algebra, an abelian algebra, and a Hilbert space. Theorem 4.1 extends this to maps: for $t\\in\\{\\mathrm{loc},\\mathrm{q},\\mathrm{qa},\\mathrm{qc}\\}$, a bisynchronous density in $C^{bs}_t(n,n)$ is exactly one whose associated completely positive map $\\Phi_p(E_{x,y})=\\sum_{a,b}p(a,b|x,y)E_{a,b}$ is $t$-factorizable via a quantum permutation. The proof mechanism is the trace formula itself, which converts multiplicativity of the density into an ancilla implementation of the channel.","pith_inferences":["One testable extension is the rectangular case $|I|\\neq|O|$: the paper's counting argument gives only the necessary condition $n\\le k$ for a perfect $C^*$-strategy, and the trace characterization likely needs a non-square quantum-permutation analogue; constructing such an analogue would generalize Theorem 2.2.","The paper's Slofstra-type example suggests a route to factorizable maps that genuinely require infinite-dimensional ancilla: use a graph-isomorphism density that is qa but not q, and check whether the factorizable map it defines admits any finite-dimensional factorization, not just one with quantum-permutation unitary.","Because every synchronous game can be transformed into a bisynchronous game by making players return the question, any structural result for bisynchronous games automatically transfers back to synchronous games whenever the construction preserves the relevant parameters; the obstacle in the paper is that this transform doubles the output set, so the square equality $n=k$ is lost."],"forward_implications":["When $n=k$, bisynchronous games with perfect quantum-commuting strategies are exactly those whose densities are trace evaluations on the quantum permutation group, so the theory of such games is a chapter of the theory of $O(S_n^+)$.","For $t\\in\\{\\mathrm{loc},\\mathrm{q},\\mathrm{qa},\\mathrm{qc}\\}$, the classes of bisynchronous densities and of $t$-factorizable maps via quantum permutations coincide; in particular, local bisynchronous correlations correspond exactly to mixed permutation maps.","Graph isomorphism $G\\sim_t H$ holds if and only if there is a quantum $t$-permutation intertwining the adjacency matrices, and equivalently if and only if there is a $t$-factorizable map $\\Phi$ with $\\Phi(A_G)=A_H$ and $\\Phi^*(A_H)=A_G$.","The fixed-point algebra of the channel $\\Phi_p$ equals the commutant of any representing quantum permutation, and this fixed-point algebra is closed under the Schur product.","The flip of a bisynchronous density is again bisynchronous for $t\\in\\{\\mathrm{loc},\\mathrm{q},\\mathrm{qa},\\mathrm{qc},\\mathrm{vect}\\}$, so the transpose density belongs to the same correlation class."],"supporting_citations":[{"why":"Supplies the trace characterization of synchronous correlations on $C^*(F(n,n))$ that the proof of Theorem 2.2 starts from and lifts to $O(S_n^+)$.","marker":"[19]"},{"why":"Supplies the analogous synchronous correlation characterizations for qa, q, loc and the identity $C^s_{qs}=C^s_q$, used for the parallel bisynchronous cases.","marker":"[10]"},{"why":"With [15], provides the vectorial characterization of synchronous correlations used in the vectorial case of Theorem 2.2.","marker":"[14]"},{"why":"With [14], provides the convergent-hierarchy vectorial characterization of synchronous correlations used for $C^{bs}_{vect}(n,n)$.","marker":"[15]"},{"why":"Provides the equivalence between graph isomorphism and existence of a quantum t-permutation (Theorem 4.9), used in Theorem 4.3, and background on $O(S_n^+)$.","marker":"[2]"},{"why":"Defines factorizable maps, gives the ancilla characterization of mixed unitary maps, and supplies the factorization theorem used in the fixed-point algebra proof.","marker":"[5]"},{"why":"Introduces the completely positive map $\\Phi_p$ from a correlation and its composition rule, which the bisynchronous map analysis extends.","marker":"[16]"},{"why":"Studies the CP map for graph isomorphism games and its coherent-algebra consequences, providing the map-based viewpoint that Section 4.1 develops for factorizability.","marker":"[12]"}],"fun_headline_variants":["Bisynchronous games tie to quantum permutation traces","Square bisynchronous games yield quantum group traces","Bisynchronous densities are factorizable via quantum permutations","When inputs equal outputs, bisynchronous traces emerge","Quantum group traces classify bisynchronous correlations"],"cache_read_input_tokens":20224,"weakest_assumption_plain":"The characterization inherits previously established characterizations of synchronous correlations—the proof leans on those results for qc, qa, q, loc, and vectorial correlations—and on the known equivalence between graph isomorphism and existence of a quantum t-permutation; if any of those prior results is incomplete, the bisynchronous theorems built on them collapse.","fun_headline_variants_meta":{"raw":{"variants":["Bisynchronous games tie to quantum permutation traces","Square bisynchronous games yield quantum group traces","Bisynchronous densities are factorizable via quantum permutations","When inputs equal outputs, bisynchronous traces emerge","Quantum group traces classify bisynchronous correlations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000197,"raw_usage":{"total_tokens":1346,"prompt_tokens":905,"completion_tokens":441,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":521,"completion_tokens_details":{"reasoning_tokens":372}},"tokens_in":521,"tokens_out":441,"duration_ms":4825,"temperature":1.0,"reasoning_tokens":372,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T14:01:13.919099+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a bisynchronous density in $C^{bs}_{qc}(3,3)$ whose entries cannot be written as $\\tau(u_{x,a}u_{y,b})$ for any tracial state on $O(S_3^+)$; since the set of such trace evaluations is convex and semidefinite-representable, a concrete counterexample would be a finite matrix violating the feasibility problem. Alternatively, exhibit a local bisynchronous density whose associated map is not a mixed permutation map; Theorem 4.2 predicts no such density exists.","supporting_citations":[{"cited_title":"P aulsen, S","cited_arxiv_id":null,"evidence_quote":"Supplies the trace characterization of synchronous correlations on $C^*(F(n,n))$ that the proof of Theorem 2.2 starts from and lifts to $O(S_n^+)$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the analogous synchronous correlation characterizations for qa, q, loc and the identity $C^s_{qs}=C^s_q$, used for the parallel bisynchronous cases."},{"cited_title":"Navascu ´es, Y","cited_arxiv_id":null,"evidence_quote":"With [15], provides the vectorial characterization of synchronous correlations used in the vectorial case of Theorem 2.2."},{"cited_title":"Navascu´es, S","cited_arxiv_id":null,"evidence_quote":"With [14], provides the convergent-hierarchy vectorial characterization of synchronous correlations used for $C^{bs}_{vect}(n,n)$."},{"cited_title":"Bigalois extensions and the graph isomorphism game","cited_arxiv_id":"1812.11474","evidence_quote":"Provides the equivalence between graph isomorphism and existence of a quantum t-permutation (Theorem 4.9), used in Theorem 4.3, and background on $O(S_n^+)$."},{"cited_title":"Haagerup and M","cited_arxiv_id":null,"evidence_quote":"Defines factorizable maps, gives the ancilla characterization of mixed unitary maps, and supplies the factorization theorem used in the fixed-point algebra proof."},{"cited_title":"Ortiz and V","cited_arxiv_id":null,"evidence_quote":"Introduces the completely positive map $\\Phi_p$ from a correlation and its composition rule, which the bisynchronous map analysis extends."},{"cited_title":"Manˇcinska; D","cited_arxiv_id":null,"evidence_quote":"Studies the CP map for graph isomorphism games and its coherent-algebra consequences, providing the map-based viewpoint that Section 4.1 develops for factorizability."}],"review_version":1}