Remarks on some infinitesimal symmetries of Khovanov--Rozansky homologies in finite characteristic

Emmanuel Wagner; Joshua Sussan; Louis-Hadrien Robert; You Qi

REVIEW 1 major objections 5 minor 22 references

Remarks on some infinitesimal symmetries of Khovanov--Rozansky homologies in finite characteristic

T0 review · 1 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

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

Pith's one-line read Reduced Khovanov–Rozansky homology in characteristic $p$ is base-point independent, and its $\mathfrak{sl}_2$-symmetries force slice knots to contain a Steinberg summand and split links to admit an $\mathfrak{sl}_2\times\mathfrak{sl}_2$…

desk verdict A clean Morita-theoretic proof of base point independence, with Section 3 applications that are real but rest on imported and partly unproved machinery. read the letter →

arxiv 2412.08401 v1 pith:UTKPJ4UF submitted 2024-12-11 math.GT math.QA

You Qi , Louis-Hadrien Robert , Joshua Sussan , Emmanuel Wagner This is my paper · ORCID

classification math.GTmath.QA MSC 57K1857K1017B50

keywords Khovanov–Rozanskyhomologycharacteristicpbasepointindependencep-DGalgebrasl2actionSteinbergmodulesliceknotsplitlinkdetection

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

This paper establishes that reduced Khovanov–Rozansky homology of a knot in characteristic $p$ does not depend on the chosen base point, giving a new proof of a known theorem: the ambient $p$-differential graded algebra $A = k[x]/(x^p)$ is Morita equivalent to a matrix algebra, and reducing by the base-point action is exactly passing to the reduced complex. The paper then studies consequences of an $\mathfrak{sl}_2$-action on the non-equivariant characteristic-$p$ homology. Homology of a link diagram with a thickness-one edge has a filtration by dual baby Verma modules, which yields a parity-unimodality statement for its bigraded coefficients. A slice knot must contain the Steinberg module as a direct summand in cohomological degree zero. Finally, a link is split exactly when its homology carries an $\mathfrak{sl}_2\times\mathfrak{sl}_2$-action whose two factors annihilate each other's base-point variables.

What carries the argument

The load-bearing object is the $p$-differential graded Frobenius algebra $A = k[x]/(x^p)$ with differential $B_- = -\partial/\partial x$; its smash product with $H = k[B]/(B^p)$ is the $p \times p$ matrix algebra $M_p(k)$, graded by $\deg E_{i,j} = 2(j-i)$. Reducing a chain complex by the base-point module is Morita reduction to the ground field, and this recovers the reduced Khovanov–Rozansky complex. In Section 3 the same operators $B_- = -\partial/\partial x$, $B_0 = -2x\,\partial/\partial x$, $B_+ = x^2\,\partial/\partial x$ form a restricted $\mathfrak{sl}_2$-action; the simple Steinberg module $L(p-1) = \Delta(p-1) = \nabla(p-1)$ is the homology of the unknot, and dual baby Verma modules are the filtration pieces that organize the homology of a general link.

What would settle it

Compute the reduced $\mathfrak{gl}(3)$-homology (characteristic 3) of a nontrivial slice knot such as the stevedore knot $6_1$ and check whether cohomological degree zero contains the Steinberg summand; if it does not, Theorem 3.11 is false. To test Theorem 3.12, look for a two-component link with components not separated by an embedded sphere whose characteristic-$p$ homology admits an $\mathfrak{sl}_2\times\mathfrak{sl}_2$-action satisfying (3.18).

Watch

Extended reading notes

Core claim

The central discovery is that, in characteristic $p$, the local base-point algebra $A = k[x]/(x^p)$ with differential $B_- = -\partial/\partial x$ has smash product $A\#H \cong M_p(k)$, where $H = k[B]/(B^p)$. Because modules over $M_p(k)$ are direct sums of shifted column modules, reducing a chain complex at either base point gives isomorphic reduced complexes, so the reduced $\mathfrak{gl}(p)$-Khovanov–Rozansky homology is independent of the base point. The paper further argues that the $\mathfrak{sl}_2$-action on equivariant $\mathfrak{gl}(N)$-foams descends, after killing the equivariant parameters in characteristic $p$, to the finite-dimensional non-equivariant homology. With that action, the homology of any link with a thickness-one edge is filtered by dual baby Verma modules, the unknot homology is the Steinberg module, a slice knot must contain that module as a degree-zero summand, and an $\mathfrak{sl}_2\times\mathfrak{sl}_2$-action of the specified form exists exactly when the link splits.

