{"id":"1cb00c62-5a5b-431c-928d-3d1e218390a3","arxiv_id":"hep-th/9706110","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Geometric engineering plus local mirror symmetry yields exact Coulomb-branch solutions for asymptotically free SU-product N=2 gauge theories, with matter graphs classified by ADE and affine ADE Dynkin diagrams.","lead":"This paper gives exact solutions for the Coulomb branch of four-dimensional N=2 gauge theories built from products of SU groups with bi-fundamental and fundamental matter, using mirror symmetry of type II string compactifications. It shows that asymptotic freedom forces the matter graph to be an ADE or affine ADE Dynkin diagram, and identifies the S-duality groups with moduli of flat ADE bundles on a torus.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The decoupling limit (5.6) is asserted rather than derived, and the ordinary D/E cases are deferred in §7.2; but the A-case match to Witten and the Looijenga check of the affine ADE geometry supply enough independent support that this does not land as a fatal objection.","rationale":"The reader's weakest_assumption correctly identifies the decoupling limit (5.6) as the least-secured step, and the rationale fairly credits both the Witten match for the A-chain and the Looijenga match for the affine ADE cases. I would only sharpen two points. First, the Looijenga check is stronger than 'no independent field theory computation' suggests: it is an independent mathematical confirmation that the mirror geometry reproduces the known Coulomb-branch moduli space, even if it does not check the periods. Second, the ordinary D/E cases are not merely unchecked but explicitly deferred in §7.2, which is a genuine (if minor) gap in a paper whose abstract claims 'all' such theories. Neither point overturns the central claim: the Perron-Frobenius argument of §2.2 is a clean, self-contained proof, the A-case construction is validated against an independent string result, and the E-case geometry is validated against Looijenga. The base-fiber exchange duality (9.2) is explicitly speculative and is not load-bearing for the exact-solution claim. The paper is transparent about its heuristic section and its open directions. I therefore see no concern that lands as a fatal objection, and I would leave the verdict at ACCEPT with the reader's MODERATE confidence, noting that the E-case prepotential is the one place where an explicit period computation would most strengthen the paper.","tokens_in":43369,"tokens_out":11403,"duration_ms":203766,"concrete_test":"For affine E6 with k=1, solve the Picard-Fuchs system (4.18) with the Mori vectors read off from (7.6), compute the period integrals of the holomorphic form Omega in the limit (5.6), and verify that the resulting metric on the Coulomb branch agrees with the natural metric on the moduli space of flat E6 bundles on T^2 (the Looijenga description used in §8.3). Agreement validates the decoupling limit for the E case; disagreement would show that the threefold periods do not reduce to the field-theory prepotential.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (abstract:s0) is that the mirror geometry's periods give the exact Coulomb-branch prepotential for every theory in the ADE/affine-ADE class. The step that carries the most weight and is least secured is the decoupling limit of eq. (5.6), a_{r,l} ~ epsilon^{l-k_r-1}, epsilon -> 0. The paper checks that the moduli z scale as epsilon^0 and the couplings as epsilon^{-b_r} (correct AF behaviour), but it does not exhibit the period integrals of Omega in this limit and show that they reduce to rigid special geometry. For the A-chain this is validated indirectly: the resulting curve (6.5) matches Witten's M-theory curves (hep-th/9703166), so the method is anchored. For the affine E and D cases the only independent check is the Looijenga match of the moduli space of flat G-bundles on E (§8.3), which confirms the geometry but not the periods/metric; and the ordinary D/E cases are explicitly deferred in §7.2 ('we will not go into any further detail'). So the exact prepotential for the E/D entries is asserted by uniformity rather than demonstrated. This is a real soft spot, but it is a standard geometric-engineering step applied uniformly, and the geometry itself is independently confirmed, so the concern is modest rather than fatal.","agreement_with_reader":"agree"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The manuscript applies geometric