{"id":"08938936-ad0c-40d9-899b-a584da0e2b86","arxiv_id":"2608.12484","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The quiver mutation class of a malleable divide determines the integral monodromy module, and hence the topological type, of an irreducible plane curve singularity.","lead":"This paper proves that the quiver mutation class of a malleable divide from a real Morsification uniquely determines the integral monodromy module of a plane curve singularity. It establishes the algebraic-to-topological direction of the Fomin-Pylyavskyy-Shustin-Thurston conjecture for irreducible malleable cases, using new representation-theoretic invariants from Ginzburg algebras and Euler pairings.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bridge from malleable divides to plabic fences (Lemma 6.4) is asserted rather than proved; if the double-to-single plabic fence reduction fails, the Main Theorem is not established.","rationale":"The reader identifies Lemma 4.14 as the weakest assumption, and that is indeed a delicate fixed-point lemma inside the proof of Theorem 4.1. On inspection, Lemma 4.14 appears plausible and its order-by-order induction is a standard argument for pro-unipotent groups under a torus action; the vanishing of H^1(G_m,M) for rational representations is correct, though the proof would benefit from explicitly using H^2(G_m,M)=0 to justify the nonabelian induction. I therefore do not regard Lemma 4.14 as the most likely point of failure. The more exposed step is Lemma 6.4, which is the only place where malleable divides are brought into the plabic-fence framework. The proof of that lemma is a short sketch relying on moves from [16], without a complete reduction algorithm or a proof that the stated restrictions never obstruct the conversion. Since the central theorem covers all malleable divides, and since Theorem 5.4 and the cokernel invariant are only available for ordinary plabic fences, an unproved failure of Lemma 6.4 would leave the main result unsubstantiated. A computational enumeration of small double plabic fences is a concrete way to test the conversion claim. If the enumeration finds a counterexample, the Main Theorem must be restricted; if it succeeds, the remaining gap is a matter of supplying a rigorous reference or proof for the reduction. The reader's verdict of ACCEPT is therefore reasonable but slightly premature; the paper should either prove Lemma 6.4 fully or explicitly state it as an imported theorem with a precise location in [16]. I would accept the article after that condition is met.","tokens_in":51262,"tokens_out":44813,"duration_ms":421761,"concrete_test":"Enumerate all double plabic fences with at most four horizontal lines and at most six vertical edges that are not already ordinary plabic fences, including at least one with an interior black-on-top vertical edge. For each, attempt the greedy reduction: take the leftmost misoriented edge, slide it to a level boundary using only allowed sliding moves, reflect it, and slide back; if stuck, try all square-move sequences of bounded length. Check that the resulting graph is an ordinary plabic fence, that its quiver is mutation-equivalent to the original quiver, and that the associated links are smoothly isotopic (for example by verifying that the positive braid word is preserved up to braid relations). If any instance cannot be reduced, Lemma 6.4 fails; if all reduce, the bridge step is supported.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"Lemma 6.4 is the unique bridge connecting the geometric hypothesis (a malleable divide D) to the representation-theoretic machinery, which is developed only for ordinary plabic fences as in Definition 4.4. Its proof asserts that any double plabic fence arising from a scannable divide can be converted to an ordinary plabic fence by sliding, reflection, and square moves from [16], with the quiver changing only by mutation and the link type preserved. However, no algorithm, termination argument, or precise citation of a reduction theorem is supplied. The restrictions are real: reflection moves are allowed only for the leftmost or rightmost edge of a given level, and sliding moves require a vertically opposite neighbor. It is not obvious that an arbitrary interior black-on-top/white-on-bottom vertical edge can be slid to a boundary, reflected, and slid back, nor that any sequence of square moves used in the process preserves the smooth link type of the associated divide. If some double plabic fence cannot be reduced, or if reduction changes the link, then the Main Theorem does not follow for all malleable divides even if Theorems 4.1 and 5.4 are correct. The paper flags no limitation here, but the proof is a sketch rather than a demonstration.