{"id":"66b547ce-cb3b-4c1d-81ad-1ee101511767","arxiv_id":"1908.08171","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Tropical moduli spaces of stable curves are simply connected, Delta_3 is homotopy equivalent to S^5, and Delta_4 has 3-torsion in H_5 and 2-torsion in H_6 and H_7.","lead":"This paper proves that spaces of stable tropical curves are simply connected, that the genus-3 space is a 5-sphere up to homotopy, and that the genus-4 space has torsion in its homology. It also builds general tools for computing fundamental groups and integral homology of spaces made from symmetric quotients of simplices.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.3 hinges on three sphere-quotient homology computations that are asserted but not pinned or audited; an error in any of them would change the torsion claims.","rationale":"The reader's weakest assumption identifies exactly the right spot: the proof of Theorem 1.3 is a spectral-sequence argument whose input consists of machine-generated homology groups and a graph enumeration, none of which are auditable from the text alone. I agree that this is the most load-bearing assumption. The rest of the paper's machinery—Theorem 3.1, Theorem 4.2, Proposition 5.1, and Theorem 6.1—is independent of the disputed numerical inputs and appears sound, so I see no reason to reject the paper. However, because the headline torsion result depends on un-pinned computations, acceptance should be conditional on an independent verification of those concrete inputs. The 'source larger than target' step is robust once the E1 page is accepted, so the fix is not to redo the spectral sequence but to audit the homology computations and the enumeration. If independent recomputation confirms the listed groups and the three-cell enumeration, the proof of Theorem 1.3 goes through as written. This is a genuine but bounded caveat, not a demonstrated error, so the verdict should move from ACCEPT to CONDITIONAL rather than to REJECT or UNVERDICTED.","tokens_in":11351,"tokens_out":29629,"duration_ms":315268,"concrete_test":"Pin the linked GitHub repository to a specific commit, then independently recompute the reduced integral homology of S^6/Aut(G), S^7/Aut(G'), and S^7/Aut(K3,3) using a second implementation (for example Sage) on the simplicial chain complex of the barycentric subdivision of the boundary simplex, and compare the Smith normal forms with the values listed in Section 6. In parallel, re-enumerate the stable genus-4 graphs modulo isomorphism (for instance with nauty/geng or a brute-force degree-sequence search) and confirm that exactly the three listed cells lie outside Delta^bm_4. If the homology groups match and the enumeration agrees, Theorem 1.3 is supported; any discrepancy requires recomputing the E1 page and the resulting torsion conclusions.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim most exposed is Theorem 1.3. Its proof in Section 6 uses the relative spectral sequence of Theorem 4.3 with Z = Delta^bm_4. The E1 page is assembled from three asserted computations: exactly three genus-4 stable graphs (the square-pyramid graph G, the triangular-prism graph G', and K3,3) lie outside Z, and their sphere quotients have the listed reduced integral homology groups. The 3-torsion in H5 depends entirely on the entry E^8,-3_1 = Z/3 coming from \\tilde H4(S^7/Aut(K3,3)); the 2-torsion in H6 and H7 depends on E^8,-2_1 and E^8,-1_1. The text says these were computed with a Python/Magma script and gives a GitHub URL, but it provides no commit hash, no generated boundary matrices, and no output tables. The later 'each source is larger than its target' step only shows that the d1 kernels are nontrivial given those E1 terms; it cannot detect a wrong E1 entry. If, for instance, \\tilde H4(S^7/Aut(K3,3)) were zero or a different cyclic group, the claimed Z/3 in H5 would disappear or change, and similarly for the 2-torsion groups. The graph enumeration is also asserted without a reproducible list. Thus Theorem 1.3, the paper's headline torsion result, is only as reliable as these unaudited computations.