{"id":"939713e0-e3a9-4a51-a592-b7bfcd28972e","arxiv_id":"2501.00409","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"For every d ≥ 3, d-party d-level supersinglet states are self-tested by d-partite d-dimensional perfect quantum strategies based on rigid Kochen-Specker sets.","lead":"The authors show that for any number of parties d ≥ 3, there is a nonlocal game that is won perfectly only by the d-party d-level supersinglet state, and the game's correlations uniquely certify that state. This answers two open questions: whether supersinglets enable a task no other state can do, and whether they have a unique high-dimensional nonlocal signature.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Appendix A's proof that perfect strategies force projective KS measurements relies on an unjustified disjoint-events-to-orthogonality inference; this is the load-bearing bridge for Theorems 1 and 2 and is not established in the text.","rationale":"The paper's central claim is Theorem 2: for every d>=3 there is a d-partite, d-dimensional perfect quantum strategy that self-tests the d-d supersinglet. The proof route is Proposition 1 plus Theorem 1 plus the rigidity-to-state argument in Appendix B. The single most load-bearing step is Theorem 1's claim that any perfect quantum strategy must realize the measurements as a projective KS set: only then can KS-set rigidity give unitary equivalence of the measurements. The reader's weakest assumption identifies exactly this step, and my reading agrees. The clearest defect in Appendix A is the inference from disjoint outcome sets to orthogonal supports of conditional reduced states; in POVM-based strategies this is generally false. The surrounding argument could potentially be repaired, because S_{x,y'} is contained in \\tilde S_{x,y} and the perfect-strategy conditions already force B_y to act as the identity on S_{x,y} and zero on \\tilde S_{x,y}, which may imply the needed orthogonality. But the paper does not supply that argument, and Theorem 1 as written depends on the invalid inference. Appendix B is terse, especially the recursive Levi-Civita sign argument for general d, but it is not the first point of failure: even if Appendix B is correct, it cannot compensate for an unproven Theorem 1. The unsupported classical success-probability values in the conclusions are a minor issue compared with this gap. I therefore agree with the reader's conditional verdict: the result is plausible and likely repairable, but the proof is not complete as written.","tokens_in":19092,"tokens_out":28909,"duration_ms":315623,"concrete_test":"Re-derive Eqs. (A8) and (A9) from Eqs. (A5)-(A7) without using the sentence that disjoint outcome sets imply orthogonal subspaces. Check whether the derivation forces B_y B_{y'}=O for all y,y' in a context using only 0<=B_y<=I, the support inclusions S_{x,y'} subset of \\tilde S_{x,y}, and the zero-probability conditions implicit in Eq. (A2). If the re-derivation requires an additional lemma about POVM effects on zero-probability events, or fails for a finite-dimensional POVM example where disjoint outcomes have overlapping support, then the proof of Theorem 1 is incomplete and Theorem 2 is unproven as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Appendix A (around Eq. A8) asserts that because the outcome sets P(Cx\\{y}) and P(Cx\\{y'}) are disjoint, the subspaces S_{x,y} and S_{x,y'} are orthogonal. This inference is not valid for general POVMs: disjoint outcome labels do not imply orthogonal supports of the corresponding conditional reduced states. The step is then used to conclude B_y B_{y'}=O and hence that {B_y} and {A_{a|x}} form a projective KS set, without which rigidity cannot be invoked to self-test measurements. The proof also moves from Eq. A11 to Eqs. A15-A16 and from Eq. A13 to Eqs. A18-A19, converting zero-probability events into operator identities and zeros on the relevant supports without proving the required positivity/support argument. As written, Theorem 1, and therefore Theorem 2, do not follow. The gap may be repairable, e.g. using S_{x,y'} subset of \\tilde S_{x,y} together with B_y=0 on \\tilde S_{x,y} rather than the orthogonality claim, but that repair is substantive and is not present in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a family of d-partite, d-outcome nonlocal games, one for each d≥3, built from complete Kochen–Specker (KS) sets in C^d. It claims that these games are won perfectly by the d-party d-level supersinglet and, moreover, that any perfect quantum strategy for them must be locally isometric to the reference strategy consisting of the supersinglet and rank-one projective measurements given by a rigid KS set. The proof strategy is to show from the perfect winning condition that all uncharacterized local measurements must be projective and form a rigid KS set, and then to use rigidity together with the antisymmetric structure of the state to fix