{"id":"a2bcac22-88d6-4781-97be-d06e0bad54af","arxiv_id":"2608.11946","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors propose modular S and T matrices and boundary RCFT characters for non-unitary TQFTs from generalized S-fold SCFTs, matching Haagerup-Izumi data for special parameter values.","lead":"This paper builds new examples of non-unitary quantum field theories by applying topological twisting to generalized S-fold superconformal theories. It writes explicit modular data and boundary conformal field theory characters, and identifies special cases with the exotic Haagerup-Izumi modular matrices.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Bethe-vacuum enumeration in §4.1 is not shown exhaustive; §5 concedes the §4.2 matrices cannot yet be checked via complete simple lines, so the Haagerup identification rests on an unproven truncation.","rationale":"The reader's weakest assumption was the completeness of the simple-line sets in §3.2; my concern is the analogous but distinct exhaustiveness problem for the Bethe-vacuum ansätze in §4.1, on which the §4.2 modular matrices directly depend. Both are completeness claims about state counting, so there is partial agreement, but the load-bearing step for the stated central claim is not the §3.2 line sets but the §4.1 Bethe-vacuum inventory. The paper is honest about this: Section 5 explicitly says that identifying the complete simple-line set would be needed to verify the §4.2 proposal, and the §4.2 construction is labeled a proposal with only necessary-condition checks. I therefore do not think the concern demands rejection; it is exactly the kind of gap that makes the result conditional rather than established. The proposed numerical root-count of the exact Bethe equations is a concrete way to test whether the ansatz inventory is complete, and would either retire the concern or demonstrate a genuine truncation. The reader's CONDITIONAL verdict remains appropriate.","tokens_in":30225,"tokens_out":10972,"duration_ms":112180,"concrete_test":"Numerically solve the exact Bethe equations (A.3) for n=2 at a Table 1 pair such as (k1,k2)=(-3,27), using homotopy continuation or interval root-counting on the torus (Re Zi, Re Xi, Im Zi, Im Xi mod 2π) at small nonzero ϵ (e.g. 10^-6 and 10^-8), after quotienting by the Weyl and 2πi equivalences. Compare the total number of isolated solutions with the count implied by the ansätze (4.13)–(4.19). Any excess solution invalidates the dimension of the proposed S matrix (4.29).","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central identification in Table 1 is only as secure as the claim that the Bethe vacua found in §4.1 exhaust the full solution set of the exact Bethe equations (A.3). The paper enumerates solutions using the global linear ansatz (A.7), the local linear ansatz (A.18), and the global square root ansatz (A.28), but nowhere proves that no other limiting forms exist. The proposed modular matrices (4.29)–(4.32) are built from exactly this inventory and are tested only against necessary conditions: the SL(2,Z) relations (4.27), a positive row, and integral fusion coefficients via (4.28). These conditions do not uniquely determine modular data. The paper's own Discussion states that identifying the complete set of simple lines 'would allow us to check whether the modular matrices proposed in section 4.2 are correct' — an explicit concession that correctness is presently unchecked. If additional Bethe vacua outside the three ansätze exist for the p±-odd cases, the S matrix (4.29) is a truncation, and the Galois-conjugation match to D^ω H_{g_{2n+1}} in Table 1 would not establish that the generalized S-fold SCFT realizes those modular data.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript studies 3d N=4 rank-0 SCFTs obtained by gauging multiple T[SU(2)] blocks, called generalized S-fold SCFTs S(k-vector), and their topologically A-twisted TQFTs. It uses two complementary descriptions: abelian Chern-Simons-matter DGG theories for φ=LR^n and L^2R^n, and duality-wall S(k-vector) theories for φ=ST^{k1}ST^{k2}...ST^{kn}. In §3 it proposes complete sets of simple lines, RCFT characters, conformal weights, central charges, and modular S matrices for the LR^n and L^2R^n families, with numerical checks of modularity. In §4, via Bethe-vacuum analysis (global linear, local linear, and global square root ansätze), it derives HF data and proposes modular S and T matrices for n=2 with odd k1,k2. For the parameter choices in Table 1 these are claimed, up to Galois