{"id":"320e0ccc-946d-4570-86b7-240fb527a7c1","arxiv_id":"2411.13285","paper_version":1,"verdict":"REJECT","confidence":"HIGH","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"The claimed exact LYJ constant for the Banas-Fraczek space is false; explicit unit vectors exceed the stated value.","lead":"This paper claims an exact formula for a geometric constant of a two-dimensional normed space called the Banas-Fraczek space. The formula is wrong: a concrete pair of unit vectors gives a value larger than the formula allows.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed exact value is false: for λ=6/5, ξ=1, η=3 the unit vectors (4/5,3/5) and (5/6,-√11/6) give LYJ quotient 1.18358 > 71/60; Lemma 2.2's corner maximum is also violated.","rationale":"The reader's rejection is sustained. I re-derived the critical step and found the same weak spot: Lemma 2.2(a) asserts a corner maximum that is numerically false. More importantly, the claimed theorem is contradicted directly by an admissible pair of unit vectors in the Banaś-Frączek space. With λ = 6/5, ξ = 1, η = 3, the vectors x = (4/5, 3/5) and y = (5/6, -√11/6) both lie on the unit sphere: x satisfies x1² + x2² = 1 and |x1| = 4/5 ≤ 1/λ, and y satisfies y1² + y2² = 1 and |y1| = 1/λ. A direct norm computation gives an LYJ quotient of about 1.18358, which exceeds the claimed 71/60 by about 2.45 × 10⁻⁴. This is an internal inconsistency with the paper's own definitions, not a matter of disagreeing with an external consensus. The proof's use of Lemma 2.2 is therefore not merely inefficient; the conclusion it is meant to establish is false. I would keep the reader's REJECT verdict; the counterexample is a clean falsification of the central claim.","tokens_in":8807,"tokens_out":19058,"duration_ms":193245,"concrete_test":"Independently compute the LYJ quotient for λ = 6/5, ξ = 1, η = 3 with unit vectors x = (4/5, 3/5) and y = (5/6, -√11/6), using the norm ||(a,b)|| = max(6|a|/5, √(a²+b²)). The quotient is (13551 + 375√11)/12500 ≈ 1.18358, while Theorem 2.1 gives 1 + (2ξη/(ξ²+η²))(1 - 1/λ²) = 71/60 ≈ 1.18333. Since 1125√11 > 3722, the quotient is strictly larger, so the theorem fails as stated.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The load-bearing point is the upper-bound half of Theorem 2.1. For 1 ≤ λ < √2, the proof of inequality (1) depends entirely on Lemma 2.2, which asserts that f and g are maximized at (1/λ, 1/λ). That assertion is false: with λ = 6/5, ξ = 1, η = 3, t = 1/2, the function f from Lemma 2.2(a) satisfies f(5/6, 7/10) ≈ 11.1308 > f(5/6, 5/6) ≈ 11.0833. Thus inequalities (5) and (6), and hence (7), are not justified. The failure is not merely a proof gap: direct evaluation of the defining LYJ quotient at the same λ, ξ, η with the unit vectors x = (4/5, 3/5) and y = (5/6, -√11/6) gives (13551 + 375√11)/12500 ≈ 1.18358, which exceeds the theorem's claimed value 71/60 ≈ 1.18333. The lower-bound half of the theorem is fine, but the exact-value formula is contradicted by the paper's own definition.