Load-bearing premise

The topological consequences all rest on the assumption, imported from earlier work, that the $\mathfrak{sl}_2$-action on the equivariant foam homology still acts on the ordinary finite-dimensional homology after the equivariant variables are killed in characteristic $p$; the slice-knot claim additionally assumes that the cobordism map from unknot to knot homology is nonzero.

Editorial extensions

If this is right

Base-point independence follows from Morita reduction: the reduced complex is the image of a canonical functor, so no choice of base point remains in the construction.
For any link diagram with a thickness-one edge, the characteristic-$p$ homology has a dual baby Verma filtration; the Poincaré polynomial modulo $q^{2p-1}$ has parity-unimodal coefficients.
A slice knot must contain the Steinberg summand in cohomological degree zero, giving a computable sliceness obstruction in characteristic $p$.
A link splits precisely when its characteristic-$p$ homology admits an $\mathfrak{sl}_2\times\mathfrak{sl}_2$-action compatible with the two base-point variables, recovering split-link detection from the truncated-polynomial module structure.
Mirror symmetry holds at the graded module level: the homology of a link is the dual of the homology of its mirror.

Reading between the lines

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

The Morita-reduction proof should transfer to any $p$-DG link homology whose local algebra is this same contractible Frobenius algebra, so colored or other characteristic-$p$ variants may inherit base-point independence.
The slice obstruction is one-directional as stated, but it is directly computable: checking degree-zero $\mathfrak{gl}(p)$-homology for a Steinberg summand across small knots in characteristics 3 and 5 would show when the condition actually binds.
The only unchecked step in the slice theorem is the nonzero map from unknot homology; making that map explicit would convert the obstruction into a fully verified theorem and would clarify whether the converse might hold.
The parity-unimodality pattern looks like a footprint of Steenrod operations on link homology; comparing these coefficient patterns with Steenrod-square data would give an independent check of the $\mathfrak{sl}_2$-action's geometric meaning.

Share X Bluesky LinkedIn Reddit HN

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

Short version: the new Morita-theoretic proof of Shumakovitch–Wang is the real content and it is worth reading; the applications in Section 3 are a mixed bag, with one genuinely new structural consequence (Theorem 3.10) and two topological claims that lean on unproved imported machinery and an unchecked nonzero map.

What the paper does well. The p-DG algebra A = k[x]/(x^p) with the derivative differential is a standard gadget, but the observation in Lemma 2.4 that the smash product of A⊗A with H splits as A_y ⊗ M_p(k) is a clean way to see why reducing at either base point gives the same complex. The proof of Theorem 2.7 is short and mostly checkable; the only spot I would ask the authors to expand is the compatibility of the Morita reduction with the topological differential, which is asserted in a sentence rather than shown. That is a minor gap, not a flaw.

The genuinely new applications are Theorem 3.10 (parity unimodality from the baby Verma filtration) and Theorem 3.12 (sl2 × sl2 split detection, which repackages Wang's theorem in a neat way). Theorem 3.11, the Steinberg-summand slice obstruction, is the least supported: the cobordism map from unknot homology to the homology of a slice knot is asserted to be nonzero with 'one can verify', and the injectivity and summand conclusion depend on that. That needs a proof or a reference.

The larger soft spot is structural. Section 3.2 imports an sl2-action on finite-dimensional non-equivariant gl(p) homology from the authors' earlier papers [QRSW23, QRSW24]. The claim that the maximal ideal m = (E_1, ..., E_N) is preserved by the action, and that the p-th power relations survive after killing m, is cited rather than proved. If that descent is not already written down carefully in those papers, Theorems 3.10–3.12 are conditional. The stress-test note on this point holds up: preservation under B_+ is not formal from the equivariant construction. The authors should either pin down the precise statement in their prior work or add a short proof here.

