Irreducible fast sets of bump homeomorphisms generate copies of Thompson's groups $F_n$

Gili Golan

Irreducible fast sets of bump homeomorphisms generate copies of Thompson's groups $F_n$

T0 review · 0 major / 4 minor · reviewed 2026-07-14 · grok-4.5

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read Every irreducible fast set of n positive bumps generates a group isomorphic to Thompson's group F_n.

desk verdict Uniform proof that every irreducible fast n-bump group is F_n, settling the strong Oberwolfach question that specialists expected to fail for n>3. read the letter →

arxiv 2607.10961 v1 pith:WISO2OZV submitted 2026-07-12 math.GR

Gili Golan This is my paper · ORCID

classification math.GR MSC 20F6520F3857M07

keywords Thompsongroupsbumphomeomorphismsgeometricallyfastsetscrossinggraphdiagramdirected2-complexespeelablediagramsGuba-Sapirmoves

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper settles a question about groups of interval homeomorphisms built from simple rightward 'bumps.' When a finite collection of such bumps can be marked so that their feet stay pairwise disjoint (a geometrically fast set) and their crossing graph is connected (irreducible), the group they generate is always the same: the n-ary Thompson group F_n. Earlier work had confirmed the pattern only for small n and left open the possibility of exotic isomorphism types for larger n. By introducing a local swap move that rearranges feet while preserving both fastness and the generated group, the author reduces every irreducible diagram to a peelable one, then realises the group as a diagram group over an explicit directed 2-complex that can be transformed into a dilated core of F_n. The result completely determines the class C_n and shows that the suspected counter-examples do not exist.

What carries the argument

The swap move (Gap Swap Lemma): conjugating the owners of one block of a closed cut of a bump's gap word by that bump yields a new fast realisation of the rearranged dynamical diagram that generates exactly the same group; repeated swaps produce a peelable diagram on which an inductive reduction of the associated diagram complex to a dilated core of F_n can be carried out.

What would settle it

Exhibit a single irreducible fast set of n ≥ 5 positive bumps whose dynamical diagram cannot be transformed by any sequence of closed-cut swaps into a peelable diagram, or whose diagram group over the associated complex is not isomorphic to F_n.

Watch

Extended reading notes

Core claim

For every n ≥ 2, every group generated by an irreducible geometrically fast set of n positive bumps is isomorphic to Thompson's group F_n. Equivalently, the class C_n consists of a single isomorphism class represented by F_n.

Load-bearing premise

The swap conjugation always produces a new set of pairwise-disjoint feet that realises the rearranged diagram without changing the generated group; if that geometric preservation failed for some configurations, the peelability reduction would not go through.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

This paper settles the strong form of the Brin–Zaremsky question from the 2018 Oberwolfach report: every group generated by an irreducible geometrically fast set of n positive bumps is isomorphic to F_n. Specialists had expected a negative answer for n>3; the result was previously known only for n=2,3,4. That is the headline.

What is new is the uniform argument. The dynamical half introduces swap moves at closed cuts of a gap word. The Gap Swap Lemma shows that conjugating the owners of one block by the bump produces another fast realization of the swapped diagram that generates the same group. Using those moves, every irreducible diagram is swap-equivalent to a peelable one (remains irreducible after successive deletion of the rightmost bump). The algebraic half realizes the Belk–Stott presentation as a directed 2-complex, then reduces it by Guba–Sapir moves, via a lifting lemma and truncation, to a dilated core of F_n whose diagram group is F_n. Both halves are written out with base cases, recognition lemmas, and explicit marker transport.

The math looks solid. The Gap Swap argument (the only candidate soft spot) is careful about closedness, relative order, and pairwise disjointness of feet; the induction on components of the deleted crossing graph closes cleanly. Dilated cores reduce to the standard Stallings core of F_n without redefining anything. Citations to Bleak–Brin–Kassabov–Moore–Zaremsky, Belk–Stott, and Guba–Sapir sit at the right points; the author’s earlier core work is used as background, not as a circular definition. No free parameters, no data selection.

Soft spots are minor: the inductive bookkeeping is dense, and a reader who has never seen diagram groups will need the preliminaries. That is ordinary for this literature, not a flaw in the argument.

This is for people who work on Thompson groups, diagram groups, or subgroups of Homeo+(I). It deserves a serious referee and should be accepted after ordinary polishing. I would cite it and bring it to reading group.