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the LYJ constant L_YJ(ξ,η,X) for the Banas-Fraczek space R²_λ, defined by the norm ‖(a,b)‖ = max{λ|a|, √(a²+b²)}. The main result, Theorem 2.1, asserts that for every λ ≥ 1, L_YJ(ξ,η,R²_λ) equals 1 + (2ξη/(ξ²+η²))(1 − 1/λ²). The proof proceeds by a case analysis and relies on Lemma 2.2, which claims that two auxiliary functions f and g on [0,1/λ]² attain their maxima at the corner (1/λ,1/λ). The paper also derives corollaries on the von Neumann–Jordan constant, super-reflexivity, and weak normal structure.","tokens_in":9074,"tokens_out":6818,"duration_ms":53636,"significance":"An exact formula for a geometric constant of a concrete family of norms would be a useful contribution to the literature on geometric constants in Banach spaces, and the lower-bound construction in equations (8)–(9) is valid and does establish the claimed lower bound. However, the upper-bound proof depends entirely on Lemma 2.2, which is false, and the claimed formula is directly contradicted by an explicit numerical example using the paper's own definitions. Since the central result is incorrect, the paper cannot be accepted in its current form.","major_comments":[{"comment":"Lemma 2.2 is false. For λ = 6/5, t = 1/2, ξ = 1, η = 3, the function f defined in part (a) satisfies f(5/6, 7/10) ≈ 11.1308 > f(5/6, 5/6) ≈ 11.0833, contradicting the claimed corner maximum. The proof's inference that the absence of an interior critical point forces the maximum to occur at one of the four corners is invalid; the maximum can occur at a non-corner boundary point, as this example shows. Consequently, inequalities (5) and (6) in the proof of Theorem 2.1 are not justified.","section":"Section 2, Lemma 2.2"},{"comment":"The algebraic step after multiplying equations (i) and (ii) is erroneous: from the displayed equality involving tξ²η² and positive terms, the manuscript claims equivalence to 0 = 2tξηλ² + ξ²λ² x/y + t²λ²η² y/x. The right-hand side is a sum of positive terms, so it cannot be zero. The claimed contradiction is therefore not established, and the exclusion of interior maxima is unsupported.","section":"Section 2, proof of Lemma 2.2"},{"comment":"Theorem 2.1 is false. For λ = 6/5, ξ = 1, η = 3, the unit vectors x = (5/6, √11/6) and y = (−4/5, 3/5) in R²_λ satisfy ‖x‖ = ‖y‖ = 1, and the LYJ quotient (‖ξx + ηy‖² + ‖ηx − ξy‖²)/((ξ² + η²)(‖x‖² + ‖y‖²)) equals (13551 + 375√11)/12500 ≈ 1.18358, which strictly exceeds the claimed value 1 + (2ξη/(ξ²+η²))(1 − 1/λ²) = 71/60 ≈ 1.18333. This is a direct contradiction to the main formula using the paper's own definition of the constant.","section":"Section 2, Theorem 2.1"}],"minor_comments":[{"comment":"The displayed value of f(1/λ,1/λ) is inconsistent between the statement and the proof: the statement gives 4tξη·(λ²−1)/λ² + t²ξ² + η², while the proof's comparison uses 4tξη·(λ²−1)/λ + t²ξ² + η².","section":"Lemma 2.2"},{"comment":"The expression for ‖ηx − ξty‖² contains a typesetting error: (ηx₁ − ξty₁)²(ηx₂ − ξty₂)² should be (ηx₁ − ξty₁)² + (ηx₂ − ξty₂)².","section":"Theorem 2.1, Case(3)"},{"comment":"The term '2tξηsin √(1−x²)√(1−y²)' should be '2tξη√(1−x²)√(1−y²)'.","section":"Theorem 2.1, equations (5) and (6)"},{"comment":"The abstract contains the typo 'meticilous'; the phrase 'Bana´s-Fr ˛ aczek' is also hyphenated inconsistently throughout.","section":"Abstract"}],"recommendation":"reject","confidential_remarks":"The counterexamples in the major comments are straightforward to verify and use only the paper's own definitions. The failure of Lemma 2.2 is not a minor gap but a false statement, and the main theorem is contradicted by an explicit computation. The lower-bound part of the proof appears correct, but the upper bound fails. The algebraic error in the proof of Lemma 2.2 suggests the lemma's derivation was not checked carefully."