{"id":"c89fab89-cdae-4bc0-adb8-caa624dc2624","arxiv_id":"2508.21264","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Graph Houghton groups form a genuinely new family of Houghton-type groups with finiteness type F_{r-1} but not FP_r, and with explicit presentations of their pure subgroups.","lead":"The authors introduce graph Houghton groups, infinite symmetry groups of locally finite graphs, and prove they have prescribed finiteness properties and are distinct from all known Houghton-type groups. They also give an explicit finite presentation of the pure subgroup, making the group accessible for further algebraic and geometric study.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem 1.1 rests on unproved Theorem 5.20: the injective tethered handle complex connectivity is only asserted by analogy with [ABKL23, Thm 5.7].","rationale":"The reader's weakest-assumption analysis identifies the same load-bearing concern: Theorem 5.20 is asserted without proof, and it is essential to condition (c) of Brown's criterion for both the doubled handlebody and graph Houghton groups. This concern is central rather than peripheral. The paper contains detailed proofs of the presentation theorem, the non-commensurability theorem, and the Brown-criterion setup, so credit is due. The Dehn function gap is explicitly acknowledged and does not affect Theorem 1.1; Theorems 8.4 and 8.5 are peripheral to the main finiteness claim. Therefore the final verdict should remain conditional, exactly as the reader concluded.","tokens_in":40350,"tokens_out":3610,"duration_ms":38485,"concrete_test":"Write out the full proof of Theorem 5.20. Concretely: define the 'bad simplices' in TH(Z,Q) following [ABKL23, Thm 5.7], and verify that for each bad d-simplex σ, its link is isomorphic to TH_1(Z',Q') with g(Z') ≥ 4(k−d−1)+4 and |Q'| ≥ (k−d−1)+2 whenever g(Z) ≥ 4k+4 and |Q| ≥ k+2. Also verify that the good subcomplex is exactly TH_1(Z,Q). If this verification fails because some bad-simplex link has too few boundary components or too low rank to satisfy the induction hypothesis, the proof of Theorem 5.20 collapses. An appendix or independent note supplying this induction would settle whether the concern lands.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 1.1 uses Brown's criterion on the Stein–Farley complex. The only deep connectivity input is Theorem 5.20, which asserts that the injective tethered handle complex TH_1(Z,Q) is k-connected under rank and boundary bounds. Its proof is not supplied: the paper says 'This follows as in the proof of [ABKL23, Theorem 5.7]' and 'we do not repeat it here.' The authors also explicitly skip the discussion of bad simplices in §5.1, saying they are only used in proving Theorem 5.20. Theorem 5.20 feeds directly into Theorem 5.15 and Corollary 5.21, producing weakly Cohen–Macaulay descending links (Corollary 5.22), which is condition (c) of Brown's criterion. For the graph case, Lemma 5.10 transfers this connectivity from doubled handlebodies to graphs. Thus, if the bad-simplex induction from [ABKL23] cannot be adapted to doubled handlebodies with boundary spheres—for example, if the link of a bad simplex is not again an injective tethered handle complex with the required rank and boundary bounds—the descending links may not be sufficiently connected, and Theorem 1.1 is not established. This is an internal gap in the central argument, not a disagreement with external consensus.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper introduces asymptotically rigid mapping class groups of locally finite infinite graphs and their associated doubled handlebodies, and studies the resulting 'graph Houghton groups' and 'doubled handlebody Houghton groups'. The main results are: Theorem 1.1, finiteness properties F_{r-1} but not FP_r for graphs/doubled handlebodies with r finitely many ends all accumulated by loops/genus, and F_∞ for a Cantor set of ends; Theorem 1.2, non-commensurability of graph and doubled handlebody Houghton groups with classical, braided, surface, and handlebody Houghton groups; and Theorem 1.3, an explicit finite presentation of the pure graph Houghton group P B_r for r ≥ 3. The paper also studies ends, the Tits alternative, BNSR-invariants, the word problem, and the Dehn function, and gives a stable homology result in the Cantor case.","tokens_in":40703,"tokens_out":5894,"duration_ms":60842,"significance":"If the results hold, this is a natural and valuable contribution: it extends the Houghton-group phenomenon to asymptotically rigid mapping class groups of graphs, gives a concrete new family of groups with explicit presentations, and proves they are not commensurable with previously known Houghton-type groups. The paper is notably transparent about its limitations: Theorem 5.20 is deferred to an analogy with [ABKL23], Theorems 8.4 and 8.5 are stated without proof, and Section 8.4 explicitly records a failed strategy for bounding the Dehn function. This honesty is a strength, but those deferrals are load-bearing for the main theorem and must be addressed.","major_comments":[{"comment":"The proof of Theorem 1.1 runs