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper develops a new invariant of quiver mutation classes and applies it to plane curve singularities. For a graded quiver with potential, the author defines a graded Euler matrix E(t) in the simple basis of the derived category of a bigraded Ginzburg algebra, proves that its cokernel is invariant under graded quiver mutation, and shows that for plabic fences the relevant arrow-grading is unique up to gauge. Theorem 5.4 identifies this cokernel with the torsion part of the Alexander module of the link associated to the plabic fence. The final section bridges malleable divides to plabic fences via Lemma 6.4, yielding the Main Theorem: the quiver mutation class of a malleable divide determines the integral monodromy module of the singularity, and in particular the complex topological type in the irreducible case. This establishes the algebraic-to-topological direction of the Fomin–Pylyavskyy–Shustin–Thurston conjecture for malleable irreducible divides.","tokens_in":51493,"tokens_out":6369,"duration_ms":63240,"significance":"If the results are correct, this is a substantial advance: it provides the first proof of one direction of the Main Conjecture in a broad class of singularities, introduces a genuinely new representation-theoretic invariant of quiver mutation classes, and gives an explicit formula for the Alexander polynomial from any quiver in the mutation class. The paper is carefully written, with many worked examples, a detailed table for Milnor number at most 16, and a counterexample to a related conjecture. The central invariant is not fitted to the desired output: the equality between the cokernel of the Euler form and the Alexander module is a theorem proven through Seifert matrices, and the uniqueness of the dilation for plabic fences is a nontrivial result. The framework of dilation classes and bigraded Ginzburg algebras is likely to be useful beyond this specific application.","major_comments":[{"comment":"The reduction from a double plabic fence to an ordinary plabic fence is the unique bridge from the geometric hypothesis of a malleable divide to the plabic-fence machinery of Sections 4 and 5, but the proof is a sketch. It asserts that sliding moves, reflection moves, and square moves can transform any double plabic fence into one whose vertical edges all have white on top and black on bottom, without giving an algorithm, a termination argument, or a precise citation of a theorem that guarantees such a reduction. The restrictions are real: reflection moves are only allowed for the leftmost or rightmost edge of a level, and sliding moves require a vertically opposite neighbor. It is not demonstrated that an arbitrary interior vertical edge can be moved to a boundary position, reflected, and moved back, nor that the intermediate square moves preserve the smooth link type of the associated divide. Since Lemma 6.4 is load-bearing for the Main Theorem, this needs to be either proved in detail or replaced by a precise cited reduction theorem.","section":"Section 6.2, Lemma 6.4"},{"comment":"The uniqueness of the dilation for plabic fences, Theorem 4.1, depends on the fixed-point lemma for Gm-actions on unitriangular torsors. The proof of Lemma 4.14 is only an order-by-order sketch: it asserts that lifts modulo m^{N+1} exist because the torsor is compatible and that H^1(Gm,M)=0 provides fixed lifts, but the compatibility hypothesis is not verified in the text for the specific torsor T appearing in Proposition 4.13. A failure of this lemma would make the cokernel invariant ill-defined on the mutation class and would invalidate Corollary 4.2. The authors should expand this induction, spell out the compatibility condition, and explain why the torsor T satisfies it.","section":"Sections 4.3–4.4, Lemma 4.14 and Proposition 4.13"}],"minor_comments":[{"comment":"The phrase 'the wordend in Figure 10 (left)' appears to contain a typo; it should likely read 'the word end' or 'the end word', and the diagram would benefit from a clearer indication of which edge is the boundary edge.","section":"Figure 10 caption"},{"comment":"The proof of Lemma 4.15 would be easier to follow if the two local cycles in Figure 7 were described in terms of the face ordering from (4.2), so that the claim 'each cycle contains exactly one downward arrow' is visibly tied to the definition of d_G.","section":"Section 4.5, Lemma 4.15"},{"comment":"The table is useful, but for reproducibility it would be good to indicate which Alexander polynomial was computed by each of the methods (a), (b), (c) listed just before the table, or at least to add a remark that the entries were checked