{"id":"11d6057d-4372-4ccb-8350-17ccdc80c2ee","arxiv_id":"2505.04706","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The vacuum Schur-index modular orbit is proposed as the full VOA module-character space for several a=c theories, with a conjectured dimension formula 1+3ℓ(2+ℓ) for the T_{2,2ℓ+1} series.","lead":"The paper computes modular data, including character spaces and S,T matrices, for vertex operator algebras attached to the a=c family of 4d N=2 superconformal theories T_{p,N}. It proposes a general dimension formula for one series and a map from N=4 SU(N) module characters to these theories.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The orbit span V0 is identified with the full module-character space via the unproved equality n0 = nmin = nord; for the infinite T2,2ℓ+1 family this is explicitly conjectural, so the central claim is not yet established.","rationale":"The paper makes concrete and partly independent computational contributions: exact closed forms for Schur indices, explicit S,T matrices and Jordan block data for low examples, MLDEs for T2,3, T2,5, and T4,3, and a count of affine Springer fixed varieties. However, the strongest claim — that the modular orbit span is the full space of V[Tp,N] module characters — is not established. The authors themselves flag the conjecture for T2,2ℓ+1, the missing MLDEs for T3,4 and T2,7, and the mismatch of fixed-variety dimensions in Section 4. The reader's weakest assumption, n0 = nmin = nord, is exactly the load-bearing point: without it, the S,T matrices and Jordan block sizes describe only the orbit of the vacuum character, not necessarily the representation theory of the VOA. A CONDITIONAL verdict is therefore the right outcome, and my stress-test does not move it. The proposed check for T2,7 would materially reduce the gap, while a failure would force the conjecture to be downgraded or restricted.","tokens_in":52668,"tokens_out":5052,"duration_ms":55823,"concrete_test":"Construct, for T2,7 (ℓ = 3), the order-46 unflavored MLDE annihilating ch0 (e.g., from the null-state argument or by solving for modular-form coefficients), and check that its solution space is exactly the 46-dimensional orbit span V0 with the predicted indicial roots. If such an MLDE exists, nmin = n0 = 46 and the chain (2.74) is verified for the first nontrivial new case; if no MLDE of order 46 exists, or its solution space strictly contains V0, the identification fails. As a cheaper ancillary check, numerically evaluate the 46 proposed basis functions from (3.52) to high q-order for ℓ = 3 and confirm rank 46, isolating the dimension claim from the full-character claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim — that the modular orbit of the unflavored Schur index spans the full space of V[Tp,N] module characters — rests on the chain n0 = nmin = nord in the inequalities (2.74). The paper computes n0 = 1 + 3ℓ(2 + ℓ) for T2,2ℓ+1, but nmin is constructed only for low cases (T2,3, T2,5, T4,3) and not for the infinite family; for T3,4 no MLDE was found, so even nmin is unknown. The final paragraph of Section 3.2 states explicitly: \"We conjecture that this span is the space of V[T2,2ℓ+1]-characters.\" If additional ordinary or logarithmic modules exist outside V0, the computed S,T matrices and Jordan block counts describe only a subspace, and the match with the number of affine Springer fixed varieties in Section 4 would be a count coincidence. Section 4's opening sentence claims these discussions \"establish ... the full space of simple and logarithmic modules characters,\" which overstates the evidence. The dimension formula also assumes without proof that the three families in (3.52) are linearly independent for all ℓ; explicit bases are exhibited only for ℓ = 1 and ℓ = 2.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the modular properties of the vertex operator algebras V[T_{p,N}] associated to the infinite series of 4d N = 2 SCFTs with a = c, focusing on SU(N) gauge group. Starting from the exact closed-form Schur index of N = 4 SU(N) SYM and the specialization b = q^{p/2-1}, q -> q^p, the authors express I_{T_{p,N}} as a polynomial in twisted Eisenstein series. For the infinite family T_{2,2ℓ+1} they compute the dimension 1 + 3ℓ(2 + ℓ) of the span V0 of the SL(2,Z)-orbit of the vacuum character, construct a basis, compute S and T matrices for low ℓ, and conjecture that V0 is the full space of V[T_{2,2ℓ+1}]-characters. For T_{3,2}, T_{3,4}, and T_{4,3}, they construct orbit spans of dimensions 5, 33, and 13 respectively, finding explicit non-monic MLDEs for T_{3,2} and T_{4,3} but not for T_{3,4}. They then compare the number of non-logarithmic characters with the number of fixed varieties in the affine Springer fiber for the class-S cases