{"id":"a507243b-80ba-49c5-b907-34b47480004f","arxiv_id":"2505.15144","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The E8 non-supersymmetric heterotic string on T^2 has exactly 22 maximal gauge enhancements, found by two independent methods, with massless spectra listed for each.","lead":"This paper maps out the largest possible gauge symmetries that arise when the non-supersymmetric E8 heterotic string is compactified on a two-dimensional torus, finding 22 maximal groups. The list fills a gap in the string theory landscape and, assuming the cobordism conjecture, doubles as a catalog of non-supersymmetric branes in eight dimensions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Completeness of the 22 is inherited, not proven: the paper never shows that every E8-string maximal enhancement on T^2 occurs at a point where the parent E8×E8 theory is maximally enhanced, and both approaches start from the same parent list.","rationale":"The reader's CONDITIONAL verdict is appropriate, and my concern refines rather than replaces the reader's weakest-assumption. The reader flagged Table-12 completeness and the symmetric-Wilson-line-locus requirement; the latter is actually automatic (the E8-string moduli space is the R-fixed locus, so A=(a,a) is not a restriction but the definition of the theory), while the sharper gap is surjectivity: every E8-maximal point would have to sit at a parent-maximal point on that locus. I cannot find a proof of this in the manuscript. Section 5.4.1 states the selection criterion; Section 5.5 asserts the folded-EDD procedure (showing 20 cases and leaving six to 'the above procedures'); neither demonstrates exhaustiveness on the E8-string side. Because both approaches share the same parent input, their agreement is a genuine check of the lattice computations and spectra but cannot settle completeness. I credit the paper's concrete evidence: the explicit invariant and dual lattices I and I* for 14 cases in Section 4, the massless spectra in Table 2, the two approaches matching at all 22 entries, and the parameter-free nature of the construction. The brane interpretation in Section 4.15 is explicitly conditional on the no-global-symmetry/cobordism conjecture, which the paper states, so I do not count that as a defect of the central claim. Secondary issues, not load-bearing: the group-theory expressions in Section 4 (e.g., (4.5), (4.70)) include U(1)×U(1) factors that give apparent rank 12, inconsistent with the rank-10 invariant lattice and with Table 2; several typos ('brnae', 'theor is') and the unshown EDD cases for #5,6,25,28,54,87 should be corrected. The proposed exhaustive long-root scan is the single check that would settle whether the 22 are complete; absent that check, CONDITIONAL remains the right verdict.","tokens_in":29395,"tokens_out":47913,"duration_ms":395020,"concrete_test":"Run a Table-12-independent exhaustive scan of the E8 string on T^2: enumerate all moduli (a1,a2,E) that can support long roots — a1,a2 over the finite set of rational Wilson-line values obtained from the Kac-mark/shift algorithm (as in [17,18]), and E over all enhanced (2,2) Narain lattices (including E1,E2 and any further R-invariant enhanced tori) — and solve the invariant-sector condition (5.17) with π+ ∈ 2Γ_E8, P_L²=2, p_R=0. Collect every point whose root system has rank 10 and compute the gauge group from the invariant lattice. If the outcome is exactly the 22 groups at the Table-5 moduli, the surjectivity gap is closed; if an additional rank-10 group appears or a Table-5 point is missed, the classification is incomplete. This scan is finite and mechanical, mirroring the exhaustive procedure applied to the SUSY side in [17], and it tests the E8-string completeness claim directly.