the state coefficients. Explicit coefficient computations are given for d=3 and d=4, and a recursive argument is given for all d≥4. If the proof of Theorem 1 is valid, the paper answers Questions 1 and 2 from the introduction affirmatively.","tokens_in":43,"tokens_out":20529,"duration_ms":324222,"significance":"If the central proof gap is repaired, this would be a significant contribution: it would provide a self-test for a family of multipartite high-dimensional states and would simultaneously produce d-partite d-dimensional perfect quantum strategies with a unique quantum realization. The construction is explicit and parameter-free, the state is certified from the correlations rather than assumed, and the d=3 and d=4 coefficient systems are written out in enough detail to be checked. The use of existing rigid-KS-set results [35,52] is appropriate rather than circular, and the paper clearly identifies the open questions it addresses. The main caveat is that the proof of Theorem 1 relies on an invalid inference from disjoint outcome labels to orthogonal supports for uncharacterized POVMs; until that step is repaired, the central claim is not established.","major_comments":[{"comment":"The bridge from the perfect winning condition to projectivity is invalid. The text asserts that because P(C_x\\setminus\\{y'\\}) and P(C_x\\setminus\\{y\\}) are disjoint outcome sets, the subspaces S_{x,y'} and S_{x,y} are orthogonal. For uncharacterized POVMs this does not follow: distinct outcome labels can have overlapping supports, and the conditional reduced states ρ_{a|x} for different a need not have orthogonal ranges. This is not a minor gap: the orthogonality claim is used to derive B_yB_{y'}=O, and together with the completeness relation (A9) it is what makes {B_y} a projective KS realization. Without that, the rigidity assumption cannot be invoked, so Theorem 1, and with it Theorem 2, do not follow from the proof as written. A repair would need to replace this step with a support-based argument, for example using S_{x,y'}\\subseteq \\tilde S_{x,y} together with B_y=0 on \\tilde S_{x,y} and a separate justification that B_{y'} is the projection onto S_{x,y'}; that argument is not present in the text.","section":"Appendix A, Eq. (A8)"},{"comment":"The same disjoint-outcome-to-orthogonality issue appears for the first d-1 parties. From Eq. (A14) the trace identities give only A_{x,y}=1 on the support of σ_y and \\tilde A_{x,y}=1 on the support of \\tilde σ_y. The additional assertions A_{x,y}|_{\\tilde σ_y}=O (Eq. (A16)) and the decomposition A_{x,y}+\\tilde A_{x,y}=1^{[d-1]} (Eq. (A19)) require that A_{x,y} and \\tilde A_{x,y} have orthogonal or complementary supports; this is not implied by disjointness of the outcome sets in Eqs. (A12)–(A13). Consequently the conclusion that each A_{a_i|x_i} is a projective element of a KS realization is not established, and the subsequent application of rigidity is unsupported.","section":"Appendix A, Eqs. (A14)–(A19)"}],"minor_comments":[{"comment":"The displayed chain of equalities contains coefficients such as α_{2134}, α_{2143}, α_{2314}, α_{2341}, α_{2413}, and α_{2431}, none of which appear in the variable ordering in Eq. (B20) and which involve an index 4 outside {0,1,2,3}. This is presumably a typesetting error, but as printed the relation is undefined and should be corrected.","section":"Appendix B, Eq. (B21)"},{"comment":"The contrapositive proof considers two complete KS sets associated with the same orthogonality graph G, but the rigidity notion for a non-complete graph G compares realizations of G that need not be complete, while the game is defined by an extension G_c. As written, the 'only if' direction is not demonstrated. Since the forward direction is what is used for Theorem 2, this does not block the main result, but the statement should be proved or weakened to an 'if' statement.","section":"Theorem 1, converse direction"},{"comment":"The claimed classical success probabilities 35/36 for the 18-vector set and 59/60 for the 24-vector set are stated without proof or reference; please provide a derivation or citation.","section":"Conclusions"},{"comment":"There are several language issues that should be corrected in a revision, including 'none any of these applications' in the abstract, 'impossible any assignment satisfying' in the caption of Fig. 1, and 'The named-qudit-supersinglets follows' in the introduction.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The main concern is the invalid POVM-to-projective step in Appendix A; if the authors can repair that step, the result is publishable. I also suggest carefully verifying that the d-dimensional KS sets from [35] used in Appendix B are indeed rigid in the sense required by Theorem 1, since the dimension extension of Theorem 2 depends on that reference. The reliance on two papers co-authored by the present authors is not itself a problem, but it is worth a careful check of those references."