Alcove walk models for parabolic Mirković-Vilonen intersections and branching to Levi subgroups

Basic Information

• Paper ID: 2405.17174
• Title: Alcove walk models for parabolic Mirković-Vilonen intersections and branching to Levi subgroups
• Author: Thomas J. Haines
• Classification: math.RT (Representation Theory), math.AG (Algebraic Geometry), math.CO (Combinatorics)
• Publication Date: May 2024 (arXiv preprint, revised December 31, 2024)
• Paper Link: https://arxiv.org/abs/2405.17174

Abstract

This paper establishes an alcove walk model for the intersections of Schubert cells in the affine Grassmannian of split reductive groups with certain semi-infinite orbits (called parabolic Mirković-Vilonen intersections). More precisely, the article describes explicit cellular decompositions of these intersections, indexed by specific positively-folded alcove walks. It proves a parametrization of irreducible components of maximal possible dimension in terms of alcove walks of maximal dimension. Subsequently, it derives a new combinatorial description of the branching of irreducible highest weight representations to Levi subgroups, and in particular provides a new algorithm for computing characters of such representations.

Research Background and Motivation

Problem Background

1. Importance of Mirković-Vilonen intersections: Classical Mirković-Vilonen intersections play a central role in the geometric Satake equivalence, serving as an important bridge connecting representation theory of algebraic groups with the geometry of the affine Grassmannian.
2. Need for parabolic generalization: While the classical case (P = B, Borel subgroup) is well understood, there is a lack of explicit combinatorial descriptions for the general parabolic subgroup P case.
3. Computational challenges: Existing methods lack effective combinatorial algorithms for computing branching multiplicities and representation characters, particularly for branching to Levi subgroups.

Research Motivation

1. Unification of geometry and combinatorics: To describe complex geometric intersections using alcove walks, a purely combinatorial object.
2. Algorithmic requirements: To provide new algorithmic tools for computing branching multiplicities in representation theory.
3. Theoretical completion: To generalize the gallery method of Gaussent-Littelmann to more general parabolic cases.

Core Contributions

1. Explicit cellular decomposition: Provides an explicit cellular decomposition of parabolic Mirković-Vilonen intersections, indexed by positively-folded alcove walks.
2. Dimension characterization: Proves a bijection between irreducible components of maximal dimension and alcove walks of maximal dimension.
3. Branching formula: Establishes the relationship between branching multiplicities $[V^{\hat{G}}_\mu : V^{\hat{M}}_\lambda]$ and the counting of specific alcove walks.
4. New algorithm: Provides new combinatorial algorithms for computing characters of highest weight representations and branching multiplicities.
5. Tensor product generalization: Extends results to the computation of tensor product multiplicities.

Detailed Methodology

Problem Setup

Study parabolic Mirković-Vilonen intersections: $K_P t^\lambda K/K \cap K t^\mu K/K$ where:

• $K = L^+G$ is the positive part of the loop group
• $K_P = L^+M LN$ is the subgroup corresponding to the parabolic subgroup
• $\mu \in X_*(T)^+$, $\lambda \in X_*(T)^{+M}$ are dominant weights

Core Technique: Alcove Walks

Alcove Walk Definition

Definition 4.2: Given a base alcove $b$ and a word $s_\bullet = s_1s_2\cdots s_r\tau \in W$, an alcove sequence $a_0, a_1, \ldots, a_r$ is a $b$-positively-folded alcove walk if it satisfies:

• $a_i \in \{a_{i-1}, w_{i-1}s_i w_{i-1}^{-1} a_{i-1}\}$
• If $a_{i-1} = a_i$, then the crossing $a_{i-1} \xrightarrow{s_i} w_{i-1}s_i w_{i-1}^{-1} a_{i-1}$ is in the $b$-negative direction

Labeling System

Each step $a_{i-1} \xrightarrow{s_i} a_i$ is labeled as:

• $c^+_{s_i}$: $b$-positive direction crossing
• $c^-_{s_i}$: $b$-negative direction crossing
• $f^+_{s_i}$: positive fold

Dimension Definition

Definition 4.7: The dimension of an alcove walk $a_\bullet$ is defined as: $\dim(a_\bullet) = c^+(a_\bullet) + f^+(a_\bullet)$

Main Theorems

Theorem A (Theorem 5.8)

The parabolic Mirković-Vilonen intersection admits a cellular decomposition: $K_P t^\lambda K/K \cap K t^\mu K/K \cong \bigsqcup_{a_\bullet \in P^{a_{IP}}_\mu(\lambda)} \mathbb{A}^{c^+(a_\bullet)}_k \times (\mathbb{A}^1_k - \mathbb{A}^0_k)^{f^+(a_\bullet)}$

where the set of alcove walks is: $P^{a_{IP}}_\mu(\lambda) = \bigsqcup_{w \in W_0/W_{0,-\mu}} P^{a_{IP}}_{(t_{-w(\mu)})_0}(-\lambda_w)$

Theorem B (Theorem 5.13)

