REVIEW 4 minor 13 references
Unbounded Gaps Between Ordinary and Equivariant Dehn Surgery Numbers
T0 review · 0 major / 4 minor · reviewed 2026-08-05 · deepseek-v4-flash
Pith's one-line read Every integer k occurs as the gap between ordinary and equivariant surgery numbers.
desk verdict Genuinely new result: resolves two K3 problems with a clean, checkable proof; the eigenspace-sum lower bound is a reusable trick. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
The load-bearing object is the basic block: the standard Hopf link in S^3 with both components framed +4, swapped by the unitary half-turn (z1,z2)↦(z2,z1). Surgery on this two-component link yields L(15,4), and the induced involution τ acts as multiplication by −4 on H_1≅Z/15. This number is chosen because −4 ≡ 1 (mod 5) and −4 ≡ −1 (mod 3), so the same geometric involution displays entirely different homology eigenspaces over different primes. The argument then uses a signed-permutation involution—an involution that permutes a basis up to sign, as a half-turn does to the meridians of a periodic link exterior—and proves (Lemma 4.1) that the dimensions of its +1 and −1 eigenspaces are the sam
What would settle it
Check the k=1 case directly: the paper claims that (L(15,4), τ) has EDS=2. A search over one-component periodic surgery descriptions, meaning periodic knots in S^3 of increasing crossing number that induce τ, would settle it: if any one-component description existed, then EDS=1 and Theorem 1.1 would be false. The paper's own Lemma 2.10 predicts none exists because τ acts on H_1≅Z/15 as multiplication by −4, which is neither +1 nor −1; so the concrete check is to verify that computed action independently. More broadly, implementing Proposition 4.2 on a different pair of primes and searching for
Extended reading notes
Core claim
The central result is Theorem 1.1: for each k≥1 there is a pair (Y_k, τ_k) with DS(Y_k)=k and EDS(Y_k, τ_k)=2k, so the difference EDS−DS is unbounded even for involutions. Up to orientation, Y_k is the connected sum of k copies of L(15,4), and τ_k is the involution induced by a half-turn that swaps the two components of a split union of k copies of a (+4,+4)-framed Hopf link. On first homology (Z/15)^k the action is multiplication by −4, which is +1 modulo 5 and −1 modulo 3, so the positive eigenspace over F_5 and the negative eigenspace over F_3 each have dimension k. The key inequality, Proposition 4.2, states that EDS(Y,τ) ≥ b^+(Y,τ)+b^−(Y,τ), where b^± are the maxima over odd primes of t
Load-bearing premise
The whole lower bound depends on the claim that an involution which merely permutes the meridians of a surgery link, possibly sending a meridian to its negative, has +1 and −1 eigenspace dimensions that do not change when the coefficient field is changed; if that field independence failed, the dimensions detected modulo 5 and modulo 3 could not be added, and the inequality EDS ≥ b^+ + b^− would collapse.
Editorial extensions
If this is right
- For every positive integer k, there is an explicit closed oriented 3-manifold Y_k and an involution τ_k with DS(Y_k)=k and EDS(Y_k,τ_k)=2k, so the gap can be prescribed exactly as any positive integer k.
- This answers Problems 1.15(b) and 1.15(c) from the K3 problem list: the difference between the equivariant and ordinary surgery numbers can be arbitrarily large, and strict inequality already occurs for involutions.
- The lower bound EDS(Y,τ) ≥ b^+(Y,τ)+b^−(Y,τ) is a new general obstruction: any involution that acts as +id on homology modulo one odd prime and as −id on homology modulo another odd prime needs at least one surgery component for each detected direction.
- There are infinitely many pairwise nonhomeomorphic irreducible lens spaces Z with DS(Z)=1 < EDS(Z,σ), so the strict inequality phenomenon is not limited to connected sums.
- The large-gap examples are reducible for k>1, while the irreducible examples have a fixed small gap; together the two theorems give a complete answer to the posed questions.
Reading between the lines
- The field-independence lemma suggests a general principle beyond involutions: for any finite-order diffeomorphism whose action on a lattice is a signed permutation, eigenspace dimensions over different primes could be combined, giving lower bounds for equivariant surgery numbers of higher-order symmetries when homology has root-of-unity actions over different primes.
- Because the lower bound is purely homological, one could try to strengthen it by bringing in Floer-theoretic or 4-dimensional invariants; the paper itself does not need such tools, but richer invariants might detect eigenspace contributions not visible modulo primes, yielding sharper bounds in cases where homology eigenspaces vanish.
- The irreducible lens-space family leaves open whether the gap can be unbounded within irreducible manifolds; a natural next test is to seek lens spaces with involutions whose EDS−DS grows with the order of the manifold.
