# Gersten’s conjecture and virtual compact specialness of one-relator groups > Claims one-relator groups without Baumslag–Solitar subgroups are hyperbolic and virtually compact special. Plain-English explainer with sources. Canonical: https://closedform.nishachay.tech/e/openai-math-2026-258/ Last updated: 2026-10-14 **Claims one-relator groups without Baumslag–Solitar subgroups are hyperbolic and virtually compact special.** **Trust:** Claimed, not checked · **Papers:** 2 · **First paper:** September 25, 2026 · **Field:** Group theory · **Entry:** 258 *Explained by AI.* Drafted by one AI model and checked sentence by sentence against the paper by a different one. Spot a mistake? Report an error. A one-relator group is built from free generators and a single rule. Some such groups contain special pieces called Baumslag–Solitar groups, which stretch distances badly. Gersten conjectured these are the only thing stopping hyperbolicity, a large-scale negative curvature. The first paper claims to prove this. The second claims every hyperbolic one-relator group has a finite cover with a strong cube geometry. ## At a glance - **First step:** 1992 (Steve Gersten, who asked whether Baumslag–Solitar subgroups are the only obstructions) - **Significance:** Major. Gersten's conjecture and the cubulation of hyperbolic one-relator groups are central open problems for one-relator groups. - **Trust:** Claimed, not checked - **Field:** Group theory (2 papers) ## The explanation Take two generators a and t with one rule: t a t⁻¹ = a². This is the Baumslag–Solitar group BS(1, 2). Apply the rule twice: t² a t⁻² = t a² t⁻¹ = a⁴. After k steps, t^k a t^−k = a^(2^k). For k = 10 that gives a¹⁰²⁴. The left side has only 10 + 1 + 10 = 21 letters. So a power of a that seems 1024 letters long has a 21-letter shortcut. Such exponential shortcuts cannot happen in a hyperbolic group. A one-relator group has finitely many generators and one defining rule. BS(m, n) is the group with rule t a^m t⁻¹ = a^n, for nonzero m and n[1]. These groups are the basic obstructions to hyperbolicity among one-relator groups[1]. A word-hyperbolic group is one whose distance graph has thin triangles. The flat grid Z² is BS(1, 1), with rule t a t⁻¹ = a, so the conjecture excludes it too[2]. Gersten asked whether Baumslag–Solitar subgroups are the only obstructions[3]. Partial answers came from the hierarchy method. Magnus showed in 1930 that dropping a generator from the rule leaves a free subgroup[4]. Newman's spelling theorem gave hyperbolicity for relators that are proper powers[5]. Linton showed a one-relator group is hyperbolic and locally quasiconvex exactly when its primitivity rank is not two[6]. The obstacle was hyperbolicity alone in the primitivity-rank-two case. There the splittings need not stay geometrically controlled. The first paper claims the full conjecture. If a one-relator group contains no BS(m, n) subgroup, then it is word-hyperbolic[7]. The proof uses Magnus splittings[8]. A combination theorem then gives hyperbolicity when annuli in a graph of groups flare out[9]. The second paper claims that every hyperbolic one-relator group is virtually compact special[10]. That means a finite-index subgroup is the loop group of a finite cube complex. That complex maps nicely to the model space of a right-angled Artin group. A right-angled Artin group has one generator per vertex of a graph, with commuting rules along edges. This structure implies residual finiteness: every nonidentity element survives in some finite quotient[11]. Combined with work of Kielak and Linton, it answers Wise's question: hyperbolic one-relator groups are virtually free-by-cyclic[12]. How sure is this? Neither paper has a Lean formalization. No specialist review of either paper has been published. These are claims by the papers, not checked results. **What it doesn't show.** This does not have a Lean formalization, and it does not cover groups with more than one defining relation. ## Where it came from This result comes from an unreleased OpenAI model, published in the openai/math release on 6 October 2026. ## Hype check Major if it holds. Neither paper has a Lean formalization or a published specialist review. ## Words used - **One-relator group:** A group with finitely many generators and a single defining rule. - **Baumslag–Solitar group:** The group BS(m, n) with generators a and t and the one rule t a^m t⁻¹ = a^n. - **Word-hyperbolic group:** A finitely generated group whose distance graph has uniformly thin triangles. - **Primitivity rank:** The least rank of a free subgroup containing the relator where the relator is not part of a free basis. - **Virtually compact special:** A finite-index subgroup is the loop group of a finite cube complex that maps locally isometrically to a right-angled Artin model. - **Residually finite:** Every nonidentity element stays nonidentity in some