{"id":"95229d5e-0d31-4b1d-ad09-6717633e1d57","arxiv_id":"2502.03306","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Complex structures on nilpotent almost abelian Lie algebras are unique up to isomorphism, giving full control of Dolbeault cohomology, Frölicher degeneration, and deformations of the corresponding nilmanifolds.","lead":"This paper proves that a nilpotent almost abelian real Lie algebra admits at most one complex structure up to isomorphism. It then derives complete cohomology tables and deformation statements for the associated complex nilmanifolds.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The main theorem inherits all of its force from the external classification [ABD+24, Thm. 3.10]; a missing case or wrong coefficient in (3.4) would break the normal form and every cohomological/deformation consequence.","rationale":"I read the paper in good faith and found the internal argument largely coherent. The central claim is Proposition 3.10, and the proof reduces an arbitrary integrable complex structure on a nilpotent almost abelian Lie algebra to the explicit pair (A,J_0). The algebraic steps in the proof, including the use of the real Jordan basis in the complex vector space b = a ∩ Ja, are consistent with the stated structure equations in Corollary 3.12. The cohomological and deformation arguments then follow from the normal form in a standard way: Theorem 1.1(i) from the stable torus bundle series constructed in Corollary 3.14, Theorem 1.1(ii) from the nilpotence of the complex structure, Theorem 1.1(iii) from the dimension comparison in Theorem 4.5, and Theorem 1.2 from the cited deformation theorems together with uniqueness. I did not find a concrete internal error that would invalidate these deductions. The genuinely load-bearing point is the reliance on the external classification theorem [ABD+24, Thm. 3.10], restated as Theorem 3.2. Everything in the paper depends on that classification being complete and on the partition formula (3.4) being exactly right; the paper itself supplies no independent verification. This is a legitimate concern for a research preprint, and it matches the reader's identified weakest assumption. However, reliance on a cited theorem is normal mathematical practice, and the classification is a recent preprint by a partly overlapping group rather than a claimed new result of this paper. In addition, the injectivity of the map from (q,j) to the Jordan partition m in (3.4), which is needed for the overlapping index to be well defined, is true and easy to check from the parity of the coefficients. Therefore the concern does not justify changing the ACCEPT verdict; it does justify a specific verification step, especially since the paper has no machine-checked proofs and the classification has not yet been independently vetted in the literature.","tokens_in":17711,"tokens_out":25844,"duration_ms":252045,"concrete_test":"Independently verify [ABD+24, Thm. 3.10] for small dimensions: for every partition n = Σ q_i i with n ≤ 8 and every admissible j, construct A via (3.8), compute its Jordan partition, and check it equals the partition given by (3.4); then check that each resulting partition m has a unique preimage (q,j). Additionally, compare against the low-dimensional classifications of [Sal01] and [COUV16] to confirm that every nilpotent almost abelian complex Lie algebra of real dimension at most 8 has a Jordan partition of the form (3.4). If all cases match and no extra isomorphism class appears, the external dependence is benign.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Proposition 3.10, the core uniqueness result, is proved by passing to an arbitrary complex structure J and constructing a basis in which (g,J) equals the explicit model (g_A,J_0) built from the partition (3.3) and the overlapping index j. This construction cannot get off the ground unless [ABD+24, Thm. 3.10], quoted as Theorem 3.2, is both complete and correctly translated: every nilpotent almost abelian Lie algebra admitting a complex structure must have Jordan partition given by (3.4) for some pair (q,j), and the index j must be well defined for the Lie algebra. The present paper does not reprove that classification; it cites a partly overlapping preprint. Moreover, the step in the proof of Proposition 3.10 that concludes 'B = B′ up to reordering of the Jordan blocks except the first one' is terse and presumes, without saying so, that the map (q,j) ↦ m defined by (3.4) is injective. That