🔢 mathematics

Bump-Friedberg type periods beyond the cuspidal spectrum

This paper extends Bump-Friedberg type periods beyond the cuspidal spectrum to automorphic functions of uniform moderate growth on $\textnormal{GL}_{2n}$ and $\textnormal{GL}_{2n+1}$ , characterizing them via Whittaker-type zeta integrals and evaluating specific cases as sums of $L$ -function special values that align with the Ben-Zvi-Sakellaridis-Venkatesh numerical conjecture under the global Langlands correspondence.

Original authors: Shenghao Li

Published 2026-07-14

📖 6 min read🧠 Deep dive

CC BY 4.0

Original authors: Shenghao Li

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 the universe of numbers and shapes as a giant, cosmic orchestra. For decades, mathematicians have been trying to understand the relationship between the music the orchestra plays (called automorphic forms) and the hidden sheet music that dictates the notes (called L-functions). Usually, they've only been able to study the orchestra when it's playing a very specific, quiet, and perfectly isolated solo: the cuspidal spectrum. It's like listening to a single violinist in a soundproof room. The math works beautifully there, but it's a bit lonely.

In this paper, author Shenghao Li asks a bold question: What happens when we let the whole orchestra play, not just the soloists? Can we still find the connection between the music and the sheet music when the sound is louder, messier, and includes the booming drums and brass sections (known as Eisenstein series)?

The Big Breakthrough: Extending the Music

The paper's main finding is that yes, we can! Li shows that we can take a specific type of musical measurement, called a Bump–Friedberg period, which was previously only defined for those quiet soloists, and stretch it out to cover the entire, noisy orchestra.

Think of the original method as a delicate net designed to catch only the smallest, most elusive fish (the cuspidal forms). If you tried to use that net on a whale (an Eisenstein series), it would tear. Li's innovation is like reinforcing that net with a special, invisible elastic material. He proves that even though the "whale" is huge and heavy, the net can now hold it without breaking, provided the whale follows certain rules of behavior (what the paper calls regularity conditions).

The Magic Trick: The "Unfolding"

How does he do it? He uses a mathematical sleight of hand called unfolding.

Imagine you have a complex, folded origami crane (the period integral). When you try to look at it directly, it's a tangled mess. Li's method involves carefully unfolding the paper step-by-step.

The Problem: When you unfold the paper for a noisy orchestra, you expect to see extra flaps of paper (extra terms) that shouldn't be there. In the old days, these extra flaps would ruin the whole picture.
The Solution: Li proves that if the orchestra is playing in a specific "regular" way, those extra flaps magically vanish into thin air. They don't just disappear; they cancel each other out perfectly.
The Result: Once the extra flaps are gone, the paper unfolds into a clean, simple shape: a Whittaker-type zeta integral. This is a new, continuous way to measure the music that works for both the quiet soloists and the loud orchestras.

The New Discovery: The $GL_{2n+1}$ Puzzle

The paper doesn't just stop at fixing the old net; it builds a brand new one for a specific, tricky instrument: the group $GL_{2n+1}$ (imagine a musical ensemble with an odd number of sections, like 3, 5, or 7).

Li introduces a new way to measure this ensemble by integrating over a subgroup called $SL_{n+1} \times GL_n$ . It's like listening to the orchestra while focusing only on the rhythm section and the brass, ignoring the rest.

When he applies this new measurement to the "whales" (the Eisenstein series), something fascinating happens. The result isn't just one single number. Instead, it breaks down into a finite sum of products.

The Analogy: Imagine the total sound of the orchestra isn't one big chord, but a recipe made of several specific ingredients mixed together.
The Ingredients: Each ingredient in the recipe is a special value of an L-function (a mathematical number that encodes deep secrets about the music) multiplied by a local "taste test" (a normalized local zeta integral).