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Hey,\n\nBottom line: this is a genuinely new result—the first self-test for supersinglets and the first family of d-partite d-dimensional perfect quantum strategies for every d≥3—but the proof of the central theorem has a hole in Appendix A that needs to be patched before the result is fully rigorous.\n\nWhat's good: the idea of using rigid Kochen-Specker sets as the measurement skeleton is clever and likely to be reused. The d=3 and d=4 computations are explicit and checkable; I verified the structure of the 31-vector and Peres-24 arguments and they are sound. The construction in Appendix B for general d, while terse, is plausible and follows the same pattern. The two open questions are both answered yes, assuming the proof holds. The citations to the rigid KS sets are appropriate—those are independent results with their own proofs, so the self-referential nature isn't a problem.\n\nWhere it's soft: the stress-test note is right. In Appendix A, the paper claims that because the outcome sets P(C_x\\{y}) and P(C_x\\{y'}) are disjoint, the subspaces S_{x,y} and S_{x,y'} are orthogonal. That inference is false in general for POVMs: disjoint sets of outcome labels don't force the supports of the corresponding conditional reduced states apart. The authors use this to get B_y B_{y'}=O, which is the bridge to projectivity and then to KS rigidity. The good news is that the conclusion is repairable by a nearby argument: S_{x,y'} is actually contained in \\tilde S_{x,y}, and B_y was already shown to vanish on \\tilde S_{x,y}, so B_y B_{y'}=O follows without orthogonality. But that repair is not in the text. As written, Theorem 1 doesn't follow. The rest of the proof for the first d-1 parties uses a similar but valid containment argument, so the patch is local. Minor issues: the classical success probabilities (35/36 and 59/60) are stated without derivation, and the notation tr_{[d-2]} in Appendix A is wrong—it should trace out the first d-1 parties, not d-2.\n\nWho's this for: anyone working on self-testing, nonlocal games, or multipartite high-dimensional entanglement. It answers a clean open question and introduces a technique that will likely generalize.\n\nRecommendation: this deserves serious peer review. The gap is real but repairable, and the core contribution is strong enough that the paper should be sent to referees rather than desk-rejected. I'd ask for a revision that fixes Appendix A and adds derivations for the classical bounds.","headline":"Novel and important result, but the main proof has a repairable gap in Appendix A that should be fixed before publication.","tokens_in":19836,"tokens_out":8904,"would_cite":true,"duration_ms":79457,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"For every d≥3 there is a d-party, d-dimensional nonlocal game whose only perfect quantum realization is the d-level supersinglet, up to local unitaries.","keywords":["self-testing","supersinglets","perfect quantum strategies","Kochen-Specker sets","rigidity","nonlocal games","pseudo-telepathy","multipartite entanglement"],"falsifier":"Search the d=3 game built from the 31-vector rigid KS set for a perfect strategy that uses a non-projective POVM and is not unitarily equivalent to the KS projectors; even one such example would break Theorem 1 and Theorem 2.","tokens_in":18825,"feed_emoji":"🕹️","tokens_out":9174,"duration_ms":89326,"temperature":0.7,"pith_summary":"The paper claims that every d-particle, d-level supersinglet can be self-tested for d≥3. It builds, for each d, a d-party nonlocal game with a perfect quantum strategy whose statistics uniquely pin down the shared state as the supersinglet, up to local unitaries. No classical strategy can win perfectly, because winning would require a 0/1 assignment on a Kochen-Specker set. The result gives a concrete task achievable only with supersinglets and a maximal d-partite, d-dimensional nonlocal signature, answering the two questions the paper poses.","feed_headline":"Every d-party supersinglet can be self-tested by a game","feed_subtitle":"A d-dimensional nonlocal game with perfect win rate certifies the supersinglet uniquely, for all d≥3.","key_machinery":"The load-bearing object is a rigid, complete Kochen-Specker (KS) set in the local Hilbert space $\\mathbb{C}^d$: a finite set of rank-one projectors admitting no consistent 0/1 assignment, with every orthogonal pair lying inside some basis ('complete') and every realization of the same orthogonality graph unitarily equivalent to the reference set ('rigid'). The paper's game hands the same basis (context) to d-1 parties and a single vector from it to the last party; winning requires the first d-1 outputs to be a permutation of the basis and the last output to be 1 exactly when his vector was the one left out. The supersinglet's invariance under $U^{\\otimes d}$ yields a perfect quantum strategy, while rigidity plus the perfect win condition forces local measurements to be the KS projections and the shared state's amplitudes to have the alternating signs of $\\frac{1}{\\sqrt{d!