🔢 mathematics

Non-admissibility of some universal supersingular representations

This paper proves that for $n \ge 3$ , the universal supersingular representation of a sufficiently generic weight $\sigma$ over an unramified extension $K/\mathbf{Q}_p$ is non-admissible and of infinite length, thereby generalizing prior results for $n=2$ through a weight cycling argument and recent advances in Serre weight conjectures.

Original authors: Zachary Feng, Heejong Lee, Ray Li, Vaughan McDonald, Nischay Reddy

Published 2026-05-19

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zachary Feng, Heejong Lee, Ray Li, Vaughan McDonald, Nischay Reddy

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 trying to organize a massive, chaotic library of mathematical objects called "representations." These objects are like complex, multi-layered machines that describe how numbers and symmetries interact. The authors of this paper are investigating a specific, very tricky type of machine called a "universal supersingular representation."

Here is the story of what they found, explained without the heavy math jargon.

The Problem: A Library That Never Ends

In the world of these mathematical machines, there is a rule called "admissibility." Think of an admissible machine as one that is well-organized: it has a finite number of distinct parts, and you can describe its entire structure with a manageable list of instructions.

For a long time, mathematicians knew that for small, simple machines (specifically when the dimension $n=2$ and the field is the standard rational numbers), these "universal supersingular" machines were well-behaved. They were admissible.

However, when the machines get slightly more complex (either by increasing the dimension to $n \ge 3$ , or by using a slightly more complicated number system), things get messy. The authors wanted to know: Are these complex machines still well-organized, or do they spiral into infinite chaos?

The Discovery: Infinite Chaos

The paper proves that for these more complex machines ( $n \ge 3$ ), under certain specific conditions, the answer is chaos.

They show that these "universal supersingular representations" are non-admissible. In plain English, this means:

Infinite Length: The machine has an infinite number of layers. You can never finish counting its parts.
Infinite Variety: If you try to break the machine down into its simplest, irreducible pieces, you will find an infinite number of different types of pieces, not just a finite list.

It's like trying to build a tower out of blocks, but every time you add a block, the instructions tell you to add two more, and those two tell you to add four more, forever. The tower never stabilizes.

The Method: The "Weight Cycling" Detective Game

How did they prove this? They used a clever strategy they call "weight cycling."

Imagine you have a mystery box (the universal representation). You want to see what's inside, but you can't open it directly. Instead, you have a set of special keys (called Hecke operators).

The Cycle: You use a key to unlock a door that leads to a smaller, related box.
The Connection: Inside this smaller box, you find a specific "weight" (a type of symmetry).
The Twist: The authors realized that if they could find a specific "Galois representation" (think of this as a secret code or a blueprint from a different branch of mathematics) that matches this weight, they could force the machine to reveal its infinite nature.

They didn't just guess; they used a high-tech map called the Emerton–Gee stack. Imagine this map as a topographical chart of a mountain range where every peak represents a different mathematical object.

They looked for a specific intersection point on this map where three different "territories" (mathematical conditions) overlapped.
They proved that this intersection point exists and is not empty.
At this intersection, they found a "Galois representation" that acts like a master key. When they used this key on their machine, it forced the machine to split into an infinite number of different principal series (a specific type of simple machine).

The "Special" Condition

The paper notes that this chaos doesn't happen for every machine. It only happens when the machine's "highest weight" (its core identity) falls into a "special p-alcove."

Think of the "alcoves" as different rooms in a house.

Some rooms are "ordinary" (safe, predictable).
Some rooms are "special" (dangerous, prone to infinite expansion).
The authors proved that if your machine is built in one of these "special rooms," it will inevitably explode into infinite complexity. If it's in a regular room, the rules might be different.

Why This Matters (According to the Paper)

The authors emphasize that their method is a new way of doing math. Instead of trying to build the machine from the bottom up (which failed for complex cases in the past), they used a "top-down" approach. They used the Galois representations (the blueprints from the other branch of math) to guide them and prove that the machine must be infinite.

They view this as a showcase: Galois representations can act as a compass to navigate the confusing landscape of p-adic group representations.

Summary

The Subject: Complex mathematical machines called "universal supersingular representations."
The Finding: When these machines are complex enough ( $n \ge 3$ ) and built in "special" configurations, they are infinite and unmanageable (non-admissible).
The Tool: A "weight cycling" argument that uses a map of Galois representations to force the machine to reveal its infinite nature.
The Analogy: It's like proving that a specific type of fractal pattern, when zoomed in far enough, never repeats and never ends, using a secret code from a different dimension to unlock the view.

The paper does not discuss medical applications, engineering uses, or future technologies. It is purely a theoretical breakthrough in understanding the fundamental structure of these mathematical objects.

Technical Summary: Non-Admissibility of Some Universal Supersingular Representations

Problem Statement
The paper addresses the structure of mod $p$ representations of $p$ -adic reductive groups, specifically $G = \mathrm{GL}_n(K)$ where $K/\mathbb{Q}_p$ is an unramified extension of degree $f$ . While the mod $p$ local Langlands correspondence is well-understood for $\mathrm{GL}_2(\mathbb{Q}_p)$ , the case where $n \ge 3$ or $K \neq \mathbb{Q}_p$ exhibits significantly more complex behavior. The central object of study is the universal supersingular representation $SS(\sigma)$ associated with an irreducible Serre weight $\sigma$ of $\mathrm{GL}_n(k)$ (where $k$ is the residue field of $K$ ). Defined as the quotient of the compactly induced representation $c\text{-}\mathrm{Ind}_{KZ}^G \sigma$ by the ideal generated by Hecke operators $T_1, \dots, T_{n-1}$ , $SS(\sigma)$ is expected to classify supersingular representations.

