{"id":"4d9e8dfc-db12-4f07-8ce1-86f4c5612f5f","arxiv_id":"2607.14931","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Finite Hardy-type and Kochen-Specker configurations rule out noncontextual hidden-variable models whose outcome probabilities are drawn from any finite subset of [0,1].","lead":"This paper constructs finite sets of quantum measurements, called nested Clifton graphs, whose statistics cannot be reproduced by any noncontextual hidden-variable model that assigns outcome probabilities from a finite list. It matters because it sharpens the Kochen-Specker no-go theorem into experimentally feasible tests that rule out discrete nondeterministic alternatives to the Born rule.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Printed vectors in (B11)-(B12) are not unit: ||v3||=||v4||=||v6||=||v7||=sqrt(2/(k+2)) != 1, so the claimed qutrit Hardy construction is not an orthonormal representation as written.","rationale":"The paper's central no-go logic via Theorem 2 is clean: for any O-valued empirical model, P(1|v2^(1)) ≤ w_q under the Hardy constraints. The main risk is the geometric realizability of the claimed quantum value. The reader's weakest assumption identified the explicit representation (B11)-(B12) as the fragile premise. My specific check confirms that the printed vectors are not unit, so they do not constitute an orthonormal representation as defined. This is a concrete failure of the central construction as written. However, the failure is readily repairable: renormalizing v3,v4,v6,v7 by sqrt((k+2)/2) preserves all orthogonality relations and makes each maximum clique an orthonormal basis, while leaving the overlap |⟨v1|v2⟩| = k/(k+2) unchanged. Similarly, swapping c^2 and s^2 in Lemma 1 fixes the stated parametrization. Thus the errors are typographical/expository rather than fatal to the underlying claim. The verdict remains CONDITIONAL: the manuscript needs corrections before the construction can be used as stated, but the argument's core is sound. No change to the reader's verdict is needed.","tokens_in":35505,"tokens_out":12477,"duration_ms":103311,"concrete_test":"Write a short script (or do hand algebra) for k=1 and k=2: (1) compute the norms of v3^(1), v4^(1), v6^(1), v7^(1) from (B11) and verify they equal sqrt(2/(k+2)), not 1; (2) check all required orthogonality relations, including v5^(1)⊥v8^(1) and the identifications v5^(m)=v1^(m+1), v8^(m)=v2^(m+1); (3) renormalize the four vectors by factor sqrt((k+2)/2) and re-check that the graph is satisfied and that |⟨v1^(1)|v2^(1)⟩| = k/(k+2). If renormalization restores all constraints, the issue is typographical and the quantum value stands; if not, the explicit construction is invalid.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim depends on the explicit orthogonal representation in (B11)-(B12) being a valid orthonormal representation of the k-nested Clifton graph. For each layer, the vectors |v3>, |v4>, |v6>, |v7> are printed with prefactor 1/sqrt(1+k). Computing their squared norm: |v3|^2 = (A^2 + sin^2θ + cos^2θ)/(1+k) = (1 + k/(k+2))/(1+k) = 2/(k+2), where A = sqrt(k/2)(-cosθ+sinθ). Thus ||v3|| = sqrt(2/(k+2)) != 1 (similarly for v4,v6,v7). Definition 1 and the SDP constraints (B7) require unit vectors. As printed, the maximum cliques {v3,v4,v5} and {v6,v7,v8} are not orthonormal bases, so the Hardy normalization conditions in (2) cannot be satisfied by the stated projective measurements. The c^2/s^2 swap in Lemma 1's parametrization (B8) is a related concrete error: with the printed definitions c^2=(1+x)/2, s^2=(1-x)/2, the vectors in (B8) are not mutually orthogonal; swapping the definitions fixes the orthogonality. These are not mere cosmetics — without correction, the paper's existence proof for a quantum realization achieving P^(q)(1|v2^(1)) = (k/(k+2))^2 is unsupported. The errors appear fixable (renormalize v3,v4,v6,v7 by sqrt((k+2)/2) and swap c^2/s^2), but as written the construction fails.