}}\\sum_{\\text{perm}} \\epsilon_{a_0\\ldots a_{d-1}}|a_0\\ldots a_{d-1}\\rangle$.","core_discovery":"The core claim is Theorem 2: for every finite d≥3, there exists a Kochen-Specker set in H=C^d such that the corresponding perfect quantum strategy self-tests the d-party d-level supersinglet. That is, any unknown state and measurements producing the perfect input-output statistics of the constructed game must be related, by local unitaries, to the supersinglet shared among d parties and the projectors of the rigid KS set. Because a classical perfect strategy would have to assign 0/1 values to the KS set in a way that the KS theorem forbids, the same construction gives a d-partite, d-dimensional perfect quantum strategy and a task achievable only with supersinglets.","pith_inferences":["Going beyond the paper, the same rigid-KS-to-self-testing transfer may apply to any multipartite high-dimensional state for which all but one party can predict all KS-set observables from perfect statistics; that would give a whole family of self-tests, not just supersinglets.","A natural extension the paper leaves open is a noise-tolerant version of these games; if found, it would turn experimental preparations of 3- and 4-level supersinglets into device-independent certificates.","One could also use the certified states' genuine high-dimensional, genuinely multipartite entanglement to probe whether supersinglets are the maximally entangled states in their class; the paper does not settle that question."],"forward_implications":["The observed perfect statistics of these games certify the supersinglet without trusting the measurement devices, for every d≥3.","These games are d-partite, d-dimensional perfect quantum strategies, so supersinglets yield a pseudo-telepathy phenomenon in arbitrary local dimension.","Since any perfect classical strategy would assign 0/1 values to the KS set, no classical strategy wins every round, making the quantum advantage maximal.","Question 1 and Question 2 both receive affirmative answers: there is a task achievable only with supersinglets, and this task gives a unique d-partite d-dimensional nonlocal signature."],"supporting_citations":[{"why":"Proves existence of rigid KS sets in every dimension and supplies the d-dimensional construction that Theorem 2 extends to all d≥4.","marker":"[35]"},{"why":"Establishes rigidity of the 31-vector set in C^3, giving the d=3 case of Theorem 2.","marker":"[52]"},{"why":"Supplies the 31-vector KS set whose rigidity is verified in [52].","marker":"[51]"},{"why":"Provides the 24-vector KS set used for the d=4 case in Appendix B.","marker":"[36]"},{"why":"Defines supersinglets and their U^{⊗d} invariance, the property that yields the perfect quantum strategy.","marker":"[1]"}],"fun_headline_variants":["Self-testing all supersinglets via perfect games","Perfect quantum strategies self-test supersinglets","Supersinglets uniquely certified by perfect games","All supersinglets self-tested with perfect strategies","Perfect games self-test every supersinglet"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise, used in Appendix A, is that a perfect win rate forces each local measurement to be a projective measurement on orthogonal vectors rather than an arbitrary positive-operator-valued measure; without projectivity, Kochen-Specker rigidity does not immediately imply self-testing.","fun_headline_variants_meta":{"raw":{"variants":["Self-testing all supersinglets via perfect games","Perfect quantum strategies self-test supersinglets","Supersinglets uniquely certified by perfect games","All supersinglets self-tested with perfect strategies","Perfect games self-test every supersinglet"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000231,"raw_usage":{"total_tokens":1436,"prompt_tokens":848,"completion_tokens":588,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":464,"completion_tokens_details":{"reasoning_tokens":517}},"tokens_in":464,"tokens_out":588,"duration_ms":5042,"temperature":1.0,"reasoning_tokens":517,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:53:13.798511+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Search the d=3 game built from the 31-vector rigid KS set for a perfect strategy that uses a non-projective POVM and is not unitarily equivalent to the KS projectors; even one such example would break Theorem 1 and Theorem 2.","supporting_citations":[{"cited_title":"Trandafir and A","cited_arxiv_id":null,"evidence_quote":"Establishes rigidity of the 31-vector set in C^3, giving the d=3 case of Theorem 2."},{"cited_title":"Peres, Two simple proofs of the Kochen-Specker theorem, J","cited_arxiv_id":null,"evidence_quote":"Provides the 24-vector KS set used for the d=4 case in Appendix B."}],"review_version":1}