REVIEW 4 major objections 4 minor 17 references
Knot invariants from representations of braids by automorphisms of a free group
T0 review · 4 major / 4 minor · reviewed 2026-08-07 · deepseek-v4-flash
Pith's one-line read This paper argues that a Fox-derivative construction on braid automorphisms recovers the Alexander polynomial from the Burau representation, and that applying the same recipe to Wada's representation yields invariants that are…
desk verdict Modest but honest paper: Wada invariants are likely Alexander(-1), and the real value is the explicit Markov-invariance computation for Wada's representation. 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 central object is the Jacobian matrix of Fox derivatives $J_\varphi$ of an automorphism $\varphi$ of a free group, together with the chain rule $J_{\varphi\psi} = J^\psi_\varphi J_\psi$, where $J^\psi_\varphi$ means applying $\psi$ to the entries of $J_\varphi$. To turn this into a matrix representation of a braid group, one chooses an abelianization $\alpha$ of the free group such that $J^{\psi\alpha}_\varphi = J^\alpha_\varphi$ for all automorphisms in the image; for Artin's representation this gives the Burau representation, and for Wada's representation it gives $2\times 2$ cells of the form $\begin{pmatrix}1+t & t^2\\ -1 & 1-t\end{pmatrix}$ for odd indices and the $t^{-1}$ analogue for even indices. The invariant is the chain of elementary ideals generated by the minors of $J^\alpha_\varphi - I$, normalized by clearing negative powers and fixing sign.
What would settle it
Compute the Wada invariant for the $5_2$ knot, whose Alexander polynomial is $2t^2-3t+2$; the conjecture predicts the normalized Wada polynomial is $7$, so any other value refutes the central claim. Alternatively, symbolically test the compatibility identity for $\varphi=\sigma_1$ and $\psi=\sigma_2$ in Wada's representation with the parity abelianization, since a single failure would make the Wada polynomial ill-defined.
Extended reading notes
Core claim
The central claim is that one uniform procedure produces knot invariants from any representation of braid groups by automorphisms of a free group: compute the Fox-derivative Jacobian matrix $J_\varphi$, choose an abelianization $\alpha$ of the free group so that the chain-rule compatibility $J^{\psi\alpha}_\varphi = J^\alpha_\varphi$ holds, and then take the chain of elementary ideals of $J^\alpha_\varphi - I$, which the paper proves is Markov-invariant. For Artin's representation with $\alpha(x_i)=t$, this is the Burau representation and the resulting invariant is the Alexander polynomial. For Wada's representation with $\alpha(x_i)=t$ for odd $i$ and $t^{-1}$ for even $i$, the paper obtains the Wada polynomial and conjectures that this polynomial is the specialization of the Alexander polynomial at $t=-1$, supported by examples including the unknot, Hopf link, trefoil, torus knots, figure-eight, square knot, and granny knot.
Load-bearing premise
The load-bearing premise is that the parity abelianization $\alpha(x_i)=t$ for odd $i$ and $t^{-1}$ for even $i$ satisfies $J^{\psi\alpha}_\varphi = J^\alpha_\varphi$ for every $\varphi,\psi$ in Wada's image; if that identity fails, the Wada polynomial is not a well-defined link invariant.
Editorial extensions
If this is right
- The chain of elementary ideals of $J^\alpha_\varphi - I$ is invariant under Markov moves for both the Artin and Wada representations, so each produces isotopy invariants of the closed braid.
- For the Artin/Burau case, the construction recovers the classical Alexander polynomial of the knot or link.
- The computed Wada polynomials match $\Delta_K(-1)$ in every example: $1$ for the unknot, $2$ for the Hopf link, $3$ for the trefoil, $k$ for the $(2,k)$ torus knot or link, $5$ for the figure-eight, and $9$ for both the square and granny knots.
- Conjecture 1 states that the Wada polynomial is always the Alexander polynomial evaluated at $t=-1$; if true, Wada-based invariants do not provide new knot invariants beyond this specialization.
