{"id":"f1775489-f384-44d6-94d6-a7af08dccfcd","arxiv_id":"2411.13102","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"New upper bounds are claimed for Hankel determinants and |a4|-|a3| in the univalent class S, but the proof of the a3=0 third Hankel determinant bound contains an algebraic error.","lead":"This paper derives sharper upper bounds for Hankel determinants and coefficient differences of univalent functions in the unit disk, under the extra conditions that either the second or third coefficient vanishes. The proofs use Grunsky coefficient inequalities, but one key algebraic step in the a3=0 case, Eq (23), is incorrect.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (23) in Theorem 5(v) is algebraically inconsistent: substituting the paper's own a4 and a5 into H3(1) = -a4^2 - a5 a2^2 gives -4*omega15^2 - 4*omega11^3*omega15 - 8*omega11^2*omega17, not -omega15^2 - ...; the bound 0.6647958756 is not established.","rationale":"The reader's strongest_claim and weakest_assumption pointed directly at Eq. (23); my independent expansion confirms that the missing factor of 4 is real. This is the load-bearing step because the abstract advertises improved third Hankel bounds for both a2=0 and a3=0 cases; Theorem 4(ii) handles the former, but Theorem 5(v) is the sole support for the latter. However, the defect is localized: the a3=0 formulas for a4 and a5 are correct, and the corrected triangle-inequality bound still appears to improve on [9], so the appropriate remedy is a conditional revision, exactly as the reader concluded. I did not find an additional independent objection in the remaining sections; the Grunsky-coefficient machinery and the other estimates are not machine-checked, but no further concrete error surfaced.","tokens_in":8051,"tokens_out":11181,"duration_ms":93710,"concrete_test":"Recompute Theorem 5(v) symbolically: substitute a4=2*omega15-5*omega11^3 and a5=2*omega17+6*omega11*omega15-(25/4)*omega11^4 into H3(1)=-a4^2-a5*a2^2 with a2=2*omega11, then maximize the resulting function over x=|omega11| in [0,1/2] using the bounds (18) and (21). If the coefficient of omega15^2 is 4 rather than 1, and the corrected maximum exceeds 0.6647958756, the printed theorem needs a revised numerical value and proof.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The advertised improvement in the a3=0 case is Theorem 5(v), and its proof hinges entirely on Eq. (23). With a3=0, Eq. (3) is H3(1) = -a4^2 - a5 a2^2, which is Eq. (22). Using the paper's own formulas, a2=2*omega11, a4=2*omega15-5*omega11^3 and a5=2*omega17+6*omega11*omega15-(25/4)*omega11^4, this becomes H3(1) = -(2*omega15-5*omega11^3)^2 - 4*omega11^2*(2*omega17+6*omega11*omega15-(25/4)*omega11^4) = -4*omega15^2 -4*omega11^3*omega15 -8*omega11^2*omega17. The printed Eq. (23) has -omega15^2 instead of -4*omega15^2. The subsequent F5 bound uses 1/5 where the corrected expression would require 4/5, so the computed maximum 0.6647958756 is not a consequence of the displayed proof. Re-running the same triangle-inequality estimates with the corrected coefficient gives a maximum on the order of 1.02; this would still improve the previous bound 1.114596 from [9] but does not support the printed numerical claim.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper derives new upper bounds for |a5| and |H3(1)| for functions in the class S of univalent functions on the unit disk, in the two special cases a2=0 and a3=0, together with an improved estimate for |a4|-|a3|. The proofs use the Grunsky coefficient representation of the initial coefficients and Grunsky's inequality, reducing the bounds to one- or two-variable extremal problems. Theorems 4 and 5 present the improved Hankel determinant bounds; Theorem 6 gives the coefficient-difference bound and an odd-function result.","tokens_in":8387,"tokens_out":34491,"duration_ms":247215,"significance":"If valid, the results are modest but genuine improvements over the previous bounds from [9], and the paper correctly identifies a misprint in [9] for |a5| when a2=0. The method is standard, and most of the derivations are straightforward. However, one of the advertised improvements, the bound for H3(1) in the a3=0 case, is invalid as written because of an algebraic error, so the paper's central claim needs substantial revision.","major_comments":[{"comment":"The expansion of H3(1) in Eq. (23) is