Referee Report

0 major / 4 minor

Summary. The paper proves that for every n ≥ 2, every group generated by an irreducible geometrically fast set of n positive bumps is isomorphic to the n-ary Thompson group F_n. Equivalently, the class C_n consists of a single isomorphism class. The argument has two halves: first, every irreducible dynamical diagram is swap-equivalent (via closed-cut conjugations that preserve fastness and the generated group) to a peelable diagram; second, the dynamical directed 2-complex of a peelable set is reduced by Guba–Sapir moves to a dilated core of F_n, whose diagram group is shown to be F_n by reduction to the standard Stallings core. This settles the strong form of a question of Brin–Zaremsky from the 2018 Oberwolfach report, previously known only for n ≤ 4.

Significance. The result completely determines the isomorphism types of groups generated by irreducible fast sets of n positive bumps, answering a problem that was open for n > 4 and for which a negative answer had been suspected. The techniques combine interval dynamics (swap moves, peelability) with the theory of diagram groups over directed 2-complexes in a clean inductive way; the Gap Swap Lemma, Lifting Lemma, and recognition of dilated cores are reusable tools. The identification of diagram groups of dilated cores with F_n is parameter-free and rests on the classical presentation of F_n, not a redefinition. The paper therefore both closes a concrete open question and supplies a flexible method for recognizing Thompson groups among closed subgroups of Homeo+(I).

minor comments (4)

In Definition 2.5 the span of a bump is written (L_i, R_i); later (e.g., Lemma 3.6) the same notation is used for open position intervals. A brief clarifying sentence would prevent any momentary confusion with the geometric support.
Figure 3 caption refers to “red” feet; if the journal prints in monochrome the colour reference should be replaced by a pattern or label.
The arXiv identifiers of the author’s related preprints [9] and [10] appear with future dates (2026); once those papers are posted the citations should be updated for permanence.
In the base-case proof of Theorem 6.4 the zipper edge s is introduced without an explicit figure reference; a short pointer to the right panel of Figure 6 would help the reader.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: pure combinatorial/algebraic isomorphism proof reducing fast-set groups to F_n via explicit GS moves and dilated cores.

full rationale

The Main Theorem is established by two independent inductive halves that are fully written out. First, every irreducible dynamical diagram is swap-equivalent to a peelable one (Theorem 3.13) by the Gap Swap Lemma 3.4, which explicitly constructs a conjugate fast realization that preserves both the generated group and pairwise disjoint feet under a closed cut. Second, the dynamical 2-complex of a peelable set is transformed by a finite sequence of Guba–Sapir moves into an admissible dilated core (Theorem 6.4); the diagram group of any dilated core is then identified with F_n by reduction to the standard core K_n of the classical presentation (Corollary 4.10, using the known isomorphism F_n ≅ D(⟨x|x=x^n⟩,x) of Guba–Sapir). Self-citations to the author’s earlier work on cores and generation problems supply independent lemmas (e.g., the Stallings-core construction) that do not encode the target isomorphism for general n; they are used only as tools. No step equates a claimed prediction with a fitted input, redefines F_n in terms of the fast-set generators, or imports a uniqueness theorem that forces the result by construction. The argument is therefore self-contained against the classical definition of F_n.

Assumptions & free parameters 0 free parameters · 4 assumptions · 2 invented entities

Pure mathematical theorem. No free parameters or fitted constants appear. Background results used as black boxes are standard theorems of diagram-group theory and the representation theorem of Belk–Stott; the new combinatorial notions (swap, peelable diagram, dilated core) are definitions introduced for the proof rather than postulated physical entities.

assumptions (4)

standard math Guba–Sapir moves (M1),(M1)^{-1},(M2) preserve diagram groups of directed 2-complexes (Theorem 2.23, citing [15]).
Used throughout Section 6 to transform the dynamical complex into a dilated core without changing the diagram group.
domain assumption Belk–Stott representation: a geometrically fast set generates a diagram group over the semigroup presentation read from its dynamical diagram (Theorem 5.3 / 5.9, citing [4]).
Converts the geometric generators into a diagram group so that the subsequent GS reduction applies.
domain assumption The Stallings core of F_n (as a subgroup of itself) is the complex K_n^+ whose diagram group is F_n (Theorem 4.2, citing [9]).
Supplies the target model complex to which dilated cores are reduced.
standard math Two fast sets with identical dynamical diagrams generate isomorphic groups via the natural bijection of generators (Theorem 2.6, citing [1]).
Allows the combinatorial swap equivalence to be transferred to group isomorphism.

invented entities (2)

swap move / closed cut of a gap word
purpose: Local rearrangement of a dynamical diagram implemented by conjugation that preserves fastness and the generated group, enabling reduction to peelable diagrams.
Definition 3.3 and Lemma 3.4; purely combinatorial device internal to the proof, no external existence claim.
dilated core of F_n
purpose: Family of tree-like directed 2-complexes (active cycle plus dummy edges and sliding attachments) whose diagram groups are all isomorphic to F_n; the inductive reduction lands on an admissible dilated core.
Definition 4.4 and Corollary 4.10; again a definitional tool, not an independent object of nature.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Irreducible fast sets of bump homeomorphisms generate copies of Thompson's groups $F_n$." pith.science (2026). https://pith.science/paper/WISO2OZV