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper develops tools for studying fundamental groups and integral singular homology of symmetric Delta-complexes (and the more general class of symmetric CW-complexes), then applies them to the links Delta_g and Delta_{g,n} of moduli spaces of stable tropical curves. The main results are Theorem 1.1 (Delta_g is simply connected for g >= 1), Theorem 1.4 (Delta_{g,n} is simply connected except for (0,4) and (0,5)), Theorem 1.2 (Delta_3 is homotopy equivalent to S^5), and Theorem 1.3 (the reduced integral homology of Delta_4 has nontrivial 3-torsion in H_5 and nontrivial 2-torsion in H_6 and H_7, and vanishes otherwise). The technical core consists of Theorem 3.1, proving that the fundamental group of a finite symmetric CW-complex is generated by loops in its 1-skeleton; Theorems 4.2 and 4.3, giving spectral sequences for homology from the skeletal filtration and a relative version; Proposition 5.1, showing that quotients of spheres by finite groups containing a reflection are contractible; and Theorem 6.1, establishing contractibility of the subcomplex Delta^{bm}_{g,n}. The paper also provides counterexamples showing that higher-dimensional cellular approximation fails for symmetric CW-complexes.","tokens_in":11664,"tokens_out":6769,"duration_ms":64089,"significance":"If the results are correct, they are significant. The paper settles simply-connectedness for these tropical moduli spaces, identifies Delta_3 as a sphere, and gives the first torsion in the integral homology of Delta_g in low degree, below the range where the rational homology is governed by known cohomological vanishing. The introduced framework of symmetric CW-complexes and the relative spectral sequence are natural and likely to be useful in future work on tropical and toroidal moduli spaces. The proofs of Theorems 3.1, 4.2, 6.1, and Proposition 5.1 are clean, mostly self-contained, and careful about the failure of naive cellular approximation in higher dimensions. The main caveat is that Theorem 1.3 depends on explicit computer-assisted computations that are not sufficiently pinned down in the manuscript; this is a reproducibility gap in a load-bearing part of the paper rather than a flaw in the geometric arguments.","major_comments":[{"comment":"The integral homology groups of the three sphere quotients S^6/Aut(G), S^7/Aut(G'), and S^7/Aut(K_{3,3}) are load-bearing for the headline torsion result. The 3-torsion in H_5(Delta_4; Z) depends entirely on the asserted entry E_1^{8,-3} = Z/3 coming from \\tilde H_4(S^7/Aut(K_{3,3})), and the 2-torsion in H_6 and H_7 depends on the entries E_1^{8,-2} and E_1^{8,-1}. The manuscript states that these were computed with a Python script and Magma and gives a GitHub URL, but it provides no commit hash, no generated boundary matrices, and no output tables. A wrong or mistranscribed entry would change the torsion claims. Please include a pinned version of the code, the computed reduced homology groups, and enough intermediate data (for instance the chain matrices for the simplicial chain complexes) to make these three computations auditable from the text.","section":"Section 6, Proof of Theorem 1.3"},{"comment":"The assertion that enumerating stable graphs of genus 4 yields exactly three cells outside Delta_4^{bm}, namely the square-pyramid graph G, the triangular-prism graph G', and K_{3,3}, is also not accompanied by a reproducible enumeration. Since the E_1 page of the relative spectral sequence is empty except for the contributions of these three cells and E_{0,0} = Z, an omission or misclassification would change the conclusion. Please provide the enumeration in a table or as script output, or otherwise document the classification of the genus-4 stable graphs outside Delta^{bm}.","section":"Section 6, Proof of Theorem 1.3"}],"minor_comments":[{"comment":"The