{"id":"fa58c497-4e24-45c9-8521-76d986356aae","arxiv_id":"hep-th/9605200","paper_version":2,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":8.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Tate's algorithm applied to elliptic Calabi-Yau fibrations gives an explicit dictionary between geometric singularity types and enhanced gauge symmetries in six-dimensional F-theory, matched case by case against heterotic Higgs-branch dimensions.","lead":"Physicists show that the singularities of the elliptic Calabi-Yau manifolds used in F-theory compactification encode the gauge symmetries of the dual heterotic string, including F4, G2 and other non-simply-laced groups that appear only after monodromy shrinks a larger singularity. The dictionary predicts gauge symmetries that are partly perturbative and partly non-perturbative, and yields the Calabi-Yau threefolds dual to four-dimensional Coulomb branches.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"Dimension matching is necessary but not sufficient: for non-simply-laced groups (F4, G2, SO(7), Sp(k)) matter is not localized and is fixed by duality, so those assignments rest on the section 3 monodromy assertion plus counting subtractions that are asserted rather than derived.","rationale":"The reader's flagged assumption (section 2's concession that the Kodaira map is 'a priori less clear' on a multi-dimensional base) is actually the paper's main technical contribution to resolve, not an unguarded extrapolation: section 3 supplies the split/non-split refinement and the outer-automorphism mechanism of eq. (3.1), and sections 4.1-4.20 test it. So I partially disagree with the reader's framing of where the soft spot lies. The sharper concern is the non-simply-laced sector, where matter localization fails and the matter content is consistency-fixed rather than derived, and where the dimension counts in 4.16 and 4.17 depend on an asserted subtraction of two reparameterizations. That said, the concern is bounded: the rank counting from invariant divisors supports the monodromy assignments (D4 has four exceptional divisors; S3, Z2, trivial monodromy give ranks 2, 3, 4), the small-instanton Sp(k) identification has independent support from Witten [5] and the codimension-21 match in section 5, and the paper explicitly acknowledges the localization failure. The simply-laced sector, which carries most of the dictionary, has independent matter-localization evidence. The acknowledged limitations (unexplained factor of 2 in section 6, sketched deficit argument, non-simply-laced matter) are real but do not undermine the central claim. ACCEPT stands; the concern is worth recording but does not move the verdict.","tokens_in":70594,"tokens_out":41070,"duration_ms":354686,"concrete_test":"Resolve the D4 fiber for the three cases I*ns0, I*ss0, I*s0 using the toric data in table 4, and compute the monodromy permutation of the four exceptional divisors as z2 encircles the discriminant. Verify the invariant divisor count is 2 (G2), 3 (SO(7)), 4 (SO(8)). If the non-split case yields 3 or 4 invariant divisors, the section 3 outer-automorphism mechanism does not produce the claimed gauge group and the non-simply-laced assignments lose their justification.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's primary quantitative evidence is exact dimension matching between F-theory moduli counts and heterotic Higgs-branch dimensions (sections 4.1-4.20). Dimension matching is a single integer per group and is necessary, not sufficient: a different group with the same dimension locus would match equally. For simply-laced groups the paper supplements it with matter localization at discriminant zeroes (4.1, 4.2, 4.5, 4.10), which is independent evidence. For non-simply-laced groups (F4 in 4.3, SO(7) in 4.8, G2 in 4.9, Sp(k) in 4.19) the authors state the matter cannot be localized, and the matter content is instead fixed by requiring duality compatibility. So the group identification there rests on (a) the section 3 assertion that outer-automorphism monodromy reduces the gauge group per eq. (3.1), and (b) dimension counts that in cases like SO(12) (4.16) and SU(6) (4.17) require subtracting 'two reparameterizations' whose number is asserted rather than derived. If either the monodromy action or the subtraction count is off by one, the match fails. This is the least secure link in the dictionary. It is acknowledged in the text (4.3, 4.8) and does not by itself overturn the central claim, since the rank counting from invariant divisors supports the monodromy assignments and the small-instanton Sp(k) result has independent physics support (Witten [5], codimension 21 match in section 5).","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The paper builds a dictionary between the Kodaira/Tate type of a degenerate elliptic fiber along a