Pith. sign in

REVIEW 2 major objections 5 minor 26 references

Andrews-Curtis groups

T0 review · 2 major / 5 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read For torsion-free non-elementary hyperbolic groups, the full Andrews-Curtis transformation group acts faithfully on every nontrivial orbit of G^k, making it isomorphic to the classical Andrews-Curtis group.

desk verdict The theorem is a real extension of Roman'kov's result to hyperbolic groups, but the proof of the key equation lemma has a genuine gap around a simultaneous non-commutation choice; still deserves refereeing. read the letter →

arxiv 2506.23031 v2 pith:7ZLRMGMH submitted 2025-06-28 math.GR

Robert H. Gilman , Alexei G. Myasnikov This is my paper · ORCID
classification math.GR MSC 20F6720F0520E36
keywords Andrews-Curtisconjecturegrouphyperbolicgroupsbalancedpresentationsequationsoverbigpowerspropertynormalgeneratingtuplesfaithfulaction
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 studies two permutation groups built from the same moves: the full Andrews-Curtis group acting on all k-tuples of a group G, and the classical Andrews-Curtis group acting only on tuples that normally generate G. The main theorem says that when G is torsion-free and non-elementary hyperbolic, the full group is faithful on every nontrivial orbit, so the natural map from the full group to the classical group is an isomorphism. This matters because it shows that for a broad class of groups, including nonabelian free groups, the complicated generating-tuple set N_k(G) can be replaced by the much simpler set G^k without losing any information about the transformations.

What carries the argument

The engine of the proof is Theorem 3.3, an equation-theoretic triviality criterion: over a torsion-free non-elementary hyperbolic group, if an equation E(x_1,...,x_m) in the free product G*X is satisfied by every tuple in G^m, then E is the trivial word. The argument uses the big powers property of hyperbolic groups to force, from a supposed nontrivial equation, a substitution that violates the equation; the CSA property of centralizers then reduces commuting of large powers to commuting of the underlying elements. This theorem, applied to the free product G*X where X is free, lets the authors pass from a transformation fixing all conjugates of a tuple to an identity in the free product, which yields faithfulness.

What would settle it

Exhibit a torsion-free non-elementary hyperbolic group G, a nontrivial α in FAC_k(G), and a tuple with all entries nonidentity such that α fixes every conjugation of that tuple. Equivalently, find an equation E over such a G that is not the identity in G*X yet is satisfied by every tuple from G; Theorem 3.3 asserts none exists.

Watch

Extended reading notes

Core claim

Theorem 2.1 establishes that if G is a torsion-free non-elementary hyperbolic group, then FAC_k(G) acts faithfully on every nontrivial orbit in G^k. The proof shows that if an Andrews-Curtis transformation fixes all conjugates of a tuple with at least one nonidentity entry, then it must fix every tuple in G^k and hence be the identity. Corollary 2.2 then identifies the full Andrews-Curtis group with the classical Andrews-Curtis group: the restriction epimorphism λ: FAC_k(G) → AC_k(G) is an isomorphism.

Load-bearing premise

The proof depends on the big powers property of torsion-free non-elementary hyperbolic groups: for any sequence of elements in which consecutive entries do not commute, sufficiently large powers of those elements multiply to a nontrivial element. If that external property failed, the contradiction argument in Theorem 3.3 would not go through, and with it the faithfulness theorem would lose its main support.

Editorial extensions

If this is right

  • For every torsion-free non-elementary hyperbolic group G and integer k ≥ 2, FAC_k(G) and AC_k(G) are isomorphic, so the kernel of λ is trivial and General Problem 1 is solved for this class.
  • The result extends the earlier isomorphism theorem for free nonabelian groups to all torsion-free non-elementary hyperbolic groups, unifying the known case with a single argument.
  • Since G^k is computable when the word problem for G is decidable, the full Andrews-Curtis group can be studied algorithmically on a computationally tractable domain, unlike N_k(G) whose computability is open.
  • The faithfulness theorem provides a concrete sense in which AC-transformations on hyperbolic groups are rigid: any transformation that looks trivial on one orbit must be trivial globally.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The same proof scheme may work for any group satisfying CSA, the big powers condition, and a suitable noncommutation condition; the paper's Remark 3.5 already lists these properties, but the faithfulness conclusion for such groups is a natural extension the authors do not spell out.
  • If the isomorphism between FAC_k(G) and AC_k(G) holds in broader classes of groups, then search algorithms for Andrews-Curtis trivializations could be run on the full product G^k rather than only on normal-generating tuples, potentially making counterexample searches more tractable.
  • The triviality of the radical of the affine space G^n, noted as a remark, suggests that equations over torsion-free hyperbolic groups behave like equations over free groups in a strong model-theoretic sense; one could test whether this radical triviality has consequences for solving systems of equations over such groups.
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.

