{"id":"8e24b98b-1f54-4923-9692-1cfe18079834","arxiv_id":"2411.15680","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Infinite families of graph manifolds and Seifert fibered spaces contain essential tori that are not detected by any ideal point of the SL2-character variety over any algebraically closed field.","lead":"This paper constructs infinite families of 3-dimensional spaces where a particular donut-shaped surface is invisible to a standard algebraic detection method, regardless of the field used. These examples show a limitation of a key tool in 3-manifold topology and reveal that changing the field's characteristic can change which surfaces are detectable.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Undetection theorems are conditional on the unverified positive-characteristic Culler–Shalen theory of [11]; if Cor. 7 or Prop. 10 fails over characteristic 2, the main results do not follow.","rationale":"The reader's weakest assumption identifies the same load-bearing concern that I find: the entire detection/undetection machinery is imported from the self-authored preprint [11], and the paper gives no independent verification of Corollary 7, Lemma 8, or Proposition 10 in positive characteristic. The explicit matrix computations in Sections 3–5 are extensive and plausible; the main theorems would be settled if [11] is correct. But the claim 'not detected by the variety of SL2(F)-characters for any algebraically closed field F' is a universal negative statement, so it demands certainty that every ideal point of every curve in X(NPhi,F) has been accounted for and that the ideal-point-to-splitting construction works in characteristic 2. A single failure of Lemma 8 or Proposition 10 over F2 could introduce a detecting curve or alter the detected surface, invalidating Theorems 1, 2, 3, and 5. Thus the verdict should remain CONDITIONAL, pending external verification of [11]. I see no reason to change the reader's assessment.","tokens_in":20787,"tokens_out":35562,"duration_ms":295721,"concrete_test":"Independently re-prove Proposition 10 from first principles over F = the algebraic closure of F2: take the irreducible character curve of the trefoil complement, X_irr(M,F) = {(s,s,1) | s in F}, form the function field, construct the associated Bass–Serre tree from the ideal point at s = infinity, and verify that the induced splitting of pi1(M) has the vertical annulus R as the detected surface (as asserted in Section 4). If the splitting differs from the claimed one, the foundation fails and the undetection conclusions do not follow.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central negative claim—that the exhibited tori are not detected by the variety of SL2(F)-characters over any algebraically closed field—relies on the full force of the extension of Culler–Shalen theory to arbitrary characteristic developed in the authors' preprint [11]. Section 2.1 concedes that when p>0, the trace ring TΓ may be a proper subring of the invariant ring F[R]^{SL2}, so X(Γ,F) is not literally the GIT quotient R//SL2. The paper's Proposition 10 is stated as [22, Property 5.4.2] with the note 'the proof applies verbatim', but that proof is written over C and assumes the standard identification of the character variety with the quotient. The construction of the tautological representation over the function field of a curve, and the conclusion that each ideal point produces a splitting with the stated vertex and edge stabilizers, is precisely what [11] must establish in positive characteristic. Lemma 8 (the commutator trace criterion) is likewise cited from [11] and is used in Sections 3 and 4 to identify the irreducible components; a failure of this criterion in characteristic 2 would change the curve decomposition in Section 5.2 and could introduce or eliminate curves that detect the splitting torus T. Because Theorems 1–6 all depend on this foundation, this is the most load-bearing assumption in the paper.