Counting multigeodesics and the Selberg-Delange method
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Mathematics. The day in mathematics: new theorems, new proofs, and what they open.
Ada: We're looking at the paper: "Counting multigeodesics and the Selberg-Delange method".
Emmy: Here's the gist of the paper. We study the counting function Pm(X, x) for m-multigeodesics of length at most x on a compact hyperbolic surface X where m is a positive integer and prove a large x asymptotic for Pm(X, x).
Ada: They also study the distribution of ω1 in multigeodesics and show that as x → ∞, this distribution is close to that of a Poisson random variable with parameter log x.
Emmy: The Selberg-Delange Method Application The study adapts the Selberg-Delange method to the context of the Selberg zeta function and investigates the asymptotics of Pm(X, x) 0 such that log x ≥ max log m − 1, log 1/ (1−s1), Pm(X, x) = e x / x(log m − 1) ν0,X(m − 1) / log x + O(x(-log x)(m − 3)!!), where the implied constant depends only on X.
Ada: The dependency on X in the error term can be made explicit through asymptotic expansions in powers of 1/x.
Title and authors: Ada: So we're looking at this paper "Counting multigeodesics and the Selberg-Delange method," which is all about counting these kinds of curves on a compact hyperbolic surface X. Emmy, what's the main idea behind this approach?
Emmy: Well, essentially, they are adapting the Selberg-Delange method. It's a technique originally used for counting integers with exactly m prime factors in an interval one x <ref:2611.08165#pg1,for counting integers with exactly m prime factors in>. They are using that framework to find a large x asymptotic for counting these m multigeodesics on the surface X.
Kolya: That sounds like it connects number theory and geometry. So, what does this mean concretely for the count P m(X, x)? Does it give us a formula?
Ada: It gives us a specific form for that asymptotic behavior. They show that for fixed m as x gets really big, the result is independent of the surface X. But if we look at cases where m times log x stays bounded away from zero, then the count actually depends on an infinite Euler product over the lengths of those primitive closed geodesics on X.
Emmy: That Euler product dependence is interesting because it ties directly into the geometry of the surface itself, specifically through those geodesic lengths. It suggests that for certain values of m, you can't just have a single number; you need to sum up information from all the primitive geodesics available on X.
Kolya: So, if we are looking at someone just listening to this show, what does that mean for the raw counting function P m(X, x)? Is there a clean formula we can actually calculate?
Ada: There is a leading order term derived by looking at the m-th coefficient of a Taylor series for some function FX,x(z). They state that P m(X, x) is approximately e to x divided by x raised to the power of log m minus one times something related to X and log x.
Emmy: That leading order term itself simplifies quite nicely when you look at the contribution from the Selberg zeta function. It turns out that this main term boils down to something like lambda raised to the power of m minus one divided by log x times some other factors, which is related to the determinant of the Laplacian on X.
Kolya: That sounds like a pretty precise approximation. What about those random variables they study? I saw mentions of omega one in the abstract and introduction; what's that about?
Title and authors: Ada: They look at the function omega one defined on these multigeodesics, where if you have a multigeodesic with k components, omega one is k minus one. The paper shows that as x gets larger, this distribution of omega one starts behaving like a Poisson random variable with parameter log x.
Emmy: That's the central result for the component counts. They show that the normalized random variables for omega one converge in probability to one and they can be approximated by Poisson random variables with parameter log x as x goes to infinity <ref:2611.08165#pg1>. It's an analogue of something called the Erdős-Kac Theorem from analytic number theory.
Kolya: So, it’s not just that the count grows fast; it’s about how those components are distributed randomly. What does that tell us about the structure of these multigeodesics?
Ada: It tells us that for large x, you can predict the component structure quite well using Poisson statistics. This is based on studying a generating function FX,x(z) which sums over all primitive multigeodesics of length at most x on X.
Emmy: The way they connect that generating function to the Selberg zeta function is crucial here. They observe that in the region where the zeta function behaves nicely, it acts like ZX(s) times z, which lets them use a Perron formula to get an asymptotic for FX,x(z).
Kolya: So, if we're thinking about this from a computational standpoint, how does that help someone who just wants to know how many multigeodesics there are? Does it give us something practical?