engineering to type IIA string theory on local Calabi-Yau threefolds, passes to the local type IIB mirror, and reads off the exact Coulomb-branch geometry of four-dimensional N=2 gauge theories built from products of SU groups with bi-fundamental matter between selected pairs and arbitrary fundamental matter for each factor. The central structural result (abstract:s1) is that asymptotic freedom forces the bi-fundamental graph to be an ADE or affine ADE Dynkin diagram, argued via Perron-Frobenius on the generalized Cartan matrix M = 2I - N. The mirror geometry is a Riemann surface for the A chains and a threefold for the E cases (abstract:s3). The paper also identifies the S-duality group with the fundamental group of the moduli space of flat ADE bundles on an elliptic curve, and applies the construction to three-dimensional N=4 dualities, including a check of the Intriligator-Seiberg small-instanton conjecture.","tokens_in":43660,"tokens_out":10859,"duration_ms":175623,"significance":"If the construction holds, this is a foundational contribution: it gives a uniform, algorithmic solution for a large class of N=2 theories and explains the ADE pattern of the bi-fundamental sector from a single eigenvalue condition. The independent cross-checks are strong: the A-chain curves reproduce Witten's M-theory curves (hep-th/9703166), and the affine ADE mirror geometries reproduce Looijenga's moduli spaces of flat G-bundles on an elliptic curve (SS8.3). The observation that the E cases require a threefold rather than a Riemann surface is novel and has been influential. The three-dimensional application in SS10 provides a nontrivial check of a conjecture that was not obviously accessible by other means.","major_comments":[{"comment":"The ordinary D and E cases are dismissed with 'we will not go into any further detail,' and the reader is referred to the companion paper [22], which is listed as 'to appear.' Since abstract:s0 advertises exact solutions for the ADE class, and the affine cases are worked out explicitly in SS7 and SS8, the ordinary D/E entries are the one part of the headline claim that is asserted by uniformity rather than written down. Either the explicit curves (or at least the defining polynomials and the moduli) should be included, or the abstract and introduction should state that the ordinary D/E solutions are deferred to [22].","section":"SS7.2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The decoupling limit a_{r,l} ~ epsilon^{l-k_r-1}, epsilon -> 0, is asserted rather than derived. The paper checks that the algebraic coordinates scale as epsilon^0 and the couplings as epsilon^{-b_r}, which is the correct asymptotic-freedom behaviour, but it does not exhibit the period integrals of Omega in this limit and show that they reduce to rigid special geometry. For the A chains this is validated indirectly by the match to Witten's M-theory curves; for the affine E and D cases the only independent check is the Looijenga match of SS8.3, which fixes the geometry but not the metric or the periods. The paper should say explicitly which entries are cross-checked and which rely on uniformity, and ideally exhibit one period computation in the limit for a non-A case.","section":"SS5.2, Eq. (5.6)","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The Perron-Frobenius statement is imprecise: N has non-negative, not positive, entries, and the equality case in the Rayleigh-quotient sketch requires irreducibility of N (i.e. a connected quiver). The conclusion lambda >= 0 implies ADE or affine ADE also relies on Kac's classification, which is cited only in a footnote. The phrase 'M k is a positive semi-definite vector' should read 'non-negative vector.'","section":"SS2.2","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sum over alpha in R_G appears to run over all roots, which double counts relative to the standard positive-root one-loop prepotential; and W_G is used for the matter weights, which conflicts with the usual notation for the Weyl group. Please state the conventions explicitly.","section":"SS4.4, Eq. (4.19)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The moduli are printed as products (for example z^{(c)}_{r,i} = a_{r,i-1} a_{r,i+1} a^2_{r,i}) where the mirror map (4.16) requires a quotient by a_{r,i}^2. The fractions should be written explicitly to avoid ambiguity.","section":"SS5.2, Eq. (5.5); SS6.2, Eq. (6.2); SS7.1, Eq. (7.6)","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The term sum_i