{"id":"a4424a20-7a10-4993-9b7b-1d82d07b48a0","arxiv_id":"2502.10366","paper_version":1,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Two 2-braid groups over circumference-one graphs are quasi-isometric exactly when their quasi-minimal representatives are isometric, and the comparison is algorithmic.","lead":"This paper classifies, up to large-scale equivalence, the 2-braid groups of graphs made of a tree with small loops attached, and gives an algorithm to tell when two such groups are equivalent. It also shows that the same machinery classifies 4-braid groups over trees.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Concern: Corollary 6.6 rests on Lemma 4.10's Helly-type intersection claim for lifts of maximal product subcomplexes; a hidden counterexample would collapse the flag-complex/developability step and the completeness direction. Direct check on the tripod example should settle it.","rationale":"The reader's weakest_assumption identifies exactly the same load-bearing point, and I agree with that assessment. The entire classification is engineered so that the intersection complex is a complete invariant, and the completeness proof depends on the Helly-type intersection control in Lemmas 4.9 and 4.10. If that control failed, I(UP2(Γ)) would not be a flag complex, the quotient would not be developable, and the reconstruction of Γ_min from an isomorphism of intersection complexes would break. I found no actual counterexample for bunches of grapes: the normal, finite-tree structure of the stem makes the colinearity assertion plausible, and the tripod case is consistent with Lemma 4.9 because the three pairwise intersections are disjoint corners inside any fixed lift. The main reason this is still a live concern rather than a settled objection is that the paper's proof of Lemma 4.9 is concise and the lift version relies on two structural facts—convexity of the lifts and Helly in CAT(0) cube complexes—without spelling out the local-to-global argument in detail. A second point worth checking while doing this is the reduction at the start of Section 6.2, where the authors assume two grapes at each non-leaf vertex and justify it by Theorem 5.5; verifying that adding a grape at an internal bivalent vertex does not change the isomorphism type of I(UP2(−)) and that the quasi-minimal representative behaves as claimed would remove residual doubt about that reduction. These are finite, local checks, and they do not change the reader's verdict unless they fail.","tokens_in":65907,"tokens_out":20516,"duration_ms":217828,"concrete_test":"Run the check on the minimal normal tripod Γ: stem S3 with one 3-cycle at each leaf, and one at the center if needed to stay in Grapelarge_min. The three twigs correspond to maximal product subcomplexes M1, M2, M3 with pairwise intersections Γ_i ∘× Γ_j for distinct leaves i,j. Enumerate all lifts of M1, M2, M3 in the universal cover UP2(Γ) and test whether any triple of pairwise-intersecting lifts exists. Lemma 4.10 predicts none: the three pairwise corner flats in any lift of M1 are disjoint, so Helly would force an empty triple intersection. Equivalently, verify directly that the 1-skeleton of I(UP2(Γ)) is flag by checking that every triangle of pairwise-adjacent vertices has a label given by Lemma 4.9.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing point is the pair Lemmas 4.9/4.10. Lemma 4.9 controls intersections of maximal product subcomplexes in the base UP2(Γ): a family has nonempty intersection exactly when the corresponding twigs are colinear. Lemma 4.10 then uses CAT(0) Helly to lift this to the universal cover, where pairwise-intersecting lifts are forced to have a common standard product subcomplex and hence the twigs are colinear. This is what makes I(UP2(Γ)) a flag complex (Theorem 4.13), what makes RI(UP2(Γ)) developable, and what allows the iterative glueing in Theorem 5.5 and the reconstruction of Γ_min in Theorem 6.5. If three non-colinear twigs could have pairwise-intersecting lifts without a common point, I(UP2(Γ)) would contain an empty triangle and Theorem 4.13 would fail; the quotient RI would not be developable, and the completeness direction (c)⇒(b) of Corollary 6.6 would collapse. The proof of Lemma 4.9 is one short paragraph, and Lemma 4.10 is justified in two sentences after invoking Helly; the convexity/standardness of the relevant lifts is not fully detailed. This is internally identified as a class-specific point: Remark 2.28 and Example 2.30 show that the analogous property fails for general weakly special square complexes. I do not see a concrete false step for bunches of grapes, but this is the assumption on which the central claim depends.