T_{3,2} = (A2,D4), T_{4,3} = (A3,E6), and T_{6,5} = (A5,E8). Finally, using a difference operator and a specialization map, they propose a relation between module characters of V[T_{SU(N)}] and V[T_{p,N}], with explicit checks for T_{2,5}, T_{2,7}, and T_{3,4}.","tokens_in":52928,"tokens_out":6421,"duration_ms":64707,"significance":"If the central identification is established, the paper provides a concrete infinite family of non-rational quasi-lisse VOAs with explicit modular data: dimension of the character span, S and T matrices, and Jordan block structure of T. It also connects this modular data to the Coulomb branch geometry through affine Springer fibers, and proposes a systematic map from V[T_{SU(N)}] modules to V[T_{p,N}] modules. The computations are explicit and reproducible, no free parameters are fitted, and the paper is honest in labeling the T_{2,2ℓ+1} span-to-character-space identification as a conjecture. These strengths make the paper valuable even though the full characterization of the module-character space is not yet proven.","major_comments":[{"comment":"The central claim that the modular orbit span V0 equals the full space of V[T_{p,N}]-module characters rests on the chain n0 = nmin = nord in the inequalities (2.74), but this chain is not derived. The paper itself states, in the final paragraph of Section 3.2, \"We conjecture that this span is the space of V[T_{2,2ℓ+1}]-characters,\" and for T_{3,4} in Section 3.3 it states \"we have not constructed an order-33 MLDE to verify nmin = n0.\" Since nmin is not known for the infinite family and no MLDE is known for T_{3,4}, the identification of V0 with the full module-character space is not established. The authors should either prove that all module characters lie in V0 by an independent argument, or consistently present the full-character-space claim as conjectural throughout, including in the abstract and the introduction.","section":"Section 2.3 and Section 3.2"},{"comment":"The opening sentence of Section 4, \"The previous discussions establish the modularity properties of the T_{p,N} theory, providing the full space of simple and logarithmic modules characters of the associated VOA V[T_{p,N}],\" overstates what has been shown. For T_{2,2ℓ+1} the statement is explicitly conjectural, and for T_{3,4} no MLDE has been found, so the space of module characters has not been determined. Even for T_{4,3}, the existence of an order-13 MLDE satisfied by the vacuum character, together with dim V0 = 13, shows that V0 is the full solution space of that MLDE, but it does not by itself show that all V[T_{4,3}]-module characters satisfy this MLDE. The authors should either supply a VOA-level argument (for example, a null-state construction or a flavored MLDE analysis) that identifies the module characters with solutions of the unflavored MLDE, or soften the claim accordingly.","section":"Section 4, opening paragraph"},{"comment":"The dimension formula dim V0 = 1 + 3ℓ(2 + ℓ) assumes that the three families of objects in (3.52) are linearly independent for all ℓ. Explicit bases are exhibited only for ℓ = 1 and ℓ = 2, and accidental linear relations among twisted Eisenstein series are known to occur in this paper (for example, in the T_{4,3} analysis in Section 3.4). The authors should provide an argument that no such relations occur for the (3.52) families, for instance by a leading-order q-expansion analysis or a modular-forms dimension count, before the dimension formula can be regarded as established for the whole infinite series.","section":"Section 3.2, Eqs. (3.52)-(3.53)"},{"comment":"The geometric match in Section 4 is only a match of the number of allowed translations with the number of non-logarithmic Jordan blocks, not a match of the dimensions of the fixed varieties. The text acknowledges this for T_{3,2}: \"Unfortunately, these dimensions do not match with the Jordan block structure of the T matrix,\" and for T_{4,3} the naive dimensions [16,16,16,8,5,0] do not match the Jordan block sizes [3,3,2,2,2,1]. If V0 is not the full character space, the numerical match of counts would be a formal coincidence. The authors should clarify what precise statement about the geometric side is being compared with which modular datum, and should explain why only the count, and not the dimensions, is expected to match.","section":"Section 4, affine Springer calculations"}],"minor_comments":[{"comment":"The sentence \"Since V[T_{p,N}] has no residual flavor symmetry, we expect only ordinary modules\" appears to conflict with the later use of logarithmic modules and non-logarithmic solutions; the authors should clarify which modules are ordinary, which are logarithmic, and how the absence of flavor symmetry constrains this distinction.","section":"Section 