{"id":"0d535f5e-eb37-4a02-bbac-b77b199b83a8","arxiv_id":"2506.15149","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The hexablock is a new C4 domain tied to mu-synthesis for upper triangular 2x2 matrices, with distinguished boundary described by |a|^2+|x1|^2=1 over the tetrablock boundary.","lead":"Mathematicians define a new domain in four-dimensional complex space, the hexablock, coming from a structured stability problem for 2x2 matrices. It completes a known family of domains, and the paper computes its boundary, automorphisms, and a Schwarz lemma.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified: the hexablock construction and its Shilov-boundary theorem are internally supported; the only soft spot is the routine extremal-point computation in Proposition 3.1, which I did not find to be an actual gap.","rationale":"The reader's strongest claim is the complete description of the distinguished boundary bH = {(a,x1,x2,x3) in C x bE : |a|^2+|x1|^2=1}, together with H = int(H_mu) = polynomial hull of H_N. I traced the proof through the relevant theorems. Proposition 3.1 is the foundational computational step, and the reader is right to single it out as the least externally verified assumption; however, the supplied derivation is detailed enough that I did not find a concrete failure mode. The discriminant is strictly positive exactly because |beta1|+|beta2|<1, the choice of root is justified by an explicit contradiction, and the degenerate cases are covered. The continuity of K* then follows from the explicit formulas. Theorem 8.21's use of the density of peak points in the Shilov boundary of a uniform algebra on a compact metric space is a standard theorem, and the inclusion bH subset Hp is obtained by exhausting the three boundary components. No circularity appears in Theorem 8.23: it uses bH subset H_N to control polynomials on H_N from their values on bH, which is legitimate. The only caveats are the missing overline distinctions in the OCR text and the summarized algebraic verifications in Chapter 7-8; these are presentation issues rather than load-bearing errors. I therefore see no basis to change the ACCEPT verdict.","tokens_in":1022,"tokens_out":1495,"duration_ms":345296,"concrete_test":"Independently re-derive equations (3.7)-(3.9) and verify, for a pseudorandom sample of at least 10^4 points x uniformly drawn from the tetrablock E with respect to the parameterization of Theorem 2.5(7), that the point (z1(x),z2(x)) given by (3.8)-(3.9) is the unique global maximizer of |kappa(z1,z2,x)| over D^2, by comparing with a dense numerical grid. Also check the claimed discriminant positivity directly from |beta1|+|beta2|<1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"I examined the chain leading to the central claims: Proposition 3.1 gives the unique maximizer of |kappa(z1,z2,x)| and the continuous function K*; Theorem 6.1 and Theorem 6.3 identify H with int(H_mu) and establish the closed form of H; Theorem 8.21 characterizes bH as H_p; Theorem 8.23 identifies the polynomial hull of H_N with the closure of H. I found no internal inconsistency or unsupported step that would overturn these results. The reader's flagged Proposition 3.1 is indeed the least machine-verified computational premise, but the proof is materially complete: the discriminant factorization follows from |beta1|+|beta2|<1 for x in E, the rejected root contradicts |z1|<1, and the cases beta1=0, beta2=0 are consistent with the limiting formulas (3.8)-(3.9). The later arguments that use K* and the peak-point density in bH are standard and do not introduce a circularity. Some notation in the provided text has lost overlines (e.g., H_mu versus its closure, H_N versus its closure), but the surrounding arguments make the intended meaning clear and the claims remain coherent.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces the hexablock H, a bounded domain in C^4 defined by a supremum condition over D^2 involving the fractional linear maps ψ_{z1,z2}(a,x1,x2,x3)=a√((1-|z1|^2)(1-|z2|^2))/(1-x1 z1 - x2 z2 + x3 z1 z2), with (x1,x2,x3) ranging over the tetrablock E. It also defines two companion sets, the μ-hexablock H_μ and the normed hexablock H_N, as images of the structured-singular-value unit ball and the operator-norm unit ball of M_2(C) for the upper triangular subspace under the coordinate map π(A)=(a21,a11,a22,det A). The central structural claims are that H is a domain, that H = int(H_μ) and H = H_μ ∪ ({0}×E), and that the closure of H is the polynomial hull of