- The paper's approach also gives a more delicate analysis of how Burau matrices change under Markov moves, potentially distinguishing knots with isomorphic fundamental groups in ways Alexander matrices cannot.
Reading between the lines
- If Conjecture 1 is true, the Wada polynomial reduces to a single integer $\Delta_K(-1)$, so it cannot detect chirality, orientation, or any property invisible to the Alexander polynomial at $t=-1$.
- The paper does not verify the chain-rule compatibility condition for Wada's parity abelianization; until that identity is checked for all pairs of automorphisms in the representation's image, the Wada polynomial should be regarded as an empirically motivated invariant whose well-definedness rests on an unproven assumption.
- The two-variable representation in Section 7, whose leading invariant appears to be $\mathrm{AL}_\beta(st)$, suggests that Burau, Wada, and the two-variable invariants may all sit in one family of specializations; one could test whether the elementary ideal of $M_\beta - I$ equals the Alexander ideal under the substitution $s=t$.
- A natural extension would be to compute Wada polynomials for knots outside the listed examples, such as two-bridge or pretzel knots with known Alexander polynomials, and compare the normalized invariant to $\Delta_K(-1)$; any mismatch would refute Conjecture 1.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper proposes an approach to link invariants by taking Fox Jacobian matrices of automorphism representations of braid groups, applying an abelianization homomorphism α, and then taking elementary ideals of minors of Jβ−I. For the Artin representation this recovers Alexander polynomials, and the Markov-invariance of the ideal chains is shown in Section 3.1. The paper then applies the same scheme to Wada's representation, using a parity abelianization α(x_i)=t for odd i and t^{-1} for even i. Section 5 establishes Markov invariance of the elementary ideals for this parity abelianization, and Section 6 computes Wada polynomials for several knots and links (unknot 1, Hopf 2, trefoil 3, figure-eight 5, square and granny 9), which match the corresponding Alexander polynomials specialized at t=−1. The paper states Conjecture 1 that this equality holds in general, with a footnote referring to [17] and [14] for a proof, and Section 7 supplements the discussion with a two-variable local representation whose 'leading invariant' is claimed to equal ALβ(st).
Significance. If Conjecture 1 were established, the paper would show that Wada's representation does not produce new invariants beyond the Alexander(−1) specialization; the explicit computations and the Markov-invariance argument for the parity abelianization are useful and reproducible. The paper is honest in labelling the main comparison as a conjecture, and the numerical evidence is consistent. The computations of the examples are straightforward and check out, and the chain-rule compatibility of the parity abelianization, once verified, is a neat observation. However, the paper's central comparative claim is not proved in the text, and two technical justifications are missing, so the contribution is more a computational and conjectural note than a proved theorem.
major comments (4)
- [Section 5, definition of α] The paragraph defining α(x_i)=t for odd i and t^{-1} for even i asserts, but does not demonstrate, the chain-rule compatibility condition J^{ψα}_φ = J^α_φ for all φ,ψ in the image of Wada's representation. This condition is load-bearing: every Wada invariant in Sections 5 and 6 is computed from the resulting matrix representation, and if it failed the 'Wada polynomial' would not be a well-defined link invariant. The verification is short and should be included: for odd i, α(σ_i(x_i))=α(x_i^2x_{i+1})=t^2t^{-1}=t=α(x_i) and α(σ_i(x_{i+1}))=α(x_{i+1}^{-1}x_i^{-1}x_{i+1})=t t^{-1} t^{-1}=t^{-1}=α(x_{i+1}), and the even-i case is symmetric.
- [Section 5.1] The definition of the Wada polynomial as the generator of the smallest nonzero E_k presupposes that the full n×n determinant of J^α_φ−I is identically zero for Wada's representation; otherwise a braid with nonvanishing determinant would produce a degree-0 invariant, and Conjecture 1 would not be comparable to the standard Alexander(−1) normalization, which is extracted from codimension-one minors. The text neither proves nor cites this vanishing. A short proof is available from the Fox derivative identity ∑_i ∂_i(φ(x_j))^α (α(x_i)−1)=α(x_j)−1, which shows that the nonzero vector (α(x_1)−1,...,α(x_n)−1)^T is a left eigenvector of J^α_φ with eigenvalue 1, so det(J^α_φ−I)=0. Adding this argument (or a reference for it) would remove a gap in the definition.
