{"id":"26df993c-951c-4338-ac67-a8e682025004","arxiv_id":"2412.03437","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Norms that are sums of absolute values of rational linear functionals are exactly the nonvanishing Thurston norms of closed oriented graph manifolds, which can be chosen to fiber over the circle or not.","lead":"A new theorem characterizes which geometric norms can appear as the Thurston norm of a graph manifold: exactly those built from sums of absolute values of rational linear functionals. The result gives a general construction of 3-manifolds with prescribed Thurston unit balls, including every rational symmetric polygon in dimension two.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The corrected Neumann plumbing lemma (Lemma 2.17) is load-bearing for Proposition 3.2 and Theorem 4.1, and its proof is still only a sketch; verifying the rational-extension step would settle the main residual risk.","rationale":"The central construction is otherwise in good shape: Lemma 4.2 is elementary and correct, Lemma 4.3 is a standard plumbing/surgery construction that can be checked entry by entry, and Lemma 3.1/Proposition 3.3 are standard additivity arguments. The one place where a false statement would propagate through the whole paper is the exactness of the sequence in Proposition 3.2, which is exactly the content of the corrected Neumann lemma. The paper itself shows the original lemma is false, so the correction is not merely cosmetic; the sketch relies on a reduction of the plumbing intersection form whose details are not supplied. My proposed check isolates that reduction and the rational-extension/lifting step. A minor additional imprecision is that the non-fibering branch of Theorem 4.1 appends a zero vector β_{n+1}=0, which increases the number of JSJ pieces beyond the prescribed genera; this is easily repaired and does not affect the realizability claim. Overall the reader's conditional verdict is appropriate.","tokens_in":18965,"tokens_out":18078,"duration_ms":184121,"concrete_test":"Work out the unproved step of Lemma 2.17 in exact rational arithmetic for a concrete non-torus graph manifold whose JSJ decomposition has two pieces and whose plumbing refinement adds at least one extra vertex per piece. Enumerate all integer tuples l with A l = 0 up to a fixed norm bound and check that each admits a rational tuple l' on the added coordinates annihilating the full intersection matrix S; then verify that after clearing denominators the lifted tuple lies in the image of H2(M;Z). If every test case passes and the claimed direct-sum reduction of S to A ⊕ D is reproduced, the concern is resolved; the first counterexample would refute Lemma 2.17 and collapse Theorem 4.1.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The realizability theorem depends on Proposition 3.2, which identifies H_nv(M;Q) with ker A ∩ Q^n via the corrected Neumann plumbing lemma (Lemma 2.17). The original lemma is explicitly false, and the correction introduces a quantifier over a natural number n0 that is essential to the argument. In the sketch, an integer tuple l with A l = 0 is extended to a rational tuple annihilating the plumbing intersection matrix S, then multiplied by n0 to clear denominators. This requires two unproved facts: (i) the rational extension exists for every l in ker A, and (ii) after clearing denominators, the resulting tuple lies in the image of H2(M;Z) → H2(X;Z), so it is represented by a surface. Both facts rest on the claimed reduction of S to a direct sum of A with a diagonal block by rational row and column moves; if that reduction fails, the image of ψ(H2(M;Q)) can be a proper subspace or sublattice of ker A ∩ Q^n. Since Theorem 4.1, Corollary 4.4, and Corollaries 5.7–5.8 all build on this identification, a gap here is the single most load-bearing risk.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the Thurston norm of closed oriented graph manifolds. It proves that for a good graph manifold the nonvanishing Thurston norm is a sum of absolute values of linear functionals, with either all or no top-dimensional faces of the unit ball fibered. The main theorem (Theorem 4.1) asserts a converse: every norm on R^d expressible as a sum |<β_i,·>| with rational β_i is realized as the nonvanishing Thurston norm of some closed oriented good graph manifold with prescribed base genera, and the construction can be arranged so that the manifold either fibers over the circle or does not. A final section analyzes the unit polyhedra of