🔢 mathematics

The HOMFLY-PT polynomial and HZ factorisation

This paper introduces the HZ factorisation of the HOMFLY-PT polynomial, demonstrating how full and Jucys-Murphy twists generate infinite families of hyperbolic knots with polynomials encoded by integer exponents linked to Khovanov homology, while establishing a new relation between HOMFLY-PT and Kauffman polynomials that implies the vanishing of two-crosscap BPS invariants in topological string theory.

Original authors: Andreani Petrou, Shinobu Hikami

Published 2026-07-16

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Andreani Petrou, Shinobu Hikami

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are a detective trying to solve the mystery of a tangled ball of string. In the world of mathematics and physics, these tangled strings are called knots, and they aren't just for sailing or shoelaces; they are fundamental shapes that help scientists understand the very fabric of space and time. To crack the code of a knot, mathematicians use special "magic formulas" called polynomials. Think of these formulas as a unique fingerprint or a DNA test for a knot. If you twist or turn the knot, the formula changes in a predictable way, allowing you to tell if two knots are actually the same shape just by looking at their numbers.

One of the most famous of these fingerprints is the HOMFLY–PT polynomial. It's a complex recipe involving two variables that can describe almost any knot. However, calculating this recipe for complicated knots is like trying to solve a Sudoku puzzle with a million squares—it gets messy and hard to read. Recently, scientists discovered a clever trick called the Harer–Zagier (HZ) transform. You can think of this transform as a high-tech lens or a magic filter. When you look at a knot's messy polynomial through this lens, the chaos often simplifies into a neat, organized fraction. Sometimes, this fraction breaks down into a simple list of building blocks, a property the researchers call "factorisation." Why does this matter? Because when a knot's fingerprint simplifies this way, it hints at a deep, hidden order in the universe, connecting the shape of the knot to the behavior of tiny particles and strings in theoretical physics.

This paper is a treasure hunt for knots that play nice with this magic lens. The authors, Andrei Petrou and Shinobu Hikami, set out to find new families of knots that, when viewed through the HZ transform, reveal these beautiful, simple patterns. They didn't just look at the knots that were already known to be simple (like the classic "torus knots" that look like pretzels wrapped around a donut); they went hunting for more exotic, "hyperbolic" knots that twist and turn in wilder ways.

The researchers discovered that they could generate infinite families of these special knots by performing specific twisting operations. Imagine taking a knot and giving it a full spin (a "full twist") or a specific kind of twist named after mathematicians Jucys and Murphy. They found that if you start with a knot that already has this "factorised" property and apply these twists, the resulting knot keeps the property. It's like a genetic trait that never fades, no matter how many times you twist the DNA. This allowed them to create endless lists of new knots that are mathematically "clean" and easy to describe.

But the paper goes deeper than just finding these knots. The authors proved a surprising connection between two different types of knot fingerprints: the HOMFLY–PT polynomial and the Kauffman polynomial. For a long time, mathematicians knew these two formulas were related for the simple "donut" knots, but they suspected the link broke for the wilder, hyperbolic knots. This paper proves that for the new families they discovered, the link still holds true. It's as if they found a secret handshake that works even for the most chaotic knots, provided those knots have this special "factorised" structure.

The authors also suggest a fascinating reason why this happens, linking it to a concept from string theory called BPS invariants. In the language of string theory, these invariants count the number of special, stable states a system can be in. The paper suggests that for these specific knots, a particular type of state (involving "two crosscaps," which are like weird, twisted surfaces in higher dimensions) simply vanishes. It's as if the universe decides that for these knots, certain complex possibilities are forbidden, leaving only the simple, clean ones.

While the paper provides solid proofs for specific families of knots, it also offers a bold guess, or conjecture, for the rest of the world. They propose that any knot that has this neat, factorised HZ transform must also have this special relationship between its HOMFLY–PT and Kauffman fingerprints. Conversely, if a knot doesn't have this relationship, it probably won't have the neat factorisation either. They tested this idea on knots with up to 12 crossings and found no exceptions, but they admit that for knots with even more twists, the mystery remains open.

Finally, the paper connects these mathematical patterns to a field called Khovanov homology, which is like a 3D version of the knot's fingerprint. They found that the numbers in the neat HZ formula correspond directly to the "layers" or "grades" in this 3D structure. It's like finding that the ingredients in a simple cake recipe perfectly match the layers of a complex, multi-tiered wedding cake. This suggests that the "magic lens" of the HZ transform isn't just a mathematical trick; it's revealing the actual, underlying architecture of the knot's shape.

In short, this paper is a map to a hidden garden of knots. It shows us how to grow infinite families of them, proves they share a secret language with other mathematical tools, and hints that their simplicity is a sign of a deeper, silent order in the physics of the universe. While some questions remain for future explorers, the authors have successfully shown us that even in the most tangled messes, there is a pattern waiting to be found.

Technical Summary: The HOMFLY–PT Polynomial and HZ Factorisation

Problem Statement
The paper addresses the structural properties of the HOMFLY–PT polynomial, a fundamental knot invariant defined via skein relations and linked to $SU(N)$ Chern–Simons gauge theory. Specifically, the authors investigate the conditions under which the Harer–Zagier (HZ) transform of the HOMFLY–PT polynomial admits a "factorised" form. The HZ transform maps the polynomial into a rational function of parameters $q$ and $\lambda$ . While it is known that torus knots $T(m,n)$ possess factorised HZ transforms, the existence of such factorisation for hyperbolic knots and links was an open question. Furthermore, the paper seeks to establish a rigorous connection between this factorisation property, the relationship between HOMFLY–PT and Kauffman polynomials, and the vanishing of specific BPS invariants in topological string theory.