by two independent methods.","section":"Section 7.4, Table 4"},{"comment":"The two Euler matrices are asserted to have the same determinant, and this is correct, but the reader would benefit from an explicit statement that they are related by the congruence of Proposition 3.12, since determinant equality alone does not exhibit the cokernel isomorphism.","section":"Section 8.2, equations (8.6) and (8.11)"}],"recommendation":"major_revision","confidential_remarks":"I recommend major revision rather than rejection. The central strategy is coherent and the main invariant is well motivated, but the proof as written depends on the unproved reduction in Lemma 6.4, which is essential for the passage from divides to plabic fences. If the authors can supply a complete proof of that reduction or state and prove a precise theorem from their prior work that guarantees it, and if they expand the proof of Lemma 4.14, the paper would be suitable for publication. I do not see evidence of circularity or parameter fitting; the concern is purely about the completeness of a load-bearing proof."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things you should know. The main result is as significant as claimed: the quiver mutation class of a malleable divide determines the integral monodromy module, proving one direction of the FPST conjecture in the irreducible malleable case. The machinery—dilation classes, the Adams-graded Ginzburg algebra, the cokernel of the equivariant Euler pairing—is genuinely new and carefully developed. But the bridge from malleable divides to plabic fences, Lemma 6.4, is a sketch. It asserts that any double plabic fence from a scannable divide can be converted to an ordinary plabic fence by sliding, reflection, and square moves, with the quiver changing only by mutation and the link preserved. No algorithm, termination argument, or precise citation is given. This is not a nitpick: everything in Sections 2–5 applies to plabic fences, and without Lemma 6.4 the Main Theorem simply does not reach all malleable divides.\n\nWhat the paper does well: the invariant is defined from the quiver and a canonical dilation, with no parameters fitted to the Alexander polynomial; the equality coker E(t) ≅ Alexander module is proved, not assumed. Section 5's Seifert matrix argument and Theorem 5.4 make the topology explicit. Section 7's table and the counterexample to [28, Conjecture 6.17] are useful. The writing is dense but honest, and the author flags dependencies on prior work. The reader's weaker assumption, Lemma 4.14, is internal and looks reasonable; the order-by-order fixed-point argument with H^1(G_m, -)=0 is standard.\n\nThe stress-test note landed on the paper: Lemma 6.4's proof is a paragraph of assertions. It may well be true—the author is one of the authors of [16] and the moves are from there—but the present manuscript does not demonstrate the reduction. The main theorem is conditional on it. That is the biggest issue; I do not see other load-bearing flaws.\n\nWho this is for: specialists in plane curve singularities, cluster algebras, and quiver representations. They will get a lot from Sections 2–5 and the examples, and the obstruction to the full conjecture is now localized. It deserves a serious referee, but the referee should demand a complete proof of Lemma 6.4 or a precise reference to a reduction theorem. I would send it to peer review, expecting heavy revision on that lemma. If the reduction holds, this is an important paper.","headline":"The result is significant and the machinery is real, but the bridge from malleable divides to plabic fences (Lemma 6.4) is asserted, not proved; the paper needs revision before acceptance.","tokens_in":52022,"tokens_out":3714,"would_cite":true,"duration_ms":33483,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H20","14B05","16G20","16E45","13F60","32S55","57K10"],"pacs":[],"model":"deepseek-v4-flash","headline":"The quiver mutation class of a malleable divide determines the integral monodromy module of the associated plane curve singularity.","keywords":["plane curve singularities","quiver mutation","integral monodromy module","Alexander module","plabic fences","Ginzburg dg algebra","dilation classes","Morsifications"],"falsifier":"Find two plabic fences $G$ and $G'$ such that $[Q(G)] = [Q(G')]$ but the torsion Alexander modules of $L(G)$ and $L(G')$ are not isomorphic. By Corollary 4.2 and Theorem 5.4 that would force the cokernel of the graded Euler matrix to be ill-defined on the mutation class, directly contradicting the main invariant chain.","tokens_in":51046,"feed_emoji":"🪢","tokens_out":8087,"duration_ms":63861,"temperature":0.7,"pith_summary":"This paper proves that the quiver