3, first paragraph"},{"comment":"The notation \"eq\" in equations (2.66)-(2.69) seems to be used without definition; the reader is left to infer that it denotes a flavored MLDE or a set of equations. Please define this notation explicitly.","section":"Section 2.3, Eqs. (2.66)-(2.69)"},{"comment":"The statement \"Also, T ch16 = 0\" cannot hold because T is an invertible linear operator on the space spanned by the characters; this is likely a typo for something like (T - id) ch16 = 0 or T ch16 = ch16. Please correct it.","section":"Section 3.3, after Eq. (3.80)"},{"comment":"Equation (3.58) writes I_{3,2} in terms of E1 at q^{1/6} after the identity (3.57), whereas equation (3.12) writes I_{3,2} = E1[-1/√q](3τ). The equivalence is presumably the identity (3.57) applied with p = 3, but the reader must reverse-engineer this; please make the relation explicit.","section":"Section 3.3, T_{3,2} formulas"},{"comment":"The difference operator in equation (1.4) uses the shorthand ch(bq,q), but later applications such as Δ(5) ch0(b,q^2)|_{b0} in Section 5 mix the notations b and q in a way that is hard to follow; please add a sentence explaining the convention for evaluating the b-expansion after the specialization.","section":"Section 1 and Section 5"},{"comment":"Figure 2 is referenced in the T_{3,2} discussion, but the figure itself is not included in the text; either include the figure or delete the reference.","section":"Figure 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is honest about the conjectural status of the T_{2,2ℓ+1} character-space identification, and the explicit computations are a real asset. However, the abstract and Section 4 claim more than is proven: the equality n0 = nmin = nord is not established for the main infinite family, and no MLDE is known for T_{3,4}. I recommend major revision with the expectation that the authors either supply the missing supporting arguments or reformulate the central claims as conjectures throughout, including the abstract and the opening of Section 4."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Let me give you the short version. Pan and Yang work out the modular orbit of the unflavored Schur index for the T_{p,N} theories with a=c, give a dimension formula 1+3ℓ(2+ℓ) for T_{2,2ℓ+1}, compute S,T matrices for the low cases, find lower-order MLDEs for T_{2,3} and T_{4,3}, and count affine Springer fixed varieties for T_{3,2}, T_{4,3}, T_{6,5}. They also propose a character map from N=4 SU(N) modules via a difference operator Δ(N). The concrete computations are new and, as far as I can tell, correct at the level of algebra. I would not be surprised if this becomes a standard reference for these examples.\n\nThe soft spot is the one the authors themselves flag: the orbit span V0 is identified with the full module-character space of V[T_{p,N}] only for a few low cases, where an MLDE of matching order exists. For the infinite T_{2,2ℓ+1} series this is a conjecture (end of Section 3.2), and for T_{3,4} and T_{2,7} no MLDE was found, so even nmin is unknown. If extra ordinary or logarithmic modules sit outside V0, then the S,T matrices and Jordan block counts describe only a subspace. Section 4 opens by saying the previous discussion 'establish[es] the full space of simple and logarithmic modules characters' – that overstates what is proven; for T_{3,2} and T_{6,5} the dimensions of fixed varieties do not match the Jordan block sizes, only the counts do. The paper honestly notes the mismatch in a footnote, but the opening sentence should be softened.\n\nTwo smaller points. First, the dimension formula 1+3ℓ(2+ℓ) assumes the three families in (3.52) are linearly independent for all ℓ; explicit bases are only shown for ℓ=1,2. That is probably true, but it is an unproved assumption. Second, no code or notebook is shipped, and several MLDEs are asserted 'by direct computation'. Given how large the coefficients are, a referee should ask for the computation to be reproducible.\n\nNet: this is a useful, honest paper with real new computations and a clearly labeled conjecture. The math is not circular and no parameters are fitted. If I were the editor I would send it out; the referee should push for either a proof of n0=nmin for at least one infinite family, or an explicit statement in the abstract and intro that the full-character-space identification is conjectural. Also fix the Section 4 overstatement and either ship the computer algebra or describe the algorithm.","headline":"Concrete new modular-orbit computations for a=c SCFTs, with the full-character-space identification honestly labeled but not yet proven.","tokens_in":53529,"tokens_out":2761,"would_cite":true,"duration_ms":27146,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"One index's modular