H_N. The main geometric result, Theorem 8.21, identifies the distinguished boundary bH with H_p = {(a,x1,x2,x3) : |a|^2+|x1|^2=1, (x1,x2,x3)∈bE}, and gives eight equivalent characterizations involving unitary matrices, peak points, and boundary points of H_N. The paper also constructs a large subgroup G(H) of Aut(H), proves that any automorphism of H extends past the closure, studies rational H-inner functions and a Schwarz lemma, and shows that B_2, G_2, E and P are analytic retracts of H. The later chapters apply the theory to recover results for the pentablock.","tokens_in":95059,"tokens_out":10580,"duration_ms":114058,"significance":"If the results stand, the paper is a substantive contribution to the μ-synthesis family of domains: it produces a new bounded domain in C^4 with a complete Shilov-boundary description, establishes polynomial convexity and linear convexity, and connects the hexablock with the symmetrized bidisc, tetrablock and pentablock through analytic retracts. The argument is concrete and parameter-free: no fitted constants enter, and the chain from Proposition 3.1, through Theorems 6.1, 6.3, 8.21 and 8.23, is internally coherent rather than imposed. I specifically checked the least machine-verified point, the uniqueness of the maximizer in Proposition 3.1; the discriminant factorization and the rejection of the second root under |β1|+|β2|<1 are valid, and the limiting cases β1=0 or β2=0 are handled consistently. The cited facts about E, G_2 and P are prior results, so I found no circularity. The main caveat is that the copy of the manuscript supplied to me cuts off in Chapter 10, so the pentablock applications in Chapters 11-14 were not independently audited; the central claims of Chapters 1-10 do not depend on those applications.","major_comments":[],"minor_comments":[{"comment":"The set H(r) is displayed as the preimage of the open set H under the anisotropic scaling map, which is an open set, yet the proof then asserts that H(r) is compact and treats it as a compact polynomially convex set. The intended meaning is evidently that an overline on H or H(r) has been lost; please redefine H(r) as a compact exhaustion of H explicitly, for instance as the closure of the scaled set, and adjust the proof accordingly.","section":"Proposition 6.9"},{"comment":"The overlines distinguishing H, H_μ, H_N from their closures are missing in several key statements, for example in the statements H = int(H_μ), H_μ = H, and H = int(cH_N). Since the distinction between open hexablock and its closure is load-bearing for these identities, a short notational table or consistent overline placement would remove real ambiguity.","section":"Throughout, especially Theorems 6.3, 5.12, 8.23"},{"comment":"In the last sentence of the first paragraph of the proof, the text says 'Hence, π(A)∈H_N and so, {π(A) : A∈M_2(C), ||A||≤1}⊆H_μ'; the final inclusion should read '⊆H_N' (or '⊆overline H_N') to follow from the preceding sentence.","section":"Proposition 5.7"},{"comment":"The sentence 'Since q is a peaking function for H' should read 'Since q is a peak point of H'; the surrounding argument is correct but the terminology is nonstandard.","section":"Proposition 8.10"},{"comment":"The derivation of the quadratic equation (3.7) from the critical-point equations (3.6) is summarized as 'some laborious but routine calculations'. Because the uniqueness of the maximizer is the basis for the function K* and hence for all later characterizations of H, including a few displayed intermediate steps or a reference to a companion computation would substantially improve verifiability.","section":"Proposition 3.1"},{"comment":"The step from density of peak points in the Shilov boundary to bH⊆H_p should explicitly use that H_p is compact; as written the conclusion 'bH = P(H)' is stronger than the cited density theorem justifies without the closure argument.","section":"Theorem 8.21, proof of (7)=>(1)"}],"recommendation":"minor_revision","confidential_remarks":"The copy of the manuscript provided to me contains Chapters 1-10 in full and then breaks off in Chapter 10, so the advertised applications to the pentablock in Chapters 11-14 were not independently checked. My assessment is based on the central claims in Chapters 1-10, which are sound and well supported. If the full manuscript is sent out, the remaining chapters should receive a separate verification, but nothing in the visible portion gives me reason to doubt them."