Ada: It helps by giving us a way to calculate the leading order term with precision. Theorem four point six shows that P m(X, x) is approximated by a polynomial of degree m minus one evaluated at log n, which is essentially what we need to count these things efficiently for large values of x.
Emmy: And the error terms they provide are actually quite explicit. They give bounds in terms of spectral and geometric data of X, which means you can see exactly how much the approximation deviates from the true count based on properties like the eigenvalues of the Laplacian.
Kolya: That is helpful because it grounds these abstract counting problems in actual measurable quantities about the surface X. But I wonder, what's actually left out? What's a limitation they admitted?
Title and authors: Ada: They mention that their method relies on good bounds for the Selberg zeta function in that region where its real part is greater than one-half. That means their analysis stops working if you need to go into those parts of the complex plane where the zeta function doesn't behave as nicely.
Emmy: So they can handle this specific regime well, but that opens up questions about what happens when you look at the entire spectrum of zeros. It’s a limitation on the domain where their asymptotic formulas hold true.
Kolya: Okay, so to wrap up, we've covered how this paper "Counting multigeodesics and the Selberg-Delange method" uses known methods to count these curves on surfaces, providing both an asymptotic formula and Poisson approximation for their components. What does that mean for the broader field of geometry and number theory?
Ada: It shows a powerful application of the Selberg-Delange method to this new area. It demonstrates that you can translate techniques from prime factor counting directly into geometric counting on hyperbolic surfaces, linking the spectral data of X to arithmetic properties in a deep way.
Emmy: And the convergence to Poisson distributions for component counts is interesting because it suggests a random structure underlying these geometric objects that is statistically well-behaved in the limit. It gives us tools to model complexity arising from interactions between these geodesic paths.
Kolya: For someone just listening, this paper shows you how deep connections are hidden between counting curves on surfaces and the distribution of numbers with prime factors. It's a nice piece of machinery for those who want to see how these seemingly separate fields interact.
Ada: Exactly. We've looked at the title, the method used, and what it actually achieves in terms of asymptotics for P m(X, x) and the distribution of omega one on multigeodesics. That's what this paper is about.
Emmy: It really highlights how much structure is hidden within these counts when you apply the right analytic machinery to something as complex as a hyperbolic surface. We can't help but see that connection between the zeta function and the geometry of X through this work on counting multigeodesics and the Selberg-Delange method.
Kolya: Yeah, it gives us a solid foundation for how we might approach similar counting problems in other areas where we have similar analytic tools. It's a good piece of machinery for future research.
Ada: We'll leave it there for today, focusing on this paper about counting multigeodesics and the Selberg-Delange method. We’ve covered the main results and what they mean for understanding these curves on surfaces, but we’ll pick up another paper next time.
The paper's summary: Ada: So, to recap, this paper uses the Selberg-Delange method to count these multigeodesics on a hyperbolic surface X and shows that as x gets bigger, the count settles into a predictable large x formula that depends on something related to X.
Emmy: Exactly. They are taking a counting problem from number theory—like finding integers with m prime factors—and mapping it onto geometry by looking at paths on the surface. The main result is that for a fixed number of components, say k, the distribution of those components behaves like a Poisson random variable with parameter log x.
Kolya: That’s interesting because normally in these counting problems, you get very rough estimates. But this paper gives us something much better on the component level. So if we think about randomly picking a multigeodesic and looking at how many pieces it has, we can use Poisson statistics to predict that count with reasonable accuracy for large lengths.
Ada: Right. And they do this by studying a generating function called FX,x(z). They find that in the right part of the complex plane where the zeta function is well-behaved, this function behaves nicely enough to allow them to extract an asymptotic formula for P m(X, x).
Emmy: The structure they uncover is that the counting isn't just about a single number; it’s governed by a relationship with the Selberg zeta function, which itself encodes all the lengths of primitive closed geodesics on X. This ties the arithmetic counting directly to how many paths exist geometrically on that surface.
Kolya: So, for someone who only listens to this show, what does this mean for their work? It means that when you're trying to count complicated geometric objects on surfaces, you can use tools from analytic number theory like the Selberg-Delange method to get a concrete formula instead of just a vague idea.