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces strong O-valued contextuality, a hierarchy extending the sheaf-theoretic notions of Abramsky and Brandenburger, and claims two main results. First, for any finite O={0,w1,...,wq,1} in [0,1], a k-nested Clifton graph yields a state-dependent Hardy-type test: every O-valued noncontextual model must satisfy P(1|v2^(1)) ≤ wq, while a qutrit realization is claimed to achieve P(1|v2^(1)) = (k/(k+2))^2, so choosing k > 2√wq/(1−√wq) falsifies O-valued noncontextual models. Second, for any O* with (O*)^3∩Δ2 ⊆ ∂Δ2, a finite state-independent KS-type set is claimed to admit no global O*-valued assignment. The paper also discusses applications to randomness, no-signaling polytopes, and finite many-valued logics. The central no-go logic is plausible and appears to extend prior work, but the printed proof of the quantum value contains explicit vector-normalization and algebraic errors that must be fixed.","tokens_in":35932,"tokens_out":8644,"duration_ms":74503,"significance":"If the construction is repaired, the state-dependent result would be a significant advance: it gives finite, experimentally feasible Hardy tests that rule out arbitrary finite nondeterministic noncontextual alternatives to the Born rule, directly addressing a question of Pitowsky and generalizing the author's earlier {0,p,1−p,1} result. The state-independent theorem for the class (O*)^3∩Δ2 ⊆ ∂Δ2 also broadens previous KS-type no-go results. The paper is constructive and, apart from the identified normalization errors, does not appear to smuggle in the conclusion through fitted parameters or normalization choices. The LP proof of Theorem 2 is clean and the explicit overlap formula k/(k+2) appears to be the correct target after renormalization. However, as printed, the existence proof for the quantum realization is invalid, and the state-independent proof delegates a load-bearing step to prior work.","major_comments":[{"comment":"The printed vectors |v3^(m)>, |v4^(m)>, |v6^(m)>, |v7^(m)> are not unit vectors. Their squared norm is 2/(k+2), not 1, because the first component contributes k/(k+2) in the numerator. This violates Definition 1 and the constraints M_{x,x}=1 in the SDP (B7), so the claimed qutrit projective measurements do not satisfy the Hardy normalization conditions (2). The apparent fix is to multiply these four vectors by sqrt((k+2)/2), which preserves their mutual orthogonality as written. This must be corrected before the claimed quantum value P^(q)(1|v2^(1))=(k/(k+2))^2 is established.","section":"Appendix B, Eqs. (B11)-(B12)"},{"comment":"The parametrization of |v3^(k)> and |v4^(k)> has c^2 and s^2 exchanged. With the printed definitions c^2=(1+x_k)/2 and s^2=(1−x_k)/2, the two vectors have inner product −2x_k/(1+x_k), not zero. Swapping the definitions to c^2=(1−x_k)/2 and s^2=(1+x_k)/2 restores orthogonality. Since Lemma 1 is the engine of the induction in Theorem 3, this is a load-bearing error, although a local one.","section":"Appendix B, Lemma 1, Eq. (B8)"},{"comment":"The induction display after Lemma 1 is garbled. Lemma 1 gives x_k = (x_{k+1}+1)/(3−x_{k+1}), with x_{k+1}=|<v5^(k)|v8^(k)>|. Starting from the innermost edge v5^(k)⊥v8^(k), i.e. x_{k+1}=0, the recurrence produces x_{k-r} = r/(r+2) after r steps, yielding x_1=k/(k+2). The printed text instead substitutes x_{k+1}=k/(k+2) and obtains (k+1)/(k+3), which inverts the induction direction and is algebraically incorrect. The final formula is consistent with the explicit representation, but the proof as written does not establish it.","section":"Appendix B, Theorem 3 induction"},{"comment":"The proof of the generalized KS theorem is incomplete in the manuscript. Step three is described as 'strictly analogous' to the author's prior [18] and not proved, while the gadget vectors in the displayed list are unnormalized and no explicit verification of the orthogonality graph of Fig. 4 is given. Since Theorem 1 is a central claim, the adaptation of the step-three gadget to the full class O* satisfying Definition 19 must be written out or at least reduced to [18] with a precise dictionary. As printed, the state-independent result rests on an assertion.","section":"Appendix E, Theorem 10 / Step three"}],"minor_comments":[{"comment":"In the formula for |v4^(m)> the superscript is printed as |v4^(1)>; it should be |v4^(m)>. Also several lines in the vector display have inconsistent