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper does what it says: it introduces a new domain in C4, the hexablock, shows it arises from mu-synthesis in M2(C) with upper-triangular structure, and proves a full Shilov-boundary description. The headline result—H = int(H_mu), with distinguished boundary bH = {(a,x1,x2,x3) in C x bE : |a|^2 + |x1|^2 = 1}—is new and is proved by a coherent chain: unique maximizer in Proposition 3.1, identification of H_mu/H_N interiors in Chapters 4–6, peak-point analysis in Chapter 8, polynomial hull in Theorem 8.23. No circularity: the equality is derived from definitions and prior tetrablock theorems, not imposed. The paper also gives a substantial automorphism subgroup, rational H-inner functions, a Schwarz lemma, and analytic retracts of G2, E, P. The proofs are mostly detailed and the arguments in the visible chapters track.\n\nCredit where due: this is not a repackaging. The hexablock family is genuinely new, and the transfer from tetrablock to H is used rather than assumed. The authors also are careful about deviations from earlier cases (H_mu and H_N are not open, {0} x E not contained in them) and resolve them cleanly via H = int(H_mu).\n\nSoft spots are modest. The extremal argument in Proposition 3.1 is the least machine-verified step: uniqueness of the maximizer requires checking a discriminant and rejecting a root. The proof is materially complete—the factorization and root rejection are spelled out—but it is long, computational, and load-bearing for K*, so a referee should verify it line by line. The later chapters (rational inner functions, Schwarz lemma, pentablock applications) were truncated in the text I got, so I could not fully re-audit them; nothing so far suggests a gap. Some overline/closure notation is lost in the version we have, which is annoying but the intended meaning is clear. A few algebraic verifications in Chapters 7 and 9 are summarized as 'routine'; they involve long fraction compositions but the pattern is consistent.\n\nBottom line: if you work on mu-synthesis domains or the tetrablock family, read this and cite it. It deserves a serious referee, and I would accept it pending a close check of Proposition 3.1 and the later chapters.","headline":"Hexablock is a genuinely new C4 domain in the mu-synthesis family with a clean Shilov-boundary theorem, and the visible proofs hold up; the main risk is the computational uniqueness step, not the architecture.","tokens_in":95587,"tokens_out":2210,"would_cite":true,"duration_ms":24640,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["30C80","32F45","93B36","93B50"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves the hexablock, a bounded C⁴ domain born from µ-synthesis on upper-triangular 2×2 matrices, is polynomially convex and has distinguished boundary {(a,x₁,x₂,x₃) ∈ C×bE : |a|²+|x₁|² = 1}.","keywords":["Hexablock","Symmetrized bidisc","Tetrablock","Pentablock","Automorphism","Distinguished boundary","Schwarz lemma","Rational H-inner function"],"falsifier":"Take a generic tetrablock point such as x = (0.3, 0.2, 0.05) in E ∩ R³, compute the two roots of the quadratic (3.7), and check numerically whether the discarded '−' root also lies in D² and gives the same value of |κ| as the '+' root; a second interior critical point at equal or greater height would break the uniqueness on which K* rests. The cheapest decisive test is to numerically maximize |κ(z₁,z₂,x)| over D² for a spread of x in E ∩ R³ and compare with the closed-form value |κ(z₁(x),z₂(x),x)| from Corollary 3.2; any mismatch refutes Proposition 3.1 and hence Theorem 8.21.","tokens_in":94640,"feed_emoji":"🎛️","tokens_out":13109,"duration_ms":109934,"temperature":0.7,"pith_summary":"The paper introduces a bounded domain in four complex dimensions, the hexablock H, and argues it is the right finite-dimensional representative of the µ-synthesis problem for 2×2 upper-triangular matrices — the last subspace case in M₂(C), after the symmetrized bidisc, the tetrablock, and the pentablock. The two most naively defined companions, Hμ (image of the structured-singular-value unit ball) and H_N (image of the operator-norm unit ball), turn out to be connected but non-open subsets of C⁴, and a large part of the paper's work is showing that H = int(Hμ) = int(polynomial hull of H_N): the hexablock is the open, bounded, tractable object carrying the same interpolation data. The main payoff is a complete description of the distinguished (Shilov) boundary, bH = {(a,x₁,x₂,x₃) ∈ C × bE : |a|²+|x₁|² = 1}, where E is the tetrablock, along with an explicit subgroup of