sentence 'Each source is larger than its target' is telegraphic. A homomorphism from a finite abelian group to a strictly smaller group has a nontrivial kernel, but one must also note that no higher differential can hit or leave the surviving classes; expanding this one sentence would make the argument easier to check.","section":"Section 6, Proof of Theorem 1.3"},{"comment":"There is a typo: 'the barycenter of the opposite face t' should be 'the barycenter of the opposite face of t' or similar.","section":"Example 3.6"},{"comment":"Several references are cited by arXiv identifiers alone; adding journal or publication data where available would improve the reference list.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The computational reproducibility gap is the only substantive obstacle to acceptance. If the authors can supply pinned code and outputs, or otherwise document the three sphere-quotient homology computations and the graph enumeration, the paper should be acceptable. The paper fits the journal's scope well."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: genuinely new integral topology for tropical moduli spaces, and the proofs are worth taking seriously. The headline results hold up as far as I can see—Delta_g and Delta_{g,n} are simply connected except for the two known small cases, Delta_3 is homotopy equivalent to S^5, and the paper finds the first torsion in low-degree integral homology of these spaces: Z/3 in H_5(Delta_4), Z/2 in H_6 and H_7.\n\nWhat is new here is a mix of useful machinery and concrete computation: the skeleton spectral sequence for symmetric CW-complexes with integral coefficients (Theorem 4.2), its relative version (Theorem 4.3), and the van Kampen argument Theorem 3.1 showing pi_1 is generated by the 1-skeleton, with Examples 3.2 through 3.7 showing cellular approximation fails in all higher dimensions. Proposition 5.1 and Theorem 6.1 are short and clean: multiple edges give a reflection, so the sphere quotients are contractible, and the subcomplex Delta^bm is strictly larger than Delta^br yet still contractible. I checked the main arguments and found no gaps.\n\nThe exposed point is Theorem 1.3. The proof depends on an asserted enumeration of the three genus-4 graphs outside Delta^bm_4 and three asserted homology computations of S^6/Aut(G), S^7/Aut(G'), and S^7/Aut(K3,3), all done by script. The GitHub link is there, but no commit hash and no output tables. The differential argument only preserves nontriviality given those E1 entries—a wrong entry, say in H_4(S^7/Aut(K3,3)), would change the torsion claims. To be clear, these are finite, well-specified computations a referee can redo in an afternoon, so I treat this as a verification requirement, not a suspected error.\n\nThe citation pattern is healthy; reliance on [CGP18] and [CGP19] is legitimate use of prior theorems, not circularity. No parameter fitting, no hidden data.\n\nThis is for people working on tropical moduli, graph complexes, or the topology of moduli spaces. I would bring it to reading group and cite the spectral sequence idea. It deserves a serious referee; accept, but require the code to be pinned or the homology outputs included before publication.","headline":"Genuinely new integral topology for tropical moduli spaces with a useful spectral-sequence toolkit; only the unaudited computer computations behind Theorem 1.3 need referee verification.","tokens_in":12204,"tokens_out":5487,"would_cite":true,"duration_ms":47294,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["14H10","55N10","55P10"],"pacs":[],"model":"deepseek-v4-flash","headline":"Tropical moduli spaces are simply connected except for two small marked cases, and genus 4 has integral torsion.","keywords":["tropical moduli space","symmetric Delta-complex","symmetric CW-complex","fundamental group","integral homology","spectral sequence","moduli of curves","torsion"],"falsifier":"Independently recompute the reduced integral homology of $S^6/\\mathrm{Aut}(G)$, $S^7/\\mathrm{Aut}(G')$, and $S^7/\\mathrm{Aut}(K_{3,3})$ and re-enumerate the stable genus-$4$ graphs to check that exactly three cells of $\\Delta_4$ lie outside $\\Delta^{\\mathrm{bm}}_4$; any difference would overturn Theorem 1.3. A direct computation of $H_5(\\Delta_4;\\mathbb{Z})$ by another method would also settle the presence of $\\mathbb{Z}/3\\mathbb{Z}$.","tokens_in":11155,"feed_emoji":"🌴","tokens_out":9547,"duration_ms":80255,"temperature":0.7,"pith_summary":"This paper develops tools for computing the fundamental group and integral singular homology of symmetric $\\Delta$-complexes, spaces built by gluing quotients of simplices by finite groups. It applies them to $\\Delta_g$ and $\\Delta_{g,n}$, the links of the moduli spaces of stable tropical curves of unit volume. The main results are that $\\Delta_g$ is simply connected for $g \\geq 1$, and that $\\Delta_{g,n}$ is simply connected whenever $(g,n)$ is not one of the two excluded pairs $(0,4)$ or $(0,5)$. The paper also proves that $\\Delta_3$ is homotopy equivalent to the $5$-sphere and that $\\Delta_4$ has nontrivial $3$-torsion in $H_5$ and nontrivial $2$-torsion in $H_6$ and $H_7$. These claims matter because the spaces are the links of tropical moduli spaces whose rational homology is tied to graph homology and to the top-weight cohomology of the classical moduli space of curves.","feed_headline":"Tropical moduli spaces are simply connected; genus 4 has torsion","feed_subtitle":"A new spectral sequence computes integral homology, showing Delta_3 is a 5-sphere and Delta_4 carries 2- and 3-torsion.","key_machinery":"The load-bearing object is the skeleton filtration spectral sequence (Theorems 4.2 and 4.3): for a symmetric CW-complex $X$ whose $p$-cells are quotients $(B^p)^\\circ/G_i$ by finite groups $G_i \\subset O(p)$, the $\\mathrm{E}^{1}$ page is $E^1_{p,q} = \\bigoplus_i \\widetilde{H}_{p+q-1}(S^{p-1}/G_i; A)$, converging to filtration quotients of $H_{p+q}(|X|; A)$. The companion mechanism is the contractible subcomplex $\\Delta^{\\mathrm{bm}}_{g,n}$ of tropical curves with bridges, cut vertices, loops, repeated markings, positive-weight vertices, or multiple edges; Proposition 5.1 shows it is contractible because any quotient of a sphere by a finite group containing a reflection is contractible. Together these reduce the homology of $\\Delta_g$ and $\\Delta_{g,n}$ to finitely many sphere quotients attached to graphs with none of those features.","core_discovery":"On the paper's own terms, the central discovery is that the skeleton filtration of a symmetric CW-complex gives a usable spectral sequence for integral homology, and that for tropical moduli spaces the $\\mathrm{E}^{1}$ page can be computed explicitly by discarding a large contractible subcomplex. The theorems establish that $\\Delta_g$ is simply connected for every $g \\geq 1$; that $\\Delta_{g,n}$ is simply connected for all $(g,n)$ except $(0,4)$ and $(0,5)$, where $\\Delta_{0,4}$ is disconnected and $\\Delta_{0,5}$ is connected but not simply connected; that $\\Delta_3$ is homotopy equivalent to $S^5$; and that the reduced integral homology of $\\Delta_4$ vanishes outside degrees $5,6,7$, with a $\\mathbb{Z}/3\\mathbb{Z}$ class in $H_5$ and $\\mathbb{Z}/2\\mathbb{Z}$ classes in $H_6$ and $H_7$.","pith_inferences":["The same computational pipeline—enumerate graphs without bridges, cut vertices, loops, multiple edges, or positive weights, then compute sphere-quotient homology—could be run for $\\Delta_5$ or for marked spaces $\\Delta_{g,n}$ with small $g$ and $n$ to look for further torsion; the paper stops at genus $4$.","Because $\\Delta_g$ is rationally identified with top-weight cohomology of $\\mathcal{M}_g$, the $3$-torsion in $H_5(\\Delta_4)$ suggests—but the paper does not establish—that the integral top-weight cohomology of $\\mathcal{M}_4$ may carry a matching $3$-torsion class.","The reflection argument behind the contractible subcomplex suggests that toroidal