P_{y i k}(w) z^{4-i} should presumably be sum_i P_{z i k}(w) z^{4-i}, since y appears only quadratically in the E7 equation y^2 + x^4 + z^4 + a x y z.","section":"SS3.3, E7 display","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The caption says 'SU(2) Nf=1' while the surrounding text says 'the SU(3) gauge theory with Nf = 1 matter in fig. 6 a).' Please reconcile the caption with the text.","section":"Fig. 6 caption","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The sentence 'In section 6 we show how arbitrary matter ... In section 6 we mainly concentrate on the A case' repeats the section number; the second reference should presumably point to SS6.1 or SS6.2.","section":"Introduction","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"'Three dimensional complex manifold' should read 'three complex dimensional manifold' or 'threefold.'","section":"Abstract","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"Reference [41] gives the year as 1998 for Nucl. Phys. B324, which should be 1989; references [17] and [21] are the same entry (Bershadsky et al., Nucl. Phys. B481 (1996) 215) listed twice; and [24] has 'Nucl. Phys. BB 462.'","section":"References","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The identification of the S-duality group with pi_1(M_G) is stated rather than argued. A sentence on why the coupling-constant space coincides with the base moduli space, and how this interacts with the base-fiber mixing discussed in SS9, would help the reader.","section":"SS8.4","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The 'duality' SU(n)^m <-> SU(m+1)^{n-1} is introduced as a suggestion ('This suggests a duality'). Please state explicitly whether it is a conjecture or a derived statement.","section":"SS9","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"The paper is a strong candidate for the journal and the central construction is sound. The main editorial issue is scope: the abstract's 'all' claim is qualified in the body, and the ordinary D/E entries are deferred to a companion paper listed as 'to appear.' The authors should either close that gap or scope the abstract accordingly before publication."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"This is the paper that turned geometric engineering from a collection of examples into a classification. The structural result in §2.2 — asymptotic freedom on a product of SU groups with bi-fundamentals forces the quiver to be ADE or affine ADE, via Perron-Frobenius on M = 2I − N — is short, self-contained, and correct as far as I can check. That alone is worth the read. The rest is a systematic toric mirror construction that produces explicit curves for the A chains and threefolds for the affine E cases, with fundamental matter added by the trivalent vertex in §6. The affine E solutions in §7 are genuinely new relative to [8][12], and the identification of S-duality groups with the fundamental group of the moduli of flat ADE bundles on the elliptic curve (§8.3, matching Looijenga) is a nice piece of structure that I have not seen assembled this way before. The paper is also honest about its debts: §3 is flagged as heuristic, and the overlap with Witten's M-theory curves is stated rather than buried. That is good practice and I would not hold it against them. The soft spot is the one the stress-test names, and it is real but modest. The decoupling limit (5.6), a_{r,l} ~ ε^{l−k_r−1}, is asserted; the paper checks that the moduli scale as ε^0 and the couplings as ε^{−b_r}, which is the right AF behaviour, but it does not exhibit the periods of Ω in that limit and show they reduce to rigid special geometry. For the A chain this is fine, because the resulting curve (6.5) matches Witten's independent M-theory derivation. For the affine E and D entries the only independent check is the Looijenga match of the moduli space, which confirms the geometry but not the metric. And §7.2 explicitly defers the ordinary D and E cases ('we will not go into any further detail'), so those prepotentials are asserted by uniformity rather than demonstrated. That is a limitation the authors acknowledge, not a hidden flaw, and it does not break the central claim. The base-fiber exchange duality (9.2) is proposed without derivation and should be read as a conjecture. Who should read it: anyone working on N=2 Coulomb branches, quiver gauge theories, or the geometric-engineering dictionary. It deserves a serious referee, and I would cite it.","headline":"A