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper constructs closed orientable 3-manifolds N_Φ by gluing the twisted I-bundle over the Klein bottle to the complement of the right-handed trefoil along a torus via Φ ∈ SL_2(Z). For each such manifold the authors classify the connected essential surfaces up to isotopy and study which of them are detected by ideal points of curves in the variety of SL_2(F)-characters for arbitrary algebraically closed fields F. The main results exhibit infinite families in which: a single essential torus is never detected in any characteristic (Theorem 1); a single essential torus is detected exactly in characteristic 2 (Theorem 2); a torus and a genus-two surface are both detected exactly in characteristic 2 (Theorem 3); a torus is never detected while a non-separating genus-two surface is always detected (Theorem 4); and Seifert fibered examples contain vertical tori with characteristic-dependent or uniform detection behaviour (Theorems 5 and 6).","tokens_in":20999,"tokens_out":17556,"duration_ms":154343,"significance":"If the main results are correct, they provide the first infinite families of closed 3-manifolds with essential surfaces that evade detection by SL_2-character varieties over every algebraically closed field, and also examples where switching to characteristic 2 improves detection. This is a meaningful contribution to the study of Culler-Shalen theory in positive characteristic. The paper is built on explicit matrix computations, a complete classification of essential surfaces in the constructed manifolds, and concrete families of gluing matrices, which are valuable assets. The main caveat is that the detection arguments rely on the authors' companion preprint [11] for the positive-characteristic extension of Culler-Shalen theory, so the unconditional status of the theorems depends on that work.","major_comments":[{"comment":"The graph-manifold condition stated in §5.4, 'm ≠ ±6 ≠ n', is inconsistent with the families used in Theorems 1 and 2. The family Φ_k = [[1+6k, k], [6, 1]] has m = 6, and the family Φ'_k = [[k, 1+6k], [-1, -6]] has n = -6; both are claimed to be graph manifolds. The correct condition should exclude precisely the two Seifert fibered families, namely Φ = [[k, ±1], [∓1-6k, ∓6]] and Φ = [[±1, l], [∓6, ±1-6l]]. As written, the condition misclassifies the examples and therefore the graph-manifold assertions in Theorems 1-4 need to be corrected.","section":"§5.4 and §5.3"},{"comment":"The matrix family for the Seifert fibered case with base S^2(2,2,2,3) is misstated. The displayed family Φ = [[k, ±1], [∓1, -6k ∓6]] is not in SL_2(Z) for arbitrary k; for example, with k=1 and one choice of signs the determinant is -11. Moreover, the proof of Lemma 14 itself says that the gluing equation 'requires n = -6l', which gives n = ∓6 and m = ∓1 - 6k, not the displayed entries. The correct family should be Φ = [[k, ±1], [∓1-6k, ∓6]]. This mistake propagates to Lemmas 14, 15, and 17 and to Theorem 6, even though the computations inside the proofs appear to use the correct condition n = -6l.","section":"§5.1 (S4), §5.3, §5.5"},{"comment":"The detection results in Section 5 depend on the full positive-characteristic Culler-Shalen framework of the preprint [11], specifically Corollary 7, Lemma 8, and Proposition 10. Proposition 10 is quoted from [22, Property 5.4.2] with the note that the proof applies verbatim, but [22] works over C and uses the identification of the character variety with the GIT quotient R//SL_2, which §2.1 explicitly notes may fail in positive characteristic. If Proposition 10 or Lemma 8 fails in characteristic 2, the conclusions that the exhibited tori are undetected in every characteristic would not follow. The authors should either include a proof of the positive-characteristic statements needed here or clearly state that the main theorems are conditional on the companion preprint [11] being available and correct.","section":"§2.3, Proposition 10"}],"minor_comments":[{"comment":"The sentence 'We write F = F_p if we want to emphasise the characteristic' is confusing because F already denotes an algebraically closed field; the finite field of p elements is usually denoted F_p and should not be conflated with the algebraic closure.","section":"§2.1"},{"comment":"After Equation (3.5), the text says the matrix pair 'splits into infinitely many conjugacy classes'; it would be helpful to state explicitly that these conjugacy classes are not distinguished by the character, since the character variety is the quotient by closure equivalence rather than by conjugacy.","section":"§3"},{"comment":"In the proofs of Lemmas 14 and 15, the phrase 'so it is of is of type (C3)' contains a duplicated word and should read 'so it is of type (C3)'.","section":"§5.3"},{"comment":"The 2-dimensional component C_Φ appears in both Lemma 14's proof and Lemma 17's proof with identical definitions; the paper could