{"id":"7b051f4f-3f22-41eb-8b91-1579565c8677","arxiv_id":"hep-th/9704170","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Intersecting (p,q) five-brane webs are proposed as a diagrammatic construction of five-dimensional superconformal field theories, and three-branes stretched between two such webs give new three-dimensional superconformal theories.","lead":"Using webs of intersecting five-branes in type IIB string theory, the authors build five-dimensional supersymmetric gauge theories and argue that a web collapsed to a point defines a new five-dimensional superconformal field theory. Stretching three-branes between two such webs then yields three-dimensional superconformal theories with several mutually non-local Coulomb branches.","discovery_kind":"unclear","skeptic_critique":{"model":"selfhosted/deepseek-v4.1-flash","headline":"The 5D (p,q)-web construction is well supported, but the abstract's 3D SCFT claim (s3) rests on the classical brane moduli space; §4 never checks whether an Affleck-Harvey-Witten superpotential splits the Coulomb branches it says meet at one point.","rationale":"The reader's verdict (CONDITIONAL, moderate confidence) is right, and the 5D half of the paper holds up: the junction rules, the area-equals-tension identification, and the collapse limit reproduce the SU(2)/SU(3) prepotentials and the Morrison-Seiberg flow pattern, and the framework became standard. I only partially agree with the reader's weakest_assumption. The reader targets the 'no corrections to (3.3)' sentence after (3.2). That worry is mislocated: the piecewise-linear profile is fixed by charge conservation plus the preserved supersymmetry, not by a dynamical modulus, so it is exact for a BPS web; the paper's phrasing is loose but the conclusion is sound. The genuine unacknowledged gap is in §4, where the abstract's headline claim lives. The reader does note that the paper cannot decide whether different asymptotic webs give the same fixed point, but that is an acknowledged limitation (pp. 25-26), so per the rules I do not present it as a new finding. What is not acknowledged is that the 3D 'branches meet at a point' statement is purely classical, while the paper's own §2.5 demonstrates that AHW instanton effects lift Coulomb branches in these very constructions. That is the load-bearing soft spot for s3. It does not overturn the paper — the 5D construction is the durable contribution and the 3D proposal was later developed — but it means the 3D SCFT claim should be read as a proposal, which is consistent with the reader's CONDITIONAL. Hence UNCHANGED.","tokens_in":29922,"tokens_out":13630,"duration_ms":217591,"concrete_test":"In the simplest non-trivial case — Nc=2 D3-branes at the E0 or E1 junction of §3.4 — compute the 3D N=2 Coulomb-branch moduli space independently, using the 3D mirror/duality technology of [14,21] or the M-theory lift of the web, and check whether the two Coulomb branches meet at a single point or are split by an AHW-type superpotential. If split, the abstract's s3 claim weakens to a classical-geometry statement.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract's headline (s3) is that the 5D webs 'describe new superconformal fixed points of three dimensional N=2 supersymmetric theories, which have both electric and magnetic Coulomb branches.' The evidence in §4 is the classical D3-brane moduli space: a 3-brane slides along any web segment and can turn at a junction, so the branches meet at the intersection. The paper then asserts 'at the intersection point we thus expect to get some non-trivial superconformal field theory,' citing only the 4D Argyres-Douglas analogy [28,29,30]. No quantitative check is done: no scaling dimension, no central charge, no operator spectrum, and no verification that the meeting survives quantum corrections. This is not a peripheral worry, because the paper itself invokes the Affleck-Harvey-Witten instanton superpotential in §2.5 as lifting Coulomb branches in 3D N=2 brane constructions. The 3D N=2 Coulomb branch is a Kähler manifold whose complex structure is not protected (unlike the N=4 case), so the classical statement that the (p,q) branches meet at one point is exactly the feature an AHW-type superpotential could split. §4 does not address this. If the branches are split, 'electric and magnetic Coulomb branches meeting at a point' is a classical artifact and the claimed SCFT is not what the brane picture suggests.","agreement_with_reader":"partial"},"referee_report":{"model":"selfhosted/deepseek-v4.1-flash","summary":"The paper has three parts. In §2 the authors generalize the Hanany–Witten 3D N=4 construction by rotating the 5-branes in the 45–89 plane, obtaining 3D N=2 (and, in §2.7, 4D N=1) gauge theories in which the x6 positions of the flavor D5-branes — 'hidden parameters' in [1] — are argued to affect the low-energy superpotential. For various orderings of D and D' branes they propose quartic