Referee Report

2 major / 5 minor

Summary. The paper studies the relationship between the full Andrews-Curtis group FAC_k(G), acting on all of G^k, and the original Andrews-Curtis group AC_k(G), acting on the subset N_k(G) of tuples that normally generate G. Its main theorem states that when G is a torsion-free non-elementary hyperbolic group, FAC_k(G) acts faithfully on every nontrivial orbit in G^k, so the natural epimorphism lambda: FAC_k(G) -> AC_k(G) is an isomorphism. The proof reduces faithfulness to a statement about equations over G: if an equation has all tuples of G as solutions, then it is trivial in G*X (Theorem 3.3). That theorem is proved using Olshanskii's big powers property and the cyclic malnormal centralizer structure of torsion-free hyperbolic groups, after which Section 4 applies it to the word equations describing an AC-transformation.

Significance. If the proof is completed, the result is a valuable extension of Roman'kov's theorem for free groups to all torsion-free non-elementary hyperbolic groups. It gives the first broad class of groups for which the kernel of lambda is trivial, thereby justifying the study of the more tractable group FAC_k(G) as a tool for the Andrews-Curtis conjecture. The reliance on deep external results (big powers, CSA centralizers) is appropriate, and the paper is concise. However, the central proof of Theorem 3.3 contains a nontrivial gap: the existence of a tuple avoiding all adjacent commutations is asserted without proof, and this step is load-bearing for the main theorem.

major comments (2)
  1. [Theorem 3.3, proof after Eq. (3.2)] The sentence "But clearly we can choose g_{i_j} in G such that g_{i_j} and a_j g_{i_{j+1}} a_j^{-1} do not commute" is not justified as written. A single variable may occur in several adjacent pairs, so the requirement that every adjacent pair u_j(g_{i_j}) and u_{j+1}(g_{i_{j+1}}) be non-commuting is a system of inequalities coupling different variables. The big powers property applies only to a fixed sequence with no commuting adjacent pair, so the proof needs a lemma asserting the simultaneous existence of a tuple (g_1,...,g_m) with [g_{i_j}, a_j g_{i_{j+1}} a_j^{-1}] != 1 for every j. Such a tuple does exist: order the variables arbitrarily, and when choosing each variable avoid the finite union of cyclic centralizers imposed by the already-chosen neighbors; the union is proper because a non-elementary hyperbolic group is not a finite union of cyclic subgroups (and condition 3 of Remark 3.5 is exactly designed for this). The manuscript must supply this argument, since Theorem 3.3 is the only route to Eq. (4.1) in the proof of Theorem 2.1.
  2. [Section 4, proof of Theorem 2.1, after Lemma 4.1] The phrase "Since G*X is non-elementary torsion-free hyperbolic" before Eq. (4.1) misidentifies the group to which Theorem 3.3 is applied. The equation W_i(u_1^{x_1},...,u_k^{x_k}) = u_i^{x_i} is an equation over the coefficient group G with indeterminates x_i; the ambient free product G*X is where the equation lives, not the coefficient group. The application of Theorem 3.3 is valid because G is non-elementary torsion-free hyperbolic, but the sentence should be reworded to say so explicitly, for example "Since G is non-elementary torsion-free hyperbolic, Theorem 3.3 applied to the equation in G*X gives ...".
minor comments (5)
  1. [Theorem 3.3, proof] The proof claims to argue by induction on n, but no inductive step is used after the base cases; either remove the induction framing or state the inductive hypothesis and use it.
  2. [Throughout] There are several typographical errors: "theelementary" in Section 2, a stray "u" in the definition of u_2 in the proof of Theorem 3.3, and "endomomorphism" in Section 4. These should be corrected.
  3. [Eq. (4.1) and surrounding text] The notation u_i^{x_i} is used without an explicit definition; it should be stated that u_i^{x_i} denotes x_i u_i x_i^{-1} (or its inverse, consistently).
  4. [Section 4, first paragraph] The reduction "Without loss of generality we may assume u_i != 1 for all i" deserves a one-sentence justification: if some u_i = 1, apply elementary AC moves to replace it by a non-trivial entry from another coordinate, which is possible because the orbit is nontrivial; this would improve readability.
  5. [References] The paper should cite Roman'kov's result [25] in the introduction or in Section 2 when mentioning that the free-group case was known; currently it appears only in a note after Corollary 2.2.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the main proof relies on independent external theorems (Olshanskii's big powers property and standard hyperbolic-group facts), and the authors' self-citations are background only.

