Invariants for $\mathbb G_{(r)}$-modules

Basic Information

• Paper ID: 2505.08094
• Title: Invariants for $\mathbb G_{(r)}$-modules
• Author: Eric M. Friedlander
• Classification: math.RT (Representation Theory)
• Publication Date: October 16, 2025
• Paper Link: https://arxiv.org/abs/2505.08094

Abstract

This paper revisits the refined invariants for finite-dimensional representations of infinitesimal group schemes $\mathbb G_{(r)}$ over fields $k$ of characteristic $p > 0$, previously constructed by J. Pevtsova and the author. The research focuses on viewing generic $p$-nilpotent operators as elements in the group algebra of the group scheme $\mathbb G_{(r),X}$ over $X$, where $X$ is either the moduli space $V_r(\mathbb G)$ of height $r$ one-parameter subgroups of $\mathbb G$, or the moduli space $\mathcal C_r(\mathcal N_p(\mathfrak g))$ of $r$-tuples of pairwise commuting $p$-nilpotent elements in the Lie algebra of $\mathbb G$.

Research Background and Motivation

Problem Context

1. Development of Support Theory: This paper employs methods and results from support theory to study explicit invariants of finite-dimensional modules over infinitesimal group schemes of specific types, contrasting with previous work using support theory to establish categorical properties of representations of finite group schemes.
2. In-depth Study of Jordan Type Functions: The author further investigates the "Jordan type function" previously introduced with J. Pevtsova, which provides a continuous map from scheme points representing height $r$ one-parameter subgroups to the poset of Young diagrams with $p$ columns.
3. Computational Accessibility: The existing Jordan type function is computationally complex, necessitating the development of more readily computable variants.

Research Motivation

1. Finer Invariants: Compared to cohomological support theory $M \mapsto P_{H^*(G_{(r)},k)}M$, the Jordan type function $M \mapsto JT_{G,r,M}(-)$ captures more information about $G_{(r)}$-modules $M$.
2. Special Treatment of Exponential Groups: For exponential groups, one-parameter subgroups can be described more explicitly through $p$-nilpotent elements in the Lie algebra, thereby simplifying computations.

Core Contributions

1. Formalization of Jordan Type Functions: Formalizes Jordan type functions using multiple variants of continuous functions $JT_{G,r,M}(-): \mathbb P V_r(\mathbb G) \to \mathcal Y$, where $\mathcal Y$ is the poset of Young diagrams with $p$ columns.
2. Simplified Exponential Jordan Type: Introduces a more computationally tractable exponential Jordan type function $JT^{\exp}_{g,r,M}(-)$, particularly suited for exponential group schemes.
3. Extension of Vector Bundle Construction: Extends the vector bundle construction of Pevtsova and the author to all finite-dimensional $G_{(r)}$-modules, yielding coherent sheaves that are locally free on the fibers over $X$ related to $JT_{G,r,M}(-)$.
4. Correction of Stabilization Results: Corrects a minor error in previous work concerning the stabilization of Jordan type functions with respect to $r$.

Detailed Methodology

Fundamental Concepts and Definitions

Definition of Jordan Type

For a $k[t]/t^p$-module $M$, its Jordan type $a = (a_1, \ldots, a_p) \in \mathbb{N}^{\times p}$ is determined by the number $a_i$ of Jordan blocks of size $i$ in the Jordan normal form of the matrix $\rho_M(t)$.

Generic $p$-Nilpotent Operators

For an affine group scheme $G$ and positive integer $r$, the generic one-parameter subgroup $U_{G,r}: G_{a(r),k[V_r(G)]} \to G_{(r),k[V_r(G)]}$ induces a generic $p$-nilpotent operator $\Theta_{G,r} = (U_{G,r})_*(1 \otimes u_{r-1}) \in k[V_r(G)] \otimes kG_{(r)}$

Exponential Group Schemes

Exponential Structure

A group scheme $G$ equipped with an exponentiation structure of height $r$ means there exists a morphism $E^{(r)}: G_{a(r)} \times \mathcal{N}_p(g) \to G_{(r)}$ satisfying specific commutativity and homogeneity conditions.

Simplified Operators

For exponential groups, define $\Theta^{\exp}_{g,r} = \sum_{s=0}^{r-1} \Theta_{g,r|s} \in k[C_r(\mathcal{N}_p(g))] \otimes kG_{(r)}$ which is a linearization of the original operator $\Theta_{g,r} = \prod_{s=0}^{r-1} \Theta_{g,r|s}$.

Continuity of Jordan Type Functions

Theorem: For a finite-dimensional $G_{(r)}$-module $M$, the function $JT_{G,r,M}(-): \text{Spec}A \to \mathcal{Y}$ is continuous, where $A$ is a Noetherian commutative $k$-algebra.

The proof is based on Nakayama's lemma and the lower semicontinuity of Jordan type.