automorphisms, a full characterization of rational H-inner functions, a Schwarz lemma, and the theorem that the symmetrized bidisc, tetrablock, pentablock, and biball are all analytic retracts of H. A sympathetic reader would care because this gives the theory of the three classical domains a single ambient domain with a transparent boundary, through which results can be transferred and the fourth µ-synthesis case studied concretely.","feed_headline":"New C⁴ hexablock gets an exact Shilov boundary","feed_subtitle":"A fourth domain joins the µ-synthesis family in M₂(C); its boundary is the tetrablock boundary with |a|²+|x₁|² = 1.","key_machinery":"The load-bearing object is the two-parameter family of fractional linear maps ψ_{z₁,z₂}(a,x₁,x₂,x₃) = a√((1−|z₁|²)(1−|z₂|²))/(1−x₁z₁−x₂z₂+x₃z₁z₂) on C × E, which generalizes the one-parameter family ψ_z behind the pentablock; the hexablock is simply the set where all |ψ_{z₁,z₂}| < 1 over the tetrablock. Proposition 3.1 shows that for each tetrablock point x the modulus |κ(z₁,z₂,x)| has a unique interior maximizer (z₁(x),z₂(x)) over D², obtained by solving a quadratic whose discriminant factors as a product of four terms (1−|β₁|±|β₂|)(1±|β₁|−|β₂|), and the rejected root is excluded precisely by the tetrablock inequality |β₁|+|β₂| < 1. This yields the continuous function K*(x₁,x₂,x₃), in terms of which membership in H, Hμ, and H_N reduces to |a|K*(x₁,x₂,x₃) < 1, ≤ 1, or the explicit two-sided bounds of Lemma 5.4 respectively. The norm characterization for upper-triangular matrices (Lemma 4.1: ‖X‖ < 1 iff |z₁|,|z₂| < 1 and |w| < √((1−|z₁|²)(1−|z₂|²))) converts operator-norm data into the same shape, and everything downstream — the automorphism group, the boundary computation, the rational inner functions — rests on this reduction of the supremum to a closed form.","core_discovery":"The paper's central claim is that the hexablock H, defined by sup_{z₁,z₂ ∈ D} |a|√((1−|z₁|²)(1−|z₂|²))/(1−x₁z₁−x₂z₂+x₃z₁z₂) < 1 over the tetrablock E, is the distinguished domain associated with µ-synthesis on the subspace of upper-triangular matrices in M₂(C). Its proof strategy is to study not H directly but its two raw siblings, the µ-hexablock Hμ = {π(A) : μ_hexa(A) < 1} and the normed hexablock H_N = {π(A) : ‖A‖ < 1}, where π(A) = (a₂₁,a₁₁,a₂₂,det A); unlike the older domains, Hμ and H_N are distinct and neither is open, yet the paper proves H = Hμ ∪ ({0}×E), that the closure of H equals the closure of Hμ and equals the polynomial convex hull of H_N, and that H = int(Hμ) = int(polynomial hull of H_N). The boundary theorem (Theorem 8.21) is the centre of the paper: a point lies in the distinguished boundary bH exactly when |a|²+|x₁|² = 1 and (x₁,x₂,x₃) ∈ bE, equivalently when it is a peak point, equivalently when it lies in H_N with |x₃| = 1. From this single description the paper derives the automorphism subgroup G(H), the rational H-inner functions, the Schwarz lemma, and the retraction theorems, and in Chapter 14 re-derives existing pentablock results as consequences.","pith_inferences":["The boundary formula bH = {(a,x₁,x₂,x₃) : |a|²+|x₁|² = 1, (x₁,x₂,x₃) ∈ bE} suggests H is the natural ambient domain for transferring function-theoretic rigidity: any invariant of the tetrablock fibre could be extended to H by averaging over the circle |a|²+|x₁|² = 1, an avenue the paper does not pursue.","Because the non-Levi-flat part of ∂E is dense (Theorem 9.10), the methods used to determine the pentablock automorphism group, which rely on Levi-flat boundary strata of the symmetrized bidisc, will not transfer directly; a complete description of Aut(H) will likely require new boundary invariants beyond the Levi form.","The strict containments H_N ⊊ Hμ ⊊ H and the fact that Hμ \\ H_N contains an open set mark the first instance in the M₂(C) family where the µ-unit ball and the norm unit ball push forward to genuinely different sets; this suggests that for larger upper-triangular subspaces of M_n(C) the gap between the two images will grow, and the 'domain of the µ-synthesis problem' should generally be read as the","Proposition 3.1's uniqueness claim is directly checkable by numerical maximization of |κ| over D², so the entire tower of results could be stress-tested independently of the paper's analytic proof; a cheap verification along the real slice E ∩ R³ would either confirm or destabilize the K* formalism."],"forward_implications":["Membership in H, Hμ, and H_N reduces to a single closed-form inequality: (x₁,x₂,x₃) ∈ E with |a|K*(x₁,x₂,x₃) < 1 