I don't see a load-bearing mathematical error in the main theorem, and the citation pattern is honest: the paper credits Shumakovitch, Wang, and Rozansky where needed. The self-citations to [QRSW23, QRSW24] are legitimate if the descent really is there, but a referee should verify that before endorsing Section 3.

Verdict: this deserves a serious referee. The proof of base point independence is a genuine shortcut and the parity-unimodality application is a nice corollary. The slice-obstruction and split-detection results should be conditional on a careful check of the descent and the nonzero cobordism map. Send it to peer review; I would not desk-reject.

Referee Report

1 major / 5 minor

Summary. The paper has two main parts. In Section 2, the authors give a new proof of the Shumakovitch–Wang theorem that reduced Khovanov–Rozansky homology in characteristic p is independent of the choice of base point. The proof uses the p-DG algebra A = k[x]/(x^p) with the Shumakovitch–Wang differential B_-, identifies the smash product A # H with a matrix algebra M(p,k), and applies graded Morita reduction to show that the reduced complexes at two different base points are isomorphic. In Section 3, assuming an sl2-action on finite-dimensional gl(p)-homology in characteristic p that descends from equivariant foam evaluations after killing the equivariant parameters, the authors derive structural consequences: parity unimodality for webs with a thickness-1 edge (Theorem 3.10), a sliceness obstruction saying a slice knot must contain a Steinberg summand (Theorem 3.11), and a criterion for splitness of links in terms of an sl2 × sl2 action (Theorem 3.12).

Significance. The new proof of base point independence is a genuinely neat and self-contained contribution: it recasts the Shumakovitch–Wang result as a formal consequence of p-DG Morita reduction, and the argument in Section 2 is largely checkable and elegant. The structural results in Section 3 are potentially interesting, especially the Steinberg-summand sliceness obstruction and the splitness detection theorem, which connect sl2-representation theory in characteristic p to link invariants. The paper also gives credit to the relevant prior work and openly states the reliance on the authors' earlier construction of the sl2-action, which is a strength insofar as it delimits the scope of the present note. However, these applications are all conditional on an imported descent result that is not proved here.

major comments (1)

[Section 3.2, after Eq. (3.2)] In the proof of Theorem 2.7, after the chain of isomorphisms identifying the reduced complexes at x1 and x2, the paper says 'One can also check that the isomorphism is compatible with topological differentials by the functoriality of reductions.' This compatibility is essential: the two complexes are isomorphic as chain complexes only if the chain of isomorphisms commutes with d_T. Please spell out the verification, using explicitly the functoriality of the Morita reduction with respect to the differential d_T on C(D).

minor comments (5)

[Section 2.2, Eq. (2.8)] The notation C(D), C(D) and C ˚ , ˚ (D) is introduced in a dense paragraph; please make the homological and quantum degree gradings explicit in the display and in the definitions, especially because the Morita reduction is applied termwise in each topological degree.
[Equation (3.18) and Theorem 3.12 proof] The condition in (3.18) is written with B_{i,-}(x_j) = -δ_{i,j}, but in the proof of Theorem 3.12 it appears as B_{i,-}(x_j) = δ_{i,j}. The sign should be made consistent, or the discrepancy explained as a harmless convention.
[Examples 3.3 and 3.4] Several assertions in these examples are stated without proof: the classification of vectors annihilated by B_-, the claim that B_-^{p-1} s_{n,n} ≠ 0, and the identification with P(4i+2-2p). Since these examples are not used in the main theorems, they could be labeled as sketches, or be moved to a longer paper with full proofs.
[Proposition 3.5 proof] The sentence 'Since B_- acted upon x_b^{p-1} generates the entire A' is terse. A one-line justification using the Leibniz rule and the characteristic-p condition would help the reader verify that the generated b_+-submodule is indeed all of A.
[References] The text repeatedly cites [QRSW23, Theorem 4.3 and Section 6.1] for the sl2-action, but does not state the theorem in the text. Including the precise statement or at least a theorem number would significantly improve traceability.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the base-point-independence proof is self-contained; Section 3's use of prior self-cited results is dependence, not circularity.