conjugation, to agree with the generalized Haagerup-Izumi modular data D^ω H_{g_{2n+1}} of Evans-Gannon. The Discussion lists several open problems, including the absence of a proof that the proposed simple lines are complete and that the Bethe-vacuum ansätze exhaust the solution set.","tokens_in":30425,"tokens_out":5627,"duration_ms":58121,"significance":"If the main proposals hold, the paper provides a physical construction of generalized Haagerup-Izumi modular data and of associated non-unitary TQFTs and boundary RCFTs, extending earlier S-fold constructions. The strengths are the concreteness of the proposal: the modular data contain no fitted parameters, the fusion coefficients (4.28) are checked to be non-negative integers, the characters are given in closed Nahm-sum form, and the identification with the independent Evans-Gannon data [61] in Table 1 is an external benchmark. The paper is also careful to separate what is checked from what is assumed, and the Discussion explicitly flags the main open problems. However, the load-bearing completeness and exhaustiveness assumptions are not proved, and the modularity checks are numerical; the central identification is therefore conditional rather than established.","major_comments":[{"comment":"The paper asserts that the listed line operators form a complete set of simple lines, but only the orthonormality condition (3.4) is checked. Since the modular S matrices (3.15) and (3.20) and the character set are defined on exactly this set, the completeness assertion is load-bearing: any additional simple line would change the dimension of the modular representation and the RCFT characters. The Discussion (Section 5) itself states that finding the complete set of simple lines is a highly nontrivial task. I would like to see either a proof of completeness (for example, through the Bethe-vacuum/simple-line map or an independent counting argument) or an explicit statement that the characters and S matrices are conditional on this assumption.","section":"Section 3.2, Eqs. (3.12) and (3.17)"},{"comment":"The enumeration of Bethe vacua is based on the global linear ansatz (A.7), the local linear ansatz (A.18), and the global square root ansatz (A.28), but no argument is given that every solution of the exact Bethe equations (A.3) in the ϵ→0 limit falls into one of these classes. The modular data (4.29)–(4.32) are built from exactly this inventory and are only checked against necessary conditions: the SL(2,Z) relations (4.27), a positive row, and integral fusion coefficients (4.28). These conditions do not uniquely determine modular data. If additional Bethe vacua exist in the odd-p± cases, the proposed S matrix is a truncation of the true one. The authors should either prove exhaustiveness of the ansatz classes or provide an independent check, for example by constructing the full S matrix from the Bethe-vacuum/simple-line map once the complete line set is known.","section":"Section 4.1 and Appendix A"},{"comment":"The identification of S{φ}|A with D^ω H_{g_{2n+1}} is performed only for the specific pairs (p;k1,k2) listed in Table 1, and it is a Galois-conjugation match of data that are themselves produced by the unproven Bethe-vacuum truncation discussed above. Section 5 concedes that the proposed matrices cannot yet be checked through complete simple lines. The match is therefore a consistency test of the proposal rather than a derivation. This should be stated plainly in the main text, and the claim that the generalized S-fold SCFT realizes the Haagerup-Izumi data should be explicitly flagged as conditional on the completeness and exhaustiveness assumptions.","section":"Section 4.2 and Table 1"}],"minor_comments":[{"comment":"The heading contains a typo: 'Simplie lines' should be 'Simple lines'.","section":"Section 3.2 heading"},{"comment":"The displayed modular matrices are difficult to read in the compiled text, especially the block structures and the repeated row/column labels. Please add explicit dimension labels to each block and ensure the submatrix entries are typeset unambiguously (e.g., the entry '2a112a1D(1)' in (4.29) appears to be missing a subscript or separator).","section":"Equations (3.15), (3.20), and (4.29)"},{"comment":"The numerical modularity checks are described only as checks 'for various values of q'. Please state the range of q, the number of terms checked, and the achieved precision, or move the details to an appendix, so that the reader can assess the strength of the numerical evidence.","section":"Section 