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the quasi-isometry classification of 2-braid groups over 'bunches of grapes' (graphs of topological circumference at most one). The main result (Corollary 6.6, Theorems 6.5 and 6.7) is an algorithmic classification: for two bunches of grapes Γ, Λ, the groups B2(Γ) and B2(Λ) are quasi-isometric if and only if the quasi-minimal representatives Γ_min and Λ_min are isometric, if and only if the intersection complexes I(UP2(Γ)) and I(UP2(Λ)) are isomorphic. The proof combines the quasi-isometry invariant introduced by the second author in [Oh22], new graph operations (pruning/smoothing twigs, picking over-grown grapes, pruning over-grown substems) that preserve the quasi-isometry type, and a completeness theorem showing that the intersection complex is a complete invariant on quasi-minimal representatives. Applications include new infinite families of graph 2-braid groups that are / are not quasi-isometric to RAAGs, and an algorithm deciding quasi-isometry of 4-braid groups over trees.","tokens_in":66201,"tokens_out":23458,"duration_ms":210914,"significance":"If correct, this is a significant step in the quasi-isometric classification of graph braid groups. The completeness of the intersection complex invariant for a natural class of 2-dimensional special groups is new and nontrivial, and the algorithmic nature of the classification is a strength. The paper is carefully structured, with explicit algorithms and numerous examples. The main theorems are supported by detailed arguments; the most delicate steps, namely the Helly-type colinearity Lemmas 4.9–4.10 and the iterative gluing in Theorems 5.5 and 5.15, are sound as written, though some auxiliary lemmas (e.g., Lemma 4.18) are stated without proof. The application to 4-braid groups over trees is elegant and gives a new quasi-isometry classification for that class.","major_comments":[],"minor_comments":[{"comment":"The sentence 'we assume that Γ = (T,ℓ) and Γ' = (T',ℓ') are in Grapelarge_min but both have two grapes at each vertex which is not a leaf of the stems' is confusing and appears to contradict the definition of Grapelarge_min, where val_T(v) ≥ 2 forces ℓ(v) = 1. Please rephrase the normalization being made.","section":"Section 6.2, first paragraph"},{"comment":"The statement 'Since N_{n+1}(v)\\N_n(v) is countably infinite' is inaccurate: the sphere of radius n+1 in the locally finite complex I(UP2(Γ)) is finite. The induction still works with finite spheres; please correct the sentence.","section":"Proof of Theorem 5.5"},{"comment":"The invocation of the Helly property requires that the p-lifts M(t_i) be convex in UP2(Γ); this follows because they are images of local isometries from products of trees, but the justification should be stated explicitly.","section":"Lemma 4.10"},{"comment":"This lemma is load-bearing for the proof of Theorem 6.5 (via Lemma 6.12 and Lemma 6.15) but is stated without proof; please provide a proof or a detailed argument that canonical order is preserved by semi-isomorphisms.","section":"Lemma 4.18"},{"comment":"Lemmas 5.2 and 5.7 and the algorithms in Appendix A would benefit from explicit termination and correctness arguments; currently they are stated as obvious.","section":"Section 5.1 and Appendix A"},{"comment":"The notation 'T_min ← T_min \\ T_{v,3}\\···\\T_{v,m}' is ambiguous; use a union symbol or set-builder notation to indicate removal of the union of the specified stems.","section":"Algorithm 3"}],"recommendation":"minor_revision","confidential_remarks":"The paper is a substantial contribution and the classification seems correct. The main risk is the complexity of the completeness proof; the authors should be encouraged to expand the proofs of the auxiliary lemmas (especially Lemma 4.18) in the final version. The self-citation to [Oh22] is appropriate, as the intersection complex framework is used as a black box; however, the novelty of the present paper lies in the operations and completeness, which are proved here."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"This paper delivers. It gives the first algorithmic quasi-isometry classification for graph 2-braid groups over bunches of grapes (circumference-one graphs), and the completeness direction—that the intersection complex is a complete invariant on quasi-minimal representatives—is the genuinely new structural result. The three operations (pruning/smoothing, picking over-grown grapes, pruning over-grown substems) are natural and the proof that they preserve quasi-isometry type is carefully built. I also like the applications: the enlargement of the RAAG/non-RAAG classes and the reduction of tree 4-braid group quasi-isometry to the same algorithm.\n\nI read the soft spot the stress-test flags. Lemma 4.10 is indeed load-bearing: it converts Helly for convex subcomplexes into colinearity of the corresponding twigs, and it is what makes the flag-complex/developability step work. On reading, I think it holds. The proof is short but sound: after Helly gives a nonempty intersection, projecting down, Lemma 4.9 identifies the base intersection as a standard product subcomplex, and then the unique lift containing the intersection must sit inside each of the pairwise-intersecting lifts, since lifts of the same base product are either disjoint or equal. I don't see a hidden counterexample. The paper itself is honest that this property fails for general weakly special square complexes (Remark 2.28, Example 2.30), which is a point in its favor.