{"id":"76849449-96f4-4755-8c30-6ab027f9afe7","arxiv_id":"1908.08653","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper derives explicit formulas for LMOV invariants of the framed unknot, but the multi-hole formula (29) is wrong as written.","lead":"This paper derives closed-form formulas for LMOV invariants, integer quantities in open topological string theory, for the resolved conifold with a framed brane. It uses mirror curves and BKMP recursion to extract disc, annulus, and higher-hole invariants, but one of the advertised formulas is internally inconsistent.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Equation (29) uses the wrong Möbius-inversion exponent: formula (7) gives d^{l(μ)-3}, not d^{l(μ)-1}, so the multi-hole LMOV formula is internally inconsistent.","rationale":"The reader's rejection is justified, and the most load-bearing problem is the internal inconsistency of equation (29): it is claimed to follow from formula (7), but the Möbius inversion is performed with the wrong power of d. This directly invalidates one of the four advertised explicit formulae and the integrality statement attached to it. I did not select the annulus a=0 reduction flagged as the reader's weakest assumption as the primary concern, because that step is a gap that might be repairable; the inversion error is a demonstrable algebraic contradiction within the paper itself. The concrete calculation with μ=(2,2,2), τ=1 isolates the mismatch without relying on any external conjecture, using only the paper's own formulas (7) and (28). Since the central claim includes (29) as a headline result, the paper cannot be accepted in its present form. The reader's verdict of rejection therefore stands unchanged.","tokens_in":21829,"tokens_out":7567,"duration_ms":76350,"concrete_test":"Take μ=(2,2,2), τ=1, so l(μ)=3. Using equation (28), compute K_{(2,2,2),0,3}=108 and K_{(1,1,1),0,3/2}=-4. The correct inversion of (7) gives n_{(2,2,2),0,3} = -[K_{(2,2,2),0,3} + μ(2)K_{(1,1,1),0,3/2}] = -[108 + (-1)(-4)] = -112. Equation (29) gives -[K_{(2,2,2),0,3} + μ(2)·2^2·K_{(1,1,1),0,3/2}] = -[108 + (-1)·4·(-4)] = -124. The mismatch -112 vs -124 demonstrates that replacing d^{l(μ)-3} by d^{l(μ)-1} changes the result; re-deriving (29) from (7) with the correct exponent restores consistency.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive flaw is in §4.4. After deriving K^τ_{μ,0,|μ|/2} in equation (28), the paper states that 'By using formula (7), we get' equation (29): n_{μ,0,|μ|/2}(τ) = (-1)^{l(μ)} ∑_{d|μ} μ(d) d^{l(μ)-1} K^τ_{μ/d,0,|μ|/(2d)}. But formula (7) says K_{μ,0,Q} = ∑_{d|μ} (-1)^{l(μ)} d^{l(μ)-3} n_{μ/d,0,Q/d}. Since dividing a partition by d preserves its length, the standard Dirichlet inversion of (7) gives n_{μ,0,Q} = (-1)^{l(μ)} ∑_{d|μ} μ(d) d^{l(μ)-3} K_{μ/d,0,Q/d}. The exponent l(μ)-1 in (29) is the factor appearing in the unexpanded multiple-covering formula (6) before the factor (2 sin(dg_s/2))^{-2} is expanded; that expansion is precisely what produces the shift from l-1 to l-3. The error is not cosmetic: for l(μ)=3 the correct factor is d^0, while (29) inserts d^2, and the difference changes the numerical value of the invariants. Consequently the multi-hole formula (29), and the integrality conclusion drawn from it, are unsupported as written.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies open-string LMOV invariants for the resolved conifold with one Aganagic-Vafa brane in integer framing, i.e. for the framed unknot in S^3. Using the mirror curve obtained from the q-holonomic recursion for colored HOMFLY-PT invariants, the author derives explicit formulas for genus-zero one-hole (disc) invariants by Lagrange inversion, for genus-zero two-hole (annulus) invariants from the Bergmann kernel, for genus-zero multi-hole invariants at Q=|mu|/2 from a Hodge integral formula, and for higher-genus one-hole invariants from the Chern-Simons partition function via the large-N duality proved by Zhou [79]. The stated results are the disc formula (24), the annulus formula (27), the multi-hole formula (29), and the generating-function formula (33); integrality of the disc and annulus invariants is asserted to be proved in the companion paper [56].","tokens_in":22105,"tokens_out":24143,"duration_ms":228670,"significance":"If the formulas are correct, the paper provides