superpotentials W ∼ (Q Q̃)(q q̃) and check them against the classical brane moduli space, mirror symmetry, Seiberg duality, and the chiral-symmetry conjecture of [6]. In §3, which is independent, they build 5D N=1 gauge theories from webs of (p,q) 5-branes in the x3–x7 plane, with charge conservation at each junction fixing the piecewise-linear profile of Eq. (3.3). The Coulomb branch, the monopole tensions (from polygon areas), and the global symmetry rank n−3 are read off the web, and the SU(2) flows E2 → E1, Ẽ1 → E0 of [8,9] are reproduced. In §4 a stack of Nc D3-branes is stretched in x6 between two copies of such a web, giving 3D N=2 theories in which Coulomb branches associated with different (p,q) directions are mutually non-local and are claimed to meet at the junction.","tokens_in":30239,"tokens_out":5843,"duration_ms":92637,"significance":"The 5D web construction of §3 is the durable contribution: it gives a purely geometric handle on 5D N=1 fixed points that complements the Calabi–Yau/M-theory approach of [8,9,10,24], and the checks are genuine and parameter-free. The monopole tensions computed from polygon areas reproduce the Intriligator–Morrison–Seiberg prepotentials, including the 8−Nf coefficient for SU(2) with Nf = 1, 4 and the pure SU(3) case, and the E0/E1/Ẽ1/E2 flow pattern is recovered from a web with charges {(1,1),(−2,1),(1,−2)} that satisfies charge conservation. The §4 proposal, if correct, opens a systematic route to a large class of 3D N=2 SCFTs with non-local Coulomb branches, which is a genuinely new direction. The §2 material is more heuristic, but it identifies a real phenomenon — that the x6 ordering of flavor branes can change the low-energy theory — that the paper is honest about not controlling.","major_comments":[{"comment":"The central claim of the abstract (s3) — that the electric and magnetic Coulomb branches meet at a single point — is read off the classical D3-brane moduli space: a 3-brane slides along any web segment and can turn at a junction. The paper then asserts 'at the intersection point we thus expect to get some non-trivial superconformal field theory,' citing only the 4D Argyres–Douglas analogy [28,29,30]. This is load-bearing, because §2.5 of the same paper invokes the Affleck–Harvey–Witten instanton superpotential as lifting Coulomb branches in closely related 3D N=2 brane constructions, and the 3D N=2 Coulomb branch is a Kähler manifold whose complex structure is not protected. The paper should either give an argument why an AHW-type effect cannot split the junction, or state explicitly that the meeting is a classical statement whose survival under quantum corrections is assumed. A useful sharpening: along each branch separately the low-energy theory is N=4 (a (p,q) vector multiplet plus a decoupled chiral multiplet from x8,9), so the branch itself is protected; the open question is only the junction, and saying so would make the scope of the claim precise.","section":"§4, paragraph beginning 'Things get more interesting when the 3-brane approaches some intersection'","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The quartic superpotentials of §2 are obtained by fitting the field-theory moduli space to the moduli space read off the brane picture, and the resulting agreement is then presented as evidence that the brane configuration is described by that superpotential. The authors are explicit about this ('we will determine the low-energy superpotentials to fit the global symmetries and the moduli spaces we find', and 'we will again have to be content with guessing the results'), so this is an acknowledged limitation rather than a hidden circularity. What remains unresolved is the framing: the abstract's 'we find brane configurations which correspond to ... SQCD theories with quartic superpotentials' reads as a derived result, whereas §2.4–2.5 establish only that a particular superpotential is consistent with the brane moduli space and the global symmetries. A sentence in §2.4 stating the logic explicitly, and a softening of the abstract, would bring the presentation in line with what is actually shown.","section":"§2.4–2.5 and abstract (s1)","authors_acknowledge":null,"authors_acknowledge_p":null}],"minor_comments":[{"comment":"The sentence 'we get an infinite series of such SCFTs, one for every value of Nc' sits awkwardly with the preceding hedge that non-triviality requires 'more than one D3-brane so that there are charged particles (W bosons) under all the mutually non-local vector multiplets.' For Nc = 1 the intersection theory may be free, so the series should be stated as beginning at Nc = 2, or the Nc = 1 case should be discussed separately.","section":"§4, final paragraph before §4's discussion of chiral symmetry","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The Z2 anomaly for Nc > 2 and odd Nf is noted and dismissed with 'the anomaly is probably cancelled by higher dimensional effects (as in [27,11]), but we have not checked this carefully.' Since the brane construction is offered as a definition of these theories, one sentence on the mechanism — or an explicit statement that the construction defines a theory distinct from the anomalous 5D gauge theory — would help the reader know how to read the Nf-odd entries of §3.3.","section":"§3.3, paragraph after Figure 8","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The argument that only n−3 of the n parameters c_i are physical (two set the origin of x3, x7 and one is fixed by Σc_i = 0) is correct but compressed. Spelling out the n = 3 case, which gives rank 0 and matches the later assertion that E0 has no parameters, would make the counting easy to verify and would tie §3.2 to §3.4.","section":"§3.2, counting of deformation parameters","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The text notes that a 5-brane stays asymptotically NS only when m = n, and that otherwise 'the 5-brane looks different at the far ends.' Since the SU(Nc) constructions with flavors in §3.3 and Figure 8 use precisely such configurations, one sentence on the field-theory interpretation of the unbalanced case (as a theory with a net flavor charge, or as a fixed point without a gauge-theory limit) would connect Eq. (3.3) to the constructions that follow.","section":"§3.3, Eq. (3.3) and the m ≠ n case","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The coefficient of the quartic superpotential is said to carry a negative mass dimension in the UV and to be of order one over the string scale. It would be worth stating whether the brane construction is intended as a UV completion of the 4D theory or as an IR description in which the quartic is a higher-dimension operator, since the two readings differ in what the construction is claimed to establish.","section":"§2.7, 4D N = 1 generalization","authors_acknowledge":null,"authors_acknowledge_p":null},{"comment":"The argument that the U(1) Nf = 2 NS-D-D'-NS' theory is self-mirror, via the S1S2 superpotential in the mirror of [13,14], is clean. A short check that the quantum Coulomb branch also matches — not just the Higgs branch superpotential — would strengthen a claim that is otherwise supported only at the level of the classical moduli space.","section":"§2.6, self-mirror claim","authors_acknowledge":null,"authors_acknowledge_p":null}],"recommendation":"minor_revision","confidential_remarks":"The 5D web construction in §3 is the part of this manuscript likely to have the longest shelf life, and it is checked against published prepotentials in a way that is easy to verify. The §2 material is more speculative and the authors say so; the §4 proposal is a conjecture supported by the classical moduli space and the Argyres–Douglas analogy. My recommendation of minor revision reflects that the fixes I am asking for are local — a paragraph in §4 on quantum corrections to the junction, and a sentence in §2.4 plus an abstract adjustment on the status of the quartic superpotentials — rather than a problem with the central 5D derivation."},"author_rebuttal":null,"desk_editor":{"model":"selfhosted/deepseek-v4.1-flash","letter":"The (p,q) five-brane web in section 3 is the real contribution here, and it holds up. Everything else in the paper is downstream of it, and one section is honest guesswork the authors label as such. What is new: the junction rules — half five-branes of charges (p_i, q_i) meeting at a point with charge conservation, angles fixed by supersymmetry — plus the identification of D3-brane rectangle areas with monopole string tensions, plus the collapse-to-a-point limit as the definition of an infinite-coupling fixed point. That is a new framework, not an application of an existing one. The checks are real: the polygon areas reproduce the Intriligator-Morrison-Seiberg prepotentials for SU(2) with Nf = 0, 1, 4 and for pure SU(3), and the E0/E1/E1-tilde/E2 flow pattern of Morrison-Seiberg falls out of figure 11. The rank n-3 global symmetry counting is a clean, checkable statement. Section 2 is weaker, and the authors say so. They state outright that they cannot write the mass terms for the phi fields, that they are fitting quartic superpotentials to the moduli space, and that the x6-ordering rule is an assumption. I would call that acknowledged rather than circular — the reader's circularity flag is a bit harsh, since the text is explicit about the guess. Still, the section 2 claims are conjectures and should be read as such. On the stress-test: it lands, but proportionately. Section 4's claim that the (p,q) Coulomb branches meet at one point is read off the classical D3-brane moduli space, and the paper never checks whether an Affleck-Harvey-Witten superpotential splits them. The authors know about AHW — they use it in section 2.5 — so this is a gap in section 4's argument, not ignorance. But section 4 is explicitly framed as 'we expect', and the paper