orbit determines all VOA characters","keywords":["vertex operator algebra","Schur index","modular linear differential equation","a=c SCFT","4d mirror symmetry","affine Springer fiber","Eisenstein series","logarithmic module"],"falsifier":"Compute an explicit unflavored modular linear differential equation of order 33 for $T_{3,4}$; if its solution space has dimension different from 33, or if such an equation does not exist, the modular orbit does not span the character space. For $T_{2,7}$ the predicted order is 46: an order-46 MLDE whose solution space is not 46-dimensional would disprove the conjecture.","tokens_in":52420,"feed_emoji":"🧮","tokens_out":10491,"duration_ms":90020,"temperature":0.7,"pith_summary":"The paper aims to show that, for the infinite family of 4d $\\mathcal{N}=2$ superconformal field theories with equal central charges $a=c$, the modular orbit of the unflavored Schur index spans the full space of characters of the associated vertex operator algebra (VOA). For the $T_{2,2\\ell+1}$ series this is made quantitative: the character space would have dimension $1+3\\ell(2+\\ell)$, and the $T$-matrix would have Jordan blocks of the predicted sizes. If true, this turns the representation theory of these non-rational, quasi-lisse VOAs into explicit modular data computable from a single index. The paper also connects the counting of non-logarithmic modules to fixed loci of affine Springer fibers in the Coulomb branch, a geometric check coming from 4d mirror symmetry.","feed_headline":"One index's modular orbit determines all VOA characters","feed_subtitle":"This gives explicit S and T matrices for non-rational vertex algebras that are usually intractable.","key_machinery":"The load-bearing object is the modular orbit of the vacuum character: the set of $SL(2,\\mathbb{Z})$ transforms of the unflavored Schur index $I_{T_{p,N}}(q)$. The paper writes $I_{T_{p,N}}$ as a polynomial in twisted Eisenstein series by specializing the closed-form $\\mathcal{N}=4$ $SU(N)$ Schur index, so modular transformations become linear algebra on monomials. A modular linear differential equation (MLDE), an ordinary differential equation in $q$ whose coefficients are modular forms, constrains these characters; the paper finds non-monic MLDEs whose order equals the dimension of the orbit span whenever possible. The other key tool is the difference operator $\\Delta^{(N)}\\mathrm{ch}(b,q)=b^{-(N^2-1)}q^{-(N^2-1)/2}\\mathrm{ch}(bq,q)-\\mathrm{ch}(b,q)$, which acts on $\\mathcal{N}=4$ $SU(N)$ characters and, after the specialization $b\\to q^{p/2-1}$, $q\\to q^p$, produces solutions to the $T_{p,N}$ equations.","core_discovery":"On the paper's own terms, the central discovery is that the closed-form Schur index of $T_{p,N}$, written as a polynomial in Eisenstein series, has a finite $SL(2,\\mathbb{Z})$-orbit whose span $V_0$ is the space of $\\mathbb{V}[T_{p,N}]$ module characters. The paper conjectures this for all $T_{2,2\\ell+1}$, where it computes $\\dim V_0 = 1+3\\ell(2+\\ell)$; the $T$-matrix then has $1+3\\ell$ Jordan blocks with sizes $[2\\ell+1,\\dots,5,5,5,3,3,3,1]$, each block belonging to a non-logarithmic character. For $T_{3,2}$, $T_{3,4}$, and $T_{4,3}$ the same construction yields dimensions 5, 33, and 13 with explicit $S$ and $T$ matrices, and where a modular linear differential equation is found, its order agrees with the dimension of $V_0$. The paper further proposes a map from $\\mathcal{N}=4$ $SU(N)$ module characters to $T_{p,N}$ characters, realized through a difference operator in the flavor fugacity, and matches the number of non-logarithmic modules with the number of Coulomb-branch fixed varieties in class-S examples.","pith_inferences":["If the orbit-span conjecture holds for all $T_{2,2\\ell+1}$, the family becomes a testbed for logarithmic VOA bootstrap: a single vacuum character fixes all logarithmic module data, including Jordan block sizes, without any input from a constructed module category.","The dimensions of affine Springer fixed varieties computed in Section 4 do not match the Jordan block sizes of the $T$-matrix, even though the number of fixed varieties does; this suggests the geometric count should be refined to encode logarithmic data, a direction the paper leaves open.","The difference-operator construction could be applied with higher powers of $\\Delta^{(N)}$ for other $N$ and $p$; each iteration that lands in the modular orbit would certify a new module character, giving a practical algorithm independent of finding an MLDE.","For $T_{3,4}$, an explicit order-33 MLDE is the natural next computation; its existence would turn the 33-dimensional orbit span into a proven character space, and its failure would show exactly where the orbit span falls short of the full module-character