through Brown's criterion; condition (c) is supplied by Corollary 5.22, which relies on Theorem 5.15, which in turn relies on Theorem 5.20. The proof of Theorem 5.20 is not contained in the paper: it says 'This follows as in the proof of [ABKL23, Theorem 5.7]' and 'we do not repeat it here.' The earlier remark in §5.1 explicitly says that bad simplices are skipped because they are used only in proving Theorem 5.20. Since adapting from the surface setting to doubled handlebodies with boundary spheres is precisely where the rank and boundary-count bounds and the good-link identification must be checked, this is a load-bearing gap. Please provide the full induction or a detailed reduction, not just an analogy.","section":"§5.4, Theorem 5.20"},{"comment":"Theorems 8.4 and 8.5 are stated as results, but Section 8.3 contains no proof. The sentence 'with occasional modifications, the entire analysis of [TW24, Section 3 & 4] can be carried out unchanged' is insufficient to establish the BNSR-invariant statement in this new setting. Either provide the proof or explicitly present these as conditional claims/conjectures.","section":"§8.3, Theorems 8.4 and 8.5"},{"comment":"Theorem 7.4, the stable homology computation, is stated with a proof sketch that says 'after making the appropriate modifications, the proof of [DZ25, Theorem 6.4] can be directly adapted.' The adaptation is not described. Since this is a standalone theorem, the proof should be supplied or the statement should be downgraded to a conjecture.","section":"§7.2, Theorem 7.4"}],"minor_comments":[{"comment":"The phrase 'in each case, 2 ≤ i < j ≤ n' should use r instead of n; the symbol n is not defined in the statement.","section":"Theorem 1.3"},{"comment":"The same symbol B(g,h,r) is used for both the graph Houghton group and the doubled handlebody Houghton group. Although Remark 2.6 introduces calligraphic notation, the theorem statements do not consistently use it. Please add a clarifying sentence in Theorem 1.1 or in Notation 2.9.","section":"Notation 2.9 / Theorem 1.1"},{"comment":"Typo: 'Genovois' should be 'Genevois'; the Figure 10 caption uses 'Stein–Farely' instead of 'Stein–Farley'.","section":"§3.2 and Figure 10"},{"comment":"The section title promises a Dehn-function result, but the section ends with an open question and a statement that no bound is obtained. Consider retitling the section or making the open status explicit in the first sentence.","section":"§8.4"},{"comment":"The proof says the kernel contains all finite products of sphere twists and no infinite products, but the forward inclusion is only implicit. Please spell out why every element of the kernel of Ψ_B is a finite product of sphere twists.","section":"Lemma 2.7"}],"recommendation":"major_revision","confidential_remarks":"The paper is plausible and likely correct, but the deferred proof of Theorem 5.20 is central to Theorem 1.1. I would support acceptance after the authors supply that proof or a precise reduction, and either prove or relabel Theorems 8.4, 8.5, and 7.4. The paper's transparency about its gaps is commendable."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First thing you should know: this is a genuinely new family of groups, not just a relabeling of known examples. The graph Houghton groups B(g,h,r) and the doubled handlebody versions are new, and the non-commensurability theorem (Thm 1.2) plus the explicit presentation of P B_r (Thm 1.3) are the strongest content. The presentation work is substantial and the proof of Theorem 1.3 is detailed, using AFV generators and Lee's strategy in a way that looks correct. The Cantor-end case and stable homology result in Section 7 is also a solid extension of known techniques.\n\nBut the headline finiteness theorem (Thm 1.1) rests on Theorem 5.20, which is not proved in this paper. The authors say it follows as in the proof of [ABKL23, Theorem 5.7] and that they do not repeat it. This is not a peripheral detail: Theorem 5.20 feeds directly into Theorem 5.15, Corollary 5.21, and Corollary 5.22, which give condition (c) of Brown's criterion. The stress-test note is right to flag this. The paper even says it is skipping the discussion of bad simplices because they are only used in proving Theorem 5.20. That means the central argument has a load-bearing gap. I do not think the gap is necessarily fatal—the adaptation from surfaces to doubled handlebodies with boundary spheres is plausible and the authors clearly know the argument—but in a paper whose main theorem is a finiteness result, deferring the key connectivity input to an induction sketch is not enough.\n\nOther soft spots are more minor. Theorems 8.4 and 8.5 (CAT(0), BNSR invariants) are stated with a promise that TW24 transfers, but no proof or even a precise statement of the modifications. Section 8.4 honestly records the failure of the Dehn function strategy, which is fine as an open question, but it means the paper's claims about geometric properties should be read accordingly. The non-commensurability proof is dense and uses some external results (e.g., She25, DHK25) but seems to hold together.