e-Positivity of vertical strip LLT polynomials

Michele D’AdderioJournal of Combinatorial Theory Series A 2020 · DOI · arXiv

TL;DR

This paper proves a key conjecture about the e-positivity of a specific class of symmetric functions called vertical strip LLT polynomials. This result has broad implications, confirming several existing conjectures in the field of algebraic combinatorics and providing a new algorithmic approach to understanding these functions.

What Problem Does This Paper Address?

The paper tackles a problem in the realm of symmetric functions, which are fundamental objects in combinatorics and representation theory. Specifically, it focuses on a property called "e-positivity." A symmetric function is considered e-positive if its expansion in terms of elementary symmetric functions has non-negative coefficients. This property is desirable because it often signifies a deeper combinatorial meaning, such as counting objects with certain characteristics.

The authors are interested in the e-positivity of a particular type of symmetric function known as LLT polynomials, especially when they are "vertical strip" LLT polynomials. The e-positivity of these vertical strip LLT polynomials, when evaluated at $q+1$, was conjectured in several previous works. Proving this conjecture would not only confirm these specific conjectures but also imply many other related e-positivity conjectures. The challenge lies in finding a systematic way to prove this property for a broad class of these polynomials.

Methods

The core of the paper's method relies on a powerful result by Carlsson and Mellit, which establishes a connection between vertical strip LLT polynomials and operators acting on a specific algebraic structure called the "Dyck path algebra." This theorem states that a vertical strip LLT polynomial can be obtained by applying a sequence of these Dyck path algebra operators to the constant function 1.

The authors then leverage this theorem by:

  • Defining Dyck Path Algebra Operators: They introduce and utilize operators ($T_i$, $d_+$, $d_-$, $\phi$) that act on polynomial spaces and are inspired by the structure of Dyck paths. These operators have specific commutation relations and identities that are crucial for the proof.
  • Associating Paths to LLT Diagrams: A key step is the construction of "Schröder paths" that are associated with vertical strip LLT diagrams. These paths are then encoded using a sequence of steps (north, east, northeast), and each path corresponds to a specific composition of the Dyck path algebra operators.
  • Proving Operator Identities: The paper meticulously proves several identities involving these operators. These identities are essential for simplifying complex operator compositions.
  • Algorithmic Proof: The proof of the main theorem is constructive. It shows that any operator corresponding to a vertical strip LLT polynomial can be systematically reduced to a linear combination of simpler operators. These simpler operators, when applied to the constant 1, directly yield expansions in the elementary symmetric function basis with non-negative coefficients (in $q$ and $t$).

Datasets

Not discussed in the paper.

Key Insights and Ideas

  • Bridging Combinatorial Objects and Algebra: The paper beautifully connects geometric objects (vertical strip LLT diagrams) to algebraic structures (Dyck path algebra operators) through the intermediary of paths. This connection is the linchpin of the proof.
  • Operator Composition as Multiplication: A significant byproduct of the proof is the demonstration that the compositions of Dyck path algebra operators corresponding to LLT polynomials are, in fact, multiplication operators. This is a non-trivial insight into the structure of these operators.
  • Algorithmic Expansion: The proof provides a concrete algorithm for expanding vertical strip LLT polynomials in the elementary symmetric function basis. This is valuable for both theoretical understanding and potential computational applications.
  • Generalization of e-Positivity: The result confirms a broad class of e-positivity conjectures, unifying several previously disparate observations and conjectures in the field.

Key Results

The main result is the proof of the e-positivity of $G_v[X; q+1]$ when $G_v[X; q]$ is a vertical strip LLT polynomial. This means that the coefficients in the expansion of $G_v[X; q+1]$ in the elementary symmetric function basis are non-negative integers. This result is achieved through a constructive proof that relies on the properties of Dyck path algebra operators. As a consequence, several previously conjectured e-positivities are proven.

Limitations and Future Work

  • Specific Class of LLT Polynomials: The paper focuses on "vertical strip" LLT polynomials. While this is a significant class, it does not cover all LLT polynomials. The authors acknowledge that not all e-positivities observed in previous works follow from their results.
  • Explicit Formulas: The paper does not provide explicit formulas for the expansion of $G_v[X; q+1]$ in the elementary symmetric function basis, although it suggests that the algorithmic nature of their proof could lead to such formulas with further investigation.
  • Connection to Delta Conjecture: Corollary 6.1 mentions that some conjectured e-positivities are conditioned on the validity of the Delta conjecture. The paper proves these conjectures, but the connection to the Delta conjecture itself is not elaborated upon.
  • Commutation of Path Operators: An interesting consequence noted is that path operators commute, which is not immediately obvious from their definitions. This suggests avenues for further exploration into the structure of these operators.

Summary updated

BibTeX
@article{dadderio2020e,
  author = {Michele D’Adderio},
  title = {{e-Positivity of vertical strip LLT polynomials}},
  journal = {Journal of Combinatorial Theory Series A},
  year = {2020},
  doi = {10.1016/j.jcta.2020.105212},
}
Figure