closed-form expressions for a substantial class of LMOV invariants of the framed unknot in arbitrary integer framing, giving explicit data that could be used to test structural conjectures such as the Gaussian-distribution conjecture of Mironov-Morozov-Sleptsov [61]. The derivations are direct rather than fitting-based: the disc invariants are obtained from the mirror curve by Lagrange inversion, the annulus coefficient is read from a Bergmann-kernel expansion, and the higher-genus one-hole invariants are extracted from the Chern-Simons free energy by Möbius inversion. These are real strengths. However, the paper is not self-contained: the key integrality statements are imported from [56], Lemma 4.5 is imported from [80], and the multi-hole formula as displayed contains a Möbius-inversion error that changes the numerical values of the invariants. The annulus derivation also contains an unproved reduction from the full mirror curve to its a=0 part. Because these points affect the central formulas, the paper needs substantial revision before the claims can be accepted.","major_comments":[{"comment":"The Möbius inversion in Eq. (29) uses the wrong power of d. Formula (7) states K_{\\mu,0,Q} = \\sum_{d|\\mu} (-1)^{l(\\mu)} d^{l(\\mu)-3} n_{\\mu/d,0,Q/d}. Since l(\\mu/d)=l(\\mu), inverting this relation gives n_{\\mu,0,Q} = (-1)^{l(\\mu)} \\sum_{d|\\mu} \\mu(d) d^{l(\\mu)-3} K_{\\mu/d,0,Q/d}. The exponent l(\\mu)-1 appearing in Eq. (29) is the power that appears in the unexpanded multiple-covering formula (6) before (2\\sin(dg_s/2))^{-2} is expanded; after that expansion the correct power is l(\\mu)-3. The difference is numerically significant: for l(\\mu)=3 the correct factor is d^0, while Eq. (29) inserts d^2. Thus the displayed multi-hole formula (29) is incorrect as written. The integrality conclusion n_{\\mu,0,|\\mu|/2}\\in\\mathbb{Z} still follows from Eq. (28) with the corrected exponent, but the explicit formula and the statements in the abstract that rely on it must be revised.","section":"§4.4, Eq. (29)"},{"comment":"The reduction 'when l=(m1+m2)/2, we only need to consider the curve Y=X(1-Y)^\\tau' is asserted without proof. In the full mirror curve (20), Y = X(1-Y)^\\tau(1-a(1-Y)), and the coefficient of a^{(m1+m2)/2} in the expansion of \\log((Y_1-Y_2)/(X_1-X_2)) can in principle receive contributions from several b_{n,i} factors whose a-degrees sum to (m1+m2)/2. No degree bound or cancellation argument is given to justify keeping only the a=0 curve. Since formula (27) is obtained precisely from this reduction together with Lemma 4.5, the annulus formula and the resulting invariants n_{(m1,m2)}(\\tau) are not established as written. The authors need to prove the reduction, or alternatively compute the coefficient from the full curve and verify that the higher-a terms cancel.","section":"§4.3"}],"minor_comments":[{"comment":"The index conventions around b_{n,i} and \\tilde b_{m,l} are unclear: the displayed b_{n,i} appears to be the coefficient of X^{n+1} rather than X^n, and the definition \\tilde b_m = \\sum_l \\tilde b_{m,l} a^l with \\tilde b_{m,l}=\\sum_{i=0}^l b_{m,i} seems to introduce an extra factor 1/(1-a). Please clarify these definitions and check the resulting expansion.","section":"§4.3"},{"comment":"After correcting Eq. (29), the sentence 'since l(\\mu)\\ge 3, it is clear that n...' should be reworded, because the integrality argument with the corrected exponent uses the same integrality of K but the displayed formula changes.","section":"§4.4"},{"comment":"There are several typos and infelicities: 'Agangica-Vafa' in the Introduction, 'Revist' in the heading of §4.5.1, 'finial' in the Conclusions, and 'Möbius' is sometimes written without the umlaut. The reference [28] is incomplete as printed.","section":"Throughout"},{"comment":"Formula (33) is an algorithmic coefficient-extraction formula rather than a closed form; the paper should state clearly that for higher genus the result is a generating-function characterization, not an explicit binomial-type expression.","section":"§4.5.2"}],"recommendation":"major_revision","confidential_remarks":"The paper relies heavily on the author's companion work [56] for the proofs of the integrality claims and on [80] for the annulus lemma; the referee report should ask the author to state explicitly which