divisor and the enhanced gauge symmetry of the dual six-dimensional F-theory/heterotic vacuum, refining Kodaira's classification with Tate's split/non-split distinction so that outer-automorphism monodromy (eq. 3.1) reduces the singularity type to the actual gauge algebra. The dictionary is tested by matching, for roughly twenty groups, the dimension of the F-theory singularity locus in the moduli of the elliptic fibration over F_n against the heterotic Higgs-branch dimension obtained from the independent index formulas (4.1)-(4.2); for simply-laced groups the matter is further localized at zeroes of the discriminant. The authors then apply the dictionary to small SO(32) instantons (section 5), to mixed perturbative/non-perturbative enhancements associated with singularities along general curves Sigma_{p,q} (section 6), and, after toroidal compactification, tabulate the Calabi-Yau threefolds dual to the four-dimensional Coulomb branches (Table 5).","tokens_in":35609,"tokens_out":6305,"duration_ms":129468,"significance":"If the dictionary holds, this is a substantial paper: it converts a qualitative statement about singular fibers into a quantitative map, and the evidence is of an unusually checkable kind. The dimension matches of sections 4.1-4.20 are exact integers compared against counts that come from the heterotic index theorem rather than from the geometry, so they are not arranged by construction, and the same machinery yields falsifiable numbers: codimension 21 for a single small instanton with SO(9) x Sp(1) unbroken, and 12k(45-k)-1 for k instantons at a point, both matching expectations from Witten's small-instanton analysis. The toric data of Table 4 and Table 5 make the claimed chain of dual Calabi-Yau threefolds reproducible by anyone with the reflexive polyhedra of Candelas-Font, and the identification of non-simply-laced groups in several rows corrects a published assignment. The mixed (p,q) gauge symmetries of section 6 are a genuinely new and falsifiable prediction.","major_comments":[{"comment":"The dimension of the SO(12) locus is obtained by counting the parameters in p_{4+n-r}, q_{8+n-r}, t_r and u_{r+4} 'after subtracting two degrees of freedom which can be absorbed in reparameterizations', giving 2n+18, and the same subtraction of two recurs in the SU(6) count of section 4.17. Section 2 derives the n+2 reparameterizations z1 -> a z1 + P_n(z2) and the subtraction of one for the rescaling of z1, but the count of two in these two subsections is not derived. Since the agreement with the heterotic value 2n+18 is the entire check for these entries, please exhibit the two-parameter family explicitly (which reparameterizations survive the constraints (4.11)) or state that the count is fixed by requiring the match. An off-by-one here would remove two of the twenty supports for the dictionary.","section":"Sections 4.16 and 4.17, Eqs. (4.11) and (4.14)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The k=1 case is worked in full and the codimension 21 agrees with the expected SO(9) x Sp(1) spectrum, which is convincing. For general k, however, the codimension increment 22-k is asserted and the identification of the singularity at w1=0 with the perturbative SO(8+k) locus is supported only by the remark that 'it is easily seen in the first few cases'; the SO(8+k) x Sp(k) gauge group and the codimension 12k(45-k)-1 are then presented as established results. Because the abstract lists the small SO(32) instanton physics as one of the three headline claims, please either supply the induction or label the general-k statement explicitly as an extrapolation from k=1,2,3, so that the reader can tell which part of section 5 is derived and which is conjectural.","section":"Section 5, general k","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The assertion that the non-split A_{2k} singularity does not give Sp(k) gauge symmetry rests on a private communication (Lepowski) together with the remark about the order-four outer automorphism. A short published argument or sketch would let readers weigh this, since the exclusion is what keeps the non-split A_{2k-1} assignment unambiguous.","section":"Section 3, footnote 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The authors state in these subsections that the matter of F4, SO(7), G2 and Sp(3) cannot be localized at discriminant zeroes, and section 8 says they 'limited ourselves to what matter a given singularity must encode in order to be compatible with duality'. This is an acknowledged limitation and not a defect of the paper, but it would help to say it once at the head of section 4, so that a reader can distinguish at a glance which matter spectra are predictions of the geometry and which are inputs used to identify the group.","section":"Sections 4.3, 4.8, 4.9, 4.19 and Section 