Ada: It gives us a way to calculate that leading order term with precision. They show that the main part of this asymptotic count is given by something like lambda raised to the power of m minus one divided by log x times some other factors, which is related to the determinant of the Laplacian on X.
Emmy: And they make sure we know how much error there is. They give explicit bounds for those errors using spectral and geometric data about X, so you can see exactly how much the approximation drifts from reality based on things like the eigenvalues of that Laplacian.
Kolya: It's good to see that precision, but I wonder what happens if we go outside that well-behaved region they mentioned. They have to assume good bounds for the Selberg zeta function in a certain area, which means their counting results don't hold up everywhere on the complex plane related to the zeta function.
Ada: That’s a fair point. They flag that limitation clearly—the method works well when you are looking at those specific regions of the complex plane related to the zeros, but they can’t guarantee it for every possible input.
Emmy: So, while they get a very strong result for the main asymptotic behavior and component distribution in that specific regime, there’s still more work needed to prove that these results hold across the entire domain of interest.
Kolya: So, the implication is that we have a solid method for counting these multigeodesics when x is very large and m isn't too complicated relative to log x, but it leaves open questions about extending those counts to more difficult regimes on the surface.
The paper's improvements: Ada: So, to wrap up that section, the authors suggest some ways they can make their counting method even stronger or more useful for other kinds of problems.
Emmy: They are looking at how they can improve their ability to determine whether an asymptotic formula is specific to one particular surface or if it holds generally.
Kolya: What does that mean in practice? Does it mean they can figure out when the counting result depends on the surface X versus when it just relies on a general infinite product over all possible primitive closed geodesics?
Ada: They suggest that by looking at how the parameter m relates to log x, they can create a system where you can tell exactly which type of dependency you're dealing with.
Emmy: That connects back to the geometry; it means the method is more sensitive to the specific shape of X and less likely to be a generic formula that applies everywhere.
Kolya: So, if they improve that determination, what does that do for applying this counting machinery? Does it make it easier for other researchers who want to use Selberg-Delange methods in different areas?
Ada: Yes, it increases the capacity for computational number theory because they can use this framework to study the distribution of sums of multiplicative functions.
Emmy: It shows that these counting techniques aren't just isolated; they are a powerful way to model how prime factors and geometric paths interact.
Kolya: And regarding those error terms we talked about, do they suggest any way to handle them better? Can we get tighter bounds on the error terms in the asymptotic expansions?
Ada: They are trying to provide superior handling of spectral data integration by giving explicit upper bounds for those error terms. These bounds are given in terms of the spectral and geometric data of X, like the eigenvalues of that Laplacian.
Emmy: It means you can see exactly how much the approximation deviates from reality based on measurable properties about the surface itself, which is a big step toward more precise geometric analysis.
Kolya: So, it’s about making sure that when we use this method for things like modeling complex systems or counting objects, we get a better sense of the error involved in our prediction.
Ada: Exactly. It’s about having precision when you are using this sophisticated machinery to count these paths on surfaces.
Emmy: And they are also focusing on how to extend the results beyond just the large x limit, trying to understand the behavior at smaller scales too.
Kolya: That’s a good direction because it means their tools aren't just for theoretical limits; they can be used for more detailed analysis of the system.
Conclusion: Ada: We've covered how this paper, "Counting multigeodesics and the Selberg-Delange method," uses number theory tools to count paths on hyperbolic surfaces and gets a Poisson approximation for their component counts.
Emmy: It really shows that geometry and arithmetic are deeply linked through these zeta functions, giving us a concrete way to count geometric objects using analytic methods.
Kolya: I think the main thing is that we have a more reliable way to estimate those component counts than just guessing based on the number of components you see in an average object.
Ada: Right. And they give us that error term information tied directly to the surface's geometry, so we know exactly how much our counting formula might be off.
Emmy: It moves this from a pure theoretical exercise into something that could actually inform more applied geometric problems where you need reliable estimates.
Kolya: For me, what matters is that they show the convergence to Poisson distributions for the component counts, which gives us a statistical model for these paths when x is large.
Ada: And they did all this by studying that generating function FX,x(z) and relating it back to the Selberg zeta function in those specific regions.
