{"id":"a585f97b-294c-499e-be26-511cedbd89a0","arxiv_id":"2507.03476","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Standard cocompact discontinuous groups for reductive homogeneous spaces are classified by deformability, and for Spin(n,1) lattices a bending construction attains the maximal Zariski closure G_phi.","lead":"This paper classifies when cocompact discrete symmetry groups of pseudo-Riemannian spaces can be deformed without breaking proper discontinuity, and when such deformations become Zariski-dense. It provides a general construction that reaches the largest possible Zariski closure for representations of Spin(n,1) lattices in arbitrary real algebraic groups.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claimed completeness of The classification theorems 2.9 and 2.21 depends on the exhaustiveness of Tables 2.1–2.2, which is imported from the unpublished preprint [6] and Tojo [68] rather than proved here.","rationale":"The reader's weakest-assumption diagnosis is exactly the point I would stress: the central classification theorems claim completeness, but the exhaustiveness of the underlying list is outsourced to an unpublished preprint and a cited paper. This is not an internal inconsistency in the deformation machinery, but it is a genuine correctness risk for the headline classification results. I scrutinized the proof of the main deformation theorem, Theorem 3.9, including the bending construction in Proposition 4.9 and the linear-algebraic step involving Lemma 4.23; I did not find a clear internal gap that would undermine the construction of Zariski-closures Gφ. The proof of Theorem 3.9(3) uses Raghunathan's vanishing theorem in a way consistent with Fact 3.1, and the stability input from Fact 5.6 is standard. Thus, the most load-bearing unresolved issue is not inside the self-contained part of the paper but in the external support for Tables 2.1–2.2. If [6] is found valid and exhaustive, the classification theorems stand as stated; if not, the completeness assertions need to be weakened. Since the reader already made the verdict CONDITIONAL on this exact external input, my stress-test does not move the verdict.","tokens_in":67579,"tokens_out":15708,"duration_ms":176374,"concrete_test":"Obtain Bocheński–Tralle [6] and check whether its list of triples (G,H,L) with G absolutely simple, H non-compact reductive, and L acting properly and cocompactly on G/H coincides exactly with Tables 2.1 and 2.2, including the complex simple cases 8 and 9 and the exceptional case 12. Also verify from Tojo [68] that Table 2.1 is exhaustive for symmetric G/H. If any triple is missing, or if a row in Table 2.3 is not supported by [6], then Theorems 2.9 and 2.21 must be downgraded from complete classifications to classifications among the listed triples.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 2.9 and Theorem 2.21 are presented as complete answers to Question 2.5 for all triples (G,H,L) satisfying the properness and cocompactness condition (2.1), and Table 2.3 is explicitly labeled 'Complete answers'. The exhaustiveness of Tables 2.1 and 2.2 is therefore load-bearing for the classification claims. This exhaustiveness is not proved in the paper: Remark 2.8 states that the symmetric-space case is proved by Tojo [68] and that the general reductive case is shown in the preprint Bocheński–Tralle [6]. The paper also concedes in Remark 2.8 that it is 'plausible' that the tables list all such spaces. Since Theorems 2.9, 2.15, and 2.21 draw conclusions for every triple satisfying (2.1), a triple missing from Tables 2.1–2.2 would invalidate the word 'complete' and the classification statements. The core deformation construction, Theorem 3.9, is independent of this table and appears internally sound; the dependence is specifically at Section 2.3, the paragraph before Remark 2.8, and in the proof of Theorem 2.9 in Section 5.2, which invokes the tables as the universe of cases.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies deformations of standard discontinuous groups for homogeneous spaces G/H of reductive type with non-compact isotropy H. Its main technical result is Theorem 3.9: for any Zariski-connected real algebraic group G and any homomorphism φ from Spin(n,1) to G with n ≥ 2, there is a torsion-free cocompact lattice Γ in Spin(n,1) and a small deformation φ' of φ|Γ whose Zariski closure is exactly a specified algebraic subgroup G_φ; for n ≥ 3, G_φ is also an upper bound for all sufficiently small deformations. This theorem is proved by a three-step bending construction, with explicit representation-theoretic computations of G_φ in several classes. The paper