Methodology
The authors employ a combination of algebraic knot theory, combinatorial braid operations, and topological string theory concepts:

HZ Transform Analysis: They define the HZ transform $Z(K; \lambda, q)$ as a generating function of the unnormalised HOMFLY–PT polynomial $\bar{H}(K; q^N, q)$ . They analyze the conditions under which $Z$ can be expressed as a ratio of products of monomials $(1 \pm \lambda q^\beta)$ .
Braid Operations: To generate infinite families of knots, the authors utilize two specific braid operations: full twists ( $F_m$ ) and Jucys–Murphy twists ( $E_m$ ). They demonstrate that applying these operations indefinitely to specific seed knots preserves the factorisability of the HZ transform.
Inverse Transform: The authors derive closed-form formulas for the HOMFLY–PT and Alexander polynomials of these factorisable families by applying an inverse HZ transform (via contour integrals) to the rational function form of $Z$ .
Polynomial Relations: They test the validity of a specific relation between the HOMFLY–PT polynomial $H$ and the Dubrovnik version of the Kauffman polynomial $\tilde{KF}$ , previously known only for torus knots.
Homological Interpretation: The paper correlates the exponents in the factorised HZ transform with the gradings ( $j$ and $r$ ) found in Khovanov homology tables, specifically examining the Euler characteristics of the homology groups.
BPS Invariant Calculation: Using the gauge/string duality, they relate the HOMFLY–PT/Kauffman relation to the vanishing of two-crosscap ( $c=2$ ) BPS invariants ( $\hat{N}^{c=2}_{g,Q}$ ) in topological string theory.

Key Contributions and Results

Discovery of Infinite Hyperbolic Families: The authors identify and construct infinite families of hyperbolic knots and links that admit a factorised HZ transform. These families are generated by applying full twists ( $F_m$ ) and Jucys–Murphy twists ( $E_m$ ) to seed knots such as the pretzel knots $P(2, -3, \pm(2j+1))$ , the knot $5_2$ , and specific twisted torus knots.
Closed Formulas: They provide explicit, compact formulas (Eq. 43 and Eq. 56) for the HOMFLY–PT and Alexander polynomials of these infinite families, expressed entirely in terms of the HZ parameters (strand number $m$ and integer exponents $\alpha_i, \beta_i$ ).
HOMFLY–PT/Kauffman Relation: The paper proves (Theorem 4.1) that for the identified hyperbolic families, the HOMFLY–PT polynomial $H$ is related to the Kauffman polynomial $\tilde{KF}$ by the equation:
$H(K) = \tilde{KF}(K) := \tilde{KF}_{\text{even}} - \frac{z}{a-a^{-1}}\tilde{KF}_{\text{odd}}$
This relation, previously established only for torus knots, is shown to hold for these new hyperbolic families.
Conjecture on Factorisability: Based on exhaustive checks of knots up to 12 crossings, the authors conjecture (Conjecture 4.1) that the HZ transform of a knot is factorisable if and only if the HOMFLY–PT and Kauffman polynomials satisfy the relation above.
BPS Invariant Vanishing: The authors demonstrate that the HOMFLY–PT/Kauffman relation is equivalent to the vanishing of the two-crosscap BPS invariants ( $\hat{N}^{c=2}_{g,Q} = 0$ ). They verify that this vanishing occurs for all HZ-factorisable knots and links examined.
Homological Connection: The paper establishes a direct correspondence between the HZ exponents and the $j$ -grading in Khovanov homology. For factorisable knots, the HZ exponents correspond to the rows of the Khovanov table where the Euler characteristic is non-zero ( $\pm 1$ ). The numerator exponents correspond to odd $r$ -grading (negative Euler characteristic), and denominator exponents to even $r$ -grading (positive Euler characteristic).
Link Analysis: For 2-component links, a modified relation involving the linking number is proposed (Eq. 90). The authors note that while most HZ-factorisable links satisfy this, the link $L_{10n42}$ is an exception that admits factorisation but fails the polynomial relation, suggesting the correspondence is not strictly 1-to-1 for links.

Significance and Claims
The paper claims to extend the known scope of HZ factorisability from torus knots to infinite families of hyperbolic knots and links. By proving the HOMFLY–PT/Kauffman relation for these families, the authors provide a potential necessary and sufficient criterion for HZ factorisability.

The significance of these findings is framed in two contexts:

Mathematical: It offers a new structural insight into knot polynomials, suggesting that the factorisation of the HZ transform is deeply linked to the algebraic relationship between $SU(N)$ and $SO(N+1)$ invariants. The connection to Khovanov homology suggests that the "simplification" seen in the HZ transform reflects a specific, sparse structure in the underlying homology (specifically, the presence of "separable" tables and specific torsion patterns).
Physical: The vanishing of the two-crosscap BPS invariants implies that for these specific knots, the topological string amplitudes receive no contributions from unoriented worldsheets with two crosscaps (Klein bottles). The authors speculate that this vanishing may be a manifestation of supersymmetry within the HZ transform, viewed as a Hilbert or Poincaré series.

The authors remain modest regarding the generalizability of their results, noting that the HOMFLY–PT/Kauffman relation does not hold for all links (e.g., $L_{10n42}$ ) and that the homological interpretation for unoriented BPS invariants remains an open problem. They propose that the conjectured equivalence between factorisability and the polynomial relation requires further rigorous investigation, particularly for non-factorisable knots and links with more than two components.

More like this