statements are proved here and which are imported. There is no indication of a fitting-based circularity, but the Möbius-inversion error in Eq. (29) and the unproved reduction in §4.3 are load-bearing and need to be fixed or justified before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper has a real, load-bearing error. Equation (29) states n_{μ,0,|μ|/2}(τ) = (-1)^{l(μ)} ∑_{d|μ} μ(d) d^{l(μ)-1} K^τ_{μ/d,0,|μ|/(2d)}, but inverting the paper's own formula (7) gives d^{l(μ)-3}, not d^{l(μ)-1}. The stress-test note is right: the l(μ)-1 exponent is what appears before expanding the sine factor in (6); the expansion shifts it to l(μ)-3. This is not cosmetic—for l(μ)=3 the correct factor is d^0, while (29) inserts d^2, changing the invariants. So the multi-hole formula and the integrality conclusion drawn from it are unsupported as written.\n\nWhat the paper does well: the mirror-curve computation for the disc amplitude is coherent and self-contained. The Lagrangian inversion in Section 4.2 is clean, and the paper honestly notes that the disc formula is a special case of Panfil-Sulkowski. The annulus formula (27) follows from a cited lemma, and the higher-genus generating function (32)–(33) is taken from the author's previous paper [56]. The literature review is careful, and the paper is honest about what is imported.\n\nWhere the soft spots are, in proportion: (1) the (29) error is fatal for that formula; (2) Section 4.3's assertion that for l=(m1+m2)/2 one can reduce to the a=0 curve is asserted without proof—this is a load-bearing modeling step for the annulus formula; (3) the paper's genuinely new content is the derivation route, not the formulas themselves, since all four advertised formulas already appear in [56] or [69]. The integrality of the disc and annulus invariants is also cited from the author's own prior work, so the paper is more a synthesis-with-new-proof than a standalone source of new results.\n\nWho this is for: specialists in topological string integrality who want to see the mirror-curve/BKMP route to these formulas. It is not a paper to cite for the formulas themselves, since they are already published. A corrected version—fixing (29), justifying the a=0 reduction, and clearly separating new derivations from restatements—would be a useful contribution.\n\nRecommendation: send to peer review. The error is specific, fixable, and likely to be caught by a referee; the paper is technically serious and addresses a topic of interest. But it should not be accepted without the correction and the justification for the annulus reduction.","headline":"The paper's advertised multi-hole formula (29) inverts its own equation (7) with the wrong exponent, and the annulus reduction to the a=0 curve is unproved; the clean disc-counting derivation does not rescue the paper as it stands.","tokens_in":22694,"tokens_out":4337,"would_cite":false,"duration_ms":41819,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Closed-form formulas give the LMOV invariants of the framed unknot for every integer framing.","keywords":["LMOV invariants","framed unknot","resolved conifold","AV-brane","mirror curve","Bergmann kernel","topological strings","Chern-Simons theory"],"falsifier":"Take the full mirror curve $Y=X(1-Y)^{\\tau}(1-a(1-Y))$, compute the Bergmann-kernel coefficient of $X_1^{m_1}X_2^{m_2}a^{(m_1+m_2)/2}$ in $\\log((Y_2-Y_1)/(X_2-X_1))$ for small values such as $(m_1,m_2,\\tau)=(1,1,0)$, $(1,2,1)$, and $(2,2,-1)$, and compare with formula (27). Any nonzero contribution from the $a$-dependent factor at that degree would overturn the annulus formula; equality would support the reduction.","tokens_in":21552,"feed_emoji":"🧮","tokens_out":9628,"duration_ms":81618,"temperature":0.7,"pith_summary":"This paper derives closed-form expressions for the Labastida–Mariño–Ooguri–Vafa (LMOV) invariants of the resolved conifold with one framed brane, equivalently the framed unknot in $S^3$. The point is that, although open-string integrality invariants are usually hard to compute, for this model each sector collapses to explicit sums of binomial coefficients and Möbius functions. A reader should care because these integers count BPS domain walls, and explicit formulas make their integrality, growth, and distribution directly checkable. The formulas cover the disc (one hole, genus 