also states complete answers to deformability questions Q1–Q3 for all triples in Tables 2.1 and 2.2, discusses the group-manifold case, constructs Zariski-dense discontinuous groups of cohomological dimension 6, and surveys spectral applications on quotients X_Γ.","tokens_in":67835,"tokens_out":24790,"duration_ms":281413,"significance":"If Theorem 3.9 is correct, it is a substantial generalization of the Johnson–Millson and Kassel bending results: it gives an optimal, representation-theoretically defined Zariski closure for deformations into an arbitrary real algebraic group. The proof is detailed and the paper is unusually self-contained, with appendices covering algebraic-group topologies, Clifford algebras, torsion-freeness, HNN extensions, and deformation bounds. The classification theorems 2.9 and 2.21, together with Table 2.3, provide a useful and concrete answer to the deformability questions, and the spectral applications are a valuable addition. The core construction in Theorem 3.9 appears sound and is independent of the external classification input; the classification claims, however, depend on an exhaustiveness statement that is not proved in this paper.","major_comments":[{"comment":"The exhaustiveness of Tables 2.1 and 2.2 is load-bearing for the classification claims, but it is not proved here. Remark 2.8 says it is 'plausible' that the tables list all such spaces, and then delegates the symmetric case to Tojo [68] and the general reductive case to the preprint Bocheński–Tralle [6]. Because Theorems 2.9 and 2.21 are phrased as complete answers to Question 2.5 for all triples satisfying (2.1), and Table 2.3 is labelled 'Complete answers', a triple missing from the tables would invalidate the word 'complete'. I recommend either proving the exhaustiveness or explicitly reformulating Theorems 2.9 and 2.21 as conditional on the classification in [6] and [68], and adjusting the abstract and introduction accordingly. Theorem 3.9 is independent of this external input and is not affected by this concern.","section":"Section 2.3, Remark 2.8, and proofs of Theorems 2.9 and 2.21"},{"comment":"The proof asserts H^1(Γ, g/g_φ) = 0 by 'Raghunathan's vanishing theorem (Fact 3.1)' in a single sentence. As stated, Fact 3.1 applies to irreducible coefficient modules, while g/g_φ is in general reducible; the proof needs an explicit complete-reducibility argument. It also needs to explain why the trivial module does not appear in g/g_φ: the trivial submodule z_g(dφ(spin(n,1))) is contained in g_φ because it is contained in z_g(dφ(spin(n−1,1))), which is included in (3.2). Without this, the possible nonzero contribution Hom(Γ,R) = H^1(Γ,R) would not be excluded. The argument is standard and fixable, but it should be written out.","section":"Section 3.2, proof of Theorem 3.9(3)"}],"minor_comments":[{"comment":"The sentence 'In other words' asserts that the Lie algebra generated by dφ(spin(n,1)) and all spherical-harmonics submodules coincides with the smallest Lie subalgebra containing dφ(spin(n,1)) + z_g(dφ(spin(n−1,1))). This equivalence is used later, for example in Corollary 3.11, but no proof is given; a short proof or a reference would help.","section":"Section 3.2, Definition 3.5"},{"comment":"There are several typos that should be corrected, including 'disconitnuous' in Theorem 2.21(2)(ii), 'comological' in Corollary 5.23, and the missing spacing in 'whenastandardquotient' in the abstract.","section":"Throughout"},{"comment":"The proof of part (3) is very terse: it does not spell out how the 'no' entries in Table 2.3 imply the hypothesis (b) of Proposition 5.18, nor the role of the stability theorem (Fact 5.6) in passing from 'no Zariski-dense deformation preserving properness' to 'no Zariski-dense small deformation' for the relevant rank-one factor. A short case-check would make the argument easier to verify.","section":"Section 5.4, proof of Theorem 2.21(3)"}],"recommendation":"major_revision","confidential_remarks":"The main obstacle to acceptance is the external dependence of the completeness claims on Tojo [68] and the preprint Bocheński–Tralle [6]. If the authors are willing to state the classification results conditionally, the paper is in good shape; if unconditional completeness is required, the exhaustiveness of Tables 2.1 and 2.2 would need to be established or the claims weakened. The core Theorem 3.9 appears sound and valuable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know about arXiv:2507.03476. First, the core construction theorem, Theorem 3.9, is a genuine generalization of the Johnson–Millson and Kassel bending results: for any Zariski-connected real