mutation class of a malleable real Morsification of an isolated plane curve singularity determines the integral monodromy module of that singularity, and hence its complex topological type when the singularity is irreducible. That is the algebraic-to-topological half of the Main Conjecture of [28], established here in the malleable irreducible case. The proof works by attaching to each quiver a differential bigraded Ginzburg algebra and extracting from it a Laurent-polynomial matrix whose cokernel is invariant under quiver mutation and is isomorphic to the torsion Alexander module of the associated link. A sympathetic reader should care because it turns a purely combinatorial object, the mutation class, into a complete topological invariant for a large class of singularities.","feed_headline":"Quiver mutations determine plane-curve monodromy","feed_subtitle":"The mutation class of a malleable divide uniquely recovers the Alexander module and the topological type.","key_machinery":"The engine is the differential bigraded Ginzburg algebra $\\Gamma(Q,W,d)$ of a graded quiver with potential: a 3-Calabi-Yau dg algebra whose generators carry an internal Adams grading compatible with a degree-one potential. Its derived category of continuous finite-dimensional dg modules carries an equivariant Euler pairing; in the basis of vertex simples this pairing is the matrix $E(t)$ above, which depends only on the arrow grading. The paper proves that this matrix's cokernel is invariant under the three dilation equivalences and under graded QP mutation, then proves that plabic fences have a unique dilation class and a unique non-degenerate potential, so the cokernel is a well-defined invariant of the quiver mutation class. Finally, a Seifert-matrix comparison identifies that cokernel with the torsion Alexander module of the link associated to the fence, bridging algebra and knot topology.","core_discovery":"The central claim is stated as the Main Theorem: if $(C,0)$ is an isolated plane curve singularity, $D$ is a divide coming from a real Morsification, $Q(D)$ its quiver, and $D$ is malleable, then the quiver mutation class $[Q(D)]$ uniquely determines the $\\mathbb{Z}[t,t^{-1}]$-module $H_1(M_f;\\mathbb{Z})$ with the algebraic monodromy action, the integral monodromy module. Since the Alexander polynomial of an irreducible plane curve singularity determines its topological type, it follows that $[Q(D)]$ determines the complex topological type of an irreducible singularity with malleable divide. The paper obtains the invariant as the cokernel of the graded Euler matrix $E(t)$ with entries $(1-t)\\delta_{ij} - \\sum_{a:j\\to i} t^{d(a)} + \\sum_{a:i\\to j} t^{1-d(a)}$, where $d$ is the unique dilation grading of the associated plabic fence. A key intermediate theorem states that this cokernel is isomorphic to the torsion part of the Alexander module of the smooth link of the plabic fence, for any plabic fence, algebraic or not.","pith_inferences":["Because the key isomorphism holds for every plabic fence, the invariant $\\operatorname{coker} E(t)$ is a knot-theoretic invariant of all positive-braid links from plabic fences, not only algebraic links; this suggests the machinery could detect finer link invariants in non-algebraic settings.","The author notes that no combinatorial proof of the invariance of equation (1.1) is known; a purely graph-theoretic proof would likely expose how far the invariant reaches beyond the Ginzburg-algebra construction.","The counterexample to [28, Conjecture 6.17] shows that quiver-mutation equivalence is strictly coarser than move-and-switch equivalence for plabic graphs; the new monodromy invariant therefore measures something that survives mutation but not the stricter plabic-graph moves.","One could test the sharpness of the main theorem by searching for reducible singularities whose integral monodromy modules coincide but whose topological types differ; the theorem says mutation classes cannot see the difference, so any such pair would delimit the invariant's resolving power."],"forward_implications":["For any irreducible plane curve singularity admitting a malleable divide, two real Morsifications with mutation-equivalent quivers must have the same complex topological type.","The Alexander polynomial of the singularity can be read off from any quiver in the mutation class as the determinant of the matrix in equation (1.1), once the canonical dilation grading is applied.","All 74 isolated plane curve singularities with Milnor number at most 16 are distinguished by the invariants coming from the mutation class, so the implication holds in that range.","If all algebraic divides are malleable, as conjectured in [28], the Main