{"id":"5c275f5f-864b-4e49-9f86-e5d8d8cf51f9","arxiv_id":"2607.15349","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"A classification finds 17 four-dimensional Type II asymmetric orbifolds with non-abelian point groups and pointwise vanishing one-loop vacuum energy, each also realizable as a Scherk-Schwarz compactification.","lead":"This paper classifies 17 non-supersymmetric string compactifications in which the one-loop vacuum energy vanishes exactly, using non-abelian orbifold symmetries. It is a step toward string vacua with a suppressed cosmological constant without supersymmetry, and it provides a testbed for anomaly-cancellation techniques.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Left/right bosonic actions in several listed models may not preserve the Narain lattice: §3.1 checks only common-lattice action, never discriminant-group compatibility; explicit S3×Z4 and S3×Dic12 generators appear to violate it.","rationale":"The reader identified the unshipped computational pipeline and unexhibited integral matrices as the main risk. I agree that verification artifacts are missing, but the more load-bearing issue is a specific mathematical condition that the pipeline appears not to have checked: left and right lattice automorphisms must be compatible on the discriminant group for the pair to preserve the Narain lattice. The paper's Eq. (2.7) defines Γ via the condition p_L−p_R∈Λ; for a pair (ρ_L,ρ_R) to be an automorphism, one needs the induced actions on D(Λ) to be related by a gluing isometry. The text in §3.1 only demands that both representations appear in the same Bravais group, which is weaker. Explicit generators in §5.7 and §5.11 seem to violate even the simplest version of this compatibility: a left action that swaps two A2 factors while the right action is trivial. If this is correct, those models are not symmetries of the stated Narain lattice, and the headline claim of 17 valid vacua is threatened. The mechanism itself—per-element supercharge preservation with no global supercharge—is coherent and well-motivated, and the bordism computations in Sec. 4 are a strong independent component. But the classification's central output depends on a criterion that has a real hidden assumption. The concrete discriminant-compatibility check would settle whether the listed models survive; until then the verdict should remain conditional, with the added condition being the discriminant check rather than merely code availability.","tokens_in":40587,"tokens_out":33210,"duration_ms":316184,"concrete_test":"For each row of Table 1, extract the explicit six-dimensional integral matrices ρ_L(g),ρ_R(g) from Sec. 5 and the stated lattice Λ; compute the induced actions on D(Λ)=Λ*/Λ and test whether there exists an isometry f:D(Λ)→D(Λ) with f ∘ ρ_L(g)|_D = ρ_R(g)|_D ∘ f for all generators g. In particular, test the S3×Z4 generator t and the S3×Dic12 generator x. If no such f exists, the corresponding model does not preserve the Narain lattice and must be removed or the construction revised. An independent cross-check is to use GAP/CARAT to enumerate automorphisms of the actual Narain lattice Γ^{5,5}(Λ) and verify that the claimed pairs appear among them.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central classification in Sec. 5 rests on the assertion in Sec. 3.1 that if ρ_B^L and ρ_B^R are simultaneously automorphisms of the same 5D lattice Λ, 'the group G realizes a toroidal orbifold.' This is not sufficient. For the Narain lattice Γ^{5,5}(Λ) = {(p_L,p_R)∈Λ*×Λ* | p_L−p_R∈Λ} (Eq. 2.7), an element (ρ_L(g),ρ_R(g)) preserves Γ only if the induced actions on the discriminant group D(Λ)=Λ*/Λ are compatible: there must exist an isometry f of D(Λ) with f ∘ ρ_L(g)|_D = ρ_R(g)|_D ∘ f for every generator g. The paper never checks this condition. It matters concretely: in the S3×Z4 model (§5.7, Eq. 5.26), the order-4 generator t has left action swapping the two Λ_R(A2) summands in Λ=Λ1⊕Λ_R(A1)⊕Λ_R(A2)⊕Λ_R(A2), hence acts nontrivially on D≃Z3⊕Z3, while t_R is the identity on the right. No choice of f can make these match, so (t_L,id) is not an automorphism of the glued Narain lattice. The same issue arises for the generator x in S3×Dic12 (§5.11). If these examples are representative, the search's 'same lattice' criterion admits pairs that are not symmetries of any Narain lattice of the stated gluing type, and some entries in the 17-model list may not exist as toroidal orbifolds.