full rationale

No circularity found. The main theorem (Theorem 2.7) is proved in-paper: it uses Lemma 2.4 (A^b2#H ≅ A_y ⊗ M_p(k) ⊗ M_p(k)) and a chain of natural isomorphisms to identify the two reduced complexes C_1(D) and C_2(D); the conclusion is not assumed. Proposition 2.6, Lemma 2.5, and Corollary 2.3 are likewise established from Wedderburn theory, not from the statement being proven. Section 3 is a different matter: it imports the sl2-action on equivariant gl(p)-foams and its characteristic-p descent from the authors' earlier papers [QRSW23] and [QRSW24], citing [QRSW23, Theorem 4.3 and Section 6.1]. This is self-citation and it is load-bearing for Theorems 3.10–3.12, but it is not circular by construction: the cited papers are presented as prior proofs, and the present deductions (dual baby Verma filtrations, Steinberg summand, split-link detection) do not identify their conclusions with their inputs. Theorem 3.12 additionally rests on Wang's external theorem [Wan23]. The only flagged concerns are non-circular correctness gaps: Theorem 3.11's 'one can verify that this map is nonzero' is unproved, and the preservation of the maximal ideal under the sl2-action is delegated rather than re-derived. Because these are unproved premises rather than a reduction of the claims to themselves, the circularity score is 0.

Assumptions & free parameters 1 free parameters · 5 assumptions · 0 invented entities

The central derivation is honest: it states what comes from prior work. The main external load-bearing inputs are the foam-evaluation construction of gl(p)-homology, the self-cited sl2-action and its descent in characteristic p, Bellamy-Thiel's projective-injective criterion, and Wang's split-link theorem. The only hand-chosen constants are the t1,t2 parameters of the sl2-action, which affect the representation-theoretic decompositions in the examples.

free parameters (1)

t1, t2 with t1 + t2 = 1 = arbitrary in k; examples set (1,0) or (2,2) for p = 3
Chosen by hand to define the sl2-action and crossing resolutions in (3.11); the representation-theoretic decompositions in Examples 3.3, 3.4, and 3.8 depend on these values.

assumptions (5)

domain assumption The gl(p)-Khovanov-Rozansky chain complex C(D) is a chain complex of p-DG modules over A = k[x_b]/(x_b^p), with the Shumakovitch-Wang differential commuting with the topological differential.
Invoked in Section 2.2 and attributed to [Shu14, Wan24, QRSW24, QRSW23]; not re-derived in this paper.
domain assumption The sl2-action on equivariant gl(N)-foams descends to an sl2-action on the non-equivariant gl(p)-homology after killing the equivariant parameters in characteristic p.
Imported from [QRSW23, Section 6.1] and [QRSW24]; stated in Section 3.2 and used as the foundation for all of Section 3.
standard math Bellamy-Thiel's theorem that a graded u-module with simultaneous Delta and Nabla filtrations is projective-injective.
Used in Section 3.1 before Proposition 3.1, cited as [BT18, Theorem 5.1].
domain assumption Wang's theorem that H(L) is free over k[x1,x2]/(x1^p,x2^p) if and only if the link components are separated by an embedded sphere.
Used as the key input in Theorem 3.12, cited as [Wan23, Theorem 4.13].
standard math The usual graded Morita equivalence between a matrix algebra and the ground field identifies the reduced complex with Morita reduction.
Used in Section 2.2 via the isomorphism in (2.8).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Remarks on some infinitesimal symmetries of Khovanov--Rozansky homologies in finite characteristic." pith.science (2026). https://pith.science/paper/UTKPJ4UF

@misc{pith2026241208401,
  author       = {Pith},
  title        = {Pith review of: Remarks on some infinitesimal symmetries of Khovanov--Rozansky homologies in finite characteristic},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/UTKPJ4UF}},
  note         = {Machine review of arXiv:2412.08401}
}

abstract

We give a new proof of a theorem due to Shumakovitch and Wang on base point independence of Khovanov--Rozansky homology in characteristic $p$. Some further symmetries of $\mathfrak{gl}(p)$-homology in characteristic $p$ are also discussed.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