refer back to Equation (5.6) instead of restating the full coordinate expression in Equation (5.10).","section":"§5.5, Lemma 17"},{"comment":"The remark that Theorem 6 is not proved for all essential tori in the stated Seifert fibered manifolds is useful and should be incorporated into the theorem statement so that the scope of the claim is unambiguous.","section":"Theorem 6 and Remark after it"},{"comment":"The paper relies on the unpublished companion works [11] and [17]; a sentence in the introduction stating this dependence explicitly would help readers assess the conditional nature of the results.","section":"Introduction and References"}],"recommendation":"major_revision","confidential_remarks":"The main theorems are conditional on the authors' own preprint [11] for the positive-characteristic Culler-Shalen machinery, and the paper also cites [17] as 'in preparation'. The editor may wish to consider whether the journal accepts results whose foundational input is an unpublished companion paper. In addition, the misstated matrix family and the incorrect graph-manifold condition suggest that a careful proofreading pass focused on the displayed matrices and classification statements is needed before the paper can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe headline is that this is a genuine contribution. The paper gives infinite families of closed graph manifolds and Seifert fibered spaces with essential tori not detected by ideal points of the SL2(F)-character variety for any algebraically closed field F, and — more interestingly — families where a torus is detected only in characteristic 2. That characteristic-dependence is new and is not visible over C. The constructions are explicit, the essential surface classification in Section 5 is careful, and the paper is honest about what it does and does not prove: Theorem 6 only covers an infinite family of vertical tori, not all of them, and the conjecture is labeled as such.\n\nThe main caveat is one the authors would want to fix. Section 5.4 states that NΦ is a graph manifold iff Φ = [[k,l],[m,n]] with m ≠ ±6 ≠ n, but the family used in Theorem 2 is Φ′_k = [[k, 1+6k],[-1,-6]], which has n = -6. The paper calls these graph manifolds, and they do not fall into the Seifert-fibered families, so the condition in 5.4 is simply wrong as stated. This does not appear to break the main theorems, but it is an internal contradiction that needs correcting in revision.\n\nThe second soft spot is the reliance on the authors' preprint [11] for the positive-characteristic Culler–Shalen theory, specifically Corollary 7, Lemma 8, and Proposition 10. The paper says the proof of Proposition 10 applies verbatim from Shalen, but that proof is written over C. If any of these fail in characteristic 2, the undetection conclusions would not follow. This may be a routine extension, but it is load-bearing and not independently verified. A referee should either check that the cited preprint is solid or ask the authors to include the needed statements.\n\nThird, the case analyses in Section 5 are exhaustive in the sense of listing options, but the elimination details are often summarized as 'no other curves are found.' The explicit families are convincing, but the negative claims rest on unshown computations. This is a mild opacity, not a fatal flaw.\n\nOverall, the central theorems are plausible and likely correct. The issues are fixable. This deserves a serious referee, not a desk rejection. I would accept it and ask for the 5.4 fix and a discussion of the [11] dependence.","headline":"New explicit families show essential tori can be undetected by SL2(F)-character varieties over every algebraically closed field, with characteristic 2 playing a special role; the paper is solid but has an internal inconsistency in Section 5.4 and leans on a self-cited preprint for the foundational theory.","tokens_in":21571,"tokens_out":3619,"would_cite":true,"duration_ms":31787,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57M05","57K31","57K35","20C99"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper constructs infinitely many closed 3-manifolds whose unique essential surface is a torus, and proves this torus is not detected by ideal points of the variety of SL2(F)-characters for any algebraically closed field F.","keywords":["3-manifold","character variety","detected surface","essential torus","positive characteristic","graph manifold","Seifert fibered