@misc{pith2026260710961,
  author       = {Pith},
  title        = {Pith review of: Irreducible fast sets of bump homeomorphisms generate copies of Thompson's groups $F_n$},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WISO2OZV}},
  note         = {Machine review of arXiv:2607.10961}
}

abstract

A homeomorphism of an interval is a positive bump if its support is a single open interval on which it moves every point to the right. Choosing a fundamental domain $[m,b(m))$ for the action of a positive bump $b$ on its support splits the remainder of the support into two intervals, called the feet of $b$. A finite set of positive bumps is geometrically fast if fundamental domains can be chosen so that all the resulting feet are pairwise disjoint. The crossing graph of such a set has the bumps as its vertices, two bumps being adjacent whenever their supports overlap but are not nested, and the set is irreducible if its crossing graph is connected. We prove that for every $n\geq 2$, every group generated by an irreducible geometrically fast set of $n$ positive bumps is isomorphic to the $n$-ary Thompson group $F_n$. This answers the strong version of a problem posed by Brin and Zaremsky (Oberwolfach Rep. 15 (2018)).

Figures

Figures reproduced from arXiv: 2607.10961 by the authors.

**Figure 1.** The three possible dynamical diagrams of two fast bumps, up to reflection: disjoint, nested and crossing, together with the groups their fast realizations generate [1]. As noted in [5], a fast set of positive bumps that is not irreducible generates a group which decomposes non-trivially as a direct product or as a permutational wreath product of groups generated by smaller fast sets. For n ≥ 2, let Cn denote the cla… view at source ↗

**Figure 2.** Feet and bump arcs. Each bump and its two feet share a color, and each bump is drawn as an arc joining the left endpoint of its left foot to the right endpoint of its right foot, so that the arc spans the support of the bump. Left: a fast set of two crossing bumps, Lx < Ly < Rx < Ry; the group they generate is Thompson’s group F [1]. Right: a dynamical diagram of three bumps in which bumps 1 and 2 cross, bump 3 is n… view at source ↗

**Figure 3.** The swap XY → Y X at the bump b. The bump b itself, and hence its support (a, c) (spanned by the arc, which joins the left endpoint of Lb to the right endpoint of Rb), is unchanged. The owners of the feet of Y are conjugated by b −1 ; their feet inside the gap of b (red) are dragged into the old left-foot zone of b, while their feet beyond supt(b) do not move and appear identically before and after. The right foot o… view at source ↗

Figures from the paper (5 more)

**Figure 4.** Figure 4: The complex K+ 6 (left) and a dilated core of F6 (right). Active edges carry the relation “once around the cycle”; dummy edges carry no relation. Lemma 4.7 (Path equivalence in a dilated core). Let K be a dilated core and let p1, p2 be 1-paths with the same initial an…

**Figure 5.** Figure 5: The base line graph of the canonical two-bump configuration generating F, the identifications forced by the two bumps, and the resulting 1-skeleton, in which w = {v2, v3, v4} and the inner feet A2, A3 become loops. 5.3. The structure of the 1-skeleton. Lemma 5.10 (Ver…

**Figure 6.** Figure 6: The two-bump base case. Each Di denotes a possibly empty path of dummy edges. In the left panel, the undirected dashed curves indicate the forced vertex identifications τ (e1) ∼ ι(e3) and τ (e2) ∼ ι(e4). After adjoining the zipper edge s and deleting e2, e3, the right …

**Figure 7.** Figure 7: The two vertex configurations in the truncation step. When ℓ > 0, the discarded dummy path D joins t = τn to w; when ℓ = 0, the vertices satisfy t = τn = w. Thus Kfull is related to Ktr exactly as the larger complex in the Lifting Lemma is related to the smaller one. A…

**Figure 8.** Figure 8: The mixed complex after lifting the inductive GS sequence. In both panels the old inner cycle contains vR X −→ z Ln+1 −−−→ w Y −→ vR. When ℓ = 0, the edge Rn : vR → w = t is a chord. When ℓ > 0, the corresponding chord is subdivided as vR Rn −−→ t D −→ w. In both cases…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

16 extracted references · 3 linked inside Pith

[1]

Bleak, M

C. Bleak, M. G. Brin, M. Kassabov, J. T. Moore, and M. C. B. Zaremsky,Groups of fast homeomorphisms of the interval and the ping-pong argument, J. Comb. Algebra3(2019), no. 1, 1–40

2019
[2]