injectivity is true by inspection of the odd parts of m, but it is not stated, and the completeness half of the classification is entirely external. If a missing case or an incorrect coefficient existed in Theorem 3.2, the uniqueness statement would fail in that case, and Theorems 1.1 and 1.2, which are proved only from the normal form, would not cover all almost abelian complex nilmanifolds. This is an epistemic dependence rather than an internal contradiction, but it is the most load-bearing assumption in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that a complex structure on a nilpotent almost abelian real Lie algebra, when it exists, is unique up to Lie algebra isomorphism, and it uses this uniqueness to obtain strong geometric and cohomological consequences for the corresponding complex nilmanifolds. Working from the existence classification quoted as Theorem 3.2 from [ABD+24], the authors construct an explicit normal form (g_A, J_0) for any such complex structure (Proposition 3.10). From this normal form they prove Theorem 1.1: every almost abelian complex nilmanifold has a stable principal torus bundle series, its Dolbeault cohomology is computed by left-invariant forms, and its Frölicher spectral sequence degenerates at E1. They also prove Theorem 1.2: small deformations are again complex nilmanifolds, deformations are unobstructed, the Kuranishi family is universal, and every deformation in the large is again a complex nilmanifold. The tools are Jordan normal form, sl2-representation theory, Clebsch-Gordan rules, Hochschild-Serre spectral sequences, and standard results on nilmanifold deformations. Explicit formulas for Betti and Hodge numbers in terms of counts of Jordan blocks are derived and illustrated in several examples.","tokens_in":29,"tokens_out":23596,"duration_ms":315483,"significance":"If the external classification [ABD+24] is correct, this is a significant and novel contribution: it appears to be the first class of complex nilmanifolds with unbounded nilpotency index for which the Dolbeault cohomology and the deformation theory are fully controlled. The paper is careful in its use of standard tools, gives an explicit normal form that is used for all later computations, and provides a concrete algebraic recipe for computing Hodge and Betti numbers. The main weakness is an epistemic one: Proposition 3.10 and all subsequent statements depend crucially on Theorem 3.2, quoted from a preprint by a partially overlapping set of authors, which is not reproved here. This is a normal citation dependence rather than an internal inconsistency, but it should be stated explicitly so that readers are aware that the main results are conditional on the correctness of that classification.","major_comments":[],"minor_comments":[{"comment":"The sentence comparing the Jordan partitions of A, B and B' and concluding 'B = B′ up to reordering of the Jordan blocks except the first one' is too terse for a step on which the uniqueness claim rests; the conclusion follows from the parity structure of (3.4), but the authors should spell out that the map (q,j) ↦ m is injective, for example by noting that for j > 1 the largest part size with odd multiplicity is j, while for j = 1 all multiplicities except that of size 1 are even.","section":"§3.B, Proposition 3.10"},{"comment":"The displayed Hodge numbers h^{2,0}(X) = \\binom{n+1}{2} and h^{0,2}(X) = \\binom{n}{2} + 1 appear to be typographical errors: they should be δ(V^2 W_{n+1}) = \\lfloor (n+1)/2 \\rfloor and δ(V^2 W_n) + 1 = \\lfloor n/2 \\rfloor + 1, respectively, in order to be consistent with the Betti number computation b_2(X) = 2n + 2 and with the formula in Proposition 4.11.","section":"§4.C, Example 4.13"},{"comment":"In the proof of Theorem 1.2, the expression b_1(X) = 2δ(b^*_{1,0}) + 1 introduces the undefined symbol b^*_{1,0}; this should read δ(g^*_{1,0}), since δ(a^*) = 2δ(g^*_{1,0}) in the ε = 1 case by Lemma 4.8.","section":"§4.B, proof of Theorem 1.2"},{"comment":"In the ε = 1 case of Corollary 3.12, the formula 'd(e_2 + ie_{2+n}) = 1/(2i) α ∧ α' must be a typo, since α ∧ α = 0; the intended expression is α ∧ \\bar{α}.","section":"§3.B, Corollary 3.12"},{"comment":"The introduction states that for a nilpotent almost abelian Lie algebra of dimension 2n + 2 the step of nilpotency is at most n; this follows from the structure of the Jordan partition (3.4) but the argument is not given in the introduction, so the authors should either add a reference or move the