The "Fixed Points" Connection

Here is where the paper gets really exciting and connects to a massive, hypothetical theory called the Global Numerical Conjecture (proposed by Ben-Zvi, Sakellaridis, and Venkatesh).

This conjecture suggests that for every musical ensemble, there is a "dual" geometric shape (a dual variety) floating in a parallel universe. The conjecture predicts that the total sound of the orchestra should be determined by the fixed points of a specific "conductor's baton" (the L-parameter) on that dual shape.

Li's paper proves that for this new $GL_{2n+1}$ period:

The number of ingredients in his recipe (the sum) matches exactly the number of fixed points on the dual shape.
The value of each ingredient matches exactly the tangent space (the local geometry) at that fixed point.

It's as if Li built a bridge between the noisy orchestra and the silent geometric shape, and the bridge fits perfectly. The math on the orchestra side (the period) and the math on the geometry side (the fixed points) agree down to the last decimal.

What the Paper Rules Out

It is important to note what this paper says doesn't work.

No Magic for Everyone: The method only works if the "cuspidal datum" (the underlying structure of the music) satisfies specific regularity conditions. If the music is too chaotic or doesn't follow these rules, the extra flaps in the unfolding step won't vanish, and the net will tear. The paper explicitly states that without these conditions, the extension fails.
No Single Value: The paper argues against the idea that the result is always a single L-value. For these non-cuspidal forms, the result is explicitly a sum of multiple terms. If you tried to force it into a single number, you would be wrong.

How Sure Are We?

The paper is not just guessing or simulating; it provides rigorous proofs.

Li proves that the integrals converge (the math doesn't blow up).
He proves that the "unfolding" works and the extra terms vanish under the right conditions.
He proves that the final formula for the Eisenstein series is a finite sum of specific L-functions.
He demonstrates that this formula matches the prediction of the global numerical conjecture, assuming the hypothetical Global Langlands Correspondence holds true.

In short, Li has successfully expanded the territory of mathematical music theory. He showed that the beautiful, clean relationships we found in the quiet solos also exist in the loud, complex symphonies, provided we know how to listen for the right patterns. The bridge between the music and the geometry is not just a guess; it's a solid, proven structure, waiting for the rest of the mathematical world to cross it.

Technical Summary: Bump–Friedberg type periods beyond the cuspidal spectrum

Problem Statement
The classical theory of Bump–Friedberg periods, which provides integral representations for products of standard and exterior-square $L$ -functions, is traditionally restricted to cuspidal automorphic forms. This restriction arises because the original period integrals are only defined for cuspidal forms, and the unfolding process relies on the vanishing of proper constant terms—a property that fails for general automorphic forms of uniform moderate growth.

This paper addresses the extension of Bump–Friedberg type periods to the broader spectrum of automorphic forms, specifically targeting Eisenstein series and functions of uniform moderate growth. The work is situated within the framework of the Relative Langlands Duality proposed by Ben-Zvi, Sakellaridis, and Venkatesh (BZSV). Specifically, it investigates the "global numerical conjecture," which predicts that for non-strongly tempered spherical varieties, the period integral should correspond to a finite sum of special values of $L$ -functions, indexed by the fixed points of the Langlands parameter on the dual variety. Previous work by Lu and Xi established this phenomenon for certain Eisenstein series; this article aims to establish an analogous phenomenon for Bump–Friedberg type periods.

Methodology
The author employs a strategy of analytic continuation and regularization to extend periods from the cuspidal spectrum to the space of automorphic functions of uniform moderate growth, denoted $T([G])$ . The methodology proceeds in three main stages:

Definition on Schwartz Spaces: The periods are first defined on the space of automorphic Schwartz functions $S([G])$ , where the integrals are absolutely convergent.
Regularity Conditions and Unfolding: The author imposes specific "regularity conditions" on the cuspidal data (denoted as $(\eta, \Delta^*)$ -regularity or $\Delta^*_1$ -regularity). These conditions ensure that the "extra terms" arising during the unfolding of the period integral (specifically, integrals over non-trivial orbits or smaller periods) vanish. This allows the period to be unfolded into a Whittaker-type zeta integral.
Analytic Continuation: The unfolded zeta integrals are shown to possess absolute convergence in a suitable region and admit an entire continuation to the whole complex plane. The extended period on $T([G])$ is then defined as the value of this continued zeta integral at a specific point (typically $(0,0)$ ).