foundational geometric-engineering paper: the ADE classification is clean and the A-chain solutions are anchored to Witten's independent curves, but the E/D prepotentials rest on a decoupling limit that is asserted rather than shown.","tokens_in":44296,"tokens_out":991,"would_cite":true,"duration_ms":16168,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["11.25.Mj","11.30.Pb","12.60.Jv"],"model":"selfhosted/deepseek-v4.1-flash","headline":"Mirror symmetry is claimed to give the exact answer for a whole class of four-dimensional N=2 gauge theories built from products of SU groups with matter between them.","keywords":["geometric engineering","mirror symmetry","Seiberg-Witten curve","N=2 supersymmetric gauge theory","ADE Dynkin diagrams","Calabi-Yau threefolds","S-duality","toric geometry"],"falsifier":"Compute the periods of the mirror threefold for the affine $E_8$ quiver in the stated limit and compare the one-loop piece against the field-theory expression of equation (4.19) with the expected beta-function coefficient; a mismatch in the coefficient, or an extra period without a field-theory counterpart, would break the claim that the threefold reproduces the Coulomb branch.","tokens_in":44704,"feed_emoji":"🪞","tokens_out":3052,"duration_ms":49157,"temperature":0.7,"pith_summary":"The paper asks a concrete question: can you write down the exact low-energy geometry of the Coulomb branch — the space of vacuum expectation values of the adjoint scalars — for a product of SU gauge groups with matter shared between pairs of them, plus any number of extra fundamental flavours? The authors' answer is yes, for every such theory that is asymptotically free, and they get it by a chain of string dualities rather than by field-theory calculation. The construction puts an ADE-type singularity of a specific rank over each 2-sphere of a base, so the pattern of bi-fundamental matter becomes a graph, and then mirror symmetry converts the type IIA geometry, whose sizes receive worldsheet instanton corrections, into a type IIB geometry whose complex-structure periods are exact at string tree level. Why a reader should care: the same argument delivers a classification (asymptotic freedom alone forces the matter graph to be an ADE or affine ADE Dynkin diagram), a uniform solution for all those cases, and a prediction for the S-duality group in terms of flat ADE connections on a torus. The paper also reports that for the exceptional cases the answer is not a Riemann surface but a three-dimensional complex manifold.","feed_headline":"Mirror symmetry solves a whole family of N=2 gauge theories","feed_subtitle":"Asymptotic freedom forces the matter graph to be an ADE diagram; the mirror geometry then gives the exact answer.","key_machinery":"Geometric engineering plus local mirror symmetry. A type IIA compactification on a non-compact Calabi-Yau threefold, built by fibering an $A_{k-1}$ singularity over each 2-sphere of a base whose spheres intersect according to the matter graph, produces the gauge theory; D2-branes wrapped on the shrinking spheres give the charged vector bosons, and the base volume sets $1/g^2$. The type IIA Kähler moduli receive worldsheet instanton corrections, so the authors pass to the mirror, where the same data become complex-structure deformations and the periods of the holomorphic $(3,0)$-form are exact at tree level. The periods satisfy a GKZ/Picard-Fuchs system read off from the toric charge vectors, and the field-theory limit is a scaling $a_{r,l} ~ 1/epsilon^{k_r+1-l}$ with $epsilon -> 0$ that sends the base to infinite volume while holding the Coulomb parameters and masses fixed.","core_discovery":"The central claim is that the exact Coulomb branch of any asymptotically free $N=2$ theory built from a product of SU groups with bi-fundamental matter between chosen pairs, and arbitrary fundamental matter for each factor, is the complex-structure moduli space of an explicit local mirror geometry. Two structural results carry it. First, asymptotic freedom is encoded in the matrix $M = 2I - N$, where $N_{ij}$ counts bi-fundamentals between the $i$-th and $j$-th group; a Perron-Frobenius argument on the smallest eigenvalue of $M$ shows the matter graph must be an ADE Dynkin diagram (positive eigenvalue) or an affine ADE Dynkin diagram (zero eigenvalue), and in the affine case the ranks are forced to be proportional to the Dynkin labels, which makes the theory superconformal with no extra fundamentals needed. Second, the