3.2"},{"comment":"The lists of conformal weights should be explicitly matched, element by element, to the line operators in (3.12) and (3.17); the current notation with ranges and repeated entries is ambiguous.","section":"Equations (3.13) and (3.18)"}],"recommendation":"major_revision","confidential_remarks":"The paper is within JHEP's scope and is best read as a proposal paper. The main risk is overclaiming completeness: the language in §3 and §4.2 ('complete set', 'realize') goes beyond what is proven, and the authors' own Discussion acknowledges the missing steps. I would encourage the editor to ask for a softening of the claims and for a clearer separation between proven checks and conjectural input. The heavy reliance on the same group's earlier papers, especially [21] and [39], for cross-checks is worth noting but is not by itself disqualifying."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know about this paper. First, it supplies genuinely new explicit data: simple-line sets, characters, and modular S matrices for the phi = LR^n and phi = L^2R^n families (Section 3), and proposed S/T matrices for phi = ST^{k1}ST^{k2} with odd k1,k2 (Section 4.2). Table 1 identifies special parameter values with the Evans-Gannon generalized Haagerup-Izumi data via Galois conjugation. Second, the load-bearing gap is where the stress-test points: simple-line completeness in Section 3.2 is asserted, not proven, and the Section 4.1 Bethe-vacuum enumeration uses three ansatze with no exhaustion proof. Section 5 explicitly concedes that the Section 4.2 matrices could only be verified via complete simple-line data.\n\nWhat the paper does well. The claims are framed as proposals, and the cross-checks are real: SL(2,Z) relations, a positive row, integral fusion coefficients, the HF-data census summing to one, and the S^3 partition function computed two independent ways. The Evans-Gannon matching is a genuine external benchmark - at those points the dimensions line up with a known irreducible modular datum, so it carries weight. I also do not see a serious circularity problem: comparing characters with [21] is a check against a different paper's results, and no constants are fitted anywhere.\n\nSoft spots, in proportion. The unproven completeness is genuine and acknowledged. The modularity checks are numerical rather than proven, and no code or data files are shipped, so independent reproduction is on the reader. The local-linear-ansatz analysis is only done for n=2 with odd levels, and even tr phi is left open. For the generic odd-k family - the part with no external match - the S matrix genuinely rests on the unproven enumeration, so the truncation worry is the right one there. I would only push back on the framing that the Haagerup identification itself rests on a truncation: the Table 1 special points have independent support from the external match.\n\nVerdict: this deserves a serious referee. The gaps are explicit, the data are concrete and checkable, and this subfield operates on exactly this standard - well-grounded, multiply-checked proposals with honest caveats. A reviewer should ask for sharper completeness statements or explicit conjecture framing, broader numerical checks, and any n >= 3 evidence. I would engage with it, and I would expect publication after reasonable revision.","headline":"A solid, honest extension of the S-fold program with genuinely new explicit data; the unproven simple-line completeness is the main gap, honestly flagged, and the external Haagerup-Izumi matches carry real weight.","tokens_in":31027,"tokens_out":6684,"would_cite":true,"duration_ms":59512,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Twisted generalized S-fold SCFTs are proposed to realize non-unitary Haagerup-like TQFTs and boundary RCFTs.","keywords":["non-unitary TQFT","rational conformal field theory","S-fold SCFT","Haagerup-Izumi modular data","Galois conjugation","Bethe vacua","simple lines","3d-3d correspondence"],"falsifier":"Compute the superconformal index with a general linear combination of BPS Wilson lines and check whether any operator perpendicular to the sets (3.12) or (3.17) satisfies the simple-line criterion (3.4); finding one would make the proposed modular $S$ matrices and character sets incomplete, falsifying the completeness claim. Independently, derive the full modular $S$ matrix of $S(\\vec{k})|_A$ from all Bethe vacua through the vacuum-to-simple-line map (4.8); unless it factorizes with the $S_{\\{\\varphi\\}}|_A$ block equal to (4.29), the Section 4.2 proposal is