does not oversell it. The abstract's 'surprising way' is doing a bit of work there. Who should read it: anyone working on 5D SCFTs, brane webs, or 3D N=2 moduli spaces. The web construction is standard equipment now and this is where it comes from. Recommendation: send it out. The 5D part is solid and the 3D part is a proposal worth refereeing on its own terms.","headline":"The (p,q) five-brane web in section 3 is the real contribution and it holds up against known prepotentials; section 2 is acknowledged guesswork and section 4 is a proposal whose meeting-of-branches claim is never checked against Affleck-Harvey-Witten.","tokens_in":30869,"tokens_out":1777,"would_cite":true,"duration_ms":29781,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"pith_extraction":{"msc":[],"pacs":["11.25.-w","11.30.Pb","11.10.Kk"],"model":"selfhosted/deepseek-v4.1-flash","headline":"Junctions of charged five-branes are proposed as new superconformal field theories in five and three dimensions, with the shape of the brane web encoding the Coulomb branch and the rank of the global symmetry.","keywords":["brane webs","(p,q) five-branes","superconformal field theory","five-dimensional gauge theory","three-dimensional N=2 supersymmetry","Coulomb branch","Seiberg duality","type IIB string theory"],"falsifier":"Take the pure $SU(2)$ fixed point: field theory assigns it an $SU(2)$ global symmetry, yet a generic web with that Coulomb branch shows only a $U(1)$, and the authors find the enhancement only in the special configuration of two overlapping NS branes. A calculation showing that no choice of asymptotic charges reproduces the known global symmetry, or that two webs with the same Coulomb branch have different operator spectra, would settle whether the web geometry really determines the fixed point.","tokens_in":31256,"feed_emoji":"🕸","tokens_out":3573,"duration_ms":53429,"temperature":0.7,"pith_summary":"The paper asks what happens when several five-branes of type IIB string theory, each carrying a pair of charges $(p,q)$ and extended in five spatial directions, are brought together so that all of them meet at a single point. The authors argue that the low-energy theory at such a junction is a five-dimensional $N=1$ superconformal field theory, and that pulling the branes apart without changing their asymptotic charges traces out its Coulomb branch. If that is right, the geometry of the web becomes a dictionary: the number of independent ways to deform the junction gives the dimension of the moduli space, and the number of independent charge directions gives the rank of the global symmetry. The paper then reuses the same object, stretching three-branes between two identical copies of the web separated in a transverse direction, and argues this produces three-dimensional $N=2$ superconformal theories whose Coulomb branches are mutually non-local and meet at one point of moduli space. A separate strand revisits earlier three- and four-dimensional constructions with rotated five-branes and argues that their low-energy theories carry quartic quark superpotentials controlled by transverse brane positions that had been treated as invisible.","feed_headline":"Five-brane junctions may be new superconformal field theories","feed_subtitle":"Webs of charged five-branes meeting at a point would define fixed points in five and three dimensions.","key_machinery":"The load-bearing object is the $(p,q)$ five-brane web: a planar graph in the $x_3$-$x_7$ plane in which each edge is a half five-brane carrying $p$ units of NS charge and $q$ units of D5 charge, with charge conservation enforced at every vertex and the edge angle fixed to $tan heta = p/q$ by the requirement that the surviving supersymmetry not be broken further. The web is the five-dimensional analogue of the Seiberg-Witten curve used in four dimensions: its piecewise-linear profile is claimed to be exact, so areas of the finite faces bounded by the web give monopole string tensions and hence derivatives of the prepotential, while the ways of opening up a face without moving the asymptotic legs give the Coulomb branch. For the three-dimensional construction the same web is used twice, as two parallel copies in $x_6$, with three-branes stretched between them; moving a three-brane along the web then gives a Coulomb branch labelled by the local $(p,q)$ charge, and the branches labelled by different charges are non-local with respect to one another and can only meet at the junction.","core_discovery":"The central claim is that a configuration of half five-branes with charges $(p_i,q_i)$ all emanating from one point, subject only to charge conservation $sum_i p_i = sum_i q_i = 0$, defines a five-dimensional $N=1$ superconformal field theory in its own right, not merely a limit of some gauge theory. The evidence is geometric: the profile of the merged brane is the piecewise-linear function $x_7 = frac{1}{2}(sum_i |x_3 - a_i|) + c x_3 + d$, whose