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is exhaustive: exactly 22 maximal gauge groups for the E8 string on T^2. Both approaches feed off the same input, the classification of maximal enhancements of the parent supersymmetric E8×E8 theory on T^2 (Table 12 of [17]). Approach 2 (§5.4.1) filters Table 12 to the symmetric Wilson-line locus A=(a,a) and solves (5.17); Approach 1 (§4) folds the lattices at those same Table-12 points. The agreement of the two approaches is therefore an internal consistency check of the computations, not evidence of completeness: a maximal enhancement of the E8 string occurring at a point where the parent is not maximally enhanced would be missed by both. In the E8 string, all untwisted gauge bosons are long roots (π-=0, π+∈2Γ_E8, P_L²=2; §5.2.1 and (5.17a)). A point is E8-maximal when these long-root solutions form a rank-10 system. The parent at the same moduli has additional short-root solutions (π-²=1, P_L²=1) that are anti-invariant and hence projected out of the orbifold; the paper does not prove that moduli maximizing the invariant (long-root) sector also maximize the anti-invariant sector. The symmetric-locus restriction (footnote 3) is automatic, since the E8-string moduli space is the R-invariant subspace, so it does not supply exhaustiveness. Section 5.5's generalized-Dynkin-diagram argument is asserted (20 cases shown; six 'can also be obtained by the above procedures'), not a proof that the folded EDD enumerates all maximal enhancements. Together with reliance on [17]'s Table 12 for the starting list, the 22 entries rest on an unproven surjectivity step, so the classification's completeness is the load-bearing assumption.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the maximal gauge groups of the non-supersymmetric rank-reduced E8 heterotic string compactified on T^2. Two approaches are used. First, starting from the 8d supersymmetric E8 x E8 heterotic string at points of maximal gauge enhancement from [17], the authors orbifold by a Z2 outer automorphism combined with spacetime fermion parity and compute the resulting massless spectra. Second, they compactify the 10d E8 heterotic string directly on T^2 and solve the massless vector and matter conditions (5.17). Both approaches yield the same list of 22 maximal gauge groups, summarized in Table 2, with moduli listed in Table 5. A generalized-Dynkin-diagram method is proposed in Section 5.5, and in Section 4.15 the results are interpreted, conditional on the no global symmetry/cobordism conjecture, as a classification of codimension-two non-supersymmetric branes in 8d supersymmetric heterotic theories.","tokens_in":29787,"tokens_out":6919,"duration_ms":59735,"significance":"If the classification is complete, the paper provides the first exhaustive list of maximal rank-10 gauge enhancements of the E8 heterotic string on T^2, with explicit massless spectra and moduli. The explicit lattice constructions in Section 4, the modular-invariant partition functions of Section 2, and the agreement between two independent computational approaches are genuine strengths that make the arithmetic results highly plausible. The brane interpretation in Section 4.15 is conditional on a quantum-gravity conjecture but is clearly flagged as such. The main gap is that the exhaustiveness of the 22-entry list is inherited from the parent classification in [17] rather than proven directly for the E8 string; this affects the central claim and requires additional work.","major_comments":[{"comment":"The central claim that Table 2 classifies all 22 maximal gauge groups of the E8 string on T^2 is not proven. In Approach 2, Eq. (5.17) is solved only at the moduli listed in Table 12 of [17] that satisfy the symmetric Wilson-line condition A=(a,a); in Approach 1, the same Table-12 points are folded. The agreement of the two approaches therefore verifies the arithmetic at a common set of input points but does not exclude a maximal enhancement of the E8 string at a point where the parent E8 x E8 theory is not maximally enhanced. The symmetric-locus restriction is automatic for the orbifold construction and does not by itself supply exhaustiveness. I ask for either a direct scan of the E8-string moduli solving (5.17) without restricting to parent maximal-enhancement points, or a proof that any solution maximizing the invariant long-root sector is already a maximal enhancement of the parent theory.","section":"§5.4.1 and §1"},{"comment":"The generalized Dynkin diagram method is presented as the mechanism that yields the maximal enhancements, but its completeness is asserted rather than demonstrated. Figures 9-11 show only three representative choices of (E, a1, a2), and the text states that the six cases #5, 6, 25, 28, 54, 87 'can also be obtained by the above procedures' without displaying the deletion chains or the resulting diagrams; cases #1 and #2 are not covered by the method at all. Since the exhaustiveness of the classification rests on this