parentheses and spacing.","section":"Appendix B, Eq. (B12)"},{"comment":"The linear program uses the constraint f(v5)+f(v8) ≤ 2−1/2^k, while the text says this comes from the inductive hypothesis f(v5)+f(v8) ≤ 1−1/2^k. The LP form is the correct one; the text should be corrected.","section":"Appendix B, Eq. (B3) and surrounding text"},{"comment":"The proof of Theorem 5 is only sketched: the no-signaling box P is not fully defined for pairs of different settings x≠y, and the condition P_{ab|xy}=0 when ⟨v_{a,x}|v_{b,y}⟩=0 does not by itself determine a valid nonsignaling correlation. This application should either be proved carefully or explicitly flagged as a conjecture.","section":"Appendix C, Theorem 5"},{"comment":"There are typographical errors, e.g., 'Langrage multipliers', 'senario' in Theorem 5, and the reference to '6k+1 triangles' should be checked against the count 6k+1 measurement settings. Figures are referenced but not included in the text; the gadget graphs in Appendix E should be accompanied by explicit vector sets.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The vector errors in Appendix B are concrete and appear fixable without changing the paper's central claims; the reported overlap formula k/(k+2) is consistent with the intended construction after renormalization. The main risk is that the proof of Theorem 3 is not written carefully and the state-independent Theorem 1 depends on an omitted step. I would encourage the editor to send the paper back for a revised appendix with explicit normalized vectors, a corrected induction, and a complete or precisely reduced proof of Theorem 10. The paper's core idea is worthwhile and should not be rejected on the basis of the current, locally repairable errors."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The core idea here is worth taking seriously: a finite, state-dependent Hardy family that rules out O-valued noncontextual models for arbitrary finite O, plus a state-independent version for a class O*. The LP no-go argument (Theorem 2) is clean and does what it claims, and the framing of strong O-valued contextuality is a natural and useful extension of the Abramsky–Brandenburger hierarchy. That part of the paper deserves credit.\n\nThe problem is that the explicit quantum realization, which is the load-bearing piece of the Hardy test, is not valid as printed. I checked the vectors in (B11)–(B12): v3, v4, v6 and v7 have squared norm 2/(k+2), not 1. So the maximum cliques {v3,v4,v5} and {v6,v7,v8} are not orthonormal bases, and the Hardy normalization conditions cannot be satisfied by the stated projectors. Separately, Lemma 1's parametrization has the c^2 and s^2 definitions swapped: as printed, v3 and v4 are not mutually orthogonal; swapping the definitions fixes that. The induction display in Theorem 3 is also garbled, though the intended formula is clear. These look like fixable typos rather than a contradiction in the no-go logic, but they are not cosmetic: without correction, the claimed quantum value P(q)(1|v2) = (k/(k+2))^2 is unsupported.\n\nThere are two other soft spots. First, Theorem 10's proof is not self-contained: step three is explicitly deferred to the author's earlier [18]. That prior result is published and independent, so it's not fatal, but the paper should either include the step or state clearly that the generalized KS theorem relies on [18]. Second, the novelty assessment is fair as far as it goes: the explicit Hardy family is new, but the existence of finite no-go sets for arbitrary finite O was already implicit in Pitowsky and in Hrushovski–Pitowsky, and the paper itself uses Corollary 1 of [39] as an alternative tool in Appendix E.\n\nWho should read this? Foundational quantum information people working on contextuality as a resource and on finite Gleason-type results. They will find the framework useful, but they should not quote the Hardy construction until the appendix is corrected. I would send this to a serious referee: the question is important enough and the errors appear repairable. The referee should be asked to verify the orthonormal representation explicitly, not just check the prose.