Emmy: This paper lays a foundation for using the Selberg-Delange method to analyze more complex counting problems involving prime factors on surfaces.
Kolya: It suggests that we can use this analytic approach to study sums of multiplicative functions in ways that are very hard to tackle otherwise.
Ada: So, yeah, "Counting multigeodesics and the Selberg-Delange method" gives us a solid framework for counting these curves on surfaces and understanding their component structure statistically.
Emmy: It highlights how much structure is hidden within these counts when you apply the right analytic machinery to something as complex as a hyperbolic surface.
Kolya: It’s a piece of machinery that bridges geometry and number theory in a way that opens up new avenues for future research on spectral data and counting problems.
Aapo Karvonen
Royal Society Dorothy Hodgkin Fellowship · Isaac Newton Institute for Mathematical Sciences, Cambridge
math.GT, math.NT, math.SP
Submitted: 2026-11-05
Updated: 2026-11-05
Comments: 23 pages, 1 figure
License: http://creativecommons.org/licenses/by/4.0/
Importance score: 76/100
The gist: We study the counting function Pm(X, x) for m-multigeodesics of length at most x on a compact hyperbolic surface X where m is a positive integer and prove a large x asymptotic for Pm(X, x).
Key concepts
- m-multigeodesics
- These are paths on a hyperbolic surface composed of $m$ distinct, primitive closed geodesics joined together. The paper studies how many such paths exist up to a certain length x.
- Selberg-Delange Method
- This is a mathematical technique adapted from number theory used here to analyze the asymptotic behavior of the counting function for multigeodesics by relating it to the Selberg zeta function.
- Poisson Random Variable Approximation
- The paper shows that as x increases, the distribution of components in a random multigeodesic closely follows a Poisson distribution with mean $ ext{log} x$. This means the number of components tends to behave like this specific probabilistic model.
Terminology
Summary
We study the counting function Pm(X, x) for m-multigeodesics of length at most x on a compact hyperbolic surface X where m is a positive integer and prove a large x asymptotic for Pm(X, x). p. 0 Counting multigeodesics and the Selberg-Delange method
Aapo Karvonen
Abstract
We study the counting function P
m(X, x) for m-multigeodesics of length at most x on a compact
hyperbolic surface X where m is a positive integer. Building on the Selberg-Delange method, originally
developed for counting integers with exactly m prime factors in the interval [1, x], we prove a large x
asymptotic for P
m(X, x). , γk) a k-multigeodesic. We show that as x → ∞, the
distribution of ω1 in the set of multigeodesics of length ≤ x is close to that of a Poisson random variable
with parameter log x.Where the paper says it
Where the paper says it
The Selberg-Delange Method Application
The study adapts the Selberg-Delange method to the context of the Selberg zeta function and investigates the asymptotics of Pm(X, x) 0 such that log x ≥ max log m − 1, log 1/ (1−s1), Pm(X, x) = e x / x(log m − 1) ν0,X(m − 1) / log x + O(x(-log x)(m − 3)!!), where the implied constant depends only on X. p. 1 ν0,X
m − 1
log x
+ OX
log x
m − 1
m−3
! ,
where the implied constant depends only on X, and
ν0,X (w) = (det ∆)−we
w(2g−2)E
wΓ(w)
Y
γ∈P(X)
Y∞
n=0
(1 + we−ℓγ
)(1 − e
−(n+1)ℓγ
)
w. The dependency on X in
the error term can be made explicit, and this is the content of Theorems 4. To be precise, up
to an exponentially smaller error, we show an asymptotic expansion in powers of 1
x whose coefficients are
polynomials in log x that depend on the surface.Where the paper says it
Where the paper says it
Distribution of Multigeodesic Components