22 extracted references · 11 canonical work pages

[2]

[Cau15] S

doi:10.1016/j.aim.2018.03.011. [Cau15] S. Cautis. Clasp technology to knot homology via the afﬁne Grassmannian. Math. Ann. , DOI 10.1007/s00208-015-1196-x,

doi:10.1016/j.aim.2018.03.011 2018
[7]

Annular Khovanov homology and knotted Schur-Weyl representations

arXiv:1505.04386, doi:10.1112/S0010437X17007540. [Kho12] M. Khovanov . One-half of the Witt algebra in categor iﬁcation of quantum groups,

work page Pith review arXiv
[11]

[QRSW21] Y

arXiv:1802.04131. [QRSW21] Y . Qi, L.-H. Robert, J. Sussan, and E. W agner. A categoriﬁcation of the colored Jones polynomial at a root of unity ,

arXiv
[12]

[QRSW23] Y

arXiv:2111.13195. [QRSW23] Y . Qi, L.-H. Robert, J. Sussan, and E. W agner. Symme tries of equivariant Khovanov-Rozansky homology

arXiv
[13]

[QRSW24] Y

arXiv:2306.10729. [QRSW24] Y . Qi, L.-H. Robert, J. Sussan, and E. W agner. Symme tries of glN -foams. Quantum T opol., online ﬁrst,

arXiv
[16]

On some $p$-differential graded link homologies

arXiv:2009.06498, doi:10.1017/fmp.2022.19. [Roz23] F . Roz. An sl2 action on link homology of T p2,k q torus links,

work page Pith review arXiv 2009
[17]

[RW20a] L.-H

arXiv:2307.01910. [RW20a] L.-H. Robert and E. W agner. A closed formula for the e valuation of foams. Quantum T opol., 11(3):411–487,

arXiv
[19]

[Shu14] A

arXiv:1801.02244, doi:10.5802/jep.124. [Shu14] A. N. Shumakovitch. T orsion of Khovanov homology . Fund. Math. , 225(1):343–364,

arXiv

Show all 22 references

[21]

[W an24] J

arXiv:2112.09110, doi:10.1112/topo.12297. [W an24] J. W ang. On slpN q link homology with mod N coefﬁcients. Quantum T opol., 15(1):87–121,

arXiv
[22]

arXiv:2111.02287, doi:10.4171/qt/194. Y . Q.: Department of Mathematics, University of Virginia, Charlo ttesville, V A 22904, USA email: yq2dw@virginia.edu L.-H. R.: Universit´ e Clermont Auvergne, LMBP , Campus des C´ ezeaux,3 place Vasarely , TSA 60026, CS 60026, 63178 Aubi`...

arXiv
[1992]

[GLW18] J

doi:10.1007/BF01232042. [GLW18] J. E. Grigsby , A. Licata, and S. W ehrli. Annular Khov anov homology and knot- ted Schur–Weyl representations. Comp. Math. , 154:459–502,

doi
[1998]

[LS22] R

doi:10.1006/aima.1998.1731. [LS22] R. Lipshitz and S. Sarkar. Khovanov homology detects split links. Amer. J. Math. , 144(6):1745– 1781,

1998
[2014]

[W an23] J

arXiv:math/0405474, doi:10.4064/fm225-1-16. [W an23] J. W ang. Split link detection for slpP q link homology in characteristic P . Journal of T opology, 16(2):806–821,

arXiv
[2015]

[EQ23] B

arXiv:1207.2074, doi:10.1007/s00208-015-1196-x . [EQ23] B. Elias and Y . Qi. Actions of sl2 on algebras appearing in categoriﬁcation. Quantum T opol., 14(4):733–806,

arXiv
[2016]

[Len98] C

arXiv:1305.1642, doi:10.4171/QT/84. [Len98] C. Lenart. The combinatorics of Steenrod operation s on the cohomology of Grassmannians. Adv. Math., 136(2):251–283,

arXiv
[2017]