0), the annulus (two holes, genus 0), multi-hole genus-0 invariants at degree $|\\mu|/2$, and higher-genus one-hole invariants through a single polynomial $g_m(q,a)$.","feed_headline":"One mirror curve yields all LMOV integers of the framed unknot","feed_subtitle":"Disc, annulus, multi-hole, and higher-genus invariants all become explicit binomial sums with Möbius inversion, valid for any framing.","key_machinery":"The machinery is the mirror curve of the framed unknot, obtained from the noncommutative $a$-deformed $A$-polynomial as $y-1-a^{-1/2}(-1)^{\\tau}x y^{\\tau}(ay-1)=0$, rewritten as $Y=X(1-Y)^{\\tau}(1-a(1-Y))$. The disc count comes from expanding $Y$ by Lagrangian inversion and comparing with the multiple-cover formula; the annulus comes from the Bergmann kernel of this curve, specifically from $\\log((Y_2-Y_1)/(X_2-X_1))$, using a Hodge-integral lemma at the special degree $l=(m_1+m_2)/2$; the multi-hole genus-zero counts use the Hodge integral formula (15); and the higher-genus one-hole invariants are extracted from a Möbius-modified Chern–Simons free energy. Möbius inversion is the step that turns each generating-function coefficient into the desired integer invariant.","core_discovery":"The paper's central claim is that all LMOV invariants of the resolved conifold with one framed brane, equivalently of the framed unknot with framing $\\tau\\in\\mathbb{Z}$, are given by explicit closed formulae rather than by an infinite recursive process. For the disc sector, $n_{m,l}(\\tau)=\\sum_{d|m,\\,d|l}\\frac{\\mu(d)}{d^2}c_{m/d,l/d}(\\tau)$ with $c_{m,l}(\\tau)=-\\frac{(-1)^{m\\tau+m+l}}{m^2}\\binom{m}{l}\\binom{m\\tau+l-1}{m-1}$. For the annulus at degree $l=(m_1+m_2)/2$, $n_{(m_1,m_2)}(\\tau)=\\frac{1}{m_1+m_2}\\sum_{d|m_1,\\,d|m_2}\\mu(d)(-1)^{(m_1+m_2)(\\tau+1)/d}\\binom{(m_1\\tau+m_1)/d-1}{m_1/d}\\binom{(m_2\\tau+m_2)/d}{m_2/d}$. For three or more holes at degree $|\\mu|/2$, the invariant is a Möbius sum over the explicit product $K^\\tau_{\\mu,0,|\\mu|/2}=(-1)^{|\\mu|\\tau}[\\tau(\\tau+1)]^{l(\\mu)-1}\\prod_i\\binom{\\mu_i(\\tau+1)-1}{\\mu_i-1}(\\sum_i\\mu_i)^{l(\\mu)-3}$. For higher genus with one hole, the invariants are coefficients of $g_m(q,a)=\\sum_{d|m}\\mu(d)Z_{m/d}(q^d,a^d)$, where $Z_m$ is built from colored HOMFLY-PT invariants of the unknot. Integrality of all these numbers was proved in a separate paper, so the contribution here is the closed-form evaluation.","pith_inferences":["The same Möbius-plus-binomial structure should extend to any strip geometry behind an AV-brane, since the recently proposed strip-geometry disc formula specializes to the conifold formula used here.","If these formulas are right, the observed Gaussian/binomial distribution of LMOV numbers for large representations becomes a provable asymptotic statement for the framed unknot, obtained by analyzing these binomial sums rather than by numerical experiment.","The unproved $a=0$ reduction in the annulus sector is directly testable: one can compute the Bergmann-kernel annulus amplitude for the full curve at the same degree and check whether the $a$-dependent terms cancel.","Because the open-string partition function of the framed unknot was previously related to a quiver cohomological Hall algebra, these explicit invariants may also give closed-form characters in quiver representation theory."],"forward_implications":["For every integer framing $\\tau$, the disc invariants $n_{m,l}(\\tau)$ are computable in finitely many binomial terms, and the companion paper proves they are integers.","The annulus invariant $n_{(m_1,m_2)}(\\tau)$ at the special degree is explicit and integer for all $m_1,m_2\\ge1$ and all $\\tau\\in\\mathbb{Z}$.","Every genus-zero invariant with three or more holes at degree $|\\mu|/2$ is visibly integral because the product $K^\\tau_{\\mu,0,|\\mu|/2}$ is an integer and $l(\\mu)\\ge3$.","Higher-genus one-hole invariants can be read off coefficient by coefficient from the fixed polynomial $g_m(q,a)\\in z^{-2}\\mathbb{Z}[z^2,a^{\\pm1/2}]$, giving a finite algorithm rather than a recursion.","The explicit formulae make the LMOV integrality conjecture for the framed unknot checkable sector by sector without computing open Gromov-Witten invariants directly."],"supporting_citations":[{"why":"Supplies the Q-deformed A-polynomial method from which the mirror