Bleak, M

C. Bleak, M. G. Brin, and J. T. Moore,Complexity among the finitely generated subgroups of Thompson’s group, J. Comb. Algebra5(2021), no. 1, 1–58

2021
[3]

M. G. Brin,Elementary amenable subgroups of R. Thompson’s groupF, Internat. J. Algebra Comput. 15(2005), no. 4, 619–642

2005
[4]

Belk and L

J. Belk and L. Stott,Pseudo-F 4 is isomorphic toF 4, arXiv:2303.16868 (2023)

arXiv 2023
[5]

Burillo, K.-U

J. Burillo, K.-U. Bux, and B. Nucinkis,Cohomological and metric properties of groups of homeomorphisms ofR, Oberwolfach Reports15(2018), 1579–1633

2018
[6]

K. S. Brown,Finiteness properties of groups, J. Pure Appl. Algebra44(1987), 45–75

1987
[7]

J. W. Cannon, W. J. Floyd, and W. R. Parry,Introductory notes on Richard Thompson’s groups, Enseign. Math.42(1996), 215–256

1996
[8]

D. S. Farley and B. Hughes,Braided diagram groups and local similarity groups, in: Geometric and coho- mological group theory, London Math. Soc. Lecture Note Ser.444, Cambridge Univ. Press, Cambridge, 2018, pp. 15–33

2018

Show all 16 references

[9]

Golan and E

G. Golan and E. Sapir,On the generation problem in Thompson’s groupsF n, arXiv:2606.00863 (2026)

2026 arXiv
[10]

Golan,Higman–Thompson groupsF n all the way down, arXiv:2607.04038 (2026)

G. Golan,Higman–Thompson groupsF n all the way down, arXiv:2607.04038 (2026)

2026 arXiv
[11]

Golan-Polak,The generation problem in Thompson groupF, Mem

G. Golan-Polak,The generation problem in Thompson groupF, Mem. Amer. Math. Soc.292(2023), no. 1451

2023
[12]

Golan and M

G. Golan and M. Sapir,On subgroups of R. Thompson’s groupF, Trans. Amer. Math. Soc.369(2017), 8857–8878

2017
[13]

Guba and M

V. Guba and M. Sapir,Diagram groups, Mem. Amer. Math. Soc.130(1997), no. 620

1997
[14]

Guba and M

V. Guba and M. Sapir,On subgroups of R. Thompson’s groupFand other diagram groups, Mat. Sb.190 (1999), no. 8, 3–60; English transl., Sb. Math.190(1999), no. 8, 1077–1130

1999
[15]

Guba and M

V. Guba and M. Sapir,Diagram groups and directed 2-complexes: homotopy and homology, J. Pure Appl. Algebra205(2006), no. 1, 1–47

2006
[16]

Kilibarda,On the algebra of semigroup diagrams, Internat

V. Kilibarda,On the algebra of semigroup diagrams, Internat. J. Algebra Comput.7(1997), 313–338. Department of Mathematics, Ben Gurion University of the Negev Email address:golangi@bgu.ac.il

1997

Pith tools

Reviewed July 14, 2026 · model on record in the stance chip above.

[1] [1]

Bleak, M

C. Bleak, M. G. Brin, M. Kassabov, J. T. Moore, and M. C. B. Zaremsky,Groups of fast homeomorphisms of the interval and the ping-pong argument, J. Comb. Algebra3(2019), no. 1, 1–40

2019

[2] [2]

Bleak, M

C. Bleak, M. G. Brin, and J. T. Moore,Complexity among the finitely generated subgroups of Thompson’s group, J. Comb. Algebra5(2021), no. 1, 1–58

2021

[3] [3]

M. G. Brin,Elementary amenable subgroups of R. Thompson’s groupF, Internat. J. Algebra Comput. 15(2005), no. 4, 619–642

2005

[4] [4]

Belk and L

J. Belk and L. Stott,Pseudo-F 4 is isomorphic toF 4, arXiv:2303.16868 (2023)

arXiv 2023

[5] [5]

Burillo, K.-U

J. Burillo, K.-U. Bux, and B. Nucinkis,Cohomological and metric properties of groups of homeomorphisms ofR, Oberwolfach Reports15(2018), 1579–1633

2018

[6] [6]

K. S. Brown,Finiteness properties of groups, J. Pure Appl. Algebra44(1987), 45–75

1987

[7] [7]

J. W. Cannon, W. J. Floyd, and W. R. Parry,Introductory notes on Richard Thompson’s groups, Enseign. Math.42(1996), 215–256

1996

[8] [8]

D. S. Farley and B. Hughes,Braided diagram groups and local similarity groups, in: Geometric and coho- mological group theory, London Math. Soc. Lecture Note Ser.444, Cambridge Univ. Press, Cambridge, 2018, pp. 15–33

2018