8","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The overall normalization (u,v) = (2p,2q) is explicitly left unexplained: 'To explain the coefficient of 2 in (6.3), one would have to understand the anomaly factorization property in terms of D-branes.' The deficit-argument codimension 29(p+q)-5pq does supply an independent check of this normalization, and the text should say so at that point, so that the reader sees the factor of 2 is cross-checked by the geometry rather than assumed.","section":"Section 6, Eq. (6.3)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The claim that the mixed SO(7) and G2 theories 'cannot be obtained by Higgsing from intersecting perturbative and non-perturbative gauge groups' and that 'there must be new physics going on' is an argument from failure to construct the parent theory. Either state the obstruction or soften the phrasing to indicate that no such parent was found.","section":"Section 6, final paragraphs","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text notes that 'in the SO(12) case, the toric moduli space is only a subspace of the full moduli, but non-toric deformations can be expected to make up the difference'. It would be useful to mark in Table 5 which rows are toric-complete and which rely on this expectation, since the table is presented as a derived chain of dual threefolds.","section":"Section 7 and Table 5","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Reference [32] is listed as 'E. Witten, to appear', and the claim that progress on the heterotic explanation of matter localization 'has been recently made [32]' cannot be checked against an unpublished source. Either cite something published or phrase the remark as work in progress.","section":"Section 8, reference [32]","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The entries marked 'unconven.' (I^ns_3, I^ns_{2k+1}, IV^ns) are never given a physical interpretation anywhere in the paper. A sentence stating what these loci correspond to physically, or that they are excluded from the dictionary, would close an obvious gap for a reader using the table.","section":"Table 2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The counting of reparameterizations (n+2, minus one for the rescaling of z1) is used repeatedly downstream, including in the major comment above. A one-line derivation at its first appearance would make the later subtractions easier to follow.","section":"Section 2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The Higgs tree is difficult to parse in its linear rendering; a cleaner figure, or a table of the Higgs arrows with the broken group and the representation used at each step, would make the structure of section 4 much easier to follow.","section":"Diagram 1","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The supplied text carries numerous spacing and hyphenation artifacts ('Cambridg e', 'Piscat away', 'symme try', 'compactification'). These may originate in the source file rather than with the authors, but they should be cleaned before typesetting.","section":"Whole manuscript","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"accept","confidential_remarks":"This is a strong, checkable paper whose central dictionary is supported by exact integer matches against independent heterotic index counts, and the two major comments above concern the justification of individual counting steps rather than the soundness of the program. The overlap disclosure with Aspinwall-Gross [14] in the introduction is handled appropriately. The only citation I could not verify is reference [32], listed as 'to appear'."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"The thing to take away is the refinement of Tate's algorithm. The split, semi-split and non-split cases, and the observation that outer-automorphism monodromy around a degenerate fiber breaks the simply-laced singularity down to C_n, B_n, F4, G2 per their equation (3.1), is what everyone after this paper used. The tables in section 3 and the toric data in section 7 are the parts I would actually photocopy for someone learning the subject. That is genuine, and it is not in the literature they cite — Aspinwall-Gross [14] overlaps on monodromy and the paper says so in the introduction, which is the right way to handle it. The small-instanton result in section 5, that k collapsed SO(32) instantons show up as an A_{2k-1} singularity in z2 with codimension (1/2)k(45-k)-1, is a clean and independently checkable statement, and it lands on Witten's Sp(k) expectation. The mixed SU(2) family in section 6 with (u,v)=(2p,2q) is a nice piece of physics that the strong-weak duality argument alone could not have produced, since (1,1) is self-dual. On the central evidence: the dimension matches across twenty groups are not circular, because the heterotic side comes from the index