algebraic group G and any homomorphism φ: Spin(n,1) → G, it produces a torsion-free cocompact lattice Γ and a small deformation whose Zariski closure is exactly an explicitly defined subgroup G_φ, and for n ≥ 3 this is also the upper bound. The proof is a three-step bending construction, and the part I checked carefully – the definition of g_φ via spherical harmonics modules and the centralizer of spin(n−1,1), Lemma 3.6 that the analytic subgroup is Zariski-closed, and the use of Lemma 4.23 to avoid tower assumptions – is coherent and detailed. This is real progress: previous bending constructions only worked for specific embeddings into classical groups. The paper also gives a classification of deformation behavior for the triples (G,H,L) in Tables 2.1–2.2, with a clean dichotomy: roughly half the cases are rigid, and when non-standard deformation is possible, a Zariski-dense deformation is also possible.\n\nSecond, the classification part is conditional in a specific, honest way. The exhaustiveness of Tables 2.1 and 2.2 is not proved here. It is imported from Tojo [68] for the symmetric case and from the preprint Bocheński–Tralle [6] for the general reductive case. Remark 2.8 says it's 'plausible' that the tables are complete. Theorems 2.9 and 2.21 are framed as complete answers, so a missing triple would invalidate the word 'complete'. That is a real soft spot, but it is external: the core deformation theorem does not depend on the tables, and the authors flag the dependence. The same goes for Theorem 5.28, which leans on a forthcoming paper [25].\n\nI see no internal circularity or fitting. G_φ is defined from the φ-module structure, then shown attainable and, for n ≥ 3, to be an upper bound. The dependence on Klingler and Raghunathan is standard and correctly cited. The sign of the paper's claims matches what the proofs actually show.\n\nBottom line: this is a substantial paper that deserves a serious referee. The referee should push on the completeness question – either verify the external inputs or have the authors rephrase the classification as conditional. The construction theorem stands on its own. I'd bring it to the reading group and I'd cite it if I were working in this area.","headline":"A real bending construction for arbitrary targets, with an honest but load-bearing external dependence for the classification claims.","tokens_in":68402,"tokens_out":2568,"would_cite":true,"duration_ms":28536,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57S30","58H15","22D50","22E40","22E46","53C30","58J50"],"pacs":[],"model":"deepseek-v4-flash","headline":"For any homomorphism $\\varphi\\colon\\mathrm{Spin}(n,1)\\to G$ into a Zariski-connected real algebraic group, some torsion-free cocompact discrete subgroup can be deformed so that its Zariski closure is exactly $G_\\varphi$, and for $n\\ge 3$…","keywords":["discontinuous groups","proper actions","pseudo-Riemannian homogeneous spaces","Zariski-dense deformations","Clifford-Klein forms","Spin(n,1) lattices","local rigidity","bending construction"],"falsifier":"To break the classification, exhibit one triple $(G,H,L)$ with $G$ simple, $H$ non-compact reductive, and $L$ acting properly and cocompactly on $G/H$, that is not locally isomorphic to any row of Tables 2.1 or 2.2; running the properness and cocompactness criteria of [37] on such a candidate would settle this. To break Theorem 3.9 for $n\\ge 3$, find a small deformation of $\\varphi|_\\Gamma$ whose Zariski closure strictly contains $G_\\varphi$; the paper's argument predicts $H^1(\\Gamma,\\mathfrak g/\\mathfrak g_\\varphi)=0$ forbids exactly that.","tokens_in":67337,"feed_emoji":"📐","tokens_out":15731,"duration_ms":162552,"temperature":0.7,"pith_summary":"The paper asks how much a cocompact standard discontinuous group for a pseudo-Riemannian homogeneous space $X=G/H$ can be deformed without losing proper discontinuity, and answers it through a single algebraic object $G_\\varphi$. For any homomorphism $\\varphi\\colon\\mathrm{Spin}(n,1)\\to G$ with $n\\ge 2$, it constructs a torsion-free cocompact discrete subgroup $\\Gamma$ and a small deformation of $\\varphi|_\\Gamma$ whose Zariski closure is exactly $G_\\varphi$; for $n\\ge 3$, no small deformation of any such lattice can escape $G_\\varphi$ up to conjugacy. It then computes $G_\\varphi$ in representation-theoretic terms and uses this to give complete yes/no answers to three deformation questions for every triple in its Tables 2.1 and 2.2. A consequence is that when a listed standard quotient can be deformed at all, it can often be deformed into a Zariski-dense subgroup, including a new 7-dimensional compact example $SO(4,4)/SO(3,4)$ and Zariski-dense groups of cohomological dimension 6 in spaces with non-compact dimension up to 16.","feed_headline":"Spin lattices deform to exactly the predicted Zariski closure","feed_subtitle":"Classification tells which pseudo-Riemannian quotients stay rigid, deform non-standardly, or admit dense deformations.","key_machinery":"The load-bearing object is the real algebraic subgroup $G_\\varphi$: it is the Zariski-connected group whose Lie algebra $\\mathfrak g_\\varphi$ is generated by $d\\varphi(\\mathfrak{spin}(n,1))$ and all $\\mathfrak{spin}(n,1)$-submodules of $\\mathfrak g$ isomorphic to spherical-harmonics representations, namely the irreducible modules with a nonzero $\\mathfrak{spin}(n-1,1)$-invariant vector. The construction that reaches it is the bending deformation: choose an arithmetic lattice $\\Gamma$ in $\\mathrm{Spin}(n,1)$ whose quotient contains arbitrarily many disjoint totally geodesic hypersurfaces, express $\\Gamma$ as an iterated HNN extension, and deform the representation by exponentiating carefully chosen invariant vectors along the corresponding loops. The upper bound comes from the classical vanishing theorem for the first cohomology of cocompact lattices, which forces $H^1(\\Gamma,\\mathfrak g/\\mathfrak g_\\varphi)=0$ for $n\\ge 3$, together with a curve-selection argument showing that vanishing first cohomology confines all nearby representations to $G_\\varphi$ up to conjugacy.","core_discovery":"The central discovery is Theorem 3.9: for $n\\ge 2$, every homomorphism $\\varphi\\colon\\mathrm{Spin}(n,1)\\to G$ into a Zariski-connected real algebraic group admits a torsion-free cocompact arithmetic subgroup $\\Gamma$ and a small deformation $\\varphi'$ of $\\varphi|_\\Gamma$ whose Zariski closure is exactly $G_\\varphi$; for $n\\ge 3$, $G_\\varphi$ is also the largest possible, since any sufficiently small deformation of any torsion-free cocompact lattice lies in $G_\\varphi$ up to $G$-conjugacy. The subgroup $G_\\varphi$ is defined from the Lie algebra generated by $d\\varphi(\\mathfrak{spin}(n,1))$ together with all irreducible spherical-harmonics submodules of $\\mathfrak g$, so the largest Zariski closure is a purely representation-theoretic object. Built on this, Theorems 2.9 and 2.21 classify, for all triples in Tables 2.1 and 2.2, whether a standard compact quotient is locally rigid, can be deformed into a non-standard quotient, or admits a Zariski-dense deformation, with the group-manifold case treated separately.","pith_inferences":["Theorem 3.9 achieves $G_\\varphi$ using arithmetic lattices with many disjoint hypersurfaces; the upper-bound half already holds for every torsion-free cocompact lattice, so the natural open question is whether non-arithmetic lattices can also attain the full $G_\\varphi$ by an adapted bending construction.","If the cited exhaustive-list preprint has gaps, the word 'complete' in Theorems 2.9 and 2.21 should be read as applying to the listed triples only; the row-by-row answers would survive, but unlisted triples could behave differently.","The representation-theoretic definition of $G_\\varphi$ gives a finite algorithm for any given representation: decompose $\\mathfrak g$ under $\\mathfrak{spin}(n,1)$, collect the trivial and spherical-harmonics summands, and take the Lie algebra they generate together with $d\\varphi(\\mathfrak{spin}(n,1))$."],"forward_implications":["For every $n\\ge 3$, $G_\\varphi$ is both achievable and an upper bound up to conjugacy, so the largest Zariski closure of a small deformation is exactly $G_\\varphi$.","Whenever $\\mathfrak g_\\varphi=\\mathfrak g$, a standard cocompact quotient can be deformed to a Zariski-dense discontinuous group; Table 2.3 lists exactly which spaces do this, including the 7-dimensional space form $SO(4,4)/SO(3,4)$.","Among the listed triples, the cases with $\\mathfrak l_{ss}\\simeq\\mathfrak{so}(n,1)$ and nonvanishing first cohomology are precisely the cases where a Zariski-dense deformation exists, while the $\\mathfrak{su}(n,1)$ cases can be deformable but never non-standard.","For the group manifold $(G\\times G)/\\mathrm{diag}(G)$ with $\\mathfrak g=\\mathfrak{so}(n,1)$, some $\\Gamma\\times\\{e\\}$ deforms to a Zariski-dense subgroup of $G\\times G$, whereas for simple $G$ the direct-product discontinuous groups on the listed triples never become Zariski-dense in $G\\times G$.","The method produces Zariski-dense discontinuous groups of cohomological dimension 6 on spaces such as $SO(8,\\mathbb C)/SO(7,\\mathbb C)$, $SO(8,8)/SO(7,8)$, $SU(8,8)/U(7,8)$, and $SL(16,\\mathbb R)/SL(15,\\mathbb R)$, and for symmetric spaces satisfying the