- The construction is fully explicit, so for each k one can write down an actual 2k-component periodic link; this could be used to search for other pairs of primes and other block links that realize different exact gaps.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies the integral Dehn surgery number DS(Y) and the equivariant integral Dehn surgery number EDS(Y,phi), the minimum number of components in a periodic surgery description inducing a given finite-order diffeomorphism. The main theorem (Theorem 1.1) constructs, for each k>=1, an oriented 3-manifold Y_k—the connected sum of k copies of the lens space L(15,4)—with an orientation-preserving involution tau_k for which DS(Y_k)=k and EDS(Y_k,tau_k)=2k. This gives arbitrarily large gaps and answers Problems 1.15(b) and 1.15(c). Theorem 1.2 provides infinitely many irreducible lens spaces with DS=1<EDS. The technical core is a lower bound (Proposition 4.2) using the action of the involution on the meridian lattice of any periodic surgery link: the dimensions of the +1 and -1 eigenspaces are field-independent (Lemma 4.1), so eigenspace contributions detected over different primes can be added.
Significance. The main contribution is Proposition 4.2, a homological lower bound for EDS specifically designed to add eigenspace dimensions detected over different odd primes. The key point—Lemma 4.1—is correct: for a signed-permutation involution, each fixed basis vector contributes to one eigenspace and each two-cycle contributes one dimension to each eigenspace, so the dimensions are independent of the coefficient field. The application to Y_k, where the action is +1 mod 5 and -1 mod 3, is clean and yields the sharp values DS=k, EDS=2k. The paper also gives explicit irreducible lens-space examples. The constructions are explicit and checkable, and the proofs rely only on standard tools. If accepted, this resolves the two motivating problems and provides a reusable technique.
minor comments (4)
- [Section 4, Lemma 4.1] The field-independence lemma is central to Proposition 4.2 but is stated without proof. A two-sentence proof—fixed basis vectors contribute to one eigenspace according to their sign, and each two-cycle contributes one dimension to each eigenspace—would make the paper self-contained.
- [Section 2, Theorem 2.5 / Remark 2.6] The conversion from Moser's slope m*lambda - n*mu to the paper's N*mu + lambda is very terse. Spelling out the identification for (m,n) = (1,-N) would prevent sign errors in the applications.
- [Section 3, Proposition 3.1] The slam-dunk step 4 - 1/4 = 15/4 is used to identify Y with L(15,4), but the orientation convention for rational-surgery coefficients is not stated. The subsequent Moser argument already gives a one-component description, so the sign is not load-bearing, but a brief convention note would help.
- [Section 5, paragraph after Eq. (2)] The translation s_i(x,y,z) = (x,y,z+3ri) is easy to misread as 3r*i. Writing 3r*i or 3r*i with a thin space would improve readability.
Circularity Check
No significant circularity; the derivation is self-contained.
full rationale
The paper's central construction and bounds are derived from explicit surgery descriptions, standard homology presentations, and an elementary linear-algebraic lemma; none of the key steps assumes the result being proved. The lower bound EDS(Y,τ) ≥ b^+(Y,τ) + b^−(Y,τ) is established in Proposition 4.2 using Lemma 2.7's equivariant surjection from H_1 of the link exterior and Lemma 4.1's field-independence of eigenspaces of signed permutation involutions. Lemma 4.1 is an independent elementary fact, not a disguised version of the theorem. The upper bounds EDS(Y_k, τ_k) ≤ 2k and DS(Y_k) ≤ k come from explicit periodic Hopf-link and split-link surgeries, while the lower bounds DS(Y_k) ≥ k come from the standard generator bound d(H_1; Z). Theorem 1.2 similarly uses Moser's torus-knot surgery theorem and an explicit quotient involution. The paper does not fit parameters, rename known results, or rely on self-citations: citations to Sakuma, Moser, Lickorish–Wallace, and standard references are external background results, not load-bearing self-citations. No circular step, fitted-input-as-prediction, or imported-uniqueness argument is present.
Assumptions & free parameters
assumptions (6)
- standard math Lickorish-Wallace theorem: every closed oriented 3-manifold is obtained by integral surgery on a framed link in S^3.
- standard math Sakuma's equivariant surgery theorem: every finite-order orientation-preserving diffeomorphism of a closed oriented 3-manifold admits an integral periodic surgery description.
- standard math Moser's theorem (Theorem 2.5): surgery on a torus knot K(r,s) with slope mλ - nμ yields L(|n|, m s^2) when |rsm+n| = 1.
- standard math Linking matrix homology presentation (Lemma 2.8): H1 of the surgered manifold is the cokernel of the linking matrix.
- standard math The slam-dunk move reduces a (+4,+4) Hopf link to 15/4 surgery on an unknot.
- standard math Every lens space is irreducible.