space","Culler-Shalen theory"],"falsifier":"Take a manifold from the Theorem 1 family, for example $\\Phi=\\begin{pmatrix}1+6k&k\\\\6&1\\end{pmatrix}$ with one fixed integer $k$, and compute the full variety of $\\mathrm{SL}_2(\\mathbb{F})$-characters over $\\mathbb{F}_2$ and over $\\mathbb{C}$; if any ideal point exists whose trace valuations on both components of $N_\\Phi\\setminus T$ are nonnegative, that ideal point would detect the torus $T$ and falsify the theorem. A more direct check is to test the companion preprint's three key statements; failure of Corollary 7, Lemma 8, or Proposition 10 for some algebraically closed field of positive characteristic would also invalidate the non-detection claim.","tokens_in":20521,"feed_emoji":"🍩","tokens_out":14989,"duration_ms":121984,"temperature":0.7,"pith_summary":"The paper asks whether every essential surface in a compact 3-manifold can be detected by an ideal point of the $\\mathrm{SL}_2(\\mathbb{F})$-character variety for some algebraically closed field $\\mathbb{F}$, and answers no in a strong way. It constructs an infinite family of closed graph manifolds $N_\\Phi$ by gluing the twisted interval bundle over the Klein bottle to the complement of the right-handed trefoil along matrices $\\Phi\\in\\mathrm{SL}_2(\\mathbb{Z})$. In one subfamily, each $N_\\Phi$ contains, up to isotopy, exactly one connected essential surface, a torus, and that torus is not detected by $X(N_\\Phi,\\mathbb{F})$ for any algebraically closed field $\\mathbb{F}$. Other subfamilies show the phenomenon is sensitive to characteristic: the same torus can be detected exactly when $\\mathbb{F}$ has characteristic 2, while another family has a genus-two surface detected in every characteristic alongside a torus detected in none. These examples show that the ideal-point detection method, extended to every characteristic, still misses essential surfaces and cannot be repaired by changing the field.","feed_headline":"Every algebraically closed field misses this essential torus","feed_subtitle":"Infinite families of graph manifolds hide a torus from ideal points of every SL(2,F)-character variety.","key_machinery":"The carrying object is the variety of $\\mathrm{SL}_2(\\mathbb{F})$-characters $X(M,\\mathbb{F})$ of a 3-manifold group over an algebraically closed field $\\mathbb{F}$, together with its ideal points. The carrying mechanism is the ideal-point detection machine (Culler--Shalen theory): an ideal point $\\xi$ of a curve $C$ defines a discrete valuation $v_\\xi$ on trace functions, produces a Bass--Serre tree $T_\\xi$ on which $\\pi_1(M)$ acts, and any essential surface obtained from the corresponding splitting is said to be detected. For the two building blocks the character varieties are explicit small curves, $X_{\\mathrm{irr}}(K\\tilde{\\times}I,\\mathbb{F})=\\{(0,t,0)\\mid t\\in\\mathbb{F}\\}$ and $X_{\\mathrm{irr}}(M,\\mathbb{F})=\\{(s,s,1)\\mid s\\in\\mathbb{F}\\}$, and the gluing relations add the fourteen trace coordinates $(t_a,t_b,t_g,t_h,t_{ab},\\ldots)$ of the generated group. The decisive technical work is computing, for each family of gluing matrices $\\Phi$, which curves appear and whether $v_\\xi(I_\\gamma)<0$ for elements $\\gamma$ in the complementary pieces of a candidate surface; this determines exactly which of the surfaces (S1)--(S5) is detected by which ideal point.","core_discovery":"The central claim is Theorem 1: there are infinitely many closed graph manifolds $N_\\Phi$ each containing, up to isotopy, exactly one connected essential surface; that surface is a torus; and it is not detected by the variety of $\\mathrm{SL}_2(\\mathbb{F})$-characters for any algebraically closed field $\\mathbb{F}$. The proof classifies all connected essential surfaces in the glued manifolds, reducing them to the splitting torus $T$, the two genus-two surfaces $S_2$ and $S_3$, and the vertical tori $S_4(q,r)$ and $S_5$. It then computes the character variety of $N_\\Phi$ from those of the two building blocks, whose irreducible and reducible curves are explicit: for the Klein bottle piece $X_{\\mathrm{irr}}(K\\tilde{\\times}I,\\mathbb{F})=\\{(0,t,0)\\mid t\\in\\mathbb{F}\\}$, and for the trefoil piece $X_{\\mathrm{irr}}(M,\\mathbb{F})=\\{(s,s,1)\\mid s\\in\\mathbb{F}\\}$. Detection and non-detection are decided by trace-function valuations at ideal points: a pole of a peripheral trace function forces a boundary slope, while constant traces rule a surface