for H, with ≤ 1 for Hμ, and with the two-sided bounds of Lemma 5.4 for H_N, so the interpolation data of the µ-synthesis problem are encoded in a bounded domain rather than in the norm-unbounded unit ball.","The distinguished boundary of H is exactly the set where |a|²+|x₁|² = 1 over the tetrablock's distinguished boundary — equivalently {( −ξ z, w, ξ w, ξ) : |z|²+|w|² = 1, ξ ∈ T} — giving a complete Shilov boundary, proving H is polynomially convex, and showing H_N is not polynomially convex.","Every automorphism of the tetrablock lifts to an automorphism of H, and these lifts together with their flip variants form a subgroup G(H) that preserves Hμ, H_N, the peak-point set H_p, and the three boundary strata ∂₀H, ∂₁H, ∂₂H (Theorems 7.4–7.12 and 8.2).","Rational H-inner functions are fully described by rational E-inner functions in the last three coordinates plus an explicit boundedness condition on the first coordinate, and this description yields a Schwarz lemma for two-point interpolation into H (Chapters 12–13).","The symmetrized bidisc, the tetrablock, the pentablock, and the biball are all analytic retracts of H, and the paper reproves several known pentablock results by projecting through H (Chapters 10 and 14)."],"supporting_citations":[{"why":"Defines the tetrablock E, its characterizations, its distinguished boundary, and the µ_tetra-synthesis link; E is the base domain over which the hexablock is built.","marker":"[2]"},{"why":"Introduces the pentablock and the one-parameter family ψ_z(a,s,p); the hexablock is constructed by generalizing this family to two disc parameters.","marker":"[4]"},{"why":"Provides the symmetrized bidisc characterizations and its distinguished boundary bΓ, which the paper uses to connect G₂ with the tetrablock and hexablock.","marker":"[11]"},{"why":"Introduces the symmetrized bidisc as the first domain arising from µ-synthesis in M₂(C), the template for the hexablock construction.","marker":"[7]"},{"why":"Gives the complete automorphism group of the tetrablock, from which the subgroup G(H) of hexablock automorphisms is constructed.","marker":"[92]"},{"why":"Characterizes rational E-inner functions, the base case the paper uses to write down all rational H-inner functions.","marker":"[14]"}],"fun_headline_variants":["Exact Shilov boundary found for new C^4 hexablock","Mu-synthesis in M2 yields hexablock with known boundary","Hexablock's boundary: |a|^2+|x1|^2=1 on tetrablock edge","C^4 hexablock: a domain with explicitly characterized Shilov boundary"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the supremum of the kernel |κ(z₁,z₂,x)| over the bidisc is attained at a single interior critical point, singled out by solving a quadratic and discarding the other root with a discriminant inequality; if some tetrablock point produced a second interior critical point (or the discarded root ever became the true maximizer), the function K* and the characterizations of Hμ, H_N, and H built on it would have to be revised.","fun_headline_variants_meta":{"raw":{"variants":["Exact Shilov boundary found for new C^4 hexablock","Mu-synthesis in M2 yields hexablock with known boundary","Hexablock's boundary: |a|^2+|x1|^2=1 on tetrablock edge","C^4 hexablock: a domain with explicitly characterized Shilov boundary"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001394,"raw_usage":{"total_tokens":5864,"prompt_tokens":1392,"completion_tokens":4472,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":1008,"completion_tokens_details":{"reasoning_tokens":4382}},"tokens_in":1008,"tokens_out":4472,"duration_ms":30886,"temperature":1.0,"reasoning_tokens":4382,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T19:41:55.985351+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a generic tetrablock point such as x = (0.3, 0.2, 0.05) in E ∩ R³, compute the two roots of the quadratic (3.7), and check numerically whether the discarded '−' root also lies in D² and gives the same value of |κ| as the '+' root; a second interior critical point at equal or greater height would break the uniqueness on which K* rests. The cheapest decisive test is to numerically maximize |κ(z₁,z₂,x)| over D² for a spread of x in E ∩ R³ and compare with the closed-form value |κ(z₁(x),z₂(x),x)| from Corollary 3.2; any mismatch refutes Proposition 3.1 and hence Theorem 8.21.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the complete automorphism group of the tetrablock, from which the subgroup G(H) of hexablock automorphisms is constructed."}],"review_version":2}