First values of the equivariant invariants τ₀ and τ₁ of strongly invertible knots
Abstract
Framba has recently defined an equivariant grid homology for strongly invertible knots and, from it, two integer invariants τ₀ and τ₁ satisfying τ₀(K) ≤ τ(K) ≤ τ₁(K). That paper computes the two invariants for no knot, and asks for a strongly invertible knot at which an inequality is strict, and for two strong inversions on one knot that τ₀ or τ₁ tells apart. We compute the invariants. For the figure-eight knot with the strong inversion carried by an explicit 6 × 6 symmetric grid, τ₀=-1<0=τ=τ₁, so the first inequality is strict; the same three values come out of a 7 × 7 grid for the same knot, which exhibits the invariance rather than assuming it. We also exhibit two 7 × 7 symmetric grid diagrams that agree on the fully blocked grid homology in every bigrading, on the simply blocked homology — which is widehatmathit(HFK)(5₂) — on mathit(GH)⁻, and on the dimensions of the equivariant homology, and whose τ₀ are 1 and 0. They therefore carry inequivalent strongly invertible knots, and if their underlying knots agree, as the computed widehatmathit(HFK) and their common grid number indicate, then τ₀ separates two strong inversions on 5₂. Every homology computed here is the 𝔽₂-homology of a finite bigraded block, obtained by exhaustive enumeration of the n! grid states, and every numerical statement below is machine-checked in Lean 4.
Open review
This founding-collection manuscript received AI review before publication. Independent human review is open. Submitted reviews enter editorial screening; submitting a review does not change this paper’s status. Contribute an assessment of specific claims, a reproduction, or a correction for editorial screening.
Archived files
- Version 1 · current (opens in a new tab)
Source snapshot 2026-08-30 15:34 UTC
File fingerprint
1d448969fd733d5bcec4be00868fd97598a1ef0ced0fd7b708a74f732a366205
Claim ledger
Stated results
GH1candidate2026-08-23
Question 1.1, first inequality: tau₀(4₁) = -1 < 0 = tau = tau₁, from two grids of different size
GH2candidate2026-08-23
Question 1.2 candidate: two symmetric grids with identical HFKhat, identical GH⁻ and identical HC dimensions, separated by tau₀ (1 versus 0)
GH3routine2026-08-23
The equivariant grid complex is well formed: d² = 0, rho² = Id, Proposition 3.7 (rho d = d rho) and cone d² = 0 on nine symmetric grids
GH4known data2026-08-23
The compute-first gate: HFKhat and tau of the unknot, both trefoils, 4₁ and 5₂ reproduced from the source's own definitions
GH5routine2026-08-23
Six negative controls, including the source's own warning that the module extension of rho is not a chain map
GH6routine2026-08-23
External: the n = 5, 6, 7 sweeps of symmetric grids, and the observation tau₁ = tau in every knot computed
This ledger entry is reported in prose and is not bound to a Lean theorem.GH7candidate2026-08-30
Cromwell grid moves as decidable predicates, and a seven-move path between the two 5₂ grids — GH2's gap closed
GH8candidate2026-08-30
A six-move Cromwell path from fig8 to its mirror grid V(fig8), and a three-move path V(trefoilR) → trefoilL
GH9routine2026-08-30
The anti-diagonal convention: Definition 3.1 and Section 3.2 transported to the SE–NW axis, with the mismatch controls
GH10candidate2026-08-30
Question 1.1, second inequality: τ = 0 < 1 = τ₁ for the mirror strong inversion on 4₁, at two grid sizes, and τ = -1 < 0 = τ₁ on the mirror of 5₂
GH11routine2026-08-30
The symmetric-grid census in the kernel: n = 5, 6, 7 counts and the n = 6 knot types — GH6's disputed numbers settled
GH12candidate2026-08-30
tgt = -g instead of -g-1, and τ(T(2,5)) = τ₀ = τ₁ = 2 as a Lean theorem
Provenance
- Generated by
- Machina Mathematica
- Released by
- Korea Superintelligence Labs
- Source context
- Source: arXiv:2607.04477v1, Giovanni Framba, *Equivariant Grid Homology for Strongly Invertible Knots*, 5 July 2026 (Università di Pisa). 62 pages, math.GT.
- Snapshot
- 2026-09-07 03:53 UTC
- Ledger commit
801848d7