such norms, showing that they are 'complete' in a precisely defined sense, and that every rational symmetric polyhedron is refined by a realizable one.","tokens_in":19211,"tokens_out":26683,"duration_ms":249064,"significance":"If the main theorem is correct, it gives a complete and constructive characterization of the possible nonvanishing Thurston norms of graph manifolds, with additional control over fibering behavior and base genera. The paper also identifies and corrects a genuine error in Neumann's plumbing lemma, which is a useful contribution in itself, and it introduces a clean notion of completeness for linear cellular decompositions of spheres. The arguments are largely explicit and constructive, and the paper is written in a clear style. However, two load-bearing points are not yet fully established, so the main theorem cannot be considered proven as the manuscript currently stands.","major_comments":[{"comment":"Lemma 2.17 is the foundation of Proposition 3.2 and therefore of Theorem 4.1 and its corollaries, but only a proof sketch is supplied. The manuscript itself states that the original version of Neumann's lemma is false and gives a counterexample, so a complete proof of the corrected statement is essential. The sketch asserts without proof that the plumbing intersection matrix S can be reduced by rational row and column moves to a direct sum of the reduced plumbing matrix A with a diagonal block, and that consequently every integer tuple annihilating A extends to a rational tuple annihilating S. It further asserts that after clearing denominators this tuple is realized by an embedded surface through the exact sequence H_2(M;Z) → H_2(X;Z). These are precisely the steps needed to identify ψ(H_2(M;Q)) with ker A ∩ Q^n in Proposition 3.2; if either step fails, the image of ψ could be a proper subspace or sublattice of ker A. A rigorous proof of the corrected lemma, or a precise reference to a published proof, is required before the central identification is established.","section":"Section 2.3, Lemma 2.17 and its proof sketch"},{"comment":"After obtaining a matrix A with kernel spanned by the columns v_i of P, the proof defines vectors v'_i by scaling the coordinates of v_i by -1/χ_j and asserts that (v'_1,...,v'_d) is again a basis of ker A. This is not true in general: multiplying the coordinates of vectors in a linear subspace by a diagonal matrix does not preserve the subspace. For example, if ker A = span{(1,2)} and the diagonal scaling is diag(2,1), then (1,2) is sent to (2,2), which is not in the kernel. In the paper's setting the χ_i are outputs of Lemma 4.3 and are not prescribed in advance, so the proof gives no reason for the specific diagonal matrix with entries -1/χ_i to preserve ker A. Since Proposition 3.3 is then applied to the purported basis (v'_1,...,v'_d), the proof of Theorem 4.1 is incomplete. The argument needs to be restructured, for instance by choosing the matrix A so that its kernel is spanned by the scaled vectors, rather than by the original columns of P.","section":"Section 4, proof of Theorem 4.1"}],"minor_comments":[{"comment":"The phrase 'we can execute m (mα_i, β_i)-Dehn surgeries' is ambiguous; it should be clarified whether this means m surgeries each with coefficients (mα_i, β_i), and the effect of this operation on the orbifold Euler characteristic should be spelled out.","section":"Lemma 4.3, proof"},{"comment":"The sentence describing the CW-chain complex of the graph as the sequence 0 → GE → GV → 0 omits the augmentation to Z; writing the standard complex 0 → C_1 → C_0 → Z → 0 would make the computation of H_1(Γ) clearer.","section":"Proposition 3.4, proof"},{"comment":"The statement that non-completeness follows because 'there are vertices with odd degree' would benefit from a brief explanation of why this prevents equality in the defining inclusion of the (n−2)-skeleton.","section":"Example 5.5"},{"comment":"There is a typo in 'Aknowledgements' (should be 'Acknowledgements'), and the notation 'RP2 e×S1' in Proposition 2.13 should be typeset consistently.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"This is a promising paper with a substantial constructive result and a valuable correction to Neumann's lemma, but the two major comments above are load-bearing. The issue with the basis (v'_i) in Theorem 4.1 is not merely a missing detail; the claim is false as stated. I believe it is fixable by recasting the argument (e.g., choosing the matrix with kernel spanned by the scaled columns), but it requires real work and should be checked carefully. The incomplete proof of Lemma 2.17 is also a correctness risk that must be resolved before the central theorem is accepted. If the author can supply a complete proof of the corrected Neumann lemma and repair the basis argument, the paper would be a strong contribution."