full rationale

The paper's derivation is self-contained in the sense relevant to circularity. The central theorem (Theorem 2.1) is proved via Lemma 4.1 (words representing an AC transformation), the hypothesis that alpha fixes all conjugates of a tuple, and Theorem 3.3, which says that an equation over a non-elementary torsion-free hyperbolic group with all tuples as solutions is trivial. The engine of Theorem 3.3 is Lemma 3.1(3), the big powers property, cited to Olshanskii [21], an external and independent result; the centralizer and malnormal facts in Lemma 3.1(1) are also standard external references ([12, 3, 9, 10]). No parameter is fitted to data, and no quantity called a prediction is derived from an input that already contains it. The self-citations in the paper ([5, 6, 11, 15, 19, 20]) appear only in introductory background, in Remark 3.5 as examples of CSA and big-powers groups, and in open problems; none of these is load-bearing for the proof of Theorem 2.1 or Corollary 2.2. The proof's 'clearly we can choose g_{i_j}...' is a potential unproved existence assertion and thus a correctness concern, but it is not a circular reduction: it does not assume the conclusion, use a fitted parameter, or depend on a self-citation. Therefore the circularity score is 0.

Assumptions & free parameters 0 free parameters · 3 assumptions · 0 invented entities

No free parameters or invented entities. The paper relies on standard structural properties of hyperbolic groups and on Olshanskii's big powers theorem, all external results with independent proofs.

assumptions (3)
  • domain assumption Big powers property for non-elementary torsion-free hyperbolic groups (Lemma 3.1(3), citing Olshanskii [21])
    Used in the proof of Theorem 3.3 to assert that for any sequence, either a product of sufficiently large powers is non-trivial or some adjacent pair commutes.
  • domain assumption Centralizers of non-identity elements in G are cyclic and malnormal (Lemma 3.1(1))
    Used to infer that commuting powers imply commuting bases and to choose non-commuting elements recursively in Theorem 3.3.
  • domain assumption The free product of a non-elementary torsion-free hyperbolic group with a free group is again non-elementary torsion-free hyperbolic (Lemma 3.1(2))
    Invoked in Section 4; although the intended application of Theorem 3.3 is to G, the free product property is part of the paper's stated background.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Andrews-Curtis groups." pith.science (2026). https://pith.science/paper/7ZLRMGMH

@misc{pith2026250623031,
  author       = {Pith},
  title        = {Pith review of: Andrews-Curtis groups},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/7ZLRMGMH}},
  note         = {Machine review of arXiv:2506.23031}
}
abstract

For any group $G$ and integer $k\ge 2$ the Andrews-Curtis transformations act as a permutation group, termed the Andrews-Curtis group $AC_k(G)$, on the subset $N_k(G) \subset G^k$ of all $k$-tuples that generate $G$ as a normal subgroup (provided $N_k(G)$ is non-empty). The famous Andrews-Curtis Conjecture is that if $G$ is free of rank $k$, then $AC_k(G)$ acts transitively on $N_k(G)$. The set $N_k(G)$ may have a rather complex structure, so it is easier to study the full Andrews-Curtis group $FAC(G)$ generated by AC-transformations on a much simpler set $G^k$. Our goal here is to investigate the natural epimorphism $\lambda\colon FAC_k(G) \to AC_k(G)$. We show that if $G$ is non-elementary torsion-free hyperbolic, then $FAC_k(G)$ acts faithfully on every nontrivial orbit of $G^k$, hence $\lambda\colon FAC_k(G) \to AC_k(G)$ is an isomorphism.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

26 extracted references · 24 canonical work pages

  1. [1]

    S. I. Adyan. On algorithmic problems in effectively complete classes of groups.Dokl. Akad. Nauk SSSR, 123:13–16, 1958

    work page 1958

  2. [2]

    Akbulut and R

    S. Akbulut and R. Kirby. A potential smooth counterexample in dimension 4 to the Poincar´ e conjecture, the Schoenflies conjecture, and the Andrews-Curtis conjecture.Topology, 24(4):375–390, 1985

    work page 1985

  3. [3]

    J. M. Alonso, T. Brady, D. Cooper, V. Ferlini, M. Lustig, M. Mihalik, M. Shapiro, and H. Short. Notes on word hyperbolic groups. InGroup theory from a geometrical viewpoint (Trieste, 1990), pages 3–63. World Sci. Publ., River Edge, NJ, 1991. Edited by Short

    work page 1990

  4. [4]

    J. J. Andrews and M. L. Curtis. Free groups and handlebodies.Proc. Amer. Math. Soc., 16:192–195, 1965

    work page 1965

  5. [5]

    Baumslag, A

    G. Baumslag, A. G. Myasnikov, and V. Shpilrain. Open problems in combinatorial group theory. Second edition. InCombinatorial and geometric group theory (New York, 2000/Hoboken, NJ, 2001), volume 296 ofContemp. Math., pages 1–38. Amer. Math. Soc., Providence, RI, 2002

    work page 2000

  6. [6]

    A. V. Borovik, E. I. Khukhro, and A. G. Myasnikov. The Andrews-Curtis conjecture and black box groups.Internat. J. Algebra Comput., 13(4):415–436, 2003

    work page 2003

  7. [7]