The second main theorem concerns the limiting behavior of the function ω1(Γ) defined on primitive multigeodesics, which takes the value k − 1 for a multigeodesic with k components. The authors show that the sequence of uniform normalised random variables ωx,1 satisfies a central limit theorem and is approximable by Poisson random variables as x → ∞. p. 1 Our second main theorem is concerned with the limiting behaviour of the function ω1(Γ) defined on
primitive multigeodesics Γ, which takes the value k − 1 for a multigeodesic with k components. We show
that the sequence of uniform normalised random variables ωx,1 := ω1|GX(x)
, where GX(x) is the set of
multigeodesics of length at most x, satisfies a central limit theorem and is approximable by Poisson random
variables as x → ∞. Then the following hold for the
normalised random variables ωx,1, where x ≥ 0:
1. for any ϵ > 0,
limx→∞
P
ωx,1
log x
− 1
> ϵ
= 0,
i.Where the paper says it
Where the paper says it
Leading Order Term Derivation
The leading order term for Pm(X, x) is derived by considering the m-th coefficient of the Taylor series of FX,x(z), which is Pm(X, x). By choosing r = m − 1 log x in Cauchy's integral formula and using estimates from Proposition 3.1, Theorem 4. p. 0 The primitive length spectrum
shares many similarities with prime numbers. For example, the Prime Geodesic Theorem by Huber [6] states
that
P(X, x) = Li(x) +X
sj
Li(x
sj
) + OX
x
3/4
log x
,
where P(X, x) = {γ ∈ P(X) : ℓγ ≤ log x}, P(X) is the set of primitive oriented closed geodesics on
X, ℓγ denotes the length of the geodesic γ, and sj =
1
2 +
q
1
4 − λj are parameters corresponding to the
eigenvalues λj of the Laplace-Beltrami operator ∆X in the interval (0,
1
4
) (note that this set may be empty). For s = σ + iT with σ > 1
|ZX(s)| ≪ e
Cg
sys X+
1
(σ − 1)C
+
Y
sj
1
(σ − sj )
C
+
,
where (σ − sj )
C
+ = min{1,(σ − sj )
C }. For s = σ + iT with 1
2 + ϵ < σ ≤ 1 and |T| >
3
2
|ZX(s)| ≪ϵ log(|T|) exp
CC1(X)T
2−2σ
1−2ϵ
.Where the paper says it
Where the paper says it
Mod-Poisson Convergence and Central Limit Theorem
The final section proves Theorem 1.2 by showing mod-Poisson convergence of the random variables ωx,1 with parameters log x. p. 2 2, we study FX,x(z) as a characteristic function. To be
precise, we show that the sequence ωx,1 for x ≥ 0 of random variables satisfies what is known as mod-Poisson
convergence, which is a mode of convergence of random variables phrased in terms of their characteristic
functions. In the following subsections, we derive each of the three estimates in Proposition 3. 1 The first estimate
For the first estimate, we can bound the absolute value of the logarithm immediately. 2, we study FX,x(z) as a characteristic function. To be
precise, we show that the sequence ωx,1 for x ≥ 0 of random variables satisfies what is known as mod-Poisson
convergence, which is a mode of convergence of random variables phrased in terms of their characteristic
functions. Our second main theorem is concerned with the limiting behaviour of the function ω1(Γ) defined on
primitive multigeodesics Γ, which takes the value k − 1 for a multigeodesic with k components. We show
that the sequence of uniform normalised random variables ωx,1 := ω1|GX(x)
, where GX(x) is the set of
multigeodesics of length at most x, satisfies a central limit theorem and is approximable by Poisson random
variables as x → ∞.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Conclusion on Leading Order
The leading order term for Pm(X, x) is given by P0,m,X(x) = 1 / (2πi Z z=r ν0(z) x z z m dz for r > 0 small enough. This term simplifies to ν0(m − 1) / log x log(m − 1) x (m − 1)! p. 1 In fact, one should compare ν0,X with the function λ in (1), and interpret it
as a phenomenon arising in the regime where m−1
log x
is bounded away from zero. The dependency on X in
the error term can be made explicit, and this is the content of Theorems 4. The number theoretic analogue of P
m(X, x) is the counting function πm(x) which denotes the number of