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis is a genuine contribution to the Thurston norm realization program. The main new content is the converse direction: every norm on R^d expressible as a sum of absolute values of rational linear functionals is the nonvanishing Thurston norm of a closed oriented good graph manifold, with prescribed base genera, and you can choose whether the manifold fibers over the circle. That is a real advance, not a repackaging of known results. The completeness obstruction in Section 5 is also useful: such norms have complete unit balls, so the cube is not realizable, and Remark 5.11 shows that not every complete rational symmetric polyhedron is realizable. The fibering control (all or none of the top-dimensional faces) is a nice extra.\n\nThe paper is honest about its tools. It relies on Neumann's plumbing calculus, and importantly it corrects Neumann's Lemma 4.2: the original statement is false, and the counterexample given is convincing. The corrected Lemma 2.17 is exactly what Proposition 3.2 needs to identify the image of H_2(M;Q) in R^n with ker A ∩ Q^n, and Theorem 4.1 leans on that identification.\n\nHere is where the paper is not ready as written. Lemma 2.17 is load-bearing and the proof is only a sketch. The sketch has the right shape — reduce the plumbing intersection matrix S by rational row and column moves to a direct sum of A with a diagonal block, extend an annihilating tuple for A to one for S, then multiply by n0 to clear denominators — but it doesn't actually verify that the reduction exists for the full S, nor that the resulting integer tuple is represented by a surface in M. Those are the two facts that make the identification true, and without them the central theorem is conditional. The counterexample to the original lemma suggests the correction is genuine, and I suspect a complete proof can be extracted from Neumann's paper, but suspicion is not a proof.\n\nEverything else holds up. Lemmas 4.2 and 4.3 are clean: the symmetric integral matrix with prescribed kernel exists, and the Dehn-surgery construction for a graph manifold with that reduced plumbing matrix is convincing. The citation pattern is fair.\n\nThis paper is for 3-manifold topologists and anyone working on norm realization. It deserves a serious referee, and I would engage with it, but the referee should ask for a complete proof of Lemma 2.17 (or a precise citation to a source where the corrected version is proven) before acceptance. With that fixed, the paper is a solid accept.\n\nBest.","headline":"Fresh converse realizability theorem for Thurston norms on graph manifolds, but the paper's load-bearing Neumann lemma is only sketched and needs a full proof before I'd trust Theorem 4.1.","tokens_in":19712,"tokens_out":8771,"would_cite":true,"duration_ms":78414,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K30","57K31"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every rational absolute-value norm is a Thurston norm of a graph manifold.","keywords":["Thurston norm","graph manifolds","Seifert fibered manifolds","reduced plumbing matrix","fibered cones","polyhedral unit balls","norm realization","JSJ decomposition"],"falsifier":"Take a closed oriented good graph manifold whose reduced plumbing matrix $A$ has nontrivial kernel, compute the image of $H_2(M;\\mathbb{Q})$ under the fiber-intersection map $\\psi$ directly from the gluing data, and test whether it equals $\\ker A \\cap \\mathbb{Q}^n$; any class in the kernel whose positive multiples are never realized by algebraic intersection tuples of embedded surfaces would falsify the corrected lemma and the main theorem. Equivalently, finding a symmetric integral matrix with prescribed rational kernel that is provably not the scalar multiple of any reduced plumbing matrix would break the construction.","tokens_in":18773,"feed_emoji":"📐","tokens_out":6741,"duration_ms":63699,"temperature":0.7,"pith_summary":"The paper proves a converse statement: any norm on $\\mathbb{R}^d$ that is a finite sum of absolute values of rational linear functionals is the nonvanishing Thurston norm of some closed oriented good graph manifold, with respect to a rational basis