slopes encode the charges and whose kinks sit where branes join, and the angle of each leg satisfies $tan heta = p/q$ so that the residual supersymmetry survives. From this the authors read off that a junction of $n$ half-branes has $n-3$ real deformation parameters (the rest being absorbed by translating the origin and by the charge-sum condition), so its global symmetry has rank $n-3$, and that the Coulomb branches emanating from it correspond to separating branes while holding the asymptotic charges fixed. They then identify known five-dimensional fixed points inside this family, notably the pure $SU(2)$ strong-coupling point as the crossing of a $(1,1)$ and a $(1,-1)$ brane, and the $E_0$ theory as an intersection of three half-branes with charges $(1,1)$, $(-2,1)$ and $(1,-2)$. The second half of the claim is that placing two copies of such a web at different values of $x_6$ and stretching $N_c$ three-branes between them yields a three-dimensional $N=2$ superconformal theory in which the vector multiplets associated with different $(p,q)$ legs are mutually non-local, so the electric and magnetic Coulomb branches are different faces of one moduli space meeting at the junction.","pith_inferences":["If the web really is an exact description, the classification of five-dimensional fixed points becomes a combinatorial problem in integer charge vectors, which would be a much sharper organizing principle than the gauge-theory language it replaces.","The same geometry suggests a natural way to think about the three-dimensional mirror pairs discussed earlier in the paper: mirror symmetry would act on the web by an $SL(2,\\mathbb{Z})$ transformation plus a rotation, so mirror partners would be different presentations of one junction rather than separate theories.","The paper's own admission that it cannot tell whether webs with different asymptotic charges give the same fixed point suggests a concrete diagnostic: compare the spectra of operators of two such webs, since identical fixed points must have identical chiral rings even when the drawings differ.","The quartic superpotentials in the three- and four-dimensional sections, whose sign depends on the ordering of D and D$'$ branes along $x_6$, hint that the transverse positions are genuine parameters of the low-energy theory rather than artifacts, which would make the brane diagram a finer invariant than the field theory it is supposed to represent."],"forward_implications":["Every charge-conserving junction of half five-branes supplies a candidate five-dimensional $N=1$ superconformal field theory, giving a systematic classification scheme for such fixed points in terms of integer charge vectors rather than gauge groups.","The Coulomb branch dimension and the rank of the global symmetry of each such fixed point become computable from the number of branes and their charges, with rank $n-3$ for an $n$-brane junction.","Known fixed points are recovered as special cases, including the pure $SU(2)$ strong-coupling point and the three-brane $E_0$ junction, and the flows between $E_2$, $E_1$, $\\tilde{E}_1$ and $E_0$ can be drawn as motions of the web.","Each five-dimensional fixed point generates an infinite family of three-dimensional $N=2$ superconformal theories, one for each number $N_c$ of stretched three-branes, with electric and magnetic Coulomb branches meeting at a single point.","Constructions that would violate the field-theory bound $N_f \\le 2N_c$ still exist as webs, and the authors conjecture that these correspond to new superconformal theories rather than to the would-be gauge theories."],"supporting_citations":[{"why":"Supplies the original three-brane suspended between five-branes construction and the identification of a brane ending on a five-brane as a magnetic monopole, on which every configuration in this paper is built.","marker":"[1]"},{"why":"Supplies the four-dimensional version of the web construction via M theory, which the five-dimensional section generalises and whose treatment of deformations as parameters versus moduli is reused.","marker":"[3]"},{"why":"Supplies the field-theory claim that pure five-dimensional $SU(2)$ has a nontrivial strong-coupling fixed point, the target that the $(1,1)$/$(1,-1)$ crossing is proposed to reproduce.","marker":"[7]"},{"why":"Supplies the classification of five-dimensional fixed points and the $E_1$, $E_2$, $E_0$ flow structure that the web diagrams are matched against.","marker":"[8]"},{"why":"Supplies the discrete theta-parameter distinction between the two pure $SU(2)$ strong-coupling limits, the ambiguity the paper tries to resolve by varying asymptotic brane charges.","marker":"[9]"},{"why":"Supplies the prepotential, monopole tension formulae, global symmetry ranks and the $N_f > 2N_c$ consistency bound against which the brane-web computations are checked.","marker":"[10]"},{"why":"Supplies the three-dimensional $N=2$ quantum moduli space, including