enumeration, the paper should either provide the full list of diagrams and deletion sequences for all 22 cases or state explicitly which part of the completeness argument is inherited from [17].","section":"§5.5"}],"minor_comments":[{"comment":"The invariant sublattice displayed in Eq. (2.15) has inconsistent normalizations: from Eq. (2.14), one has alpha_4 = sqrt(2) alpha(F4)_3 and alpha_2 = sqrt(2) alpha(F4)_4, so the coefficients 1/sqrt(2) in the third and fourth terms appear to be typos, and the last term should presumably involve alpha(F4)_4 rather than alpha(F4)_1.","section":"§2.2.3, Eq. (2.15)"},{"comment":"The phrase 'non-simply raced' should read 'non-simply laced'.","section":"§5.2.1"},{"comment":"The word 'brnae' should read 'branes'.","section":"§4.15"},{"comment":"For rows #1 and #2, several entries in the E11, E12, E21, E22 columns are left blank; please indicate explicitly that these entries are zero.","section":"Table 5"},{"comment":"The paper freely switches between the covering groups and the quotient gauge groups appearing in equations such as (4.5), (4.10), and (4.15); a sentence clarifying that the classification is by the global form of the gauge group and that the fermion representations are written in the covering groups would remove ambiguity.","section":"Table 2 and text after Eq. (4.5)"}],"recommendation":"major_revision","confidential_remarks":"I am recommending major revision rather than rejection because the explicit lattice computations appear internally consistent and the gap is in the exhaustiveness argument, which is in principle fixable. The abstract's word 'classification' is stronger than what is demonstrated unless the completeness question is addressed either by a direct moduli scan for the E8 string or by an explicit proof that maximality in the invariant sector implies maximality in the parent theory."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a solid, useful paper that delivers what it promises — a list of 22 maximal gauge groups for the E8 heterotic string on T^2, with massless spectra — and it deserves a real referee. The completeness claim, however, is thinner than the abstract suggests.\n\nWhat's actually new: the 22-entry list itself, the generalized Dynkin diagram procedure for the E8 string on T^2, and the observation that the brane monodromies include D4 and E6 foldings beyond the A_n cases known from 9d. The two approaches — orbifolding the parent SUSY theory at maximal enhancement points, and direct compactification of the 10d E8 string — agree on all 22 cases. That agreement is meaningful: the computations are independent in detail, and the massless spectra are checked case by case. The paper ships explicit lattices and the modular invariance of the partition functions is standard. This is reproducible work in the best sense.\n\nThe soft spot is the completeness step. Both approaches start from Table 12 of [17], the classification of maximal enhancements of the 8d SUSY E8×E8 theory on T^2, restricted to the symmetric Wilson-line locus A=(a,a). The paper never proves that a maximal enhancement of the E8 string must sit at a point where the parent is also maximally enhanced. Since all untwisted E8-string gauge bosons are long roots from the parent's P_L^2=2 sector, it is conceivable that a locus maximizing only that invariant sector could be missed by both approaches. This is a genuine gap in the argument for exhaustiveness, and the stress-test note is right about it. I don't think it is fatal — the agreement of the two methods plus the EDD consistency makes the list very plausible — but the authors should either prove the surjectivity step or state completeness as conditional on [17] plus an unproven assumption.\n\nMinor stuff: Eq. (2.15) has a sloppy sqrt2 factor, 'non-simply raced' should be 'non-simply laced', 'brnae' appears in Section 4.15, and several per-case derivations are only sketched in prose. None of this affects the main results. The brane interpretation is explicitly conditional on the no global symmetry/cobordism conjecture, which they flag clearly.\n\nWho it's for: swampland and non-SUSY string folks. I would cite it. Send it to review, and ask for a sharper statement on the completeness logic.","headline":"The 22-entry classification is a real result worth referee time, but the completeness claim rests on an unproven surjectivity step inherited from [17].","tokens_in":30356,"tokens_out":2157,"would_cite":true,"duration_ms":19574,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["11.25.