curve of the framed unknot is derived.","marker":"[2]"},{"why":"Supplies the Lagrangian inversion formula used to expand the mirror-curve solution and obtain the disc formula.","marker":"[74]"},{"why":"Supplies the Hodge-integral identity giving the log Bergmann-kernel expansion for the reduced curve, used in the annulus formula.","marker":"[80]"},{"why":"Proves the integrality of the disc and annulus invariants and the polynomial structure of $g_m(q,a)$.","marker":"[56]"},{"why":"Provides the open Gromov-Witten definition and the Hodge integral formula used for multi-hole genus-zero invariants.","marker":"[35]"},{"why":"Establishes the BKMP topological recursion framework that interprets Bergmann-kernel amplitudes as open Gromov-Witten invariants.","marker":"[9]"},{"why":"Proves the large-N duality identity between the framed-unknot Chern-Simons partition function and the open topological string partition function.","marker":"[79]"},{"why":"Sets up the framed-knot large-N duality and the Marino-Vafa formula underlying the Chern-Simons side of the computation.","marker":"[62]"}],"fun_headline_variants":["Explicit LMOV formulas for every framing of the unknot","Closed-form LMOV invariants for the framed unknot","LMOV invariants become explicit sums for all framings","All LMOV integers of framed unknot in closed form","Explicit formulae for LMOV invariants of the framed unknot"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The annulus formula depends on the unproved step in Section 4.3 that, at degree $l=(m_1+m_2)/2$, only the reduced curve $Y=X(1-Y)^{\\tau}$ contributes and all $a$-dependent terms in the full mirror curve can be ignored.","fun_headline_variants_meta":{"raw":{"variants":["Explicit LMOV formulas for every framing of the unknot","Closed-form LMOV invariants for the framed unknot","LMOV invariants become explicit sums for all framings","All LMOV integers of framed unknot in closed form","Explicit formulae for LMOV invariants of the framed unknot"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000584,"raw_usage":{"total_tokens":2818,"prompt_tokens":1085,"completion_tokens":1733,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":701,"completion_tokens_details":{"reasoning_tokens":1645}},"tokens_in":701,"tokens_out":1733,"duration_ms":12559,"temperature":1.0,"reasoning_tokens":1645,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:37:23.300631+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take the full mirror curve $Y=X(1-Y)^{\\tau}(1-a(1-Y))$, compute the Bergmann-kernel coefficient of $X_1^{m_1}X_2^{m_2}a^{(m_1+m_2)/2}$ in $\\log((Y_2-Y_1)/(X_2-X_1))$ for small values such as $(m_1,m_2,\\tau)=(1,1,0)$, $(1,2,1)$, and $(2,2,-1)$, and compare with formula (27). Any nonzero contribution from the $a$-dependent factor at that degree would overturn the annulus formula; equality would support the reduction.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Lagrangian inversion formula used to expand the mirror-curve solution and obtain the disc formula."},{"cited_title":"Zhu, Hodge Integral Identities from the Cut-and-Join Equation o f Mariño-Vafa formula","cited_arxiv_id":null,"evidence_quote":"Supplies the Hodge-integral identity giving the log Bergmann-kernel expansion for the reduced curve, used in the annulus formula."},{"cited_title":"Luo and S","cited_arxiv_id":null,"evidence_quote":"Proves the integrality of the disc and annulus invariants and the polynomial structure of $g_m(q,a)$."},{"cited_title":"Katz, C.-C.M","cited_arxiv_id":null,"evidence_quote":"Provides the open Gromov-Witten definition and the Hodge integral formula used for multi-hole genus-zero invariants."},{"cited_title":"Bouchard, A","cited_arxiv_id":null,"evidence_quote":"Establishes the BKMP topological recursion framework that interprets Bergmann-kernel amplitudes as open Gromov-Witten invariants."},{"cited_title":"Zhou, A proof of the full Mariño-Vafa conjecture","cited_arxiv_id":null,"evidence_quote":"Proves the large-N duality identity between the framed-unknot Chern-Simons partition function and the open topological string partition function."},{"cited_title":"Mariño, C","cited_arxiv_id":null,"evidence_quote":"Sets up the framed-knot large-N duality and the Marino-Vafa formula underlying the Chern-Simons side of the computation."}],"review_version":1}