justification forward.","section":"§3.A and Remark 3.5"},{"comment":"Since all main results depend on the classification quoted as Theorem 3.2 from [ABD+24], and that classification is not proved in this paper, the authors should add an explicit sentence in Section 3.B acknowledging this dependence and pointing to the precise statement in [ABD+24].","section":"§3.B, Theorem 3.2"}],"recommendation":"minor_revision","confidential_remarks":"The paper's main theorem is conditional on the classification in [ABD+24], which is an arXiv preprint by a partially overlapping set of authors. This is not circular, but the editor may wish to verify the status of that preprint and consider whether the present paper should include a fuller account of the classification or a note that the main results are contingent on it. Additionally, Example 4.13 contains what appears to be a numerical typo in h^{2,0} and h^{0,2}; this should be corrected before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear [colleague],\n\nHere's my take after reading the full thing: this is a genuinely strong paper. The core new result is Proposition 3.10 — a nilpotent almost abelian Lie algebra that admits a complex structure has that structure unique up to isomorphism, realized by an explicit normal form built from the Jordan partition. That uniqueness was not in the literature; [ABD+24] only classified existence. The paper then plays the normal form to get a stable torus bundle series, left-invariant computation of Dolbeault cohomology, E1-degeneration of the Frölicher spectral sequence, and unobstructed deformations both small and in the large. Those consequences are new, and the machinery used to get them is standard and, from what I can see, correctly applied. The explicit Hodge/Betti formulas via sl2-representation counts in Section 4 are a nice bonus: they make the results genuinely computable.\n\nThe soft spots are moderate and mostly about what the paper leans on. The uniqueness proof starts from [ABD+24, Thm 3.10], quoted as Theorem 3.2. If that existence classification had a missing case or an incorrect coefficient in (3.4), the normal form would not cover everything, and Theorems 1.1 and 1.2 would be incomplete. The paper does not reprove that classification. That is not an internal flaw — the cited result is a theorem with stated hypotheses, and the author overlap is normal in this line of work — but a referee should check it rather than take it on faith. There is also a terse step in the proof of Proposition 3.10 where the authors conclude B = B' up to reordering; that implicitly uses injectivity of the map (q,j) ↦ m, which is true but never stated. Minor normalization slips in the proof of Corollary 3.12 (factors/signs in the definition of α) are cosmetic and do not affect the conclusions.\n\nOverall, the central claim holds up. The paper is aimed at specialists working on nilmanifolds and complex structures on Lie algebras, and those readers will get real value here. I would send it to a serious referee rather than desk-reject, with the instruction to verify Theorem 3.2 and the injectivity point.\n\nBest.","headline":"A strong, well-executed paper proving uniqueness of complex structures on nilpotent almost abelian Lie algebras; the main caveat is its reliance on an overlapping external classification that is not reproved.","tokens_in":18588,"tokens_out":3043,"would_cite":true,"duration_ms":27583,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["32Q57","22E25","17B30","32G05"],"pacs":[],"model":"deepseek-v4-flash","headline":"A complex structure on a nilpotent almost abelian real Lie algebra, when it exists, is unique up to isomorphism, and the explicit normal form this gives determines the cohomology and deformation theory of the associated complex…","keywords":["almost abelian Lie algebra","complex nilmanifold","uniqueness of complex structure","Dolbeault cohomology","Frölicher spectral sequence","deformations of complex structures","Jordan normal form","torus bundle series"],"falsifier":"Enumerate nilpotent almost abelian Lie algebras of dimension 8, corresponding to partitions of 7, and check integrability of a declared almost complex structure: finding a complex structure whose adjoint Jordan partition is not of the form (3.4), or finding a single Lie algebra carrying two non-isomorphic complex structures, would refute the classification and the uniqueness theorem built on it.","tokens_in":17456,"feed_emoji":"🧮","tokens_out":7849,"duration_ms":62423,"temperature":0.7,"pith_summary":"This