formulas (4.1) and (4.2), not from the geometry. That is a real check, and it is repeated enough times that coincidence is not a serious worry. Where I would push back on the skeptical reading: the reparameterization subtractions are not asserted. Section 2 derives the count 31n+114, identifies the n+2 shifts z1 -> az1 + P_n(z2), and subtracts one for rescaling. That is spelled out. The stress-test is right, though, that for F4, SO(7), G2 and Sp(k) the matter is not localized and the content is fixed by requiring duality to hold. The authors say this themselves in 4.3, 4.8 and again in section 8, so it is an acknowledged gap rather than a hidden one. The codimension iteration in section 5 and the deficit argument in section 6 are described rather than proved, and the factor of 2 in (6.3) is admitted to be unexplained. Those are the honest soft spots, and none of them touch the gauge-group dictionary. This is for anyone working on F-theory compactifications or on heterotic/type II duality chains; the section 7 polyhedra are still useful. It deserves referees, and I would cite it without hesitation.","headline":"The durable contribution is the split/non-split refinement of Tate's algorithm; the dimension matches are a real check, and the soft spots are the ones the authors themselves flag.","tokens_in":36357,"tokens_out":1122,"would_cite":true,"duration_ms":21373,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["11.25.Mj"],"model":"selfhosted/deepseek-v4.1-flash","headline":"Tate's algorithm turns elliptic-fiber singularities into gauge groups","keywords":["F-theory","Tate's algorithm","enhanced gauge symmetry","elliptic Calabi-Yau","F-theory heterotic duality","small instantons","Coulomb branch","non-simply laced gauge groups"],"falsifier":"Compute the number of complex-structure moduli on a locus with a specified singularity for one of the listed groups and compare it with the heterotic Higgs-branch dimension from $c_2(H)(12+n)-dim H$ and the index formula for the matter; a mismatch in any single case, or a matter spectrum at an intersection that disagrees with the four-dimensional genus count, would break the dictionary.","tokens_in":36560,"feed_emoji":"","tokens_out":3806,"duration_ms":103079,"temperature":0.7,"pith_summary":"F-theory compactified on an elliptic Calabi–Yau manifold is a type IIB vacuum in which the shape of the elliptic fiber varies over a base and may degenerate along divisors. This paper asks whether the *type* of that degeneration — the Kodaira classification, refined by Tate's algorithm into split and non-split cases — is exactly the enhanced gauge symmetry of the resulting six-dimensional theory, even when the base has more than one dimension. The test is the conjectured duality between F-theory on an elliptic threefold over a Hirzebruch surface and the $E_8$ heterotic string on K3: on the heterotic side the dimension of each Higgs branch and its charged matter content can be counted independently from instanton numbers. The authors find the two counts agree case by case, for all the classical groups including the non-simply laced $B$, $C$, $F_4$ and $G_2$ series, and then use the dictionary to read off physics that the heterotic side could not supply on its own: the small $SO(32)$ instanton gauge groups and new gauge symmetries that are neither purely perturbative nor purely non-perturbative.","feed_headline":"Tate's algorithm turns fiber singularities into gauge groups","feed_subtitle":"Moduli counts matching heterotic Higgs branches test a six-dimensional duality and predict new mixed symmetries.","key_machinery":"Tate's algorithm is the engine: a step-by-step procedure that reads the Kodaira fiber type from divisibility conditions on the coefficients $a_1,a_2,a_3,a_4,a_6$ of the generalized Weierstrass equation along a divisor $z_1=0$, and then asks whether certain polynomials factor modulo that divisor. The factorization question is what separates the split from the non-split cases, and it is what carries the physics: a non-split fiber means the vanishing cycles are exchanged by an outer automorphism of the Dynkin diagram as one moves around the base, so the gauge group is the quotient $A_{2n-1}/Z_2 = C_n$, $D_n/Z_2 = B_{n-1}$, $E_6/Z_2 = F_4$, or $D_4/S_3 = G_2$. The authors extend Tate's classification with explicit split, semi-split and non-split entries and a table of the discriminant's next branch, and later re-express the same conditions as toric data — vectors $v$ with $v·(i,j,k) ≥ -1$ — so that resolving the singularity becomes adding points to a reflexive polyhedron, which is what makes the four-dimensional chain computable.","core_discovery":"The central claim is a dictionary: the order of vanishing of the Weierstrass coefficients $f$ and $g$ along a divisor, together with the split/non-split factorization conditions of