of its second real homology. The realizing manifold can be chosen either to fiber over the circle or not to fiber, so the dichotomy forced by graph manifolds—all top-dimensional faces of the Thurston ball fibered, or none—is the only obstruction to realization. Since in dimension two every norm with a rational polygonal unit ball has the required form, every such polygon is realized, including by fibered graph manifolds. In higher dimensions the paper identifies a structural property, completeness, that all graph-manifold Thurston balls possess, and shows that every symmetric rational polyhedron can be refined by the unit ball of such a norm. The result pins down exactly which polyhedral norms can come from this large, tractable family of 3-manifolds.","feed_headline":"Every rational absolute-value norm is a Thurston norm","feed_subtitle":"Graph manifolds realize all such norms on a rational basis, with fibering or non-fibering chosen freely.","key_machinery":"The load-bearing object is the reduced plumbing matrix $A$ of a graph manifold, whose diagonal entries are Euler numbers of the JSJ Seifert pieces and off-diagonal entries are sums of reciprocals of algebraic intersection numbers of regular fibers across decomposing tori. The corrected plumbing lemma (Lemma 2.17) identifies, up to a positive integer multiple, the tuples of algebraic intersection numbers of an embedded surface with the regular fibers of the pieces with the kernel of $A$; Proposition 3.2 then identifies the nonvanishing homology $H_2^{\\mathrm{nv}}(M)$ with $\\ker A$. This reduces the realizability of a norm to finding a symmetric integral matrix with a prescribed rational kernel and then building a graph manifold whose reduced plumbing matrix is a scalar multiple of it.","core_discovery":"The central claim is Theorem 4.1: given rational vectors $\\beta_1,\\dots,\\beta_n \\in \\mathbb{Q}^d$ whose absolute-value sums define a norm, and given nonnegative integers $g_1,\\dots,g_n$, there exists a closed oriented good graph manifold $M$—meaning all its Seifert fibered JSJ pieces have orientable base orbifolds—whose pieces have base surfaces of those genera, together with a rational basis of $H_2(M;\\mathbb{R})$, such that the nonvanishing Thurston norm $x^{\\mathrm{nv}}$ coincides with $\\sum_i |\\langle \\beta_i,\\cdot\\rangle|$. The proof first shows that every good graph manifold has this form, with coefficients coming from the orbifold Euler characteristics of its Seifert pieces, and then inverts the construction: a symmetric integral matrix with a prescribed kernel is realized as the reduced plumbing matrix of a graph manifold built from surface bundles by Dehn surgeries. The same machinery gives a formula for the dimension of the null-space of the Thurston norm, shows that the fibered top-dimensional cones are either all or none depending on whether any $\\beta_i$ vanishes, and yields the two-dimensional corollary that every rational polygon is the unit ball of the nonvanishing Thurston norm of a fibering graph manifold.","pith_inferences":["Beyond the paper, the rational-basis restriction suggests a testable strengthening: if the construction could be adjusted to integral bases, the same norms would appear integrally rather than up to change of rational basis; the paper states this as an open direction.","Beyond the paper, the completeness obstruction separates graph manifolds from arbitrary 3-manifolds, since the chain-link example shows a non-complete Thurston ball exists; one could ask how much of the gap is filled by other families of 3-manifolds.","Beyond the paper, the all-or-none fibering dichotomy plus free choice of fibering suggests using these manifolds as test cases for algorithms that detect fibrations from the shape of the Thurston ball.","Beyond the paper, one could try to compute the minimal number of summands $n$ needed to realize a given polyhedral norm and ask whether it relates to the number of JSJ pieces; the paper does not address this."],"forward_implications":["The class of nonvanishing Thurston norms of closed oriented good graph manifolds is exactly the class of norms that are finite sums of absolute values of rational linear functionals.","For any such norm, the realizing graph manifold can be chosen either to fiber over the circle (all top-dimensional faces of its Thurston ball fibered) or not to fiber, at will.","Every norm on $\\mathbb{R}^2$ whose unit ball is a polygon with rational vertices occurs, so every such polygon is the unit ball of the