compactifications whose boundary strata have reflection stabilizers may admit similarly large contractible subcomplexes, which would make their low-degree integral homology computable by the same spectral sequence; this application is left open."],"forward_implications":["$\\Delta_g$ is simply connected for every $g \\geq 1$, so its fundamental group and first homology vanish.","$\\Delta_{g,n}$ is simply connected for every $(g,n)$ except $(0,4)$ and $(0,5)$; the exceptional cases are known explicitly.","$\\Delta_3$ is homotopy equivalent to $S^5$, so it has the integral homology of a $5$-sphere.","$\\Delta_4$ is not contractible; its reduced integral homology is concentrated in degrees $5,6,7$, containing $3$-torsion and $2$-torsion.","The spectral sequence and contractible-subcomplex techniques work for symmetric CW-complexes in general, not only symmetric $\\Delta$-complexes, which covers dual complexes from non-simplicial toroidal compactifications."],"supporting_citations":[{"why":"Provides the definition of symmetric $\\Delta$-complexes, the rational cellular chain complex, and the contractibility of the bridge locus $\\Delta^{\\mathrm{br}}_{g,n}$ used as a base.","marker":"[CGP18]"},{"why":"Proves contractibility of $\\Delta^{\\mathrm{br}}_{g,n}$ and the low-degree vanishing bounds that the new torsion results sit below.","marker":"[CGP19]"},{"why":"Classifies quotients of spheres by rotation groups; used to identify the quotient $S^4/S_4$ with $S^4$ in the proof for $\\Delta_3$.","marker":"[Lan16]"},{"why":"Computes $\\Delta_{0,n}$ as a wedge of $(n-2)!$ spheres of dimension $n-4$, covering the marked exceptional cases in simple connectivity.","marker":"[Vog90]"},{"why":"Supplies the Magma system used to compute the integral homology of the sphere quotients appearing in Theorem 1.3.","marker":"[BCP97]"},{"why":"Provides the chamber structure of reflection groups used in Proposition 5.1 to prove contractibility of quotients of spheres by groups containing a reflection.","marker":"[Bou68]"}],"fun_headline_variants":["Tropical moduli spaces are simply connected; genus 4 carries torsion","Simply connected tropical moduli, but Delta_4 has 3-torsion","Delta_3 is a 5-sphere, Delta_4 has 3-torsion: new spectral sequence","Tropical moduli: simply connected, Delta_3 is S^5, Delta_4 has 3-torsion"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the computer-assisted enumeration of stable genus-$4$ graphs and the computed integral homology of the three sphere quotients $S^6/\\mathrm{Aut}(G)$, $S^7/\\mathrm{Aut}(G')$, and $S^7/\\mathrm{Aut}(K_{3,3})$ are correct; the paper states the outputs but does not pin the script to a version or print its full output tables.","fun_headline_variants_meta":{"raw":{"variants":["Tropical moduli spaces are simply connected; genus 4 carries torsion","Simply connected tropical moduli, but Delta_4 has 3-torsion","Delta_3 is a 5-sphere, Delta_4 has 3-torsion: new spectral sequence","Tropical moduli: simply connected, Delta_3 is S^5, Delta_4 has 3-torsion"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002114,"raw_usage":{"total_tokens":8153,"prompt_tokens":833,"completion_tokens":7320,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":449,"completion_tokens_details":{"reasoning_tokens":7219}},"tokens_in":449,"tokens_out":7320,"duration_ms":46601,"temperature":1.0,"reasoning_tokens":7219,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:47:44.937093+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Independently recompute the reduced integral homology of $S^6/\\mathrm{Aut}(G)$, $S^7/\\mathrm{Aut}(G')$, and $S^7/\\mathrm{Aut}(K_{3,3})$ and re-enumerate the stable genus-$4$ graphs to check that exactly three cells of $\\Delta_4$ lie outside $\\Delta^{\\mathrm{bm}}_4$; any difference would overturn Theorem 1.3. A direct computation of $H_5(\\Delta_4;\\mathbb{Z})$ by another method would also settle the presence of $\\mathbb{Z}/3\\mathbb{Z}$.","supporting_citations":[],"review_version":1}