-w","11.25.Mj"],"model":"deepseek-v4-flash","headline":"The E8 heterotic string on $T^2$ has exactly 22 maximal gauge groups, with massless spectra for each.","keywords":["non-supersymmetric heterotic strings","maximal gauge enhancement","E8 heterotic string","T^2 compactification","asymmetric orbifold","Dynkin diagram folding","8d branes","cobordism conjecture"],"falsifier":"An independent scan of the $E_8$ string moduli space on $T^2$, unrestricted to the symmetric Wilson-line locus, should reproduce exactly the 22 groups of Table 2; finding a maximal gauge group of rank 16 at a configuration with $A_1\\neq A_2$, or any 23rd group, would falsify the classification. Alternatively, a direct evaluation of the orbifold partition function for each candidate should show no extra massless vectors beyond the stated spectra.","tokens_in":29225,"feed_emoji":"⚛️","tokens_out":7161,"duration_ms":64958,"temperature":0.7,"pith_summary":"This paper determines all maximal gauge groups that can arise in the $E_8$ heterotic string theory compactified on a two-torus $T^2$. The authors obtain the same list by two independent routes: orbifolding the supersymmetric $E_8\\times E_8$ heterotic string at points of maximal gauge enhancement by a $\\mathbb{Z}_2$ outer automorphism combined with spacetime fermion parity, and directly compactifying the ten-dimensional $E_8$ string on $T^2$. Both approaches give the same 22 maximal gauge groups, with the complete massless spectrum for each case. If correct, this completes the classification of maximal gauge enhancements for the $E_8$ string in eight dimensions and, via the no global symmetry/cobordism conjecture, provides a classification of non-supersymmetric codimension-two branes in 8d supersymmetric heterotic theories.","feed_headline":"E8 string on a torus yields exactly 22 maximal gauge groups","feed_subtitle":"Two independent routes agree on the full list, with massless spectra for every case","key_machinery":"The central object is the asymmetric orbifold of the Narain lattice $\\Gamma_{18,2}$ by the combined operation $g(-1)^F$, where $g$ is a $\\mathbb{Z}_2$ outer automorphism of the charge lattice that acts by folding an ADE Dynkin diagram ($A_{2n-1}\\to C_n$, $D_4\\to B_3$, $E_6\\to F_4$) and $F$ is the spacetime fermion number. The invariant sublattice and its dual carry the untwisted and twisted sectors; massless vectors and matters are read off from states with $P_L^2=1$ or $2$ and $p_R=0$. In the direct construction, the same data are organized by a generalized Dynkin diagram for $T^2$ whose nodes are the simple roots of $E_8$ plus affine and 'C' nodes encoding the torus moduli, so that maximal enhancements are identified by deleting nodes.","core_discovery":"The central claim is that the $E_8$ heterotic string on $T^2$ has exactly 22 maximal gauge groups, listed in Table 2 together with their twisted-sector massless spectra. Each group is obtained from a point of maximal gauge enhancement of the supersymmetric $E_8\\times E_8$ heterotic string on $T^2$ by an asymmetric $\\mathbb{Z}_2$ orbifold that folds part of the Dynkin diagram and includes the spacetime fermion parity, breaking supersymmetry and halving the rank. The paper shows that the same 22 groups are found by direct compactification of the 10d $E_8$ string, with Wilson lines restricted to the symmetric locus $A=(a,a)$ required by orbifold consistency, and it develops a generalized Dynkin diagram for $T^2$ that reproduces the classification by deleting nodes. The 22 entries include cases where the fold produces non-simply laced factors (C-type groups, $B_3$, $F_4$), and the massless fermions are identified as quasi-minuscule representations of the folded groups.","pith_inferences":["The 22-entry list is derived under the symmetric Wilson-line restriction $A=(a,a)$; if that restriction is not exhaustive for the $E_8$ string on $T^2$, the full moduli space could contain additional maximal enhancements outside this locus, making 22 a lower bound rather than the full count.","The same folding machinery should apply to the other non-supersymmetric heterotic strings (for example $SO(16)\\times SO(16)$) on $T^2$, potentially yielding analogous classifications, though