\n\nWho is this for? Geometric group theorists working on Houghton-type groups, big mapping class groups, and finiteness properties. They will get real value from the new construction, the presentation, and the commensurability dichotomy. The paper deserves a serious referee, but the referee should insist on a complete proof of Theorem 5.20 or an explicit reduction to ABKL23 that addresses the bad simplex issue in the doubled handlebody setting. I would not desk reject it; I would send it to review and ask for that gap to be closed.","headline":"New Houghton-type groups with a strong presentation and non-commensurability results, but the central finiteness theorem rests on an unproved connectivity result that needs to be supplied or made precise before the paper is fully convincing.","tokens_in":41139,"tokens_out":2563,"would_cite":true,"duration_ms":29224,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20F65","57M07","20F28"],"pacs":[],"model":"deepseek-v4-flash","headline":"A new family of graph Houghton groups has finiteness properties exactly controlled by the number of ends of the underlying graph.","keywords":["asymptotically rigid mapping class groups","graph Houghton groups","finiteness properties","type F_n / FP_n","Stein–Farley cube complexes","doubled handlebodies","Aut(F_n)","non-commensurability"],"falsifier":"Find a doubled handlebody Z with rk(π1(Z)) ≥ 4k+4 and at least k+2 boundary spheres whose injective tethered handle complex TH1(Z,Q) is not k-connected—for example, a non-null-homotopic k-sphere in the complex. Because Theorem 5.20 is the only unproved input in the Brown's criterion chain, such a complex would invalidate the proof of Theorem 1.1 (though not necessarily the statement).","tokens_in":40289,"feed_emoji":"♾️","tokens_out":9013,"duration_ms":84035,"temperature":0.7,"pith_summary":"This paper introduces countable subgroups of the mapping class group of a locally finite infinite graph, called graph Houghton groups, together with their doubled handlebody analogues. The central claim is that finite-ended examples with r ends accumulated by loops (or, for handlebodies, by genus) are of type F_{r-1} but not of type FP_r, and that the Cantor-ended examples are of type F∞. The finiteness bound is proved by a group action on a Stein–Farley cube complex and Brown's criterion, with descending links analyzed through piece complexes and tethered handle complexes. The paper also gives an explicit finite presentation for the pure graph Houghton group with at least three ends, shows this group contains Aut(F_n) for every n, and proves that graph and doubled handlebody Houghton groups are not commensurable with classical, braided, surface, or handlebody Houghton groups. A sympathetic reader would care because these are explicit, finitely presentable dense subgroups of 'big' mapping class groups whose finiteness properties can be read off the end space of the graph.","feed_headline":"Finiteness of graph Houghton groups depends on number of ends","feed_subtitle":"New graph Houghton groups match the classical F_{r-1}-not-FP_r pattern and get explicit presentations.","key_machinery":"The carrying mechanism is a Stein–Farley cube complex X (and its doubled handlebody analogue X), whose vertices are equivalence classes [Z,f] of suited subgraphs Z of the rigidified graph together with a group element f; edges and cubes record nesting of suited subgraphs. The group acts by left multiplication, the complexity function h([Z,f]) = rank(π1(Z)) is a discrete Morse function, and Brown's criterion turns high connectivity of descending links into finiteness properties. Descending links are analyzed through the piece complex P(Z,Q) of a compact doubled handlebody with boundary spheres and the injective tethered handle complex TH1(Z,Q); complete join maps (in the sense of Hatcher–Wahl","core_discovery":"On the paper's own terms, the discovery is a rigidity-to-finiteness dictionary for infinite graphs: once a rigid structure (a finite core plus tails that are copies of a fixed rank-h graph) is fixed, the asymptotically rigid mapping class group B(g,h,r) has the same finiteness type as the classical Houghton group H_r. For r < ∞ ends, all accumulated by loops, B(g,h,r) is of type F_{r-1} but not of type FP_r; when the end space is a Cantor set, the asymptotically rigid subgroup is of type F∞. The proof runs through a Stein–Farley cube complex on which the group acts; Brown's criterion reduces the task to showing descending links are (r-2)-connected, which is achieved by identifying the links","pith_inferences":["A natural test: use the explicit presentation of PBr as an algorithmic normal form for elements of Aut(F_n) and for end-periodic graph maps; success would give computational access to dynamics that the paper only raises as an open question.","The non-commensurability result suggests that the 'Houghton phenomenon'—finiteness type F_{r-1} but not FP_r—is not a commensurability invariant of the underlying space but a common feature of many rigid structures; quasi-isometry classification (Question 3.1) would be the sharper invariant to pursue.","One could