Excited Young Diagrams in Algebraic Combinatorics

• Excited Young diagrams are combinatorial objects defined by iterative local moves on partitions, yielding positive formulas for tableau counts and Hilbert series.
• They are in bijection with flagged semistandard tableaux, nonintersecting lattice paths, and lozenge tilings, facilitating diverse enumeration methods.
• They extend to applications in equivariant K-theory and vertex models, connecting algebraic geometry with integrable systems and q-analogues.

An excited Young diagram is a combinatorial object central to formulas for enumerative invariants and structure constants in algebraic combinatorics, representation theory, and algebraic geometry. These diagrams encode the action of local moves (excitations) on a given inner partition within an outer partition or skew shape, providing manifestly positive formulas for standard Young tableau counts, skew Schur specializations, equivariant $K$-theory restrictions, and Hilbert series. Excited Young diagrams are in bijection with flagged semistandard tableaux, sets of nonintersecting lattice paths, and lozenge tilings, and have foundational connections to vertex models and integrable systems.

1. Definition and Basic Properties

Let $\lambda$ and $\mu$ be partitions with $\mu \subseteq \lambda$, and let $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$ denote the Young diagram of $\lambda$. The skew shape $\lambda/\mu$ consists of the boxes in $\lambda$ not contained in $\mu$.

An excited Young diagram is any subset $C \subseteq D_\lambda$ obtained by applying a sequence of local moves (excitations) to the "seed" $\lambda$0, subject to the following rules (Panova et al., 2024, Graham et al., 2013, Kirillov et al., 2019):

• Excited Move: If $\lambda$1 and $\lambda$2, replace $\lambda$3 with $\lambda$4;
• Moves are iterated, each pushing boxes strictly down and to the right, preserving non-overlap and containment in $\lambda$5.

There exist variants with only Type I moves (removal/replacement), or with Type II moves (addition). The complete set of excited diagrams $\lambda$6 comprises all possible shapes produced via excitations from $\lambda$7 within $\lambda$8.

2. Combinatorial Bijections and Models

Excited Young diagrams admit bijections with several combinatorial families, giving rise to equivalent enumeration and generating formulas (Panova et al., 2024):

• Flagged Semistandard Tableaux (SSYT): Each excited diagram $\lambda$9 can be encoded by recording, for each original box $\mu$0, the row (after excitations) it occupies; this produces a tableau of shape $\mu$1 and entries subject to flag conditions determined by $\mu$2.
• Nonintersecting Lattice Paths: The complement $\mu$3 corresponds to a system of nonintersecting up/right lattice paths within $\mu$4, one for each row; these encode the data of the corresponding flagged SSYT.
• Lozenge Tilings: Each excited diagram indexes a lozenge tiling of a region formed by placing a 3D "ice pile" over $\mu$5 in a box of specified height from $\mu$6.

These bijections are leveraged for enumerative results, such as the skew hook-length formula and weighted counts via Lindström–Gessel–Viennot determinantal methods.

3. Enumeration and Skew Hook-Length Formulas

The enumeration of standard Young tableaux and other symmetrized fillings for skew shapes $\mu$7 is given by summing over excited diagrams (Panova et al., 2024, Kirillov et al., 2019):

$\mu$8

where $\mu$9 is the hook-length of cell $\mu \subseteq \lambda$0 in $\mu \subseteq \lambda$1.

Similarly, multivariate formulas emerge:

$\mu \subseteq \lambda$2

with the $\mu \subseteq \lambda$3 associated to boundary data of $\mu \subseteq \lambda$4. Contour integral and vertex-model reformulations exploit algebraic and integrable structures, leading to recursion and determinantal evaluations.

4. Excited Diagrams in Equivariant $\mu \subseteq \lambda$5-Theory and Hilbert Series

In the setting of equivariant $\mu \subseteq \lambda$6-theory for Grassmannians and cominuscule varieties (Graham et al., 2013):

• Schubert varieties $\mu \subseteq \lambda$7 indexed by permutations or partitions $\mu \subseteq \lambda$8 admit structure sheaf classes whose restriction to $\mu \subseteq \lambda$9-fixed points $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$0 is given as a sum over excited diagrams:

$$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$1

where $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$2 are torus weights.

• These formulas are termwise positive, as each factor $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$3 expands with nonnegative coefficients for positive roots $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$4, and no cancellations are required.

Hilbert series and polynomials at $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$5-fixed points are similarly expressed:

$$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$6

$$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$7

where $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$8 is the number of excited diagrams of size $$D_\lambda = \{(i,j)\in\mathbb{Z}_{>0}^2 : 1\le i\le \ell(\lambda), 1\le j\le \lambda_i\}$$9.

5. Factorization and $\lambda$0-Analogues

A striking property of formulas involving excited Young diagrams is complete factorization (Kirillov et al., 2019):

• For the skew hook-content formula:

$\lambda$1

where $\lambda$2 is the content of box $\lambda$3.

• Likewise, the $\lambda$4-analogues use $\lambda$5-integers $\lambda$6, with:

$\lambda$7

where $\lambda$8 is a rational function in $\lambda$9 with positive coefficients for appropriate shapes.

This factorization reflects the decomposition of the excitation process: from $\lambda/\mu$0 to each $\lambda/\mu$1 yields identical content factors, and from $\lambda/\mu$2 to $\lambda/\mu$3 yields hook-length inverses, with no cross-terms.

6. Vertex Models, Integrability, and Generalizations

Excited diagrams naturally correspond to configurations in six-vertex models (free-fermion point), with Boltzmann weights assigned to vertices and boundary conditions derived from $\lambda/\mu$4 and $\lambda/\mu$5. The Yang–Baxter equation yields Pieri-type recursions, and the correspondence with combinatorial bijections (paths, tableaux, tilings) is exact (Panova et al., 2024).

Contour integral formulas and vertex model interpretations generalize to factorial Grothendieck polynomials, interpolation Macdonald polynomials, spin $\lambda/\mu$6-Whittaker functions, and further symmetric functions. In each case, structure constants and linear extensions recover as weighted sums over excited diagrams or related plane partitions.

7. Open Problems and Positivity Properties

Multiple conjectures and open questions remain (Kirillov et al., 2019):

• Positivity: For fixed $\lambda/\mu$7, the normalized counts $\lambda/\mu$8 are conjectured to be nonnegative integers.
• $\lambda/\mu$9-Positivity: Rational $\lambda$0-positivity for $\lambda$1 and $\lambda$2 asserts nonnegativity of coefficients in numerators/denominators of such ratios, with open characterization of shapes for which they are genuine polynomials.
• Combinatorial Interpretation: Identification of direct combinatorial models corresponding to these normalized quantities remains an open problem, despite the bijections to tableaux and paths for standard tableau counts.

These questions suggest rich connections between excitation-based models, algebraic combinatorics, and representation theory, motivating further research.

---

Excited Young diagrams thus serve as a unifying object linking enumerative formulas, positivity phenomena, bijections, and integrable systems, with robust applications in algebraic geometry, symmetric function theory, and $\lambda$3-theoretic Schubert calculus (Graham et al., 2013, Panova et al., 2024, Kirillov et al., 2019).