ruled out.","tokens_in":29938,"feed_emoji":"🌀","tokens_out":12066,"duration_ms":100031,"temperature":0.7,"pith_summary":"Generalized S-fold SCFTs—3d N=4 rank-0 theories built from chains of $T[SU(2)]$ blocks gauged at nonzero Chern–Simons levels—are proposed to become non-unitary three-dimensional TQFTs after a topological A-twist. For the monodromy families $\\varphi=LR^n$ and $\\varphi=L^2R^n$, the paper gives complete sets of simple lines and derives the characters and modular $S$ matrices of the associated two-dimensional boundary RCFTs, with central charges $c=-6n+1$ and $c=-6n+2$. For the broader family $\\varphi=ST^{k_1}ST^{k_2}$ with $k_1,k_2$ odd, it proposes explicit modular $S$ and $T$ matrices that satisfy the $SL(2,\\mathbb{Z})$ relations, contain a positive row, and yield integral Verlinde fusion coefficients. For the parameter choices in Table 1, those matrices agree, up to Galois conjugation, with the generalized Haagerup-Izumi modular data $D^\\omega H_{g_{2n+1}}$ previously studied axiomatically. If the proposal is right, twisted generalized S-fold SCFTs give a physical construction of those exotic modular data and of the non-unitary TQFTs and boundary RCFTs they describe.","feed_headline":"Twisted S-fold SCFTs realize Haagerup-Izumi data","feed_subtitle":"Topological twists of generalized S-fold SCFTs yield non-unitary TQFTs and boundary RCFTs with Haagerup-Izumi data.","key_machinery":"The engine is the Bethe-vacuum/HF-data dictionary for A-twisted rank-0 theories: one evaluates the squashed 3-sphere partition function in the $b^2\\to 0$ saddle-point limit, and the solutions of the modified saddle equations—the Bethe vacua—each yield a Handle-gluing $H$ and Fibering $F$. These give the absolute values of the first-row $S$-matrix entries through $H^{-1}=|S_{0a}|^{-2}$ and the phases of $T$, and the vacuum-to-simple-line map fixes the full $S$ matrix. For the $\\varphi=ST^{k_1}ST^{k_2}$ family with odd $k_1,k_2$, the global linear, local linear, and global square root ansätze produce the HF data; the $s=-1$ sector of the square-root ansatz reproduces the 3-sphere partition function and is identified with the residual theory $S_{\\{\\varphi\\}}|_A$. The modular matrices are then pinned down by requiring $S^2=(ST)^3=C$, a positive row, and integral Verlinde fusion coefficients.","core_discovery":"The paper's central claim is that the A-twisted generalized S-fold SCFT $S(\\vec{k})$ carries non-unitary Haagerup-like modular data. For $\\varphi=LR^n$ with $n\\ge 2$, the operators (3.12) are proposed as a complete set of $(2n+6)$ simple lines in $T_{DGG}[\\varphi]|_A$; the associated boundary RCFT has central charge $c=-6n+1$, the characters are Nahm sums whose expansions match the earlier Haagerup RCFT characters, and the modular $S$ matrix is (3.15). For $\\varphi=L^2R^n$ with $n\\ge 1$, the analogous complete set (3.17) of $(2n+4)$ lines gives $c=-6n+2$ and modular $S$ matrix (3.20), with half-integer powers of $q$ signaling fermionic RCFTs. For $\\varphi=ST^{k_1}ST^{k_2}$ with $k_1,k_2$ odd and $p_\\pm=\\operatorname{tr}\\varphi\\pm 2$, the paper proposes the modular matrices (4.29)–(4.32) and verifies $S^2=(ST)^3=C$, $C^2=1$, the existence of a positive row, and non-negative integer fusion coefficients. After the Galois conjugation (4.33), these matrices are identified with the generalized Haagerup-Izumi data $D^\\omega H_{g_{2n+1}}$ of [61] for the choices in Table 1, including $\\omega=3,\\,n=4$ for $(p;k_1,k_2)=(31;-3,27)$.","pith_inferences":["My inference: if the line sets (3.12) and (3.17) are complete, the corresponding Nahm-sum characters are explicit vector-valued modular forms whose $q$-series could be compared with dilogarithm identities, providing an independent arithmetic check on the bulk-boundary dictionary.","My inference: the Galois-conjugation identification suggests that for each pair $(k_1,k_2)$ in Table 1 there is a whole Galois orbit of RCFT data; one could test whether the full character vector transforms covariantly under the Galois group, which would confirm these are genuine RCFTs rather than only modular data.","My inference: the obstruction to $\\operatorname{tr}\\varphi$ even and to $n\\ge 3$ is the lack of a suitable Bethe ansatz; finding one would extend the proposal to arbitrary $n$ and would make the modular $S$ matrix (3.20) a consistency check for the even-trace cases, as the paper notes.","My inference: completeness of simple lines is the sharpest testable assumption—computing the line-operator OPE or the index with all