\n\nMy recommendation: engage with it, but require a corrected appendix and a self-contained proof of Theorem 10 before accepting the central claims.","headline":"A promising generalization of Kochen-Specker to finite O with a clean LP no-go, but the printed qutrit construction is not an orthonormal representation as written; fix the appendix before relying on the central claim.","tokens_in":36453,"tokens_out":5663,"would_cite":false,"duration_ms":49414,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["03.65.Ta","03.67.Mn"],"model":"deepseek-v4-flash","headline":"For any finite set O of allowed outcome probabilities, a finite Hardy-type measurement configuration separates quantum theory from every O-valued noncontextual model, and a state-independent version rules out all O* whose triples on the sim","keywords":["contextuality","Kochen-Specker theorem","Gleason's theorem","Hardy-type tests","noncontextual empirical models","orthogonality graphs","finite-valued logics","Born rule"],"falsifier":"Numerically evaluate the Gram matrix of the 6k+2 vectors in equations (B11)-(B12) for k=2 and k=3, checking every required orthogonality, the identifications v5^(m)=v1^(m+1) and v8^(m)=v2^(m+1), and the value of |⟨v1^(1)|v2^(1)⟩|. The central claim stands only if this equals k/(k+2) and the Hardy constraints hold; a simpler signature is whether the printed inner product between v5^(k) and v8^(k) is zero.","tokens_in":35349,"feed_emoji":"⚛️","tokens_out":6750,"duration_ms":62591,"temperature":0.7,"pith_summary":"This paper claims that the Kochen-Specker no-go result can be lifted from deterministic {0,1} outcome assignments to any finite subset O of [0,1] as the allowed ontic probability values. The central construction is a family of nested Clifton orthogonality graphs, read as Hardy-type experiments: any O-valued noncontextual model forces the target box to be full with probability at most the largest element w_q of O, while a qutrit realization gives (k/(k+2))^2. Choosing the nesting depth k larger than 2√w_q/(1−√w_q) makes the quantum value strictly exceed the O-valued bound, falsifying every such discrete nondeterministic noncontextual ontology. A second, state-independent theorem rules out O*-valued assignments for any finite O* whose triples on the 2-simplex all lie on the boundary. The upshot is that finiteness of the allowed outcome set, by itself, does not rescue hidden-variable theories: quantum theory also evades global sections of all finite-valued presheaves.","feed_headline":"Finite tests beat every discrete nonquantum model","feed_subtitle":"Nested eight-vertex graphs force O-valued noncontextual probabilities below the quantum value; nesting depth sets the gap.","key_machinery":"The carrying object is the k-nested Clifton graph: k copies of the eight-vertex Clifton graph glued by identifying v5^(m) with v1^(m+1) and v8^(m) with v2^(m+1), leaving 6k+2 vertices. Its role is to amplify a small noncontextual constraint: the linear-programming induction in Theorem 2 shows that assigning probability 1 to the outer vertex v1^(1) forces the probability of v2^(1) down to at most 1−2^{−k} in any consistent model. The matching qutrit machinery is a rotation-based optimal orthonormal representation (rotation R by π/2 about (0,1,1)/√2, angles θ_j=½ arcsin(j/(j+2))) that attains the overlap bound k/(k+2) between the outer vectors; the recursion x_k=(x_{k+1}+1)/(3−x_{k+1}) connect","core_discovery":"The core claim is that for every finite O={0,w1,...,wq,1} there is a finite measurement scenario whose quantum statistics are incompatible with every O-valued noncontextual model. The scenario is the k-nested Clifton graph. For any no-disturbance assignment with f(v1^(1))=1, a linear-programming induction gives f(v2^(1))≤1−2^{−k}, so in any O-valued model f(v2^(1))≤w_q. The paper constructs a qutrit orthonormal representation of the same graph with outer-vertex overlap |⟨v1^(1)|v2^(1)⟩|=k/(k+2), and hence a Hardy test in which P(q)(1|v2^(1))=(k/(k+2))^2. For k above threshold the quantum value is larger. It also proves a constructive generalized Kochen-Specker theorem: for O* with (O*)^3∩Δ2⊆","pith_inferences":["If the construction is correct, the nesting depth needed to refute a