paper proves that a complex structure on a nilpotent almost abelian real Lie algebra—one containing an abelian ideal of codimension one—is unique up to isomorphism whenever it exists, and that the unique structure is given by an explicit normal form built from the Jordan normal form of the adjoint action. Because the normal form is explicit, the associated compact quotients, the almost abelian complex nilmanifolds, become fully accessible: they are iterated principal holomorphic torus bundles, their Dolbeault cohomology is computed by left-invariant forms, and their Frölicher spectral sequence degenerates at the first page. The same control extends to families: every small deformation is unobstructed, and every deformation in the large is again a complex nilmanifold. This gives a class of complex nilmanifolds with unbounded nilpotency index whose cohomology and deformations are completely determined by the underlying Lie algebra.","feed_headline":"Complex structures on almost abelian nilmanifolds are unique","feed_subtitle":"Uniqueness gives explicit Hodge numbers, E1 degeneration, and unobstructed deformations.","key_machinery":"The machinery is a normal form for $(\\mathfrak{g}, J)$. The classification theorem [ABD+24] encodes the Jordan partition of the nilpotent adjoint matrix $A$ in a partition $n = \\sum_i q_i i$ with a distinguished index $j$, the size of the overlapping block; the possible Jordan partitions of $A$ are listed in (3.4). With this data, $A$ is the matrix (3.8), consisting of two identical Jordan blocks from the partition of the subspace plus an off-diagonal vector, and $J_0$ is the standard block matrix (3.9). A stable principal torus bundle series—a rational, $J$-invariant filtration whose successive quotients lie in the centre—then supplies the iterated holomorphic torus bundle structure, and the Hochschild–Serre spectral sequence together with $\\mathfrak{sl}_2(\\mathbb{C})$-representation counts supplies the cohomological formulas.","core_discovery":"The central discovery is Proposition 3.10: for a nilpotent almost abelian Lie algebra $\\mathfrak{g}$ of dimension $2n+2$ with integrable complex structure $J$, there is a basis $e_0,\\dots,e_{2n+1}$ in which the codimension-one abelian ideal is $\\langle e_1,\\dots,e_{2n+1}\\rangle$, $\\operatorname{ad}(e_0)$ restricts to the fixed Jordan block matrix $A$ of (3.8), and $J$ is represented by the fixed block matrix $J_0$ of (3.9). The size $j$ of the overlapping block in the classification of [ABD+24] determines the matrix, so any two complex structures on the same Lie algebra are conjugate by a Lie algebra automorphism. From this normal form the authors derive Theorem 1.1: a stable torus bundle series, left-invariant computation of Dolbeault cohomology, and $E_1$-degeneration of the Frölicher spectral sequence. They then derive Theorem 1.2: the Kuranishi space of deformations is smooth and universal, and every deformation in the large is a complex nilmanifold.","pith_inferences":["The uniqueness result implies that the space of left-invariant complex structures on a fixed almost abelian nilmanifold is either empty or a single point up to isomorphism; continuous families of complex structures must therefore come from varying the Lie algebra or from non-left-invariant structures, a distinction the paper does not address.","The explicit $\\mathfrak{sl}_2(\\mathbb{C})$-counting algorithm for Hodge and Betti numbers is likely reusable for other cohomology theories, such as Bott–Chern or Aeppli cohomology, where $E_1$-degeneration alone does not determine the groups; testing left-invariance there is a natural next step.","Because the torus bundle series is stable, the same fibration structure could be used to study Hermitian metric functionals on all almost abelian complex nilmanifolds; the paper only notes that SKT metrics force the Lie algebra to be $\\mathfrak{h}_3 \\oplus$ abelian in the nilpotent case.","The large-deformation statement is proved via a good fibre class argument; whether it extends from nilmanifolds to the larger class of almost abelian solvmanifolds is not addressed and would be a separate question."],"forward_implications":["Every almost abelian complex nilmanifold is an iterated principal holomorphic torus bundle; the bundle structure is stable under all integrable complex structures and all rational structures, hence independent of the lattice.","Dolbeault cohomology is computed by