[QS22] Y

arXiv:1509.00438, doi:10.1090/conm/683. [QS22] Y . Qi and J. Sussan. On some p-differential graded link homologies. Forum Math. Pi , 10:Paper No. e26, 58,

arXiv
[2018]

[BT18] G

arXiv:1304.7152, doi:10.4064/fm307-3-2017 . [BT18] G. Bellamy and U. Thiel. Highest weight theory for ﬁni te-dimensional graded algebras with triangular decomposition. Adv. Math., 330:361–419,

2017 arXiv
[2020]

[RW20b] L.-H

arXiv:1702.04140, doi:10.4171/qt/139. [RW20b] L.-H. Robert and E. W agner. Symmetric Khovanov–Roz ansky link homologies. Journal d’ ´Ecole polytechnique - Math´ ematiques, 7:573–651,

arXiv
[2021]

[GK92] S

arXiv:2103.01212. [GK92] S. Gelfand and D. Kazhdan. Examples of tensor categor ies. Invent. Math., 109(3):595–617,

arXiv
[2022]

[QRS18] H

doi:10.1353/ajm.2022.0043. [QRS18] H. Queffelec, D. E. V . Rose, and A. Sartori. Annular e valuation and link homology ,

2022
[2023]

[GHM21] E

arXiv:2103.00048, doi:10.4171/qt/181. [GHM21] E. Gorsky , M. Hogancamp, and A. Mellit. T autologica l classes and symmetry in Khovanov– Rozansky homology ,

arXiv
[2024]

REFERENCES 14 [QS17] Y

arXiv:2212.10106, doi:10.4171/QT/215. REFERENCES 14 [QS17] Y . Qi and J. Sussan. Categoriﬁcation at prime roots of unity and hopfological ﬁniteness. In Categoriﬁcation and higher representation theory , volume 683 of Contemp. Math. , pages 261–286. Amer. Math. Soc., Providence, RI,

arXiv

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [2]

[Cau15] S

doi:10.1016/j.aim.2018.03.011. [Cau15] S. Cautis. Clasp technology to knot homology via the afﬁne Grassmannian. Math. Ann. , DOI 10.1007/s00208-015-1196-x,

doi:10.1016/j.aim.2018.03.011 2018

[2] [7]

Annular Khovanov homology and knotted Schur-Weyl representations

arXiv:1505.04386, doi:10.1112/S0010437X17007540. [Kho12] M. Khovanov . One-half of the Witt algebra in categor iﬁcation of quantum groups,

work page Pith review arXiv

[3] [11]

[QRSW21] Y

arXiv:1802.04131. [QRSW21] Y . Qi, L.-H. Robert, J. Sussan, and E. W agner. A categoriﬁcation of the colored Jones polynomial at a root of unity ,

arXiv

[4] [12]

[QRSW23] Y

arXiv:2111.13195. [QRSW23] Y . Qi, L.-H. Robert, J. Sussan, and E. W agner. Symme tries of equivariant Khovanov-Rozansky homology

arXiv

[5] [13]

[QRSW24] Y

arXiv:2306.10729. [QRSW24] Y . Qi, L.-H. Robert, J. Sussan, and E. W agner. Symme tries of glN -foams. Quantum T opol., online ﬁrst,

arXiv

[6] [16]

On some $p$-differential graded link homologies

arXiv:2009.06498, doi:10.1017/fmp.2022.19. [Roz23] F . Roz. An sl2 action on link homology of T p2,k q torus links,

work page Pith review arXiv 2009

[7] [17]

[RW20a] L.-H

arXiv:2307.01910. [RW20a] L.-H. Robert and E. W agner. A closed formula for the e valuation of foams. Quantum T opol., 11(3):411–487,

arXiv

[8] [19]

[Shu14] A

arXiv:1801.02244, doi:10.5802/jep.124. [Shu14] A. N. Shumakovitch. T orsion of Khovanov homology . Fund. Math. , 225(1):343–364,

arXiv