    M. R. Bridson. The complexity of balanced presentations and the Andrews-Curtis conjecture, 2015. arXiv:1504.04187

    arXiv 2015

  8. [8]

    R. G. Burns and O. Macedo´ nska. Balanced presentations of the trivial group.Bull. London Math. Soc., 25(6):513–526, 1993

    work page 1993

Show all 26 references
  1. [9]

    Coornaert, T

    M. Coornaert, T. Delzant, and A. Papadopoulos.G´ eom´ etrie et th´ eorie des groupes, volume 1441 of Lecture Notes in Mathematics. Springer-Verlag, Berlin, 1990. Les groupes hyperboliques de Gromov. [Gromov hyperbolic groups], With an English summary

    1990

  2. [10]

    Ghys and P

    ´E. Ghys and P. de la Harpe, editors.Sur les groupes hyperboliques d’apr` es Mikhael Gromov, volume 83 ofProgress in Mathematics. Birkh¨ auser Boston, Inc., Boston, MA, 1990. Papers from the Swiss Seminar on Hyperbolic Groups held in Bern, 1988

    1990

  3. [11]

    Gildenhuys, O

    D. Gildenhuys, O. Kharlampovich, and A. Myasnikov. CSA-groups and separated free constructions. Bull. Austral. Math. Soc., 52(1):63–84, 1995

    1995

  4. [12]

    M. Gromov. Hyperbolic groups. InEssays in group theory, volume 8 ofMath. Sci. Res. Inst. Publ., pages 75–263. Springer, New York, 1987

    1987

  5. [13]

    Guirardel and G

    V. Guirardel and G. Levitt. JSJ decompositions of groups.Ast´ erisque, (395):vii+165, 2017

    2017

  6. [14]

    Havas and C

    G. Havas and C. Ramsay. Breadth-first search and the Andrews-Curtis conjecture.Internat. J. Algebra Comput., 13(1):61–68, 2003

    2003

  7. [15]

    A. V. Kvashchuk, A. G. Myasnikov, and D. E. Serbin. Pregroups and big powers.Algebra Logika, 48(3):342–377, 418, 420–421, 2009

    2009

  8. [16]

    A. Lisitsa. New Andrews-Curtis trivializations for Miller-Schupp group presentations.Ex. Countex., 6:Paper No. 100168, 3, 2024

    2024

  9. [17]

    W. Metzler. On the Andrews-Curtis conjecture and related problems. InCombinatorial methods in topology and algebraic geometry (Rochester, N.Y., 1982), volume 44 ofContemp. Math., pages 35–50. Amer. Math. Soc., Providence, RI, 1985

    1982

  10. [18]

    A. D. Miasnikov. Genetic algorithms and the Andrews-Curtis conjecture.Internat. J. Algebra Comput., 9(6):671–686, 1999

    1999

  11. [19]

    A. D. Myasnikov, A. G. Myasnikov, and V. Shpilrain. On the Andrews-Curtis equivalence. InCombina- torial and geometric group theory (New York, 2000/Hoboken, NJ, 2001), volume 296 ofContemp. Math., pages 183–198. Amer. Math. Soc., Providence, RI, 2002

    2000

  12. [20]

    A. G. Myasnikov and V. N. Remeslennikov. Exponential groups. II. Extensions of centralizers and tensor completion of CSA-groups.Internat. J. Algebra Comput., 6(6):687–711, 1996

    1996

  13. [21]

    A. Yu. Olshanskii. On residualing homomorphisms and G-subgroups of hyperbolic groups.Internat. J. Algebra Comput., 3(4):365–409, 1993

    1993

  14. [22]

    Panteleev and A

    D. Panteleev and A. Ushakov. Conjugacy search problem and the Andrews-Curtis conjecture.Groups Complex. Cryptol., 11(1):43–60, 2019

    2019

  15. [23]

    A. Piggott. Andrews-Curtis groups and the Andrews-Curtis conjecture.J. Group Theory, 10(3):373–387, 2007

    2007

  16. [24]

    M. O. Rabin. Recursive unsolvability of group theoretic problems.Ann. of Math. (2), 67:172–194, 1958

    1958

  17. [25]

    Roman’kov

    V. Roman’kov. On the Andrews-Curtis groups: non-finite presentability, 2023. arxiv:2305.11838. Vol. 16:1 ANDREWS-CURTIS GROUPS 7:7

    2023 arXiv

  18. [26]

    Shehper, A

    A. Shehper, A. M. Medina-Mardones, L. Fagan, B. Lewandowski, A. Gruen, Y. Qiu, P. Kucharski, Z. Wang, and S. Gukov. What makes math problems hard for reinforcement learning: a case study, 2025. arXiv:2408.15332

    2025 arXiv

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.