The paper analyzes three specific cases:

Twisted Bump–Friedberg Period on $GL_{2n}$ : Integrating over $GL_n \times GL_n$ with a twist by a Hecke character.
Augmented Period on $GL_1 \times GL_{2n}$ : A variant involving a determinant twist.
New Period on $GL_{2n+1}$ : Integrating over the subgroup $SL_{n+1} \times GL_n$ embedded in $GL_{2n+1}$ . This case utilizes a degenerate Whittaker function associated with a degenerate character.

Key Contributions and Results

Continuous Extension of Periods:
The paper proves that under suitable regularity conditions on the cuspidal datum, the Bump–Friedberg type functionals extend continuously from the space of cuspidal automorphic forms to the space of automorphic functions of uniform moderate growth. This extension is characterized by the entire continuation of associated Whittaker-type zeta integrals.
Evaluation on Eisenstein Series:
For the new period on $GL_{2n+1}$ (integrating over $SL_{n+1} \times GL_n$ ), the author evaluates the extended period on $\Delta^*_1$ -regular Eisenstein series. The main theorem (Theorem 5.4.1) expresses this period as a finite sum indexed by the set $J_1(\pi)$ , which corresponds to the $GL_1$ blocks in the Levi component of the inducing parabolic.
The formula is given by:
$P^*(E(\phi), \Phi) = (\Delta^*_{G_n \times G_n})^{-1} L(1, \pi, \hat{n}^-_P)^{-1} \sum_{i \in J_1(\pi)} C_{i,S}(\phi_S, \Phi_S) L_S(1, \eta_i^{-1} \otimes \Pi_i) L_S(1, \eta_i \otimes \Pi_i^\vee) L_S(1/2, \Pi_i, \eta_i^{-1} \otimes \wedge^2)$
where $\Pi_i$ represents the representation induced by the remaining blocks, and $C_{i,S}$ involves normalized local zeta integrals.
Compatibility with the Global Numerical Conjecture:
Assuming the global Langlands correspondence, the paper demonstrates that the resulting formula aligns with the predictions of the BZSV global numerical conjecture.
- Fixed Points: The index set $J_1(\pi)$ of the sum corresponds exactly to the fixed points of the extended $L$ -parameter on the conjectural dual variety $\check{X} = G_{2n+1} \times G_1 \times G_{2n} (\text{std}^\vee_{G_1} \boxtimes \wedge^2)$ .
- Tangent Space Prediction: The $L$ -factors appearing in the summands for each fixed point $i$ match the tangent space contribution predicted by the conjecture. Specifically, the factors $L(1, \eta_i^{-1} \otimes \Pi_i)$ , $L(1, \eta_i \otimes \Pi_i^\vee)$ , and $L(1/2, \Pi_i, \eta_i^{-1} \otimes \wedge^2)$ correspond to the representation of the tangent space at the fixed point.

Significance
The paper claims significance primarily in extending the scope of integral representations for $L$ -functions beyond the cuspidal spectrum, thereby providing concrete evidence for the global numerical conjecture in the context of Bump–Friedberg periods. By establishing that the period of an Eisenstein series decomposes into a sum of $L$ -values indexed by fixed points on the dual variety, the work validates the structural predictions of Relative Langlands Duality for non-strongly tempered cases. The results bridge the gap between the analytic properties of automorphic periods and the geometric data of the dual spherical variety, confirming that the "fixed point" phenomenon observed in other contexts (such as by Lu and Xi) holds for these specific periods on general linear groups.