algebraically incorrect. Substituting a2=2ω11, a4=2ω15-5ω11^3, and a5=2ω17+6ω11ω15-(25/4)ω11^4 (as derived in Eqs. (17) and (20)) into Eq. (22) gives H3(1)=-(2ω15-5ω11^3)^2-4ω11^2(2ω17+6ω11ω15-(25/4)ω11^4)=-4ω15^2-4ω11^3ω15-8ω11^2ω17, not the displayed -ω15^2-4ω11^3ω15-8ω11^2ω17. Consequently the function F5 and the claimed maximum 0.6647958756 are not consequences of the preceding equations. This invalidates the proof of Theorem 5(v), one of the advertised improvements in the abstract.","section":"Theorem 5(v), Eq. (23)"},{"comment":"The numerical maximizations are not rigorously justified. For the one-variable functions in Theorems 5(iii) and 5(v), the text merely states that \"the first derivative test shows\" the maximum, without presenting the derivative or proving that the stated critical point is the unique maximizer on the interval. For the two-variable function F6 in Theorem 6(i), the paper numerically solves a system to find a stationary point but does not prove that this exhausts all stationary points in D1, so the claim that F6 attains its maximum there is not established. Please provide complete derivative sign analyses or other rigorous arguments.","section":"Theorems 5(iii), 5(v), 6(i)"}],"minor_comments":[{"comment":"In the displayed inequality chain, the term \"≤ |ω15|+4|ω11||ω13|+|ω11|^3\" should read \"≤ 2|ω15|+4|ω11||ω13|+|ω11|^3\", and the following line should end with \"|ω11|^3\" rather than \"|ω13|^3\", to match the definition of F6(x,y).","section":"Theorem 6(i), proof"},{"comment":"Reference [8] appears incomplete: \"An improvement fo the Hankel determinants...\" lacks publication details.","section":"References"},{"comment":"The stray line \"*************************** DTS\" appearing before Theorem 6 should be removed.","section":"Before Theorem 6"},{"comment":"The notation \"ω11^2 = -2/3 ω13\" is an equality of complex numbers; the biconditional \"⇔\" is misleading and should be rephrased.","section":"Eq. (14)"}],"recommendation":"major_revision","confidential_remarks":"The algebraic error in Theorem 5(v) is the main obstacle. Since the corrected expression still leads to an improvement over the previous bound from [9], the paper may be salvageable; I would encourage the authors to recompute the maximum and adjust the abstract accordingly. The numerical-maximization rigor should also be strengthened."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Obradović and Tuneski improve several coefficient bounds in S using Grunsky coefficients. The a2=0 case (Theorem 4) gives |a5| ≤ 0.75 + 1/√7 and |H3(1)| ≤ 1.026, and the proof is clean. Theorem 6's improvement of |a4|−|a3| to 1.75185 also checks out as a genuine step forward. The a3=0 parts (i)–(iv) look correct.\n\nThe problem is Theorem 5(v). Equation (23) is algebraically inconsistent with the paper's own formulas. Substituting a2=2ω11, a4=2ω15−5ω11^3, a5=2ω17+6ω11ω15−(25/4)ω11^4 into H3(1) = −a4^2 − a5a2^2 gives −4ω15^2 −4ω11^3ω15 −8ω11^2ω17, not −ω15^2 −4ω11^3ω15 −8ω11^2ω17. The printed bound F5 uses the wrong coefficient (1/5 instead of 4/5), so the maximum 0.6647958756 is not proved. Re-running the triangle estimates with the corrected coefficient yields a bound around 1.02, which still beats the old 1.114596 from [9] but does not support the advertised number. The abstract's claim of improved third Hankel bounds in the a3=0 case is therefore not established as written.\n\nThis is a localized but load-bearing error. The rest of the derivations are standard chains of inequalities, and the numerical maximizations are reported without derivative checks, but that is typical in this area and the algebra is easy to verify. The paper belongs in a revision rather than a desk reject: the authors are clearly serious and the corrected bound, if reworked, would still be an improvement. I would send it to peer review with a note to check the algebra in Theorem 5(v) and recompute F5. Specialists in Hankel determinants for S will want to read the a2=0 and coefficient-difference parts, but should not trust Theorem 5(v)'s headline number until it is fixed.","headline":"A useful but flawed manuscript: the a2=0 bounds and the |a4|−|a3| improvement look right, but the advertised a3=0 third Hankel bound rests on an algebra error.","tokens_in":8848,"tokens_out":3317,"would_cite":false,"duration_ms":28374,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["30C45","30C50","30C55"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper improves the known upper bounds on the second and third Hankel determinants of univalent functions when either the quadratic or cubic coefficient vanishes, and also improves the bound on |a4|-|a3|.","keywords":["univalent functions","Grunsky coefficients","Hankel determinant","second Hankel determinant","third Hankel determinant","coefficient difference","coefficient estimates"],"falsifier":"Directly expand $H_3(1)=-a_4^2-a_5a_2^2$ using the paper's own formulas $a_2=2\\omega_{11}$, $a_4=2\\omega_{15}-5\\omega_{11}^3$, $a_5=2\\omega_{17}+6\\omega_{11}\\omega_{15}-\\frac{25}{4}\\omega_{11}^4$. The coefficient of $\\omega_{15}^2$ is $-4$, not $-1$, giving $-4\\omega_{15}^2-4\\omega_{11}^3\\omega_{15}-8\\omega_{11}^2\\omega_{17}$; maximizing $|\\cdot|$ over $|\\omega_{11}|\\le 1/2$ with the constraints (18) and (21) gives a number to compare with $0.6647958756\\ldots$.","tokens_in":7842,"feed_emoji":"📉","tokens_out":12697,"duration_ms":99324,"temperature":0.7,"pith_summary":"This paper studies the class $\\mathcal{S}$ of functions that are univalent (one-to-one) in the unit disk and expand as $f(z)=z+a_2z^2+a_3z^3+\\cdots$. The authors improve the known upper bounds for the second Hankel determinant $H_2(2)=a_2a_4-a_3^2$ and the third Hankel determinant $H_3(1)=a_3(a_2a_4-a_3^2)-a_4(a_4-a_2a_3)+a_5(a_3-a_2^2)$ in the two special cases where either $a_2=0$ or $a_3=0$. They also improve the bound on the coefficient difference $|a_4|-|a_3|$ for the whole class $\\mathcal{S}$, and give a new estimate for $|a_5|-|a_3|$ for odd functions. The bounds are not claimed to be sharp, but each narrows a long-open gap in the coefficient theory of univalent functions.","feed_headline":"Hankel bounds improve when a coefficient vanishes","feed_subtitle":"The a3=0 case cuts the third Hankel determinant bound from 1.114596 to 0.6647958756.","key_machinery":"The central object is the set of Grunsky coefficients $\\omega_{p,q}$, defined by $\\log\\frac{f(t)-f(z)}{t-z}=\\sum_{p,q\\ge0}\\omega_{p,q}t^p z^q$. For the odd square-root transform $f_2(z)=\\sqrt{f(z^2)}$, identities (6) express $a_2,\\dots,a_5$ in terms of $\\omega_{11},\\omega_{13},\\omega_{33},\\omega_{15},\\omega_{35},\\omega_{17}$, and the Grunsky inequality (4), truncated to (7), yields the size constraints (8) on the $\\omega$'s. Substituting the identities into $H_2(2)$, $H_3(1)$, and the coefficient differences reduces each desired bound to the maximum of an explicit single- or two-variable function $F_1,\\dots,F_6$ on a compact domain; the maxima are then located by the first derivative test.","core_discovery":"The paper proves several new estimates. For $f\\in\\mathcal{S}$ with $a_2=0$, it shows $|a_5|\\le \\frac34+\\frac1{\\sqrt7}=1.12796\\ldots$ and $|H_3(1)|\\le 1.026\\ldots$, improving the earlier $|a_5|\\le 1.508\\ldots$ and $|H_3(1)|\\le 2.05$. For $f\\in\\mathcal{S}$ with $a_3=0$, it shows $|a_2|\\le 1$, $|a_4|\\le \\frac14\\sqrt{\\frac{21}{5}}+\\frac58=1.1373\\ldots$, $|a_5|\\le 1.674896577\\ldots$, $|H_2(2)|\\le 1.1373\\ldots$, and $|H_3(1)|\\le 0.6647958756\\ldots$, the last improving the previous bound $1.114596\\ldots$. For general $f\\in\\mathcal{S}$, it proves $|a_4|-|a_3|\\le 1.75185\\ldots$, improving the previous $2.1033299\\ldots$, and for odd functions $|a_5|-|a_3|\\le 2/\\sqrt7=0.7559\\ldots$. All of these constants come from a single framework: translate the coefficients $a_2,\\ldots,a_5$ into Grunsky coefficients and maximize the resulting explicit functions.","pith_inferences":["The same reduction to a finite-dimensional optimization should extend to higher Hankel determinants $H_4(1)$ or $H_5(1)$, but the number of Grunsky coefficients and the algebraic complexity grow quickly; the $a_2=0$ and $a_3=0$ cases are the natural first test.","Because the derivative tests place the maxima of $F_2,F_3,F_4,F_5$ inside the intervals, the extremal functions are likely not Koebe-type maps; identifying them exactly would settle the sharpness of the new bounds.","The corrected $a_5$ identity may propagate: other papers that quoted the erroneous term $5\\omega_{15}^2$ instead of $5\\omega_{13}^2$ will need their estimates recalculated.","Using the full Grunsky inequality rather than the truncated form (7) could in principle sharpen the constraints (8) further, turning