rank condition the constructed deformation preserves infinitely many discrete spectra."],"supporting_citations":[{"why":"It supplies the bending-deformation construction along totally geodesic hypersurfaces that the paper extends from $\\mathrm{SO}(n,1)$ to $\\mathrm{Spin}(n,1)$ and to arbitrary $G$.","marker":"[21]"},{"why":"It provides the stability theorem for properly discontinuous rank-one actions used to pass from deformations of $\\Gamma$ to deformations of quotients $X_\\Gamma$.","marker":"[27]"},{"why":"It constructs torsion-free cocompact arithmetic subgroups with arbitrarily many disjoint totally geodesic hypersurfaces, reformulated here for $\\mathrm{Spin}(n,1)$.","marker":"[56]"},{"why":"It gives the properness and cocompactness criteria and the dimension bound that define the admissible triples and control when deformations remain standard.","marker":"[37]"},{"why":"It supplies the starting list of symmetric spaces $G/H$ admitting proper cocompact reductive actions, namely Table 2.1.","marker":"[50]"},{"why":"It is cited as proving exhaustiveness of Table 2.1 for irreducible symmetric spaces, on which the completeness of the classification rests.","marker":"[68]"},{"why":"It is the cited preprint showing that Tables 2.1 and 2.2 contain all relevant reductive triples, the other pillar of completeness.","marker":"[6]"},{"why":"It supplies the vanishing theorem for first cohomology used for local rigidity and for the upper bound $H^1(\\Gamma,\\mathfrak g/\\mathfrak g_\\varphi)=0$.","marker":"[60]"},{"why":"It supplies the stability of proper discontinuity for direct-product discontinuous groups on group manifolds, used in Theorem 2.21.","marker":"[29]"}],"fun_headline_variants":["Pseudo-Riemannian quotients: when deformations stay rigid or go dense","Zariski-density predicted exactly by representation theory","Theorem: small deformations of Spin lattices have maximal Zariski closure","Rigidity vs density: classification for pseudo-Riemannian quotients","Exactly the predicted Zariski closure for Spin lattice deformations"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The completeness of the classification rests on the assumption that Tables 2.1 and 2.2 contain every admissible triple; that exhaustiveness is imported from a cited symmetric-space classification and a cited preprint, and is not proved in this paper.","fun_headline_variants_meta":{"raw":{"variants":["Pseudo-Riemannian quotients: when deformations stay rigid or go dense","Zariski-density predicted exactly by representation theory","Theorem: small deformations of Spin lattices have maximal Zariski closure","Rigidity vs density: classification for pseudo-Riemannian quotients","Exactly the predicted Zariski closure for Spin lattice deformations"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000488,"raw_usage":{"total_tokens":2413,"prompt_tokens":961,"completion_tokens":1452,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":577,"completion_tokens_details":{"reasoning_tokens":1360}},"tokens_in":577,"tokens_out":1452,"duration_ms":12966,"temperature":1.0,"reasoning_tokens":1360,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T20:08:53.864665+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"To break the classification, exhibit one triple $(G,H,L)$ with $G$ simple, $H$ non-compact reductive, and $L$ acting properly and cocompactly on $G/H$, that is not locally isomorphic to any row of Tables 2.1 or 2.2; running the properness and cocompactness criteria of [37] on such a candidate would settle this. To break Theorem 3.9 for $n\\ge 3$, find a small deformation of $\\varphi|_\\Gamma$ whose Zariski closure strictly contains $G_\\varphi$; the paper's argument predicts $H^1(\\Gamma,\\mathfrak g/\\mathfrak g_\\varphi)=0$ forbids exactly that.","supporting_citations":[{"cited_title":"Johnson and J","cited_arxiv_id":null,"evidence_quote":"It supplies the bending-deformation construction along totally geodesic hypersurfaces that the paper extends from $\\mathrm{SO}(n,1)$ to $\\mathrm{Spin}(n,1)$ and to arbitrary $G$."},{"cited_title":"Kobayashi and T","cited_arxiv_id":null,"evidence_quote":"It supplies the starting list of symmetric spaces $G/H$ admitting proper cocompact reductive actions, namely Table 2.1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the vanishing theorem for first cohomology used for local rigidity and for the upper bound $H^1(\\Gamma,\\mathfrak g/\\mathfrak g_\\varphi)=0$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"It supplies the stability of proper discontinuity for direct-product discontinuous groups on group manifolds, used in Theorem 2.21."}],"review_version":1}