space."],"forward_implications":["For every odd $N=2\\ell+1$, the VOA $\\mathbb{V}[T_{2,N}]$ is predicted to have exactly $1+3\\ell(2+\\ell)$ characters, with the $T$-matrix's Jordan block pattern fixed by $\\ell$; explicit character bases follow from the modular orbit.","The nilpotency index of these VOAs is approximated by $\\dim V_0$, and for $T_{3,2}$ this value saturates the bound $n-1\\ge \\mathrm{rank}$, supporting the use of modular orbit data as a proxy for nilpotency.","Because the $S$ and $T$ matrices are constructed explicitly, Verlinde-type fusion coefficients and modular data become available for non-rational quasi-lisse VOAs where such data is usually hard to obtain.","The proposed character map from $\\mathbb{V}[\\mathcal{T}_{SU(N)}]$ to $\\mathbb{V}[\\mathcal{T}_{p,N}]$ supplies a systematic way to generate module characters of the $a=c$ theories from the better-understood $\\mathcal{N}=4$ side.","In class-S realizations, the number of non-logarithmic modules matches the number of Coulomb-branch fixed varieties, so the modular data and the 4d mirror-symmetry geometry carry the same module count."],"supporting_citations":[{"why":"Constructs the $T_{p,N}$ theories, gives the $a=c$ condition, and supplies the Schur-index relation that the modularity analysis starts from.","marker":"[35]"},{"why":"Prior study of modular linear differential equations for the $T_{p,N}$ family; provides notation and the order-12 MLDE that the lower-order equations are compared with.","marker":"[37]"},{"why":"Establishes that quasi-lisse VOA characters satisfy modular linear differential equations, the basis for using MLDEs on these non-rational algebras.","marker":"[14]"},{"why":"Identifies the Higgs branch with the VOA associated variety and argues for MLDEs, grounding the non-rational representation-theory setup.","marker":"[15]"},{"why":"Provides the closed-form flavored $\\mathcal{N}=4$ $SU(N)$ Schur index via Fermi-gas methods that the paper differentiates and specializes.","marker":"[40]"},{"why":"Gives the exact Schur index in closed form and the elliptic and Eisenstein integration identities used to write indices as Eisenstein polynomials.","marker":"[41]"},{"why":"Develops affine Springer-fiber fixed loci for Argyres-Douglas theories, the geometric mirror-symmetry method applied to $T_{3,2}$, $T_{4,3}$, and $T_{6,5}$.","marker":"[45]"},{"why":"Defines the nilpotency index used to estimate $n$ and to check the rank inequality against the modular-orbit dimension.","marker":"[42]"},{"why":"Supplies the flavored modular differential equation and free-field realization perspective behind the difference-operator construction of module characters.","marker":"[30]"}],"fun_headline_variants":["One index orbit yields all VOA characters","Finite modular orbit gives explicit S,T matrices","Schur index orbit spans the character space","From one polynomial to all S,T matrices","Modular orbit of the index fixes characters"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the finite space spanned by modular transforms of the vacuum index already contains every module character of the associated VOA; this is checked in small examples but conjectured for the infinite family, and for $T_{3,4}$ and $T_{2,7}$ no modular linear differential equation has been found to confirm it.","fun_headline_variants_meta":{"raw":{"variants":["One index orbit yields all VOA characters","Finite modular orbit gives explicit S,T matrices","Schur index orbit spans the character space","From one polynomial to all S,T matrices","Modular orbit of the index fixes characters"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001052,"raw_usage":{"total_tokens":4489,"prompt_tokens":1084,"completion_tokens":3405,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":700,"completion_tokens_details":{"reasoning_tokens":3338}},"tokens_in":700,"tokens_out":3405,"duration_ms":23247,"temperature":1.0,"reasoning_tokens":3338,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:23:17.781073+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute an explicit unflavored modular linear differential equation of order 33 for $T_{3,4}$; if its solution space has dimension different from 33, or if such an equation does not exist, the modular orbit does not span the character space. For $T_{2,7}$ the predicted order is 46: an order-46 MLDE whose solution space is not 46-dimensional would disprove the conjecture.","supporting_citations":[{"cited_title":"Defects, modular differential equations, and free field realization of N = 4 VOAs","cited_arxiv_id":"2104.12180","evidence_quote":"Supplies the flavored modular differential equation and free-field realization perspective behind the difference-operator construction of module characters."}],"review_version":1}