the finite optimization into an infinite-dimensional one."],"forward_implications":["For $a_2=0$, the third Hankel determinant bound drops from $2.05$ to $1.026\\ldots$, and the fifth-coefficient bound from $1.508\\ldots$ to $1.12796\\ldots$.","For $a_3=0$, $|H_3(1)|\\le 0.6647958756\\ldots$ replaces $1.114596\\ldots$, and $|H_2(2)|\\le 1.1373\\ldots$ replaces $1.75088\\ldots$.","The coefficient-difference bound $|a_4|-|a_3|\\le 1.75185\\ldots$ improves the best known $2.1033299\\ldots$ for $n=3$.","Odd functions satisfy $|a_5|-|a_3|\\le 2/\\sqrt7=0.7559\\ldots$, an improvement over the earlier $|a_5|-|a_3|<1$.","The corrected formula for $a_5$ in the Grunsky expansion removes a typo from the literature, so any result built on the old expression should be re-examined."],"supporting_citations":[{"why":"Along with [5], supplies the Grunsky inequality (4) used to derive the basic coefficient estimates.","marker":"[1]"},{"why":"Contains the best previous general constant for $||a_{n+1}|-|a_n||$, the baseline that Theorem 6(i) improves for $n=3$.","marker":"[3]"},{"why":"Supplies the Grunsky-coefficient identities (6) and the Grunsky inequality on which every bound in the paper rests.","marker":"[5]"},{"why":"Contains the estimate $|a_5|-|a_3|<1$ for odd functions that Theorem 6(ii) improves.","marker":"[6]"},{"why":"States the previous bounds for the $a_2=0$ and $a_3=0$ Hankel determinants and for $|a_4|-|a_3|$ that Theorems 4-6 improve.","marker":"[9]"}],"fun_headline_variants":["a3=0: Hankel bound cut to 0.6648","Zero coefficients sharpen Hankel determinant bounds","a2=0 or a3=0: new Hankel bounds","Coefficient gap bound improved to 1.75185","Grunsky method yields sharper univalent bounds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole argument hinges on the Grunsky-coefficient identities (6) and on the algebraic expansions derived from them, especially the expression (23) that converts $H_3(1)$ into a polynomial in $\\omega_{11},\\omega_{15},\\omega_{17}$; if those identities or the expansions are not correct, the improved bounds for the third Hankel determinant do not follow.","fun_headline_variants_meta":{"raw":{"variants":["a3=0: Hankel bound cut to 0.6648","Zero coefficients sharpen Hankel determinant bounds","a2=0 or a3=0: new Hankel bounds","Coefficient gap bound improved to 1.75185","Grunsky method yields sharper univalent bounds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001512,"raw_usage":{"total_tokens":6067,"prompt_tokens":959,"completion_tokens":5108,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":575,"completion_tokens_details":{"reasoning_tokens":5026}},"tokens_in":575,"tokens_out":5108,"duration_ms":70462,"temperature":1.0,"reasoning_tokens":5026,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T16:52:10.458474+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Directly expand $H_3(1)=-a_4^2-a_5a_2^2$ using the paper's own formulas $a_2=2\\omega_{11}$, $a_4=2\\omega_{15}-5\\omega_{11}^3$, $a_5=2\\omega_{17}+6\\omega_{11}\\omega_{15}-\\frac{25}{4}\\omega_{11}^4$. The coefficient of $\\omega_{15}^2$ is $-4$, not $-1$, giving $-4\\omega_{15}^2-4\\omega_{11}^3\\omega_{15}-8\\omega_{11}^2\\omega_{17}$; maximizing $|\\cdot|$ over $|\\omega_{11}|\\le 1/2$ with the constraints (18) and (21) gives a number to compare with $0.6647958756\\ldots$.","supporting_citations":[{"cited_title":"De Branges, A proof of the Bieberbach conjecture, Acta Math","cited_arxiv_id":null,"evidence_quote":"Along with [5], supplies the Grunsky inequality (4) used to derive the basic coefficient estimates."},{"cited_title":"o, Eine Bemerkung \\","cited_arxiv_id":null,"evidence_quote":"Contains the best previous general constant for $||a_{n+1}|-|a_n||$, the baseline that Theorem 6(i) improves for $n=3$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Grunsky-coefficient identities (6) and the Grunsky inequality on which every bound in the paper rests."},{"cited_title":"Lebedev, Area principle in the theory of univalent functions, Published by\"Nauka\", Moscow, 1975 (in Russian)","cited_arxiv_id":null,"evidence_quote":"Contains the estimate $|a_5|-|a_3|<1$ for odd functions that Theorem 6(ii) improves."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the previous bounds for the $a_2=0$ and $a_3=0$ Hankel determinants and for $|a_4|-|a_3|$ that Theorems 4-6 improve."}],"review_version":1}