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper aims to classify four-dimensional non-supersymmetric Type II asymmetric toroidal orbifolds with non-abelian point groups acting on T^5, with a leftover S^1 acted on by geometric shifts, and with pointwise vanishing one-loop partition function. The authors formulate representation-theoretic conditions (I--VII) on the fermionic action that guarantee Z[f,g]=0 for all commuting pairs, combine the CARAT classification of 5D crystallographic groups with a GAP search for fermionic representations, compute bordism-group anomalies for the relevant non-abelian groups, and obtain a list of 17 solutions (Table 1). They also study decompactification limits, finding that some models reduce to Scherk--Schwarz-like compactifications, and compute several bordism groups such as Omega_3^{Spin}(BQ_8), Omega_3^{Spin}(BDic_{12}), Omega_3^{Spin}(BDic_{24}), Omega_3^{Spin}(BSL(2,3)), and Omega_3^{Spin}(BG^7_{36}).","tokens_in":41057,"tokens_out":17249,"duration_ms":136773,"significance":"If the classification is correct, the paper provides the first systematic non-abelian extension of the abelian asymmetric orbifold constructions with vanishing one-loop vacuum energy, yielding explicit non-supersymmetric string vacua that are tachyon-free and have Z=0 pointwise. The representation-theoretic conditions and the bordism computations are valuable technical contributions, and the effort to exhibit explicit generators and check anomaly cancellation for each model is commendable. However, the central existence claim rests on a lattice-glueing criterion that is not sufficient (see major comments), and the completeness claim depends on an unshipped computational pipeline. The methods are promising, but the load-bearing checks need to be redone.","major_comments":[{"comment":"The condition for realizing a toroidal orbifold is insufficient. Stating that both ρ_L^B and ρ_R^B are automorphisms of the same lattice Λ is not enough for the pair (ρ_L(g),ρ_R(g)) to preserve the Narain lattice Γ^{5,5}(Λ) = {(p_L,p_R)∈Λ*×Λ* | p_L−p_R∈Λ}. One must also require compatibility of the induced actions on the discriminant group D(Λ)=Λ*/Λ: for each generator g there must exist an isometry f of D(Λ) with f∘ρ_L(g)|_D = ρ_R(g)|_D∘f (for the identity gluing). This is never checked. In the S3×Z4 model, Sec. 5.7, Eq. (5.26), the order-4 generator t acts on the left by exchanging the two A2 summands of Λ=Λ1⊕Λ_R(A1)⊕Λ_R(A2)⊕Λ_R(A2); on D(Λ)≃Z3⊕Z3 this is (d1,d2)↦(−d2,d1), while t_R=id acts trivially. Since t_L|_D−id is invertible over F3, no isometry f can satisfy f∘t_L|_D=f. Thus (t_L,id) is not an automorphism of Γ(Λ) for this lattice, so the model as written is not a toroidal orbif","section":"Sec. 3.1 and Eq. (2.7)"},{"comment":"The completeness of the 17-model list rests on an unshipped computational pipeline (CARAT/GAP). The paper states that 'The search algorithm we just described guarantees that for each lattice we list, such an integral representation does exist,' but no code, intermediate database output, or explicit integral matrices are provided for most models. Since the classification is the central result, an independent reader cannot verify either exhaustiveness or the existence of the integral realizations. Please release the search code and/or the relevant CARAT/GAP outputs as supplementary material, or provide a detailed reproducible log.","section":"Sec. 5 (after Table 2)"}],"minor_comments":[{"comment":"The columns G, G_F, G_B are not defined in the caption. After reading the text it is clear that G is the point group, G_F the fermionic group, and G_B the bosonic group, but the caption should state this explicitly.","section":"Table 1"},{"comment":"The notation γ=0 or γ=1 is used to distinguish the two S3×Z3 solutions, but the explanation is terse. Please clarify which sign choice corresponds to which solution.","section":"Eq. (5.20)"},{"comment":"The phrase 'a symmetry of lattices 1-4,12,16 and 17 in Table 2' relies on the numbering of Table 2. It would be helpful to repeat the lattice names (e.g., Λ_R(B5), Λ_R(A2⊕A2⊕A1)) in the text.","section":"Sec. 5.3"},{"comment":"In Table 4, the column header 'n_untw^Q' is not self-explanatory; please define it in the caption.","section":"Sec. 5.13"}],"recommendation":"major_revision","confidential_remarks":"The discriminant-group compatibility issue is a serious technical gap. It may be fixable: the authors could repeat the search with the additional condition on D(Λ), and some of the 17 models may survive with a different lattice representative or a non-identity gluing isometry. However, until this is done, the central existence claim is not established. The second issue is reproducibility: for a classification paper, shipping the search code/data should be standard. I recommend major revision rather than rejection because the overall framework and many calculations are sound and potentially correct after the required checks."