given O supplies a quantitative measure of how close that discrete ontology sits to quantum theory: as w_q approaches 1, the required number of measurements diverges, suggesting a graded hierarchy between finite-valued hidden-variable theories and the Born rule.","The O-valued contextual fraction can likely be operationalized as a resource monotone for semi-device-independent tasks, since the Hardy violation gives a direct lower bound on the weight of the behavior that no O-valued model can simulate.","A natural next step the paper leaves open is to convert the nested Clifton graph into a state-independent Kochen-Specker set for arbitrary O, which would place quantum correlations arbitrarily close to extremal no-signalling points while keeping O-valued marginals finite.","The explicit vector formulas in the appendix are the part most worth re-deriving carefully: a corrected representation with the same overlap would preserve the argument, while any failure of the claimed identifications or final orthogonality would leave the central theorem without its stated quantum realization."],"forward_implications":["For any finite O, there exists an experimentally feasible Hardy-type test using 6k+2 projective measurements whose outcome statistics cannot be explained by any O-valued noncontextual model; the required number of projectors grows as 6⌈2√w_q/(1−√w_q)⌉+8.","Quantum theory evades global sections of all finite-valued presheaves, so strong O-valued contextuality strictly extends the established probabilistic-possibilistic-strong hierarchy.","Finite many-valued logical models, when identified with O-valued truth assignments via the Birkhoff-von Neumann–Łukasiewicz correspondence, are ruled out by a finite partial Boolean algebra rather than by the infinite continuity-based Gleason argument.","As the nesting depth k grows, the optimal quantum correlations approach extremal nonclassical points of the no-disturbance polytope, providing a route toward semi-device-independent randomness protocols secure against O-valued adversaries.","The generalized Kochen-Specker theorem covers a broad class of O*, including sets like {0,1/m,1} for m≥2, going beyond prior results that only handled sets of the form {0,p,1−p,1}."],"fun_headline_variants":["Nested Hardy tests beat every discrete nonquantum theory","Quantum escapes all finite-valued noncontextual models","Generalized Kochen-Specker for any finite outcome set","Strong O-valued contextuality rules out finite models","Finite configurations eliminate discrete probabilistic noncontextuality"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The no-go argument assumes the explicit list of qutrit vectors actually realizes the nested Clifton graph: that all claimed orthogonality edges, the layer identifications v5^(m)=v1^(m+1) and v8^(m)=v2^(m+1), and the innermost edge are satisfied, so that the overlap between the outer vectors is exactly k/(k+2).","fun_headline_variants_meta":{"raw":{"variants":["Nested Hardy tests beat every discrete nonquantum theory","Quantum escapes all finite-valued noncontextual models","Generalized Kochen-Specker for any finite outcome set","Strong O-valued contextuality rules out finite models","Finite configurations eliminate discrete probabilistic noncontextuality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000217,"raw_usage":{"total_tokens":1331,"prompt_tokens":859,"completion_tokens":472,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":603,"completion_tokens_details":{"reasoning_tokens":396}},"tokens_in":603,"tokens_out":472,"duration_ms":4455,"temperature":1.0,"reasoning_tokens":396,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T00:41:44.321483+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Numerically evaluate the Gram matrix of the 6k+2 vectors in equations (B11)-(B12) for k=2 and k=3, checking every required orthogonality, the identifications v5^(m)=v1^(m+1) and v8^(m)=v2^(m+1), and the value of |⟨v1^(1)|v2^(1)⟩|. The central claim stands only if this equals k/(k+2) and the Hardy constraints hold; a simpler signature is whether the printed inner product between v5^(k) and v8^(k) is zero.","supporting_citations":[],"review_version":1}