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The central claim is false. For λ=6/5, ξ=1, η=3, take x=(5/6, √11/6) and y=(-4/5, 3/5). Both are unit vectors in R²_λ. The LYJ quotient evaluates to (13551 + 375√11)/12500 ≈ 1.18358, which is strictly larger than the claimed 71/60 ≈ 1.18333. So Theorem 2.1 is not just unproved; it is contradicted by the paper's own definition.\n\nWhat is new: the exact formula for LYJ(ξ, η, R²_λ) does not appear in the earlier literature, and the lower-bound half (the choice x=(1/λ, √(1−1/λ²)), y=(1/λ, −√(1−1/λ²))) correctly attains the claimed value. The general strategy—exploiting the extreme points of the unit ball and splitting into cases according to which term in the max norm dominates—is a sensible approach that has worked for related constants.\n\nWhere it breaks: Lemma 2.2, the load-bearing upper-bound estimate for 1≤λ<√2, is false. The proof concludes that f cannot have an interior maximum because a certain equation forces 0 to equal a positive sum; that equation is the result of an algebraic error, not a property of f. Worse, the corner-max claim is numerically false: with λ=6/5, ξ=1, η=3, t=1/2, f(5/6, 7/10) ≈ 11.1308 > f(5/6, 5/6) ≈ 11.0833. Consequently inequalities (5)–(7) are unjustified. There are also typographical slips throughout, e.g. \"meticilous\", an undefined \"sin\" in (5)–(6), and an incomplete Case(3) expression in the λ≥√2 branch.\n\nThe reason this matters is not pedantry. The upper bound is the whole point of an exact value; once it fails, the paper contributes only a lower bound, which was already known for the special case ξ=η=1 from the CNJ computation.\n\nI would not send this to referees. A single explicit counterexample from the defining formula is enough to close the case. The authors should re-derive the maximum, and if there is a correct formula it will be different from the one stated. The paper is not a serious contribution in its current form.","headline":"The paper's central formula is false: explicit unit vectors give a larger LYJ quotient than the claimed exact value, and Lemma 2.2 is numerically violated.","tokens_in":9610,"tokens_out":3887,"would_cite":false,"duration_ms":35842,"reading_group":"no","serious_thinker":"no","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46B20"],"pacs":[],"model":"deepseek-v4-flash","headline":"Exact LYJ constant of Banas-Fraczek space computed","keywords":["Banach spaces","geometric constants","Banas-Fraczek space","LYJ constant","von Neumann-Jordan constant","super-reflexive spaces","weak normal structure"],"falsifier":"Check the Case(2) expression at $\\lambda=6/5$, $\\xi=1$, $\\eta=3$, $t=1/2$, $x_1=5/6$, $y_1=7/10$: the auxiliary function $f$ gives $f(5/6,7/10)\\approx 11.1308$, larger than the corner value $f(5/6,5/6)\\approx 11.0833$ used in Lemma 2.2. Since that lemma is exactly what produces inequalities (5) and (6), the discrepancy invalidates the proof's $1\\le\\lambda<\\sqrt2$ step; testing the corresponding full ratio against the claimed formula then settles whether Theorem 2.1 itself is correct.","tokens_in":8584,"feed_emoji":"📐","tokens_out":12266,"duration_ms":100633,"temperature":0.7,"pith_summary":"This paper sets out to compute the exact value of the $L_{\\mathrm{YJ}}(\\xi,\\eta,X)$ geometric constant for the Banas-Fraczek space $R_\\lambda^2$, the normed plane with norm $\\|(a,b)\\|=\\max\\{\\lambda|a|,\\sqrt{a^2+b^2}\\}$. Its central claim is that for every $\\lambda\\ge 1$ and positive $\\xi,\\eta$, the constant equals $1+\\frac{2\\xi\\eta}{\\xi^2+\\eta^2}\\left(1-\\frac{1}{\\lambda^2}\\right)$. If true, the formula gives a direct geometric handle on the space: it interpolates from the Euclidean value $1$ at $\\lambda=1$ toward $1+\\frac{2\\xi\\eta}{\\xi^2+\\eta^2}$ as $\\lambda\\to\\infty$, and it keeps the constant below $2$, which implies super-reflexivity. The paper also reads off the von Neumann-Jordan constant at $\\xi=\\eta$ and a weak-normal-structure condition.","feed_headline":"Exact LYJ constant of Banas-Fraczek space computed","feed_subtitle":"The formula 1 + (2ξη/(ξ²+η²))(1 − 1/λ²) links James-type constants, von Neumann-Jordan constants, and super-reflexivity.","key_machinery":"The central objects are the Banas-Fraczek norm and the quotient defining the constant: for nonzero $x,y$, the ratio is $\\frac{\\|\\xi x+\\eta y\\|^2+\\|\\eta x-\\xi y\\|^2}{(\\xi^2+\\eta^2)(\\|x\\|^2+\\|y\\|^2)}$, and $L_{\\mathrm{YJ}}$ is its supremum. The workhorse is a four-case split depending on whether $\\lambda|a|$ or the Euclidean part dominates each vector, which reduces the quotient to expressions in $|x_1y_1|$ and $\\sqrt{1-x_1^2}\\sqrt{1-y_1^2}$. Lemma 2.1 bounds the cross term for $\\lambda\\ge\\sqrt{2}$, and Lemma 2.2 claims that two auxiliary functions $f$ and $g$ on $[0,1/\\lambda]^2$ peak at the corner $(1/\\lambda,1/\\lambda)$, supplying the upper bound in the range $1\\le\\lambda<\\sqrt2$. The lower-bound direction uses $x=(1/\\lambda,\\sqrt{1-1/\\lambda^2})$, $y=(1/\\lambda,-\\sqrt{1-1/\\lambda^2})$, for which the quotient equals the claimed formula.","core_discovery":"On the paper's own terms, the discovery is Theorem 2.1: for $R_\\lambda^2=(\\mathbb{R}^2,\\|\\cdot\\|_\\lambda)$ with $\\|(a,b)\\|_\\lambda=\\max\\{\\lambda|a|,\\sqrt{a^2+b^2}\\}$ and $\\lambda\\ge 1$, the constant $L_{\\mathrm{YJ}}(\\xi,\\eta,R_\\lambda^2)$ equals $1+\\frac{2\\xi\\eta}{\\xi^2+\\eta^2}\\left(1-\\frac{1}{\\lambda^2}\\right)$. The proof splits the unit sphere into cases according to which branch of the norm controls each of the two vectors $\\xi x+\\eta t y$ and $\\eta x-\\xi t y$. An upper bound is obtained by applying Lemma 2.1 for $\\lambda\\ge\\sqrt{2}$ and Lemma 2.2 for $1\\le\\lambda<\\sqrt{2}$, and the lower bound is attained by the two extreme points $\\left(\\frac{1}{\\lambda},\\pm\\sqrt{1-\\frac{1}{\\lambda^2}}\\right)$. Corollaries identify the $\\xi=\\eta$ case with the von Neumann-Jordan constant and conclude super-reflexivity.","pith_inferences":["A direct numerical stress-test targets the proof's bottleneck: at $\\lambda=6/5$, $\\xi=1$, $\\eta=3$, $t=1/2$, the auxiliary function $f$ in Lemma 2.2 takes value about $11.1308$ at $(x_1,y_1)=(5/6,7/10)$, exceeding the corner value $11.0833$ asserted by the lemma.","The same case-splitting strategy could be tried on the generalized spaces $X_{\\lambda,p}$ with norm $\\max\\{\\lambda|a|,(|a|^p+|b|^p)^{1/p}\\}$, replacing the Euclidean cross-term identities by $\\ell^p$ analogues.","Because the formula depends on $\\xi,\\eta$ only through $2\\xi\\eta/(\\xi^2+\\eta^2)$, it suggests that for any two-dimensional norm whose unit ball is cut by vertical strips, the $L_{\\mathrm{YJ}}$ constant may obey a one-parameter interpolation between the Euclidean value and the anisotropic limit."],"forward_implications":["For $\\xi=\\eta$, the theorem gives $L_{\\mathrm{YJ}}(1,1,R_\\lambda^2)=2-\\frac{1}{\\lambda^2}$, recovering the known von Neumann-Jordan constant of the Banas-Fraczek space.","Because the claimed