1. There exists a bijection: $M^{a_{IP}}_\mu(\lambda) \xrightarrow{\sim} \text{Irred}_{\langle\rho,\mu+\lambda\rangle}(K_P t^\lambda K/K \cap K t^\mu K/K)$
2. The branching multiplicity equals the number of alcove walks of maximal dimension: $[V^{\hat{G}}_\mu : V^{\hat{M}}_\lambda] = \#M^{a_{IP}}_\mu(\lambda)$

Technical Innovations

1. Contraction map technique: Utilizes contraction maps $\rho_{I_P, \mathcal{A}}$ in the Bruhat-Tits building to transform geometric problems into combinatorial ones.
2. Right $W_0$-minimal elements: Introduces $(t_{-w(\mu)})_0$ as the right $W_0$-minimal element in the coset $t_{-w(\mu)}W_0$, simplifying the analysis of alcove walk types.
3. Dimension bounds: Proves dimension upper bounds $\langle\rho, \mu+\lambda\rangle$ and characterizes conditions for achieving maximal dimension.

Experimental Setup

Type $A_2$ Example

Consider $G = GL_3$, $P = B$ (Borel subgroup), $\mu = (3,1,0)$, weight orbit:

• $\lambda_1 = (1,1,2)$
• $\lambda_2 = (2,1,1)$
• $\lambda_3 = (1,2,1)$

Computational Results

For each $\lambda_i$, find all alcove walks of maximal dimension:

Case of $-\lambda_1$:

• Type $s_{012}\tau$: dimension $(A^1_k)^0 \times (A^1_k - A^0_k)^2$
• Type $s_{1012}\tau$: dimension $(A^1_k)^1 \times (A^1_k - A^0_k)^1$

Similarly compute cases for $\lambda_2, \lambda_3$, each having 2 alcove walks of maximal dimension, consistent with the known multiplicity $\dim V^{GL_3}_\mu(\lambda_i) = 2$.

Experimental Results

Main Results

1. Dimension formula verification: All computed alcove walk dimensions satisfy the upper bound $\langle\rho, \mu+\lambda\rangle$.
2. Multiplicity computation: In the Type $A_2$ example, the number of alcove walks corresponding to each weight $\lambda_i$ exactly equals the weight multiplicity 2.
3. PRV analogue: Proves the PRV analogue for branching to Levi subgroups (Proposition 5.18): If $\lambda \in W\mu \cap X_*(T)^{+M}$, then $[V^{\hat{G}}_\mu : V^{\hat{M}}_\lambda] = 1$.

Algorithm Verification

Through the new algorithm given in Theorem 5.16, one can effectively compute:

• Weight multiplicities of representations
• Branching multiplicities
• Tensor product multiplicities

Related Work

Major Prior Work

1. Gaussent-Littelmann GL05: Description of MV intersections using gallery theory; the alcove walks in this paper are special cases of their positively-folded galleries.
2. Ram Ram06: Alcove walk algebra theory; Theorem B can be viewed as a $q=1$ specialization of some of his results.
3. Kapovich-Leeb-Millson KLM08: Study of tensor product multiplicities using Hecke paths; this paper provides a pure alcove walk analogue.

Advantages of This Work

1. Unified framework: Handles both classical and parabolic cases simultaneously.
2. Explicit description: Provides completely explicit combinatorial descriptions.
3. Algorithmic practicality: Provides computable algorithms.

Conclusions and Discussion

Main Conclusions

1. Establishes a complete alcove walk model for parabolic Mirković-Vilonen intersections.
2. Proves precise correspondence between geometric objects (irreducible components) and combinatorial objects (alcove walks).
3. Provides new combinatorial formulas for branching multiplicities.

Limitations

1. Technical constraints: Requires conditions on $\mu - \lambda$ in the coroot lattice.
2. Computational complexity: For higher rank groups, enumeration of alcove walks may become complicated.
3. Geometric understanding: The precise relationship with the gallery method of Gaussent-Littelmann remains incompletely clarified.

Future Directions

1. Establish explicit bijections with galleries in GL05.
2. Generalize to more general affine flag varieties.
3. Investigate relationships with generalized saturation conjectures.

In-Depth Evaluation

Strengths

1. Theoretical depth: Organically combines profound geometric theory with combinatorial methods.
2. Technical innovation: Clever application of contraction map and alcove walk techniques.
3. Practical value: Provides new tools for representation-theoretic computations.
4. Completeness: Well-developed from theorem statements to concrete examples.

Weaknesses

1. Readability: Abundant technical details may pose comprehension difficulties for non-specialists.
2. Computational efficiency: Algorithm complexity analysis is not discussed.
3. Generalizability: Conditions for generalizing certain results are relatively restrictive.

Impact

1. Academic value: Provides new perspectives for cross-disciplinary research between algebraic group representation theory and affine geometry.
2. Application prospects: Potential applications in the local Langlands program and Shimura variety research.
3. Methodological significance: Systematic development of the alcove walk method provides inspiration for related fields.

Applicable Scenarios

1. Representation theory of split reductive groups
2. Geometric analysis of the affine Grassmannian
3. Computation of branching multiplicities and tensor product multiplicities
4. Affine Hecke algebra theory

References

The paper contains 18 important references covering classical works in core areas including geometric Satake equivalence, alcove walk theory, and building theory, providing a solid theoretical foundation for the research.