out. The non-detection theorem follows because, for the chosen families of gluing matrices, every curve in $X(N_\\Phi,\\mathbb{F})$ has the form listed in cases (C1)--(C3), and none produces the valuations needed to detect $T$; in the other families, exactly the characteristic-2 curves do.","pith_inferences":["The construction suggests a 'characteristic selector' phenomenon: choosing the gluing matrix $\\Phi$ appears to determine both which essential surfaces occur and in which characteristic they become detectable; this may generalize to other gluings of Seifert pieces with two boundary slopes.","The 2-dimensional component $C_\\Phi$ and its subcurves $C(u,v)$ indicate a systematic relation between trace valuations on $S_4(q,r)$ and the parameters $(u,v)$; if Conjecture 18 holds, the detection pattern for all vertical tori in the $S^2(2,2,2,3)$ family would be completely understood.","Because the Klein bottle group admits no faithful representation into $\\mathrm{SL}_2(K)$ over any field $K$, the examples suggest that failure of faithful linearity of the fundamental group may be the underlying reason some tori are invisible; the paper does not itself advance this explanation.","The authors' remark lists three infinite families of coprime pairs $(q,r)$ for which Conjecture 18 can be proved; checking further pairs by the same trace-valuation computation is a concrete way to test whether all $S_4(q,r)$ are detected."],"forward_implications":["For the undetected family of Theorem 1, $\\mathrm{SL}_2(\\mathbb{F})$-character detection fails in every characteristic, so the question of detecting all essential surfaces by character varieties has a negative answer without any characteristic caveat.","Theorems 2, 3, and 5 provide manifolds whose torus is detected over $\\mathbb{F}$ if and only if $\\operatorname{char}(\\mathbb{F})=2$, so positive characteristic can create detected surfaces that the complex character variety does not see.","Theorem 4 shows that in one graph manifold a non-separating genus-two surface is detected in every characteristic, because it is Poincaré dual to an epimorphism to $\\mathbb{Z}$, while the companion torus is detected in none.","Theorem 6 gives Seifert fibered spaces with infinitely many pairwise non-isotopic essential tori, all detected in every characteristic; vertical tori in these manifolds are numerous enough that the character variety detects the whole infinite family."],"supporting_citations":[{"why":"Introduces ideal points of character varieties and the construction of essential surfaces from splittings of the fundamental group; this is the detection method the paper extends to all characteristics.","marker":"[6]"},{"why":"The companion preprint that develops the variety of SL(2,F)-characters and the ideal-point machinery over arbitrary algebraically closed fields; its Corollary 7, Lemma 8, and Proposition 10 are used throughout Section 5.","marker":"[11]"},{"why":"Standard source for character varieties, ideal points, Bass-Serre trees, and the proof that an ideal point yields a splitting of the fundamental group.","marker":"[22]"},{"why":"Supplies Proposition 13, the classification of representations of the Klein bottle group over arbitrary fields, which structures the representation calculations in Section 3.","marker":"[9]"},{"why":"Hatcher's notes give the vertical/horizontal classification of essential surfaces in Seifert fibred manifolds and the boundary-slope facts used for the trefoil complement and the Klein-bottle I-bundle.","marker":"[13]"},{"why":"Schultens' correspondence between vertical essential surfaces and simple closed curves in the base orbifold is used to enumerate the tori S4(q,r) and S5.","marker":"[20]"},{"why":"Provides Proposition 12, reformulating boundary slopes in terms of valuations of trace functions of peripheral elements; this is the criterion used to decide which surface an ideal point detects.","marker":"[5]"},{"why":"Underpins the construction of character varieties in nonzero characteristic, where the coordinate ring is finite over the trace ring.","marker":"[16]"},{"why":"Stallings' construction converts a splitting of the fundamental group into an essential surface, completing the passage from ideal point to detected surface.","marker":"[23]"}],"fun_headline_variants":["Essential tori that dodge every SL2(F) character variety","No field reveals this embedded torus","Tori hidden from every algebraically closed