Technical Innovations

1. Stratified Geometric Structure

The paper establishes a stratification of $\mathbb{P}V_r(G)$, where each stratum is determined by the level sets of the Jordan type function: $(PV_r(G))_{M,=a} = \{x : JT_{G,r,M}(x) = a\}$ These strata are locally closed sets, forming a stratification ordered by the Young diagram poset.

2. Justification of Exponential Simplification

A key result shows that for the study of support varieties, $\Theta_{g,r}$ and $\Theta^{\exp}_{g,r}$ provide equivalent information:

Proposition: Let $M$ be a finite-dimensional $G_{(r)}$-module and $a$ be the maximal Jordan type of $M$. For any $B \in C_r(\mathcal{N}_p(g))$, $JT(\Theta^{\exp}_{g,r,B}, k(B) \otimes M) = a \iff JT(\Theta_{g,r,B}, k(B) \otimes M) = a$

3. $A^1$-Homotopy Equivalence of Vector Bundles

For modules of constant Jordan type, vector bundles constructed using $\Theta_{g,r}$ and $\Theta^{\exp}_{g,r}$ are $A^1$-homotopy equivalent.

Experimental Results and Examples

Example 1: The Case of $SL_2$

For $G = SL_2$, consider the irreducible $G$-module $M = S(\lambda_0) \otimes S(\lambda_1)^{(1)}$, where $0 \leq \lambda_0, \lambda_1 < p$.

Computing $JT^{\exp}_{sl_2,2,M}(B)$ where $B = (a_0E, a_1E)$ and $E$ is the standard nilpotent matrix:

• If $a_0 \neq 0, a_1 = 0$: Jordan type is $m \cdot [n]$
• If $a_0 = 0, a_1 \neq 0$: Jordan type is $n \cdot [m]$
• If both $a_0, a_1$ are nonzero: Jordan type is $[m] \otimes [n]$

where $m = \lambda_0 + 1$ and $n = \lambda_1 + 1$.

Example 2: Polynomial Representations of $GL_N$

For the $i$-th Frobenius twist $V^{(i)}$ of a polynomial representation $V$ of $GL_N$ of degree $< p$: $JT^{\exp}_{gl_N,r,V^{(i)}}(B) = JT(B_s, V)$ where $s = r - i - 1$ and $B = (B_0, \ldots, B_{r-1})$.

Related Work

Development of Support Theory

• Collaboration of Suslin, Bendel, and the author establishing support theory using varieties of one-parameter subgroups to replace cohomological varieties
• Previous work of Pevtsova and the author establishing refined invariants and vector bundle constructions
• Carlson's work on cyclic shift subgroups in identifying support varieties of $(Z/p)^{\times r}$-modules

Jordan Type Theory

• Theory development for modules of constant Jordan type
• Carlson-Friedlander's exact category of modules of constant Jordan type
• Benson-Pevtsova's realization theorem for modules of constant Jordan type and vector bundles

Conclusions and Discussion

Main Conclusions

1. Establishes a complete theoretical framework for Jordan type functions, including continuity, functoriality, and relationships with support varieties
2. Provides computationally more tractable simplified versions for exponential groups
3. Extends vector bundle construction to arbitrary finite-dimensional modules, yielding coherent sheaves that are locally free on strata

Limitations

1. Exponential simplification applies only to exponential group schemes
2. For $r > 1$, Jordan type functions rarely commute with tensor products
3. Certain computations remain challenging, particularly those involving Jordan types of sums of multiple commuting nilpotent operators

Future Directions

1. Study stabilization properties of Jordan type functions
2. Develop more efficient computational methods
3. Explore connections with other invariants in algebraic geometry
4. Investigate further properties of stratified coherent sheaves

In-Depth Evaluation

Strengths

1. Theoretical Completeness: Provides a complete theoretical framework for Jordan type functions, including fundamental properties such as continuity and functoriality
2. Computational Accessibility: Exponential simplification makes practical computation feasible
3. Geometric Insights: Establishes profound connections between representation theory and algebraic geometry
4. Technical Innovation: Extension of vector bundle construction and establishment of stratified structures

Weaknesses

1. Limited Scope of Applicability: Many results apply only to exponential groups
2. Computational Complexity: Despite simplifications, certain computations remain difficult
3. Relatively Few Examples: Could benefit from more concrete computational examples

Impact

This paper provides important geometric tools for representation theory of infinitesimal group schemes and is expected to have significant impact at the intersection of algebraic group representation theory, algebraic geometry, and homological algebra.

Applicable Scenarios

• Representation theory of algebraic groups
• Geometric analysis of module categories
• Applications of support variety theory
• Classification problems for vector bundles

References

The paper cites core literature in the field, including:

• Work of Suslin, Friedlander, and Bendel on infinitesimal one-parameter subgroups and cohomology
• Research of Friedlander and Pevtsova on π-support and generalized support varieties
• Classical work of Carlson on varieties of modules and cohomology rings
• Standard references in algebraic geometry on weighted projective spaces and coherent sheaves