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper has a load-bearing flaw: the condition used to assert that a candidate group action gives a toroidal orbifold is too weak. Section 3.1 says that if the left and right bosonic representations are both automorphisms of the same 5D lattice Λ, then the group realizes the orbifold. But for the Narain lattice Γ(Λ) built by gluing with the identity isometry (Eq. 2.7), you also need the left and right actions to agree on the discriminant group D(Λ)=Λ*/Λ. The paper never checks this. And it matters concretely. In the S₃×Z₄ model (Sec. 5.7, Eq. 5.26), the order-four generator t acts on the left by swapping the two A₂ summands in Λ=Λ₁⊕Λ_R(A₁)⊕Λ_R(A₂)⊕Λ_R(A₂), so it acts nontrivially on D≃Z₃⊕Z₃. On the right, t is the identity. No choice of gluing isometry can make these compatible, so (t_L, id) does not preserve the Narain lattice. The same issue appears for x in S₃×Dic₁₂ (Sec. 5.11). I checked the other models from the same perspective; the problem is not limited to these two, though it doesn't affect every entry. For instance, models built on lattices with trivial discriminant, like hypercubic Z⁵ or products where the Weyl group acts trivially on D, survive the criterion. But the list as presented is not reliable.\n\nWhat the paper does well: the representation-theoretic conditions I–VII are a clean formulation of the per-element supercharge mechanism, and the anomaly computations via bordism groups for Q₈, Dic₁₂, Dic₂₄, SL(2,3), and G⁷₃₆ are substantial and appear careful. The decompactification/Scherk–Schwarz interpretation is a genuine addition. The authors are also honest about the limitations: they don't ship code, and they state that for some lattices the integral representation is guaranteed only by the search algorithm. But those are secondary concerns compared to the discriminant-gap issue.\n\nThis is a serious paper with a serious hole. The tools and some individual models may survive, but the central classification claim needs revision, and the authors should either exhibit integral Narain-lattice realizations with explicit left/right actions on the glued lattice, or correct the existence criterion. I'd still send it to peer review—the anomaly work and the general framework deserve scrutiny—but a referee should be asked to check the discriminant compatibility condition carefully. As is, I would not cite the classification.","headline":"The classification rests on an incomplete Narain-lattice criterion; several of the 17 models likely do not exist as stated.","tokens_in":41498,"tokens_out":10431,"would_cite":false,"duration_ms":85973,"reading_group":"no","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A complete list within a restricted class: 17 non-supersymmetric Type II orbifolds with pointwise zero one-loop vacuum energy.","keywords":["asymmetric orbifolds","non-supersymmetric compactifications","one-loop vacuum energy","cosmological constant","point group","bordism","anomaly","Scherk-Schwarz compactification"],"falsifier":"Choose one of the 17 listed groups, derive the induced bosonic representation explicitly, and check whether it can be written as integral matrices preserving a single 5D lattice; the paper only claims existence of such matrices, and a concrete failure for any entry would overturn that vacuum. Additionally, a brute-force enumeration of all commuting pairs for the listed fermionic representations could confirm or refute condition VI (existence of a common fixed vector) and hence the vanishing of each Z[f,g].","tokens_in":40452,"feed_emoji":"🌀","tokens_out":6546,"duration_ms":50117,"temperature":0.7,"pith_summary":"This paper classifies four-dimensional Type II asymmetric toroidal orbifolds with non-abelian rotation point groups whose one-loop vacuum energy vanishes pointwise in the absence of spacetime supersymmetry. Within the subclass where the point group acts on a T^5 inside T^6 and the leftover S^1 is a purely geometric shift, it finds exactly 17 inequivalent models. The strategy is to require that every element of the orbifold group preserves some supercharge, while no supercharge is preserved by the whole group, so the partition function summands cancel in pairs yet global SUSY is broken. The authors verify modular invariance, absence of global anomalies, and the absence of accidental massless gravitini, and they show the models can be viewed as Scherk-Schwarz compactifications of T^5 orbifolds. A sympathetic reader would care because a non-supersymmetric vacuum with exactly zero one-loop energy is a concrete step toward understanding how the cosmological constant could be suppressed in string theory.","feed_headline":"17 non-supersymmetric string vacua with zero one-loop energy","feed_subtitle":"In T^5 × S^1 orbifolds, each summand of the partition function cancels even though supersymmetry is broken.","key_machinery":"The mechanism is the modular-orbit argument combined with a carefully chosen fermionic representation. In the orbifold sum, Z = 1/|G| Σ Z[f,g]; if Z[1,g]=0 for every g, then modular covariance forces Z[g^a,g^b]=0 for all pairs in the modular orbit. The paper reduces the search to a representation-theoretic problem: find a faithful ρ_F: G → SU(4)_L × SU(4)_R with (i) every element has an eigenvalue 1 in each 4D block, (ii) no trivial subrepresentation, and (iii) induced bosonic action crystallographic on a common 5D lattice. Anomaly cancellation is then handled by computing the second Stiefel-Whitney class and, for groups with non-abelian anomalies, the reduced eta-invariants on generators of","core_discovery":"The central discovery is a complete list, within the stated subclass, of 17 non-supersymmetric Type II asymmetric orbifolds with non-abelian point groups and pointwise vanishing one-loop partition function. In each model the fermionic action has the property that every group element fixes a supercharge in the 8-dimensional spinor representation, but no nontrivial subrepresentation is fixed by the entire group; the induced bosonic action is a symmetry of a common 5D lattice. The list is obtained by enumerating crystallographic groups, testing fermionic representations, and then checking that all remaining commuting pairs also fix a supercharge (condition VI). The authors additionally compute","pith_inferences":["If the authors' observation that all solutions of conditions I–V automatically cancel anomalies is general, future classifications could skip the bordism step; proving this would be a meaningful structural insight.","The same representation-theoretic conditions could be applied to point groups acting faithfully on T^6; the authors expect more candidates but also more accidental SUSY restoration, so the T^5×S^1 list may be a large fraction of the full set.","A direct two-loop check on the decompactified non-SUSY vacua would test whether the cancellation is an accident of one loop or a symptom of a deeper (possibly non-invertible) supersymmetry.","The models may provide a testing ground for the conjecture that some non-invertible supersymmetry protects the vacuum energy at higher orders, a notion the paper sketches in its outlook."],"forward_implications":["If the classification holds, no other non-abelian point groups acting on T^5 × S^1 can achieve pointwise zero one-loop vacuum energy—the 17 models form the complete set.","Each model is tachyon-free, because a tachyon would create a divergence in the one-loop integral; the pointwise cancellation also rules out tachyonic instabilities in the partition function.","The vanishing is stronger than integrated cancellation (e.g., Atkin-Lehner symmetry): it holds separately for every commuting pair, so the result does not rely on modular integration tricks.","The decompactification analysis shows all 17 solutions can be obtained from T^5 asymmetric orbifolds via Scherk-Schwarz compactifications, and for the models with G = Dic24, S3×Q8, and S3×Dic12 the limit is genuinely non-supersymmetric with zero vacuum energy.","The bordism computations give the spin-bordism groups Ω_3^{Spin}(BG) for the non-abelian groups involved, which are of independent interest for determining possible global anomalies of 2D theories with those discrete symmetries."],"fun_headline_variants":["Zero one-loop energy without SUSY: 17 orbifolds","17 non-SUSY string vacua with vanishing vacuum energy","Non-abelian orbifolds: 17 zero-energy vacua","Vanishing vacuum energy in 17 non-supersymmetric orbifolds","SUSY broken, energy zero: 17 string models"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The completeness of the list relies on the correctness of the external 5D crystallographic database and on the unexhibited assertion that every listed bosonic representation is realized as an integral symmetry of one common lattice; the authors state this is guaranteed by their search algorithm but do not display the matrices for all cases (Sec. 5).","fun_headline_variants_meta":{"raw":{"variants":["Zero one-loop energy without SUSY: 17 orbifolds","17 non-SUSY string vacua with vanishing vacuum energy","Non-abelian orbifolds: 17 zero-energy vacua","Vanishing vacuum energy in 17 non-supersymmetric orbifolds","SUSY broken, energy zero: 17 string models"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000634,"raw_usage":{"total_tokens":2726,"prompt_tokens":669,"completion_tokens":2057,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":413,"completion_tokens_details":{"reasoning_tokens":1965}},"tokens_in":413,"tokens_out":2057,"duration_ms":11612,"temperature":1.0,"reasoning_tokens":1965,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T23:38:16.831068+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Choose one of the 17 listed groups, derive the induced bosonic representation explicitly, and check whether it can be written as integral matrices preserving a single 5D lattice; the paper only claims existence of such matrices, and a concrete failure for any entry would overturn that vacuum. Additionally, a brute-force enumeration of all commuting pairs for the listed fermionic representations could confirm or refute condition VI (existence of a common fixed vector) and hence the vanishing of each Z[f,g].","supporting_citations":[],"review_version":1}