mirror of the engineered type IIA threefold is written explicitly: for a linear chain it collapses to a Riemann surface $p(X^*) = a_{0,0} + a_{M+1,0}z^{M+1} + z^r P_r(w)$ with a meromorphic one-form, while for affine $E_6$, $E_7$, $E_8$ the mirror is a threefold whose periods, not a curve, give the central charges. The authors also identify the S-duality group of the conformal cases with the fundamental group of the moduli space of flat ADE bundles on an elliptic curve, or on its degeneration for ordinary ADE.","pith_inferences":["If the classification is right, it is a statement about four-dimensional quantum field theory provable by pure linear algebra: the Perron-Frobenius step needs no string theory, so the ADE constraint could in principle be checked independently of the geometric construction.","The claim that some Coulomb branches need a threefold rather than a curve suggests a sharp diagnostic: any proposed Seiberg-Witten curve for the affine $E$ theories should fail to reproduce the full set of periods, and one could test this by comparing the curve's period lattice against the threefold's.","The S-duality identification with flat bundles on a torus is a prediction that can be confronted with the known spectrum of BPS states: the monodromies generated by loops around the discriminant locus should generate exactly the expected fundamental group.","The base-fiber exchange in the linear chain hints that the distinction between 'gauge coupling' and 'Coulomb modulus' is a choice of which sphere you call the base, which would make the superconformal cases a fixed point of a larger duality relating gauge theories with different gauge groups."],"forward_implications":["Every asymptotically free product-of-SU theory with bi-fundamental matter is classified: its matter graph is an ADE or affine ADE Dynkin diagram, so no other quiver of this type can be asymptotically free.","For the affine ADE cases with ranks proportional to Dynkin labels, the theory is automatically superconformal, and its space of gauge couplings is the moduli space of flat ADE bundles on a torus, with S-duality the fundamental group of that space.","The exact prepotential of each such theory is computable from a Picard-Fuchs system determined by the toric data, giving instanton corrections without any field-theory instanton calculation.","For affine $E_6$, $E_7$, $E_8$ (and the $D_n$ family) the Coulomb branch is not captured by a Riemann surface; the answer is a threefold, so the usual Seiberg-Witten curve is not the universal description of $N=2$ Coulomb branches.","Compactifying further to three dimensions, the Coulomb branch of the $\\prod U(s_i)$ quiver based on affine $E_8$ reproduces the Higgs branch of $k$ small $E_8$ instantons on $T^2$, extending a previously known $k=1$ statement.","Exchanging the roles of base and fiber in the linear-chain geometry suggests a duality $SU(n)^m \\leftrightarrow SU(m+1)^{n-1}$ between theories with different gauge content but the same Riemann surface."],"supporting_citations":[{"why":"Supplies the geometric-engineering prescription itself: type IIA on a local Calabi-Yau, local mirror transform, and reading the exact answer off the type IIB side.","marker":"[12]"},{"why":"Establishes the fivebrane-on-a-Riemann-surface picture and the meromorphic one-form that the linear-chain results here extend to products of groups.","marker":"[8]"},{"why":"Supplies the mechanism by which bi-fundamental matter arises from an enhanced singularity at the intersection of two base spheres, which the quiver construction rests on.","marker":"[15]"},{"why":"Provides the earlier derivation of exact $N=2$ results from ADE singularities fibered over a sphere, the template this paper generalizes.","marker":"[5]"},{"why":"Gives the M-theory derivation of the linear-chain curve with fundamentals, which the present paper recovers from perturbative string symmetries and extends to affine $E$ bases.","marker":"[29]"},{"why":"Supplies the description of the moduli space of flat ADE bundles on an elliptic curve that the mirror geometry is identified with, and hence the S-duality groups.","marker":"[52]"},{"why":"Provides the spectral-cover and del Pezzo surface representation of ADE bundles on elliptic curves that the mirror geometries are matched against.","marker":"[19]"},{"why":"The standard reference for affine Dynkin diagrams and Dynkin labels, which fix the rank assignments in the superconformal cases.","marker":"[23]"},{"why":"Supplies