nonvanishing Thurston norm of a good graph manifold, fibering if desired.","In dimension at least three, a polyhedron that is not complete—such as the cube $[-1,1]^n$—cannot be the nonvanishing Thurston unit ball of any graph manifold.","Every symmetric rational polyhedron in $\\mathbb{R}^n$ admits a good graph manifold whose nonvanishing Thurston unit ball refines the decomposition into cones over the polyhedron's faces, with fibering behavior again chosen freely."],"supporting_citations":[{"why":"Supplies the plumbing lemma, corrected here, that ties surface intersection tuples to the kernel of the reduced plumbing matrix; this is the bridge from topology to linear algebra on which Theorem 4.1 rests.","marker":"[8]"},{"why":"Defines the Thurston norm, its polyhedral unit ball, and the fibered-face criterion used throughout the paper.","marker":"[13]"},{"why":"Provides the classification of surfaces in Seifert fibered manifolds as horizontal or vertical and the Euler characteristic formulas used to compute the norm on each piece.","marker":"[6]"},{"why":"Underlies the plumbing calculus invoked in the proof of the corrected plumbing lemma.","marker":"[7]"},{"why":"Gives additivity of the Thurston norm under cutting along incompressible tori, used to split the norm across JSJ pieces.","marker":"[4]"}],"fun_headline_variants":["Graph manifolds realize every rational absolute-value norm","Every rational absolute-value norm is a graph manifold Thurston norm","Graph manifolds achieve all rational absolute-value norms","All rational absolute-value norms are Thurston norms of graph manifolds","Every rational polygon is a unit ball of a graph manifold"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The whole construction depends on a corrected technical lemma about surface intersections and the plumbing matrix, which the paper supports with a counterexample and a proof sketch rather than a complete proof.","fun_headline_variants_meta":{"raw":{"variants":["Graph manifolds realize every rational absolute-value norm","Every rational absolute-value norm is a graph manifold Thurston norm","Graph manifolds achieve all rational absolute-value norms","All rational absolute-value norms are Thurston norms of graph manifolds","Every rational polygon is a unit ball of a graph manifold"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00071,"raw_usage":{"total_tokens":3218,"prompt_tokens":986,"completion_tokens":2232,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":602,"completion_tokens_details":{"reasoning_tokens":2152}},"tokens_in":602,"tokens_out":2232,"duration_ms":15008,"temperature":1.0,"reasoning_tokens":2152,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T22:26:12.814087+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a closed oriented good graph manifold whose reduced plumbing matrix $A$ has nontrivial kernel, compute the image of $H_2(M;\\mathbb{Q})$ under the fiber-intersection map $\\psi$ directly from the gluing data, and test whether it equals $\\ker A \\cap \\mathbb{Q}^n$; any class in the kernel whose positive multiples are never realized by algebraic intersection tuples of embedded surfaces would falsify the corrected lemma and the main theorem. Equivalently, finding a symmetric integral matrix with prescribed rational kernel that is provably not the scalar multiple of any reduced plumbing matrix would break the construction.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the plumbing lemma, corrected here, that ties surface intersection tuples to the kernel of the reduced plumbing matrix; this is the bridge from topology to linear algebra on which Theorem 4.1 rests."},{"cited_title":"A Norm for the Homology of 3-Manifolds","cited_arxiv_id":null,"evidence_quote":"Defines the Thurston norm, its polyhedral unit ball, and the fibered-face criterion used throughout the paper."},{"cited_title":"An Introduction to Geometric Topology","cited_arxiv_id":null,"evidence_quote":"Provides the classification of surfaces in Seifert fibered manifolds as horizontal or vertical and the Euler characteristic formulas used to compute the norm on each piece."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Underlies the plumbing calculus invoked in the proof of the corrected plumbing lemma."},{"cited_title":"The efficient certification of knottedness and Thurston norm","cited_arxiv_id":null,"evidence_quote":"Gives additivity of the Thurston norm under cutting along incompressible tori, used to split the norm across JSJ pieces."}],"review_version":1}