try to close the Dehn function gap by tracking the quasi-isometry constants in the exponential upper bound for Aut(F_m), since the paper shows the missing ingredient is control of those constants for its particular generating set.","If Theorem 5.20's unproved induction collapses, a weakened theorem might still hold with modified bounds on rank or number of boundary components; checking the bad-simplex links in the doubled handlebody setting would separate the method's robustness from the present formulation."],"forward_implications":["For every r ≥ 2, B(g,h,r) and its doubled handlebody counterpart are finitely generated, and for r ≥ 3 finitely presented; no finite classifying space exists with finitely many r-cells, so the type exactly matches H_r.","When the end space is a Cantor set, the asymptotically rigid mapping class group is of type F∞, extending the finite-end result to a natural 'big' limit.","The pure graph Houghton group PBr, r ≥ 3, has an explicit finite presentation and contains a copy of Aut(F_n) for every n; this makes PBr one concrete finitely presented group housing all Aut(F_n).","The graph and doubled handlebody Houghton groups are not commensurable with the classical, braided, surface, or handlebody Houghton groups, so the graph construction genuinely enlarges the Houghton family.","PBr is one-ended for r ≥ 2, has solvable word problem, fails the Tits alternative, and its BNSR-invariants coincide with those of the ordinary and surface Houghton groups; the Dehn function remains an open upper-bound problem."],"supporting_citations":[{"why":"defines Map(Γ), the mapping class group of a locally finite infinite graph, the ambient group in which graph Houghton groups live.","marker":"[AB25]"},{"why":"supplies the surface Houghton group construction, the version of Brown's criterion used as Theorem 5.1, and the Theorem 5.7 whose proof Theorem 5.20 says it mirrors.","marker":"[ABKL23]"},{"why":"supplies the Cantor-manifold asymptotic mapping class group framework, complete join arguments, and the F∞ result for Cantor doubled handlebodies used in Section 7.","marker":"[ABF+24]"},{"why":"provides the braided Houghton group finiteness proof and Stein–Farley complex approach that the paper's cube complex construction mimics.","marker":"[GLU22]"},{"why":"defines handlebody Houghton groups and their F_{r-1}-not-FP_r finiteness, used as comparison and for the Cantor/stable-homology adaptation.","marker":"[DZ25]"},{"why":"gives the surjective homomorphism from doubled handlebody mapping class groups to graph mapping class groups with sphere-twist kernel, yielding the doubled handlebody Houghton group and the map Ψ_B.","marker":"[Uda24]"},{"why":"supplies complete join complexes and weakly Cohen–Macaulay complexes, the mechanism by which connectivity of descending links is transferred.","marker":"[HW10]"},{"why":"supplies the presentation of Aut(F_n) whose generators and relations are adapted into the explicit presentation of PBr in Theorem 1.3.","marker":"[AFV08]"},{"why":"provides the classical Houghton group finite presentation and geometry that the pure graph Houghton group presentation and Dehn function discussion extend.","marker":"[Lee12]"},{"why":"established that classical Houghton groups are of type F_{r-1} but not FP_r, the benchmark finiteness statement Theorem 1.1 matches.","marker":"[Bro87]"}],"fun_headline_variants":["Graph Houghton groups: finiteness follows end count","Ends decide finiteness of graph Houghton groups","New graph Houghton groups: finiteness type from ends","Graph Houghton groups: explicit presentations, distinct class"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The proof of the main finiteness theorem depends on Theorem 5.20, which asserts that injective tethered handle complexes of doubled handlebodies are highly connected under rank and boundary-sphere bounds; the paper states this follows as in the surface case but does not carry out the induction.","fun_headline_variants_meta":{"raw":{"variants":["Graph Houghton groups: finiteness follows end count","Ends decide finiteness of graph Houghton groups","New graph Houghton groups: finiteness type from ends","Graph Houghton groups: explicit presentations, distinct class"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000579,"raw_usage":{"total_tokens":2516,"prompt_tokens":646,"completion_tokens":1870,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":390,"completion_tokens_details":{"reasoning_tokens":1822}},"tokens_in":390,"tokens_out":1870,"duration_ms":14745,"temperature":1.0,"reasoning_tokens":1822,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T14:25:54.061425+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Find a doubled handlebody Z with rk(π1(Z)) ≥ 4k+4 and at least k+2 boundary spheres whose injective tethered handle complex TH1(Z,Q) is not k-connected—for example, a non-null-homotopic k-sphere in the complex. Because Theorem 5.20 is the only unproved input in the Brown's criterion chain, such a complex would invalidate the proof of Theorem 1.1 (though not necessarily the statement).","supporting_citations":[],"review_version":1}