the paper does not perform that analysis.","The brane interpretation is contingent on the cobordism conjecture; if that conjecture fails, the massless-spectrum classification would still stand as a statement about the worldsheet CFTs, but the brane count would not follow.","The node-deletion rule suggests an algorithmic shortcut: maximal enhancements of the $E_8$ string on $T^d$ might be enumerated by iterated node deletion, which could be automated for higher torus dimension."],"forward_implications":["The $E_8$ non-supersymmetric heterotic string on $T^2$ has exactly the 22 maximal gauge groups of Table 2, each with the massless spectrum listed there.","Every one of these groups is realized constructively by an asymmetric $\\mathbb{Z}_2$ orbifold that folds a Dynkin diagram of a supersymmetric maximal enhancement, not merely by enumeration.","Under the no global symmetry/cobordism conjecture, each entry yields a codimension-two non-supersymmetric brane in an 8d supersymmetric heterotic theory, with the monodromy given by the corresponding Dynkin diagram folding.","The generalized Dynkin diagram for $T^2$ provides a quick node-deletion rule that identifies maximal enhancements, extending the 9d diagram to 8d and reproducing all 22 cases from the SUSY diagram."],"supporting_citations":[{"why":"Supplies the classification of maximal gauge enhancements of the 8d supersymmetric heterotic string on $T^2$ (Table 12) that the orbifold construction starts from.","marker":"[17]"},{"why":"The authors' previous work on 9d/8d non-supersymmetric theories and branes, whose 9d results and methods this paper extends.","marker":"[38]"},{"why":"The no global symmetry/cobordism conjecture used to convert the gauge-group classification into a brane classification.","marker":"[39]"},{"why":"Sets out the asymmetric orbifold construction used to build the non-supersymmetric theories.","marker":"[41]"},{"why":"Characterizes 10d nonsupersymmetric heterotic branes by monodromy exchanging the two $E_8$s, the pattern generalized here to $D_4$ and $E_6$ foldings.","marker":"[42]"},{"why":"Gives the partition function and the moduli-independent tachyon for the $E_8$ string compactifications used in the direct construction.","marker":"[43]"}],"fun_headline_variants":["E8 string on T2: 22 maximal gauge groups","22 maximal gauge groups from E8 heterotic on torus","E8 on torus classified: exactly 22 gauge groups","Two methods agree: 22 gauge groups in E8 string"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The complete list depends on the classification of maximal enhancements of the supersymmetric $E_8\\times E_8$ heterotic string on $T^2$ being complete, and on every maximal enhancement of the $E_8$ string sitting at a Wilson line of the symmetric form $A=(a,a)$; the paper assumes both rather than proving them.","fun_headline_variants_meta":{"raw":{"variants":["E8 string on T2: 22 maximal gauge groups","22 maximal gauge groups from E8 heterotic on torus","E8 on torus classified: exactly 22 gauge groups","Two methods agree: 22 gauge groups in E8 string"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000487,"raw_usage":{"total_tokens":2405,"prompt_tokens":955,"completion_tokens":1450,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":571,"completion_tokens_details":{"reasoning_tokens":1379}},"tokens_in":571,"tokens_out":1450,"duration_ms":11198,"temperature":1.0,"reasoning_tokens":1379,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T15:22:53.027794+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"An independent scan of the $E_8$ string moduli space on $T^2$, unrestricted to the symmetric Wilson-line locus, should reproduce exactly the 22 groups of Table 2; finding a maximal gauge group of rank 16 at a configuration with $A_1\\neq A_2$, or any 23rd group, would falsify the classification. Alternatively, a direct evaluation of the orbifold partition function for each candidate should show no extra massless vectors beyond the stated spectra.","supporting_citations":[{"cited_title":"Hamada and A","cited_arxiv_id":null,"evidence_quote":"The authors' previous work on 9d/8d non-supersymmetric theories and branes, whose 9d results and methods this paper extends."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Sets out the asymmetric orbifold construction used to build the non-supersymmetric theories."}],"review_version":1}