the three-dimensional duality method (compactify, T-dualize, read the Higgs branch as a type IIA Coulomb branch) used in the final section.","marker":"[13]"}],"fun_headline_variants":["Mirror geometry gives the exact Coulomb branch for N=2 SU products","Asymptotic freedom admits only ADE or affine ADE matter graphs","For affine E6, E7 and E8 the exact answer needs a threefold, not a curve","S-duality group identified with the fundamental group of flat ADE bundles","Self-mirror K3 yields many of the exact solutions elementarily"],"cache_read_input_tokens":0,"weakest_assumption_plain":"The construction assumes that taking the scaling limit of the mirror geometry's coefficients, with the base sent to infinite volume, turns the Calabi-Yau period integrals into exactly the rigid special geometry of the four-dimensional gauge theory — an identification that is checked against known cases but not derived, and that has no independent field-theory check for the new affine $D$ and $E$ cases.","fun_headline_variants_meta":{"raw":{"variants":["Mirror geometry gives the exact Coulomb branch for N=2 SU products","Asymptotic freedom admits only ADE or affine ADE matter graphs","For affine E6, E7 and E8 the exact answer needs a threefold, not a curve","S-duality group identified with the fundamental group of flat ADE bundles","Self-mirror K3 yields many of the exact solutions elementarily"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1281,"prompt_tokens":1020,"completion_tokens":261,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":171}},"tokens_in":1020,"tokens_out":261,"duration_ms":4108,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":171,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T19:02:47.631807+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Compute the periods of the mirror threefold for the affine $E_8$ quiver in the stated limit and compare the one-loop piece against the field-theory expression of equation (4.19) with the expected beta-function coefficient; a mismatch in the coefficient, or an extra period without a field-theory counterpart, would break the claim that the threefold reproduces the Coulomb branch.","supporting_citations":[{"cited_title":"Geometric Engineering of Quantum Field Theories","cited_arxiv_id":"hep-th/9609239","evidence_quote":"Supplies the geometric-engineering prescription itself: type IIA on a local Calabi-Yau, local mirror transform, and reading the exact answer off the type IIB side."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Establishes the fivebrane-on-a-Riemann-surface picture and the meromorphic one-form that the linear-chain results here extend to products of groups."},{"cited_title":"Matter From Geometry","cited_arxiv_id":"hep-th/9606086","evidence_quote":"Supplies the mechanism by which bi-fundamental matter arises from an enhanced singularity at the intersection of two base spheres, which the quiver construction rests on."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the earlier derivation of exact $N=2$ results from ADE singularities fibered over a sphere, the template this paper generalizes."},{"cited_title":"Solutions Of Four-Dimensional Field Theories Via M Theory","cited_arxiv_id":"hep-th/9703166","evidence_quote":"Gives the M-theory derivation of the linear-chain curve with fundamentals, which the present paper recovers from perturbative string symmetries and extends to affine $E$ bases."},{"cited_title":"Looijenga, Invent","cited_arxiv_id":null,"evidence_quote":"Supplies the description of the moduli space of flat ADE bundles on an elliptic curve that the mirror geometry is identified with, and hence the S-duality groups."},{"cited_title":"Vector Bundles And F Theory","cited_arxiv_id":"hep-th/9701162","evidence_quote":"Provides the spectral-cover and del Pezzo surface representation of ADE bundles on elliptic curves that the mirror geometries are matched against."},{"cited_title":"Kac, Infinite dimensional Lie algebras , Cambridge University Press 1990","cited_arxiv_id":null,"evidence_quote":"The standard reference for affine Dynkin diagrams and Dynkin labels, which fix the rank assignments in the superconformal cases."},{"cited_title":"Non-Abelian Conifold Transitions and N=4 Dualities in Three Dimensions","cited_arxiv_id":"hep-th/9705220","evidence_quote":"Supplies the three-dimensional duality method (compactify, T-dualize, read the Higgs branch as a type IIA Coulomb branch) used in the final section."}],"review_version":1}