Tate's algorithm, determines the enhanced gauge symmetry and the matter content of the six-dimensional F-theory vacuum. The evidence is a long series of dimension matches. For example, an $E_7$ singularity requires $f$ to vanish to order three and $g$ beyond order five, leaving the coefficients $f_{8+n}$ and $g_{12+n}$ free and giving $2n+21$ moduli after subtracting one rescaling of $z_1$; the heterotic count for $12+n$ instantons in commutant $SU(2)$ is also $2n+21$. The same exercise is repeated for $E_6$, $F_4$, $SO(11)$, $SO(10)$, $SO(9)$, $SO(8)$, $SO(7)$, $G_2$, $SU(4)$, $SO(5)$, $SO(4)$, $SU(3)$, $SU(2)$, $SU(2)^2$, $SO(12)$, $SU(6)$, $SU(5)$, $Sp(3)$ and $SU(3)^2$, with the split/non-split distinction doing the work of breaking a simply laced singularity to a non-simply laced group. Applied to $n=4$, the dictionary reproduces the physics of $k$ coincident small $SO(32)$ instantons: an $A_{2k-1}$ singularity in $z_2$ forces a larger singularity in $z_1$, yielding an $SO(8+k)SU(k)$-type enhancement at codimension $k(45-k)/2-1$. Applied to the symmetric $(12,12)$ compactification, it predicts mixed symmetries $SU(2)_{(p,q)}$ with anomaly coefficients $(u,v)=(2p,2q)$, whose compact part is a curve of genus $(p-1)(q-1)$ — exactly the number of adjoints the four-dimensional topological argument requires. Compactifying further on a torus turns these loci into a chain of Calabi–Yau threefolds dual to the Coulomb branches of the heterotic string.","pith_inferences":["The genus formula $(p-1)(q-1)$ and the codimension $29(p+q)-5pq$ are two independent routes to the same number; the agreement is a check the authors could have pushed further by computing the four-dimensional spectra explicitly from the type IIA side.","Because the dictionary is stated for any divisor in any base, it should extend beyond Hirzebruch surfaces to more general elliptic threefolds and fourfolds, which would turn it into a general tool rather than a duality test.","The failure to localize matter for the non-simply laced groups is a structural clue: it suggests the outer-automorphism quotient acts on the matter curves as well as on the fiber, which a D-brane analysis could confirm or rule out.","The tensionless string appearing where two $D_4$ singularities intersect is a prediction that could be probed by looking for a light tower of states in the four-dimensional compactification at that locus."],"forward_implications":["Every enhanced gauge symmetry locus in the F-theory moduli space acquires a computational recipe, so the heterotic Higgs branches can be located geometrically rather than inferred.","The small $SO(32)$ instanton story follows from the same dictionary: $k$ instantons at a point give a perturbative $SO(8+k)$ together with a non-perturbative $Sp(k)$, at a codimension the authors compute and match against Higgs counting.","A new class of six-dimensional gauge theories appears that are neither perturbative nor non-perturbative on the heterotic side, labeled by a pair $(p,q)$ with anomaly coefficients $(u,v)=(2p,2q)$ and $(p-1)(q-1)$ adjoints.","Compactifying to four dimensions produces an explicit list of Calabi–Yau threefolds, given as reflexive polyhedra, dual to the Coulomb branches of the heterotic string, extending the web of $N=2$ dualities.","The matter representations of the simply laced cases localize at zeros of specific polynomials in the discriminant, giving a geometric handle on charged hypermultiplets."],"supporting_citations":[{"why":"Supplies Tate's algorithm, the procedure for reading a fiber type from divisibility of the Weierstrass coefficients that the whole dictionary is built on.","marker":"[10]"},{"why":"Provides the original F-theory evidence and the K3 case where the Kodaira singularity type matches the eight-dimensional gauge symmetry.","marker":"[13]"},{"why":"Supplies the F-theory/heterotic duality on elliptic threefolds over Hirzebruch surfaces and the identification of which polynomial coefficients carry which moduli.","marker":"[9]"},{"why":"Establishes the identification of A-D-E singularities with gauge symmetry enhancement, which the paper extends to a multi-dimensional base.","marker":"[1]"},{"why":"Supplies the expected non-perturbative $Sp(k)$ enhancement for $k$ coincident small $SO(32)$ instantons that section 5 must reproduce.","marker":"[5]"},{"why":"Provides the intersecting-brane picture in which matter sits at intersections of singular loci, used to interpret the extra discriminant zeros.","marker":"[3]"},{"why":"Supplies the four-dimensional statement that a curve of genus $g$ in the 7-brane world-volume yields $g$ adjoint hypermultiplets, the check on the $(p-1)(q-1)$ count.","marker":"[18]"},{"why":"Supplies the anomaly coefficients $(u,v)=(2,0)$ and $(0,2)$ for perturbative and non-perturbative gauge groups, which the mixed $(2p,2q)$ prediction extends.","marker":"[26]"},{"why":"Provides the reflexive polyhedra and the $a_H,b_H$ data against which the chain of Calabi–Yau threefolds in table 5 is compared.","marker":"[12]"},{"why":"Supplies Batyrev's reflexive-polyhedron construction that makes the toric description of the dual threefolds and their mirrors available.","marker":"[31]"}],"fun_headline_variants":["Tate's algorithm reads gauge groups from vanishing orders","Singularity orders set gauge symmetry and matter in F-theory","F-theory gauge dictionary rests on moduli counts, not a proof","Split singularities break simply laced groups to non-simply laced","Six-dimensional duality tested by moduli matching, not proven"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The load-bearing premise is that the Kodaira singularity type of the elliptic fiber continues to determine the gauge group when the base of the fibration has more than one dimension — a point the authors themselves call 'a priori less clear' in section 2, since the match is only established for the K3 case.","fun_headline_variants_meta":{"raw":{"variants":["Tate's algorithm reads gauge groups from vanishing orders","Singularity orders set gauge symmetry and matter in F-theory","F-theory gauge dictionary rests on moduli counts, not a proof","Split singularities break simply laced groups to non-simply laced","Six-dimensional duality tested by moduli matching, not proven"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":2266,"prompt_tokens":1102,"completion_tokens":1164,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":1088}},"tokens_in":1102,"tokens_out":1164,"duration_ms":16177,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":1088,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T19:16:31.016560+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Compute the number of complex-structure moduli on a locus with a specified singularity for one of the listed groups and compare it with the heterotic Higgs-branch dimension from $c_2(H)(12+n)-dim H$ and the index formula for the matter; a mismatch in any single case, or a matter spectrum at an intersection that disagrees with the four-dimensional genus count, would break the dictionary.","supporting_citations":[{"cited_title":"Algorithm for Determining the Type of a Singul ar Fiber in an Elliptic Pen- cil,","cited_arxiv_id":null,"evidence_quote":"Supplies Tate's algorithm, the procedure for reading a fiber type from divisibility of the Weierstrass coefficients that the whole dictionary is built on."},{"cited_title":"Evidence for F-theory,","cited_arxiv_id":null,"evidence_quote":"Provides the original F-theory evidence and the K3 case where the Kodaira singularity type matches the eight-dimensional gauge symmetry."},{"cited_title":"Compactifications of F-theory o n Calabi-Yau Threefolds - I,II","cited_arxiv_id":null,"evidence_quote":"Supplies the F-theory/heterotic duality on elliptic threefolds over Hirzebruch surfaces and the identification of which polynomial coefficients carry which moduli."},{"cited_title":"String Theory Dynamics in Various Dimension s,","cited_arxiv_id":null,"evidence_quote":"Establishes the identification of A-D-E singularities with gauge symmetry enhancement, which the paper extends to a multi-dimensional base."},{"cited_title":"Small Instantons in String Theory,","cited_arxiv_id":null,"evidence_quote":"Supplies the expected non-perturbative $Sp(k)$ enhancement for $k$ coincident small $SO(32)$ instantons that section 5 must reproduce."},{"cited_title":"D Strings on D Manif olds,","cited_arxiv_id":null,"evidence_quote":"Provides the intersecting-brane picture in which matter sits at intersections of singular loci, used to interpret the extra discriminant zeros."},{"cited_title":"Enhanced Gauge Symme try in Type II String Theory,","cited_arxiv_id":null,"evidence_quote":"Supplies the four-dimensional statement that a curve of genus $g$ in the 7-brane world-volume yields $g$ adjoint hypermultiplets, the check on the $(p-1)(q-1)$ count."},{"cited_title":"Evidence for Hetero tic/Heterotic Duality,","cited_arxiv_id":null,"evidence_quote":"Supplies the anomaly coefficients $(u,v)=(2,0)$ and $(0,2)$ for perturbative and non-perturbative gauge groups, which the mixed $(2p,2q)$ prediction extends."},{"cited_title":"Duality Between the Webs of Het erotic and Type II Vacua,","cited_arxiv_id":null,"evidence_quote":"Provides the reflexive polyhedra and the $a_H,b_H$ data against which the chain of Calabi–Yau threefolds in table 5 is compared."},{"cited_title":"Dual Polyhedra and Mirror Symmetry for Calabi-Yau Hypersurfaces in Toric Varieties","cited_arxiv_id":"alg-geom/9310003","evidence_quote":"Supplies Batyrev's reflexive-polyhedron construction that makes the toric description of the dual threefolds and their mirrors available."}],"review_version":1}