left-invariant forms, $H^{p,q}_{\\bar\\partial}(X) \\cong H^q(\\mathfrak{g}^{0,1}, \\Lambda^p \\mathfrak{g}^{*1,0})$, and the Frölicher spectral sequence degenerates at $E_1$, so Betti numbers are sums of Hodge numbers.","All Betti and Hodge numbers can be obtained by an explicit algorithm: decompose the relevant representations of the one-dimensional complex Lie algebra into $\\mathfrak{sl}_2(\\mathbb{C})$-irreducible summands and apply the formulas of Proposition 4.11.","Small deformations are unobstructed with a smooth universal Kuranishi family, and every fibre is again a complex nilmanifold; the same holds for deformations in the large over a connected base.","The complex structure is always nilpotent and, unless the Lie algebra is $\\mathfrak{h}_3 \\oplus \\mathbb{R}^{2n-1}$, never abelian; the class therefore gives examples of arbitrarily large nilpotency index with full cohomological control."],"supporting_citations":[{"why":"Supplies the classification of which nilpotent almost abelian Lie algebras admit complex structures, stated as Theorem 3.2 and used as the starting point for the uniqueness normal form.","marker":"[ABD+24]"},{"why":"Establishes that isomorphism classes of almost abelian Lie algebras correspond to conjugacy classes of the adjoint matrix, so Jordan partitions parametrise the Lie algebras.","marker":"[Fre12]"},{"why":"Introduces stable principal torus bundle series and proves the criterion turning such a series into a tower of holomorphic torus bundles; also supplies Theorem A used for deformations in the large.","marker":"[Rol09a]"},{"why":"Proves the inclusion of Lie-algebra Dolbeault cohomology into the manifold's Dolbeault cohomology and gives the section the paper invokes to conclude left-invariant forms compute it.","marker":"[CF01]"},{"why":"Provides the nilpotent complex structure criterion that, together with Remark 3.13, yields left-invariant computation of Dolbeault cohomology.","marker":"[RTW20]"},{"why":"Provides the small-deformation result that small deformations of complex nilmanifolds are again complex nilmanifolds, a key input to Theorem 1.2(i).","marker":"[Rol09b]"},{"why":"Shows unobstructedness for complex manifolds with trivial canonical bundle and $E_1$-degenerating Frölicher spectral sequence, used for smoothness of the Kuranishi space.","marker":"[ACRT18]"},{"why":"Observed triviality of the canonical bundle of complex nilmanifolds, a property needed in the unobstructedness argument.","marker":"[Sal01]"},{"why":"Gives the criterion for universality of the Kuranishi family via constancy of the automorphism group dimension.","marker":"[Wav69]"}],"fun_headline_variants":["Only one complex structure for almost abelian nilmanifolds","Almost abelian nilmanifolds admit at most one complex structure","Unique complex structure on almost abelian nilmanifolds","At most one complex structure per almost abelian nilmanifold","Complex structure uniqueness for almost abelian nilmanifolds"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the external classification of which nilpotent almost abelian Lie algebras admit complex structures is complete and correct; the paper builds its normal form on that classification and does not reprove it.","fun_headline_variants_meta":{"raw":{"variants":["Only one complex structure for almost abelian nilmanifolds","Almost abelian nilmanifolds admit at most one complex structure","Unique complex structure on almost abelian nilmanifolds","At most one complex structure per almost abelian nilmanifold","Complex structure uniqueness for almost abelian nilmanifolds"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000379,"raw_usage":{"total_tokens":1947,"prompt_tokens":809,"completion_tokens":1138,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":425,"completion_tokens_details":{"reasoning_tokens":1052}},"tokens_in":425,"tokens_out":1138,"duration_ms":9212,"temperature":1.0,"reasoning_tokens":1052,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-09T05:13:40.966833+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Enumerate nilpotent almost abelian Lie algebras of dimension 8, corresponding to partitions of 7, and check integrability of a declared almost complex structure: finding a complex structure whose adjoint Jordan partition is not of the form (3.4), or finding a single Lie algebra carrying two non-isomorphic complex structures, would refute the classification and the uniqueness theorem built on it.","supporting_citations":[],"review_version":1}