field","Invisible to all character varieties: new essential tori","Essential tori that no ideal point can detect"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the extension of ideal-point detection theory to arbitrary algebraically closed fields, developed in the companion preprint and quoted as Corollary 7, Lemma 8, and Proposition 10, is fully correct; if trace equality does not characterize closure-equivalence, or if the commutator trace criterion or the ideal-point splitting theorem fails in positive characteristic, the conclusion that the exhibited tori are undetected in every characteristic no longer follows.","fun_headline_variants_meta":{"raw":{"variants":["Essential tori that dodge every SL2(F) character variety","No field reveals this embedded torus","Tori hidden from every algebraically closed field","Invisible to all character varieties: new essential tori","Essential tori that no ideal point can detect"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000408,"raw_usage":{"total_tokens":2102,"prompt_tokens":911,"completion_tokens":1191,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":527,"completion_tokens_details":{"reasoning_tokens":1126}},"tokens_in":527,"tokens_out":1191,"duration_ms":8850,"temperature":1.0,"reasoning_tokens":1126,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:02:35.667723+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take a manifold from the Theorem 1 family, for example $\\Phi=\\begin{pmatrix}1+6k&k\\\\6&1\\end{pmatrix}$ with one fixed integer $k$, and compute the full variety of $\\mathrm{SL}_2(\\mathbb{F})$-characters over $\\mathbb{F}_2$ and over $\\mathbb{C}$; if any ideal point exists whose trace valuations on both components of $N_\\Phi\\setminus T$ are nonnegative, that ideal point would detect the torus $T$ and falsify the theorem. A more direct check is to test the companion preprint's three key statements; failure of Corollary 7, Lemma 8, or Proposition 10 for some algebraically closed field of positive characteristic would also invalidate the non-detection claim.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces ideal points of character varieties and the construction of essential surfaces from splittings of the fundamental group; this is the detection method the paper extends to all characteristics."},{"cited_title":"Garden and Stephan Tillmann","cited_arxiv_id":null,"evidence_quote":"The companion preprint that develops the variety of SL(2,F)-characters and the ideal-point machinery over arbitrary algebraically closed fields; its Corollary 7, Lemma 8, and Proposition 10 are used throughout Section 5."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Standard source for character varieties, ideal points, Bass-Serre trees, and the proof that an ideal point yields a splitting of the fundamental group."},{"cited_title":"Linear representations of 3–manifold groups over rings","cited_arxiv_id":null,"evidence_quote":"Supplies Proposition 13, the classification of representations of the Klein bottle group over arbitrary fields, which structures the representation calculations in Section 3."},{"cited_title":"Notes on basic 3–manifold topology","cited_arxiv_id":null,"evidence_quote":"Hatcher's notes give the vertical/horizontal classification of essential surfaces in Seifert fibred manifolds and the boundary-slope facts used for the trefoil complement and the Klein-bottle I-bundle."},{"cited_title":"Kakimizu complexes of Seifert fibered spaces","cited_arxiv_id":null,"evidence_quote":"Schultens' correspondence between vertical essential surfaces and simple closed curves in the base orbifold is used to enumerate the tori S4(q,r) and S5."},{"cited_title":"Gordon, John Luecke, and Peter B","cited_arxiv_id":null,"evidence_quote":"Provides Proposition 12, reformulating boundary slopes in terms of valuations of trace functions of peripheral elements; this is the criterion used to decide which surface an ideal point detects."},{"cited_title":"Reductive subgroups of reductive groups in nonzero characteristic","cited_arxiv_id":null,"evidence_quote":"Underpins the construction of character varieties in nonzero characteristic, where the coordinate ring is finite over the trace ring."},{"cited_title":"Stallings","cited_arxiv_id":null,"evidence_quote":"Stallings' construction converts a splitting of the fundamental group into an essential surface, completing the passage from ideal point to detected surface."}],"review_version":1}