\n\nThe real soft spots are more mundane. Several auxiliary lemmas (5.2, 5.7, 4.18) are stated without proof; they're plausible and probably routine, but in a classification paper the reader shouldn't have to reconstruct them. The gluing arguments in Theorems 5.5 and 5.15 are intricate—lots of relative quasi-isometries being patched along intersections—and deserve a careful referee's time. There's also a mild over-reliance on the authors' earlier paper [Oh22] for the intersection complex formalism, but the classification and completeness are proved here, so I don't count that as circular.\n\nWho's this for? Anyone working on quasi-isometric rigidity of special groups, graph braid groups, or RAAGs. It's a significant step, and the algorithmic angle is a nice bonus. It deserves a serious referee—send it out, and ask the referee to focus on the gluing lemmas and the omitted proofs rather than the Helly step, which I think is fine.","headline":"A genuine advance in quasi-isometry classification of graph 2-braid groups; the invariant-completeness proof holds up, and the paper deserves serious refereeing.","tokens_in":66756,"tokens_out":3205,"would_cite":true,"duration_ms":30933,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20F65","20F36","20F67","57M60"],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper proves that 2-braid groups over bunches of grapes are quasi-isometric exactly when their quasi-minimal representatives are isometric, and that the comparison is algorithmic.","keywords":["graph braid groups","quasi-isometry classification","intersection complex","bunches of grapes","circumference one graphs","special cube complexes","right-angled Artin groups","discrete configuration spaces"],"falsifier":"Build an explicit bunch of grapes and examine the universal cover of $UP_2(\\Gamma)$: if one finds a finite family of maximal product subcomplexes that meet pairwise but whose total intersection is empty, or a reduced intersection complex whose universal cover is not simply connected and flag, then the colinearity lemmas are false and the completeness proof of the main rigidity theorem breaks. The paper predicts that no such configuration exists for bunches of grapes; locating one would settle the central claim negatively.","tokens_in":65669,"feed_emoji":"🍇","tokens_out":9286,"duration_ms":83159,"temperature":0.7,"pith_summary":"The paper aims to give a complete algorithmic classification of $2$-braid groups over \"bunches of grapes\" — graphs of circumference at most one, obtained from a tree by attaching $3$-cycles at vertices. Its central claim is that two such groups are quasi-isometric exactly when the quasi-minimal representatives of the defining graphs are isometric, and equivalently when the reduced intersection complexes of the universal covers of their configuration spaces are isomorphic. This matters because it converts a coarse geometric equivalence relation into a finite combinatorial decision problem, and because the same mechanism produces algorithms for comparing $2$-braid groups with right-angled Artin groups and for deciding quasi-isometry of $4$-braid groups over trees. The proof works by showing that the intersection complex is a complete invariant on quasi-minimal inputs, and that every bunch of grapes reduces to a unique quasi-minimal one by operations that preserve the quasi-isometry type of its $2$-braid group.","feed_headline":"2-braid groups over bunches of grapes classified up to quasi-isometry","feed_subtitle":"For 2-braid groups, quasi-isometry reduces to isometry of a finite canonical graph, and the check is algorithmic.","key_machinery":"The load-bearing object is the intersection complex, a labelled almost-simplicial complex whose vertices are maximal product subcomplexes of the universal cover (top-dimensional flats), whose simplices record which finite families of such subcomplexes intersect, and whose labels record the product domains of the intersections. For bunches of grapes the paper proves a Helly-type colinearity lemma: any finite family of pairwise-intersecting maximal product subcomplexes of $\\widetilde{UP}_2(\\Gamma)$ has a common intersection that is itself a standard product subcomplex, corresponding to a colinear set of twigs (path substems of the stem tree). That lemma makes $I(\\widetilde{UP}_2(\\Gamma))$ a connected, simply connected flag complex and makes the reduced complex $RI(UP_2(\\Gamma))$ developable as a complex of groups. The quasi-minimal representative is produced by four operations that induce isomorphisms of intersection complexes and therefore quasi-isometries of the corresponding braid groups.","core_discovery":"On the paper's own terms, the discovery is Corollary 6.6: for large bunches of grapes $\\Gamma,\\Gamma'$, the groups $B_2(\\Gamma)$ and $B_2(\\Gamma')$ are quasi-isometric if and only if the intersection complexes $I(\\widetilde{UP}_2(\\Gamma))$ and $I(\\widetilde{UP}_2(\\Gamma'))$ are isomorphic, if and only if the quasi-minimal representatives $\\Gamma_{\\min}$ and $\\Gamma'_{\\min}$ are isometric. Here $\\widetilde{UP}_2(\\Gamma)$ is the universal cover of the union of all maximal product subcomplexes of the unordered discrete $2$-configuration space of $\\Gamma$, and $\\Gamma_{\\min}$ is obtained by pruning empty twigs, smoothing twigs, picking over-grown grapes, and pruning over-grown substems. The paper further proves that $\\Gamma_{\\min}$ exists, is unique up to isometry, and is computable in