- [Section 6.7 and footnote 1] Conjecture 1 is the central comparative claim of the paper, but it is not proved in the text; the footnote says a proof can be recovered from [17] and [14], yet no proof or detailed derivation is given. Because the abstract advertises a comparison with Alexander polynomials, the manuscript should either give the proof, if it is available from those references, or explicitly state that the comparison remains conjectural in this paper.
- [Section 7] The statement that the 'leading invariant' of the two-variable representation is AL_β(st) is supported only by the Hopf link and trefoil examples and is phrased as 'seems to be'. If this is intended as a theorem, a proof (or a precise definition of 'leading invariant' in the non-PID ring Z[s^±1,t^±1]) is needed; if it is intended as a conjecture, it should be labeled as such. As written, the section's conclusion that this representation 'does not seem to yield new invariants' is not established.
minor comments (4)
- [Section 5] The matrix displays in the stabilization computations would be much easier to follow if the block structure were set up explicitly, separating the n×n block of β from the newly added strand.
- [Section 4] The discussion of the right and left trefoils is explicitly speculative; the example of σ_1^{-2} versus σ_1^2 shows that the proposed obstruction cannot be a simple sign obstruction, but the paragraph could clarify that this is a motivating remark rather than a result.
- [Throughout] There are numerous typographical and OCR artifacts, including the running header 'KNOT INV ARIANTS' and missing spaces in displayed formulas; these should be cleaned up before publication.
- [Section 7] The term 'leading invariant' is used without a formal definition; the paper should state precisely which ideal or generator is meant, especially because the elementary ideals in Z[s^±1,t^±1] need not be principal.
Circularity Check
No significant circularity: Wada invariants are computed directly from a matrix representation and compared, not derived, against Alexander polynomials.
full rationale
The paper's derivation chain goes from Wada's automorphism representation to Jacobian matrices, applies an explicit parity abelianization, forms J−I, and takes elementary ideals of minors. Every Wada polynomial in Sections 5–6 is obtained by direct computation of these minors; no parameter is fitted to the Alexander polynomials that appear only in the later comparisons and in Conjecture 1. No equation in the paper identifies the Wada matrix with the Alexander matrix by construction, and the conjecture is explicitly labeled as a conjecture rather than used to compute the invariants. The footnote crediting Ito and citing [17] and [14] defers the proof to external literature; it is not a self-citation. The main technical gap is Section 5's assertion of the parity abelianization 'a little more tricky' without explicitly verifying the chain-rule compatibility condition J^{ψα}_φ=J^α_φ for all φ,ψ in the image of Wada's representation; however, an omitted verification is a correctness risk, not a circular step. The self-citations ([12], [15], [16]) are not load-bearing, and the paper itself says faithfulness [16] 'does not play a role in the present paper.' Thus no circular step can be exhibited from the text, and the appropriate finding is no significant circularity.
Assumptions & free parameters
assumptions (4)
- standard math Markov's theorem: two braids close to isotopic links iff related by Markov moves.
- ad hoc to paper The parity abelianization α(x_i)=t for odd i, t^{-1} for even i, makes the Jacobian map a representation of B_n for Wada's representation.
- domain assumption Wada's invariant equals the Alexander polynomial at t = -1 (Conjecture 1), with proof said to be in [17] and [14].
- ad hoc to paper The two-variable representation's leading invariant is AL(st).