finite image of the group. - **Free-by-cyclic:** An extension of a free group by a subgroup of the integers. ## The original papers (2) - [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf) - September 25, 2026 - [Virtual compact specialness of hyperbolic one-relator groups](https://github.com/openai/math/blob/main/preprints/Virtual-compact-specialness-of-hyperbolic-one-relator-groups-September-25-2026/paper.pdf) - September 25, 2026 ## Quoted lines from the paper (12) [1] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > The Baumslag--Solitar groups \[ \operatorname{BS}(m,n) =\langle a,t\mid ta^mt^{-1}=a^n\rangle, \qquad m,n\in\mathbb Z\setminus\{0\}, \] are the basic obstructions to hyperbolicity among one-relator groups. [2] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > In particular, the hypothesis excludes $\operatorname{BS}(1,1)=\mathbb Z^2$. [3] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > Gersten asked whether they are the only subgroup obstructions. [4] From [Virtual compact specialness of hyperbolic one-relator groups](https://github.com/openai/math/blob/main/preprints/Virtual-compact-specialness-of-hyperbolic-one-relator-groups-September-25-2026/paper.pdf): > The inductive study of one-relator groups begins with Magnus's Freiheitssatz: omitting a generator that occurs in a cyclically reduced relator leaves a subgroup freely generated by the remaining generators. [5] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > Newman's spelling theorem for proper-power relators gives a Dehn algorithm and hence hyperbolicity; see also the proof in. [6] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > His later characterization says that a finitely generated one-relator group is both hyperbolic and locally quasiconvex precisely when its relator has primitivity rank different from two. [7] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > If the one-relator group $G=\langle X\mid r\rangle$ contains no subgroup isomorphic to $\operatorname{BS}(m,n)$ for any nonzero integers $m,n$, then $G$ is word-hyperbolic. [8] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > The Freiheitssatz makes an inductive approach possible: a Magnus splitting has a smaller one-relator vertex group and free edge groups. [9] From [Baumslag-Solitar-free one-relator groups are hyperbolic](https://github.com/openai/math/blob/main/preprints/Baumslag-Solitar-free-one-relator-groups-are-hyperbolic-September-25-2026/paper.pdf): > The Bestvina--Feighn combination theorem makes annular flaring a sufficient condition for hyperbolicity when the vertex and edge data satisfy the required geometric hypotheses. [10] From [Virtual compact specialness of hyperbolic one-relator groups](https://github.com/openai/math/blob/main/preprints/Virtual-compact-specialness-of-hyperbolic-one-relator-groups-September-25-2026/paper.pdf): > Then $G$ is virtually compact special. [11] From [Virtual compact specialness of hyperbolic one-relator groups](https://github.com/openai/math/blob/main/preprints/Virtual-compact-specialness-of-hyperbolic-one-relator-groups-September-25-2026/paper.pdf): > Theorem … extends the compact-special conclusion, and hence residual finiteness, to every hyperbolic one-relator group. [12] From [Virtual compact specialness of hyperbolic one-relator groups](https://github.com/openai/math/blob/main/preprints/Virtual-compact-specialness-of-hyperbolic-one-relator-groups-September-25-2026/paper.pdf): > Combining that corollary with Theorem … gives a positive answer. ## Other sources - [openai/math release notes (2026): model provenance and review status](https://github.com/openai/math/blob/main/README.md): “manuscripts and supporting proof artifacts produced by an internal OpenAI model” ## Related - [Cannon's conjecture](https://closedform.nishachay.tech/e/openai-math-2026-246/) - Group theory - [The Benjamini–Schramm nonuniqueness conjecture](https://closedform.nishachay.tech/e/openai-math-2026-214/) - Probability and statistical mechanics - [The purely cosmetic surgery conjecture](https://closedform.nishachay.tech/e/openai-math-2026-306/) - Topology - [The Hovey–Strickland and Chai conjectures](https://closedform.nishachay.tech/e/openai-math-2026-311/) - Topology - [Seymour’s second-neighborhood conjecture](https://closedform.nishachay.tech/e/openai-math-2026-173/) - Combinatorics - [Group theory](https://closedform.nishachay.tech/science/mathematics/group-theory/) - Field · 14 results - [Named conjectures](https://closedform.nishachay.tech/topic/named-conjectures/) - Topic · 85 results - [Open since the 1990s](https://closedform.nishachay.tech/open-since/1990s/) - Era · 55 results - [Claimed, not yet checked](https://closedform.nishachay.tech/trust/claimed/) - Trust · 130 results - [Report an error](https://github.com/nishachay/closed-form/issues/new?template=report-error.yml&labels=entry-error&title=%5Bentry%20258%5D%20)