Previous work established that for $n=2$ and $K \neq \mathbb{Q}_p$ , $SS(\sigma)$ is non-admissible and of infinite length. However, existing methods for proving non-admissibility (relying on the theory of diagrams or explicit computations of pro- $p$ Iwahori invariants) were restricted to $n=2$ . The open problem was to determine the admissibility and length of $SS(\sigma)$ for $n \ge 3$ .

Methodology
The authors develop a method distinct from previous diagram-based approaches, combining a weight cycling argument with globalization techniques and the geometry of the Emerton–Gee stack.

Weight Cycling: The authors utilize a general form of the weight cycling argument (originally due to Buzzard and generalized by Emerton–Gee–Herzig). For a Serre weight $\sigma$ and a fundamental antidominant cocharacter $\mu$ , they construct a smooth representation $C_\mu(\sigma)$ of $KZ$ and a surjective morphism:
$\mathrm{cyc}_\mu: c\text{-}\mathrm{Ind}_{KZ}^G C_\mu(\sigma) \twoheadrightarrow c\text{-}\mathrm{Ind}_{KZ}^G \sigma / (T_\mu).$
Since $c\text{-}\mathrm{Ind}_{KZ}^G \sigma / (T_\mu)$ surjects onto $SS(\sigma)$ , proving that a specific subquotient of $c\text{-}\mathrm{Ind}_{KZ}^G C_\mu(\sigma)$ is non-admissible implies the same for $SS(\sigma)$ . The challenge lies in identifying the Jordan–Hölder factors of $C_\mu(\sigma)$ and understanding their images under the weight cycling map.
Globalization and Galois Representations: To resolve the difficulty of identifying relevant subquotients, the authors employ a global method. They construct a continuous Galois representation $\rho: G_K \to \mathrm{GL}_n(\mathbb{F}_p)$ and attach to it an admissible smooth representation $\pi$ of $G$ via a candidate for the mod $p$ local Langlands correspondence. By carefully choosing $\rho$ such that $\sigma$ is a "modular weight," they establish a non-zero morphism from $C_\mu(\sigma)$ to $\pi|_{KZ}$ . This forces a specific Jordan–Hölder factor $\sigma' \in \mathrm{JH}(C_\mu(\sigma))$ to appear in the socle of $\pi|_{KZ}$ .
Local Model Theory and Emerton–Gee Stacks: The existence of a suitable $\rho$ is guaranteed by studying the geometry of the reduced Emerton–Gee stack $X_{n, \mathrm{red}}$ , which serves as the moduli space of mod $p$ Galois representations. The authors analyze the intersection of irreducible components $C_\sigma$ and $C_{\sigma'}$ corresponding to $\sigma$ and $\sigma'$ .
- They utilize explicit local charts for potentially crystalline stacks developed in [LLHM+24].
- They define a "special" condition on the $p$ -alcove containing the highest weight of $\sigma$ .
- They prove that for "special" weights, the intersection of the supersingular locus of $\sigma$ , the component $C_{\sigma'}$ , and the ordinary locus of $\sigma'$ is non-empty. This ensures the existence of a $\rho$ that is supersingular with respect to $\sigma$ but ordinary with respect to $\sigma'$ .
Non-Admissibility Argument: For such a $\rho$ , the image of $c\text{-}\mathrm{Ind}_{KZ}^G \sigma'$ in $\pi$ factors through a principal series representation. By twisting the diagonal characters of $\rho$ with infinitely many unramified characters, the authors generate infinitely many non-isomorphic principal series representations as quotients of the image. By Lemma 2.11, a representation admitting infinitely many non-isomorphic irreducible quotients is non-admissible.

Key Contributions and Results

Theorem 1.1 (Theorem 5.1): Let $K/\mathbb{Q}_p$ be an unramified extension of degree $f \ge 1$ . If $n \ge 3$ and $\sigma$ is a sufficiently deep Serre weight whose $p$ -restricted highest weight lies in a "special" $p$ -restricted $p$ -alcove, then the universal supersingular representation $SS(\sigma)$ is non-admissible and of infinite length.
Generalization: This result generalizes the known non-admissibility for $n=2$ and $K \neq \mathbb{Q}_p$ to higher ranks ( $n \ge 3$ ).
Geometric Insight: The paper establishes that for special weights, the supersingular locus in the irreducible component $C_\sigma$ of the Emerton–Gee stack has codimension one (Corollary 3.28). This contrasts with the lowest alcove case, where the supersingular locus is more trivial.
Methodological Shift: The paper demonstrates that Galois representations can guide the mod $p$ representation theory of $p$ -adic groups, specifically by using the geometry of the Emerton–Gee stack to control the structure of universal representations.

Significance and Limitations
The authors claim that their method provides a new pathway for studying mod $p$ representations of $p$ -adic groups, moving beyond the limitations of the theory of diagrams which is currently unavailable for $n \ge 3$ . The result confirms that the "surprisingly complicated behaviors" observed in $\mathrm{GL}_n(K)$ for $n \ge 3$ include the non-admissibility of universal supersingular representations under generic conditions.

The paper explicitly notes limitations:

The result requires $\sigma$ to be "sufficiently deep" and its highest weight to be in a "special" $p$ -alcove. The proportion of such alcoves is $1 - (\frac{n-2}{n-1})^f$ .
The proof for $n \ge 3$ relies on the specific geometry of the intersection of local charts; the authors note that for non-special weights (e.g., in the base $p$ -alcove), the intersection required for the weight cycling argument may be empty, making the problem significantly harder.
While the argument can be adapted to show non-admissibility for $n=2$ and $f \ge 2$ , the primary focus is on the $n \ge 3$ case where previous methods failed.

The paper does not propose new experimental applications or future conjectures beyond the scope of the established non-admissibility results and the geometric properties of the Emerton–Gee stack.