Cite this review
Pith. "Pith review of Knot invariants from representations of braids by automorphisms of a free group." pith.science (2026). https://pith.science/paper/VGULABAO
@misc{pith2026250522403,
author = {Pith},
title = {Pith review of: Knot invariants from representations of braids by automorphisms of a free group},
year = {2026},
howpublished = {\url{https://pith.science/paper/VGULABAO}},
note = {Machine review of arXiv:2505.22403}
}
read the original abstract
We describe an alternative way of computing Alexander polynomials of knots/links, based on the Artin representation of the corresponding braids by automorphisms of a free group. Then we apply the same method to other representations of braid groups discovered by Wada and compare the corresponding isotopic invariants to Alexander polynomials.
Reference graph
Works this paper leans on
-
[17]
Wada,Group invariants of links, Topology31(1992), 399–406
M. Wada,Group invariants of links, Topology31(1992), 399–406. Department of Mathematics, The City College of New York, New York, NY 10031 Email address:shpilrain@yahoo.com
work page 1992
-
[14]
Sakuma,A note on Wada’s group invariants of links, Proc
M. Sakuma,A note on Wada’s group invariants of links, Proc. Japan Acad. Ser. A Math. Sci.67 (1991), 176–177
work page 1991
-
[1]
Bigelow,The Burau representation is not faithful forn= 5, Geom
S. Bigelow,The Burau representation is not faithful forn= 5, Geom. Topol.3(1999), 397–404 (electronic)
work page 1999
-
[2]
Bigelow,Braid groups are linear, J
S. Bigelow,Braid groups are linear, J. Amer. Math. Soc.14(2001), 471–486
work page 2001
-
[3]
J. S. Birman,An inverse function theorem for free groups,Proc. Amer. Math. Soc.,41(1973), 634– 638
work page 1973
-
[4]
J. S. Birman,Braids, links and mapping class groups, Ann. Math. Studies82, Princeton Univ. Press, 1974
work page 1974
-
[5]
Burau, ¨Uber Zopfgruppen und gleichsinnig verdrillte Verkettungen, Abh
W. Burau, ¨Uber Zopfgruppen und gleichsinnig verdrillte Verkettungen, Abh. Math. Sem. Univ. Ham- burg11(1936), 179–186
work page 1936
-
[6]
R. H. Crowell, R. H. Fox,Introduction to knot theory, Grad. Texts in Math., No.57. Springer-Verlag, New York-Heidelberg, 1977
work page 1977
Show all 17 references
-
[7]
R. H. Fox,Free differential calculus. I. Derivation in the free group ring, Ann. Math. (2),57(1953), 547–560
1953
-
[8]
Krammer,Braid groups are linear, Ann
D. Krammer,Braid groups are linear, Ann. of Math. (2)155(2002), 131–156
2002
-
[9]
D. D. Long, M. Paton,The Burau representation is not faithful forn≥6, Topology32(1993), 439–447
1993
-
[10]
Magnus, A
W. Magnus, A. Peluso,On a theorem of V. I. Arnold, Comm. Pure Appl. Math.22(1969), 683–692
1969
-
[11]
J. P. Mutanguha,Irreducibility of a free group endomorphism is a mapping torus invariant, Comment. Math. Helv.96(2021), 47–63
2021
-
[12]
Olshanskii, V
A. Olshanskii, V. Shpilrain,Linear average-case complexity of algorithmic problems in groups,J. Algebra668(2025), 390–419
2025
-
[13]
Pourkia,A basicn-dimensional representation of Artin braid groupB n, and a general Burau representation, Mathematics and Statistics10(2022), 145–152
A. Pourkia,A basicn-dimensional representation of Artin braid groupB n, and a general Burau representation, Mathematics and Statistics10(2022), 145–152
2022
-
[15]
Shpilrain,On monomorphisms of free groups, Arch
V. Shpilrain,On monomorphisms of free groups, Arch. Math.64(1995), 465–470
1995
-
[16]
Shpilrain,Representing braids by automorphisms, Internat
V. Shpilrain,Representing braids by automorphisms, Internat. J. Algebra and Comput.11(2001), 773–778
2001
Reviewed August 7, 2026 · model on record in the stance chip above.
Discussion (0). Sign in to comment.