finite time, yielding an algorithm that decides quasi-isometry between any two $2$-braid groups over bunches of grapes.","pith_inferences":["If the same Helly-type colinearity lemma holds for wider graph classes such as cacti, the identical scheme would likely give quasi-isometry algorithms there; a single counterexample would delimit the method sharply.","The paper's sufficient condition comparing $2$-braid groups to right-angled Artin groups invites the conjecture that, within this class, quasi-isometry to a RAAG is characterized by the shape of the quasi-minimal stem; its negative examples suggest the exact obstruction is a four-branched substem.","The construction of a quasi-isometry from an isomorphism between complexes of groups is a transferable technique: the paper's open Question 2 asks exactly how widely it holds among special square complexes satisfying the flat-intersection hypothesis.","The tree $4$-braid algorithm could plausibly be iterated, via the same edge-stabilization pattern, to compare $n$-braid groups over trees with $2$-braid groups over larger grape-like graphs; the paper stops at $n=4$."],"forward_implications":["Any two bunches of grapes can be fed to a finite algorithm that outputs whether their $2$-braid groups are quasi-isometric.","Within this class, $2$-braid groups have quasi-isometric rigidity: coarse equivalence is exactly isometry of the quasi-minimal representative.","The known families of circumference-one graphs whose $2$-braid groups are or are not quasi-isometric to right-angled Artin groups are both strictly enlarged.","Two $4$-braid groups over trees are decided by the same machinery, because each $B_4$ of a tree is isomorphic up to free factors to $B_2$ of a canonically grown bunch of grapes.","On quasi-minimal inputs the intersection complex is complete, not just invariant: isomorphic intersection complexes force isometric defining graphs."],"supporting_citations":[{"why":"Defines the intersection complex as a quasi-isometry invariant for weakly special square complexes and proves the morphism theorem used as the invariant side of the classification.","marker":"[Oh22]"},{"why":"Introduced the intersection complex for graph 2-braid groups as a quasi-isometry invariant, the starting point of the present classification.","marker":"[Fer12]"},{"why":"Supplies the free-product quasi-isometry theorem used to strip free factors and to glue relative quasi-isometries across the decomposition.","marker":"[PW02]"},{"why":"Provides the presentation and freeness criteria for graph 2-braid groups that identify the trivial small case and compute free ranks.","marker":"[KP12]"},{"why":"Establishes that unordered discrete configuration spaces of graphs are special cube complexes and that graph braid groups are their fundamental groups.","marker":"[Abr00]"},{"why":"Gives the rigidity of top-dimensional flats in CAT(0) cube complexes that lets quasi-isometries move maximal product subcomplexes to maximal product subcomplexes.","marker":"[Hua17b]"},{"why":"An earlier quasi-isometry classification for tree RAAGs provides the flat-intersection strategy adapted here.","marker":"[BN08]"}],"fun_headline_variants":["2-braid groups over bunches of grapes: quasi-isometry decidable","Quasi-isometry of 2-braid groups over bunches of grapes is decidable","2-braid groups: quasi-isometry reduces to isometry of a pruned graph","Bunch-of-grapes 2-braid groups: algorithmic quasi-isometry classification"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The classification rests on the claim that pairwise-intersecting maximal product regions in the universal cover always have a common intersection of the same standard product type; this Helly-type behavior fails for general weakly special square complexes, so if it ever failed for bunches of grapes the completeness proof would collapse.","fun_headline_variants_meta":{"raw":{"variants":["2-braid groups over bunches of grapes: quasi-isometry decidable","Quasi-isometry of 2-braid groups over bunches of grapes is decidable","2-braid groups: quasi-isometry reduces to isometry of a pruned graph","Bunch-of-grapes 2-braid groups: algorithmic quasi-isometry classification"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001045,"raw_usage":{"total_tokens":4371,"prompt_tokens":900,"completion_tokens":3471,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":516,"completion_tokens_details":{"reasoning_tokens":3381}},"tokens_in":516,"tokens_out":3471,"duration_ms":21517,"temperature":1.0,"reasoning_tokens":3381,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T18:18:37.111295+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Build an explicit bunch of grapes and examine the universal cover of $UP_2(\\Gamma)$: if one finds a finite family of maximal product subcomplexes that meet pairwise but whose total intersection is empty, or a reduced intersection complex whose universal cover is not simply connected and flag, then the colinearity lemmas are false and the completeness proof of the main rigidity theorem breaks. The paper predicts that no such configuration exists for bunches of grapes; locating one would settle the central claim negatively.","supporting_citations":[],"review_version":1}