Theorem upgrades to a complete algebraic-to-topological classification of irreducible plane curve singularities by quiver mutation classes."],"supporting_citations":[{"why":"Supplies the Main Conjecture, the definition of divides and their quivers, and the divide-to-plabic-fence moves used in Lemma 6.4.","marker":"[28]"},{"why":"Founds quivers with potential, their mutation, the splitting theorem and non-degeneracy, on which Sections 2 and 3 build.","marker":"[21]"},{"why":"Provides the Ginzburg dg algebra, the derived equivalences from QP mutation, and the computation of Ext between simples used in Proposition 3.12 and the Euler form.","marker":"[44]"},{"why":"Constructs the non-degenerate QP associated to a plabic fence and proves the properties used in Lemma 4.7 and Theorem 4.1.","marker":"[15]"},{"why":"Supplies graded QP mutation and the graded splitting theorem used in Theorem 2.16 and Definition 2.13.","marker":"[5]"},{"why":"Gives the Seifert-matrix presentation of the Alexander module used in Proposition 5.2 and Theorem 5.4.","marker":"[50]"},{"why":"Provides the Milnor fibration, monodromy, and the nonsplit property of algebraic links used to specialize Theorem 5.4 to singularities.","marker":"[52]"},{"why":"Supplies the classification of singularities with Milnor number at most 16 used in the table of Section 7.4.","marker":"[6]"}],"fun_headline_variants":["Quiver mutation class fixes integral monodromy","Mutation class recovers Alexander module torsion","Malleable irreducible: mutation class fixes topological type","Quiver mutation class: complete monodromy invariant"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the technical fixed-point lemma (Lemma 4.14) asserting that every nonempty family of unitriangular coordinate changes that is stable under the scaling action has a scaling-invariant member; the proof needs this to straighten potentials by a graded equivalence, and without it the unique dilation class of a plabic fence and the invariant $\\operatorname{coker} E(t)$ would collapse.","fun_headline_variants_meta":{"raw":{"variants":["Quiver mutation class fixes integral monodromy","Mutation class recovers Alexander module torsion","Malleable irreducible: mutation class fixes topological type","Quiver mutation class: complete monodromy invariant"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000722,"raw_usage":{"total_tokens":3240,"prompt_tokens":950,"completion_tokens":2290,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":566,"completion_tokens_details":{"reasoning_tokens":2229}},"tokens_in":566,"tokens_out":2290,"duration_ms":17565,"temperature":1.0,"reasoning_tokens":2229,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T00:07:25.717451+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find two plabic fences $G$ and $G'$ such that $[Q(G)] = [Q(G')]$ but the torsion Alexander modules of $L(G)$ and $L(G')$ are not isomorphic. By Corollary 4.2 and Theorem 5.4 that would force the cokernel of the graded Euler matrix to be ill-defined on the mutation class, directly contradicting the main invariant chain.","supporting_citations":[{"cited_title":"Fomin, P","cited_arxiv_id":null,"evidence_quote":"Supplies the Main Conjecture, the definition of divides and their quivers, and the divide-to-plabic-fence moves used in Lemma 6.4."},{"cited_title":"Derksen, J","cited_arxiv_id":null,"evidence_quote":"Founds quivers with potential, their mutation, the splitting theorem and non-degeneracy, on which Sections 2 and 3 build."},{"cited_title":"Keller and D","cited_arxiv_id":null,"evidence_quote":"Provides the Ginzburg dg algebra, the derived equivalences from QP mutation, and the computation of Ext between simples used in Proposition 3.12 and the Euler form."},{"cited_title":"Casals and H","cited_arxiv_id":null,"evidence_quote":"Constructs the non-degenerate QP associated to a plabic fence and proves the properties used in Lemma 4.7 and Theorem 4.1."},{"cited_title":"Amiot and S","cited_arxiv_id":null,"evidence_quote":"Supplies graded QP mutation and the graded splitting theorem used in Theorem 2.16 and Definition 2.13."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Seifert-matrix presentation of the Alexander module used in Proposition 5.2 and Theorem 5.4."},{"cited_title":"Milnor.Singular Points of Complex Hypersurfaces, volume 61 ofAnnals of Mathematics Studies","cited_arxiv_id":null,"evidence_quote":"Provides the Milnor fibration, monodromy, and the nonsplit property of algebraic links used to specialize Theorem 5.4 to singularities."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the classification of singularities with Milnor number at most 16 used in the table of Section 7.4."}],"review_version":1}