Cite this review
Pith. "Pith review of Unbounded Gaps Between Ordinary and Equivariant Dehn Surgery Numbers." pith.science (2026). https://pith.science/paper/FWM2XU5V
@misc{pith2026260803886,
author = {Pith},
title = {Pith review of: Unbounded Gaps Between Ordinary and Equivariant Dehn Surgery Numbers},
year = {2026},
howpublished = {\url{https://pith.science/paper/FWM2XU5V}},
note = {Machine review of arXiv:2608.03886}
}
abstract
For a closed oriented $3$-manifold $Y$ and an orientation-preserving involution $\tau$, let $\DS(Y)$ denote the minimum number of components in an integral surgery description of $Y$, and let $\EDS(Y,\tau)$ denote the corresponding minimum among periodic surgery descriptions inducing $\tau$. We prove that for every integer $k\geq 1$ there is a pair $(Y_k,\tau_k)$ such that \[ \DS(Y_k)=k, \qquad \EDS(Y_k,\tau_k)=2k. \] Consequently, the difference $\EDS(Y,\tau)-\DS(Y)$ is unbounded even when $\tau$ is an involution. This answers Problems~1.15(b) and~1.15(c) in the K3 problem list. We also construct infinitely many pairwise nonhomeomorphic irreducible lens spaces $Z$ admitting involutions $\sigma$ for which \[ \DS(Z)<\EDS(Z,\sigma). \]
Reference graph
Works this paper leans on
-
[1]
D. Auckly,Surgery numbers of3-manifolds: a hyperbolic example, in Geometric Topology (Athens, GA, 1993), AMS/IP Stud. Adv. Math., vol. 2.1, Amer. Math. Soc., Providence, RI, 1997, pp. 21–34
work page 1993
-
[2]
R. İ. Baykur, R. C. Kirby, and D. Ruberman, editors,K3—A New Problem List in Low- Dimensional Topology, Mathematical Surveys and Monographs, vol. 295, American Mathematical Society, Providence, RI, 2026, to appear
work page 2026
-
[3]
R. E. Gompf and A. I. Stipsicz,4-Manifolds and Kirby Calculus, Graduate Studies in Mathematics, vol. 20, American Mathematical Society, Providence, RI, 1999
work page 1999
-
[4]
The link surgery formula and equivariant surgeries
K. Hendricks, A. Mallick, M. Stoffregen, and I. Zemke,The link surgery formula and equivariant surgeries, preprint, 2025, arXiv:2507.12809
work page Pith review arXiv 2025
-
[5]
J. Hom, C. Karakurt, and T. Lidman,Surgery obstructions and Heegaard Floer homology, Geom. Topol.20(2016), no. 4, 2219–2251
work page 2016
-
[6]
W. B. R. Lickorish,A representation of orientable combinatorial3-manifolds, Ann. of Math. (2)76(1962), 531–540
work page 1962
-
[7]
Bounding the Dehn surgery number by 10/8
B. Liu and L. Piccirillo,Bounding the Dehn surgery number by10/8, preprint, 2024, arXiv:2405.16904
work page Pith review arXiv 2024
-
[8]
Mallick,Knot Floer homology and surgery on equivariant knots, J
A. Mallick,Knot Floer homology and surgery on equivariant knots, J. Topol.17(2024), no. 4, Paper No. e70001
work page 2024
Show all 13 references
-
[9]
Moser,Elementary surgery along a torus knot, Pacific J
L. Moser,Elementary surgery along a torus knot, Pacific J. Math.38(1971), no. 3, 737–745
1971
-
[10]
J. H. Przytycki and M. V. Sokolov,Surgeries on periodic links and homology of periodic 3-manifolds, Math. Proc. Cambridge Philos. Soc.131(2001), no. 2, 295–307
2001
-
[11]
Sakuma,Surgery description of orientation-preserving periodic maps on compact orientable3-manifolds, Rend
M. Sakuma,Surgery description of orientation-preserving periodic maps on compact orientable3-manifolds, Rend. Istit. Mat. Univ. Trieste32(2001), suppl. 1, 375–396
2001
-
[12]
Saveliev,Lectures on the Topology of3-Manifolds: An Introduction to the Casson Invariant, second revised and extended edition, De Gruyter, Berlin–Boston, 2012
N. Saveliev,Lectures on the Topology of3-Manifolds: An Introduction to the Casson Invariant, second revised and extended edition, De Gruyter, Berlin–Boston, 2012
2012
-
[13]
A. H. Wallace,Modifications and cobounding manifolds, Canad. J. Math.12(1960), 503–528. 12 QILONG GUO ∗ AND CHUNXING YAN ∗ College of Science, China University of Petroleum–Beijing, Beijing 102249, P. R. China Email address:guoqilong1984@hotmail.com School of Finance and Mathe...
1960
Reviewed August 5, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.