BPS line insertions would either confirm that no line is missing or expose a new primary absent from the proposed characters."],"forward_implications":["For $\\varphi=LR^n$, the proposed $(2n+6)$-line set makes the boundary RCFT data concrete: central charge $c=-6n+1$, explicit conformal weights, and a modular $S$ matrix whose $q$-series passes the modularity checks.","For $\\varphi=L^2R^n$, the $(2n+4)$-line set yields $c=-6n+2$ and fermionic RCFT characters, because the $q$-series contain half-integer powers after the factor $q^{h-c/24}$ is removed.","For $\\varphi=ST^{k_1}ST^{k_2}$ with odd levels, the proposed $S$ and $T$ matrices satisfy the standard consistency conditions one demands of modular data: the group relations, a positive row, and non-negative integer fusion coefficients.","For the parameter values in Table 1, the residual modular data of $S_{\\{\\varphi\\}}|_A$ equal the generalized Haagerup-Izumi data $D^\\omega H_{g_{2n+1}}$ up to Galois conjugation, so these SCFTs give a physical realization of data previously known only from operator-algebra constructions.","For odd $p_\\pm$, the full TQFT factorizes as $S(\\vec{k})|_A = S_{\\{\\varphi\\}}|_A \\otimes TFT[\\vec{k}]$, with the decoupled unitary factor of dimension $2^n$ matching the anomaly-based prediction."],"supporting_citations":[{"why":"Supplies the Bethe-vacuum/HF-data dictionary linking 3d N=4 rank-0 SCFTs to non-unitary TQFT modular data.","marker":"[2]"},{"why":"Provides the global linear ansatz and near-A-twist treatment used for the Bethe vacua of S(k-vector).","marker":"[3]"},{"why":"Gives the line-operator orthonormality criterion (3.4) and the Nahm-sum character construction used for the boundary RCFT.","marker":"[4]"},{"why":"Establishes the non-unitary bulk-boundary correspondence for S-fold SCFTs and provides the prior phi=LR^n RCFT characters that the new line sets reproduce.","marker":"[21]"},{"why":"Defines generalized non-unitary Haagerup-Izumi modular data from n=1 S-fold SCFTs, the framework extended here.","marker":"[39]"},{"why":"Provides the decoupled-TFT/anomaly analysis and the dimension formula used to factor out TFT[k-vector].","marker":"[54]"},{"why":"Supplies the DGG abelian Chern-Simons-matter description T_DGG[phi] from ideal triangulations.","marker":"[58]"},{"why":"Defines the twisted Haagerup-Izumi modular data D^omega H_{g_{2n+1}} that the proposed matrices match after Galois conjugation.","marker":"[61]"},{"why":"Supplies the Handle-gluing/fibering formalism and Bethe/gauge correspondence for extracting modular data from partition functions.","marker":"[71]"}],"fun_headline_variants":["Twisted S-fold SCFTs yield Haagerup-Izumi data","Non-unitary TQFTs from S-fold twists match Haagerup","S-fold twists produce Haagerup-like modular data","New non-unitary TQFTs from generalized S-fold SCFTs","Galois conjugation yields Haagerup-Izumi from S-fold twists"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the proposed line sets in (3.12) and (3.17) are complete lists of simple lines in the A-twisted TQFT; the paper checks their orthonormality but does not prove that no other simple lines exist.","fun_headline_variants_meta":{"raw":{"variants":["Twisted S-fold SCFTs yield Haagerup-Izumi data","Non-unitary TQFTs from S-fold twists match Haagerup","S-fold twists produce Haagerup-like modular data","New non-unitary TQFTs from generalized S-fold SCFTs","Galois conjugation yields Haagerup-Izumi from S-fold twists"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000578,"raw_usage":{"total_tokens":2793,"prompt_tokens":1083,"completion_tokens":1710,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":699,"completion_tokens_details":{"reasoning_tokens":1615}},"tokens_in":699,"tokens_out":1710,"duration_ms":11800,"temperature":1.0,"reasoning_tokens":1615,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:22:09.476475+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the superconformal index with a general linear combination of BPS Wilson lines and check whether any operator perpendicular to the sets (3.12) or (3.17) satisfies the simple-line criterion (3.4); finding one would make the proposed modular $S$ matrices and character sets incomplete, falsifying the completeness claim. Independently, derive the full modular $S$ matrix of $S(\\vec{k})|_A$ from all Bethe vacua through the vacuum-to-simple-line map (4.8); unless it factorizes with the $S_{\\{\\varphi\\}}|_A$ block equal to (4.29), the Section 4.2 proposal is ruled out.","supporting_citations":[],"review_version":1}