value is always below $2$ for $\\lambda\\ge 1$, the space $R_\\lambda^2$ is super-reflexive under the criterion supplied by the $L_{\\mathrm{YJ}}$ constant.","Taking $\\lambda=1$ makes the norm Euclidean and the formula collapses to $L_{\\mathrm{YJ}}(\\xi,\\eta,R_1^2)=1$, the minimal possible value of the constant.","For $\\eta\\le\\xi<\\frac32\\eta$ and $1\\le\\lambda<\\sqrt{4\\xi\\eta/(6\\xi\\eta-3\\eta^2)}$, the claimed value falls below the threshold in Lemma 2.3, so the space has weak normal structure."],"supporting_citations":[{"why":"defines the Banas-Fraczek space and its norm, the object whose LYJ constant is computed.","marker":"[5]"},{"why":"supplies Lemma 2.1, the cross-term estimate needed for $\\lambda\\ge\\sqrt{2}$, and the known von Neumann-Jordan constant $2-1/\\lambda^2$.","marker":"[8]"},{"why":"introduces the $L_{\\mathrm{YJ}}(\\xi,\\eta,X)$ constant and the super-reflexivity criterion used in Corollary 2.3.","marker":"[13]"}],"fun_headline_variants":["Exact LYJ constant for Banas-Fraczek space derived","Closed-form LYJ constant in Banas-Fraczek space achieved","LYJ constant of Banas-Fraczek space now explicitly known","Formula for LYJ constant in Banas-Fraczek space proved","Banas-Fraczek space: exact LYJ constant calculated"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proof for $1\\le\\lambda<\\sqrt2$ rests on Lemma 2.2, which asserts that two auxiliary functions on $[0,1/\\lambda]^2$ reach their maximum at the corner $(1/\\lambda,1/\\lambda)$; if that corner-maximum claim fails, the upper-bound inequalities (5) and (6) do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Exact LYJ constant for Banas-Fraczek space derived","Closed-form LYJ constant in Banas-Fraczek space achieved","LYJ constant of Banas-Fraczek space now explicitly known","Formula for LYJ constant in Banas-Fraczek space proved","Banas-Fraczek space: exact LYJ constant calculated"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000281,"raw_usage":{"total_tokens":1659,"prompt_tokens":932,"completion_tokens":727,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":548,"completion_tokens_details":{"reasoning_tokens":637}},"tokens_in":548,"tokens_out":727,"duration_ms":7226,"temperature":1.0,"reasoning_tokens":637,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:39:38.270415+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check the Case(2) expression at $\\lambda=6/5$, $\\xi=1$, $\\eta=3$, $t=1/2$, $x_1=5/6$, $y_1=7/10$: the auxiliary function $f$ gives $f(5/6,7/10)\\approx 11.1308$, larger than the corner value $f(5/6,5/6)\\approx 11.0833$ used in Lemma 2.2. Since that lemma is exactly what produces inequalities (5) and (6), the discrepancy invalidates the proof's $1\\le\\lambda<\\sqrt2$ step; testing the corresponding full ratio against the claimed formula then settles whether Theorem 2.1 itself is correct.","supporting_citations":[{"cited_title":"J AND FR ˛ ACZEK","cited_arxiv_id":null,"evidence_quote":"defines the Banas-Fraczek space and its norm, the object whose LYJ constant is computed."},{"cited_title":"Y ANG AND Q","cited_arxiv_id":null,"evidence_quote":"supplies Lemma 2.1, the cross-term estimate needed for $\\lambda\\ge\\sqrt{2}$, and the known von Neumann-Jordan constant $2-1/\\lambda^2$."},{"cited_title":"L IU AND Y","cited_arxiv_id":null,"evidence_quote":"introduces the $L_{\\mathrm{YJ}}(\\xi,\\eta,X)$ constant and the super-reflexivity criterion used in Corollary 2.3."}],"review_version":1}