the constraint $V_+V_- (M^{11}M^{22} - M^{12}M^{21})$ used to test which brane configurations survive instanton effects.","marker":"[14]"},{"why":"Supplies mirror symmetry in three dimensions, the operation the paper reinterprets as a transformation of the brane web.","marker":"[4]"}],"fun_headline_variants":["Five-brane webs define 5D and 3D superconformal fixed points","Half-brane junctions alone give 5D N=1 superconformal theories","SU(2) and E0 fixed points sit inside five-brane junction family","Charge-conserving five-brane junctions are superconformal theories"],"cache_read_input_tokens":0,"weakest_assumption_plain":"Everything rests on the claim that the classical piecewise-linear shape of the brane web is exact, so that the low-energy theory at a junction can be read off the drawing without corrections; the paper asserts that no object in the configuration could contribute such corrections, and it cannot decide whether webs with different asymptotic charges describe the same fixed point.","fun_headline_variants_meta":{"raw":{"variants":["Five-brane webs define 5D and 3D superconformal fixed points","Half-brane junctions alone give 5D N=1 superconformal theories","SU(2) and E0 fixed points sit inside five-brane junction family","Charge-conserving five-brane junctions are superconformal theories"]},"model":"selfhosted/deepseek-v4.1-flash","effort":"low","attempts":1,"cost_usd":0.0,"raw_usage":{"total_tokens":1708,"prompt_tokens":1048,"completion_tokens":660,"prompt_tokens_details":null,"completion_tokens_details":{"reasoning_tokens":586}},"tokens_in":1048,"tokens_out":660,"duration_ms":9300,"image_pages":[],"temperature":1.0,"reasoning_budget":50,"reasoning_tokens":586,"transport_provider":"selfhosted","cache_read_input_tokens":0,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-10-29T19:04:29.989658+00:00","model_set":{"reader":"selfhosted/deepseek-v4.1-flash"},"falsifier":"Take the pure $SU(2)$ fixed point: field theory assigns it an $SU(2)$ global symmetry, yet a generic web with that Coulomb branch shows only a $U(1)$, and the authors find the enhancement only in the special configuration of two overlapping NS branes. A calculation showing that no choice of asymptotic charges reproduces the known global symmetry, or that two webs with the same Coulomb branch have different operator spectra, would settle whether the web geometry really determines the fixed point.","supporting_citations":[{"cited_title":"Type IIB Superstrings, BPS Monopoles, And Three-Dimensional Gauge Dynamics","cited_arxiv_id":"hep-th/9611230","evidence_quote":"Supplies the original three-brane suspended between five-branes construction and the identification of a brane ending on a five-brane as a magnetic monopole, on which every configuration in this paper is built."},{"cited_title":"Solutions of Four-Dimensional Field Theori es Via M Theory,","cited_arxiv_id":null,"evidence_quote":"Supplies the four-dimensional version of the web construction via M theory, which the five-dimensional section generalises and whose treatment of deformations as parameters versus moduli is reused."},{"cited_title":"Five dimensional SUSY field theories, non-trivial fixed points and string dynamics","cited_arxiv_id":"hep-th/9608111","evidence_quote":"Supplies the field-theory claim that pure five-dimensional $SU(2)$ has a nontrivial strong-coupling fixed point, the target that the $(1,1)$/$(1,-1)$ crossing is proposed to reproduce."},{"cited_title":"Extremal transitions and five-dimensional supersymmetric field theories","cited_arxiv_id":"hep-th/9609070","evidence_quote":"Supplies the classification of five-dimensional fixed points and the $E_1$, $E_2$, $E_0$ flow structure that the web diagrams are matched against."},{"cited_title":"Small Instantons, del Pezzo Surfaces and Type I' theory","cited_arxiv_id":"hep-th/9609071","evidence_quote":"Supplies the discrete theta-parameter distinction between the two pure $SU(2)$ strong-coupling limits, the ambiguity the paper tries to resolve by varying asymptotic brane charges."},{"cited_title":"Five-D imensional Supersymmetric Gauge Theories and Degenerations of Calabi-Yau Spaces,","cited_arxiv_id":null,"evidence_quote":"Supplies the prepotential, monopole tension formulae, global symmetry ranks and the $N_f > 2N_c$ consistency bound against which the brane-web computations are checked."},{"cited_title":"Aspects of N = 2 Supersymmetric Gauge Theories in Three Dimensions,","cited_arxiv_id":null,"evidence_quote":"Supplies the three-dimensional $N=2$ quantum moduli space, including the constraint $V_+V_- (M^{11}M^{22} - M^{12}M^{21})$ used to test which brane configurations survive instanton effects."},{"cited_title":"Mirror symmetry in three dimensional gauge theories","cited_arxiv_id":"hep-th/9607207","evidence_quote":"Supplies mirror symmetry in three dimensions, the operation the paper reinterprets as a transformation of the brane web."}],"review_version":1}