integers ≤ x with exactly m distinct prime factors. This was originally studied by Landau [9], who obtained
the leading order asymptotic
πm(x) ∼
x
log x
(log log x)
m−1
(m − 1)! A later result of Sathe [15], [16] and Selberg [17] gives the asymptotic for m ≤ B log log x for
any B > 0 fixed
πm(x) = x
log x
(log log x)
m−1
(m − 1)! λ
m − 1
log log x
+ OB
m
(log log x)
2
, (1)
where the leading coefficient is given by the following function evaluated at m−1
log log x
λ (z) = 1
Γ(z + 1)
Y
p prime
1 +
z
p − 1
−1
1 −
1
p
z
,
and Γ denotes the gamma function.Where the paper says it
Where the paper says it
Where the paper says it
References
[1] A. D. Barbour, E. Kowalski, and A. p. 2 To be
precise, we show that the sequence ωx,1 for x ≥ 0 of random variables satisfies what is known as mod-Poisson
convergence, which is a mode of convergence of random variables phrased in terms of their characteristic
functions. 2 then follows immediately using a result of Kowalski and Nikeghbali [8] connecting
mod-Poisson convergence to the existence of a central limit theorem. To be
precise, we show that the sequence ωx,1 for x ≥ 0 of random variables satisfies what is known as mod-Poisson
convergence, which is a mode of convergence of random variables phrased in terms of their characteristic
functions. 2 then follows immediately using a result of Kowalski and Nikeghbali [8] connecting
mod-Poisson convergence to the existence of a central limit theorem.Where the paper says it
Where the paper says it
[2] J. Bourgain “Decoupling, Exponential Sums and the Riemann Zeta Function” in J. p. 1 λ
m − 1
log log x
+ OB
m
(log log x)
2
, (1)
where the leading coefficient is given by the following function evaluated at m−1
log log x
λ (z) = 1
Γ(z + 1)
Y
p prime
1 +
z
p − 1
−1
1 −
1
p
z
,
and Γ denotes the gamma function. Selberg’s method was later developed by Delange, and the SelbergDelange method has since found many applications in the study of partial sums of multiplicative functions.Where the paper says it
[3] P. Buser Geometry and Spectra of Compact Riemann Surfaces Birkh¨auser Boston 2010
[4] G. H Hardy and M. Riesz The general theory of Dirichlet’s series Vol 18 Cambridge Tracts in Mathematics and Mathematical Physics Stechert-Hafner, Inc. p. 3 The author is supported by Laura Monk’s Royal Society Dorothy Hodgkin Fellowship. The author would
like to thank the Isaac Newton Institute for Mathematical Sciences, Cambridge, for support and hospitality
during the programme Geometric spectral theory and applications, where some work on this paper was
undertaken.Where the paper says it
[5] D. A. Hejhal “The Selberg Trace Formula for PSL(2, R) 1” in Lecture Notes in Math. p. 4 Namely,
log ZX(s) = X
γ
X∞
k=0
log(1 − e
−(s+k)ℓγ
) = X
γ
X∞
n=1
−1
n
e
−nsℓγ
1 − e−nℓγ
(8)
and
Z
′
X(s)
ZX(s)
=
X
γ
X∞
k=1
ℓγ
2 sinh(ℓγk/2) e
−(s− 1
2
)ℓγ k
. (9)
Moreover, using the Selberg trace formula (see [5]), we have the following spectral expansion for the logarithmic derivative which holds for all s ∈ C:
1
2s − 1
Z
′
X(s)
ZX(s)
=
1
2β
Z
′
X(
1
2 + β)
ZX(
1
2 + β)
+
X
n≥0
1
r
2
n + (s −
1
2
)
2
−
1
r
2
n + β
2
+
µ(X)
2π
X∞
k=0
1
β +
1
2 + k
−
1
s + k
, (10)
where β ≥ 2 is an arbitrary real number that we shall consider fixed for the rest of this paper.Where the paper says it
[6] H Huber Zur analytischen Theorie hyperbolischen Raumformen und Bewegungsgruppen Math. Ann. 138 (1959) pp 1–26
[7] H.-K Hwang Asymptotics of poisson approximation to random discrete distributions an analytic approach Adv. p. 2 One can interpret the above theorem as an analogue of the classical Erd˝os-Kac Theorem in analytic
number theory (see [18]). In fact, similar Poisson approximation holds also for other natural distances such
as the total variational or Wasserstein distance.Where the paper says it
[8] E Kowalski and A Nikeghbali Mod-Poisson convergence in probability and number theory Int. Math Res Not IMRN 18 (2010) pp 3549–3587
[9] E Landau Handbuch der Lehre von der Verteilung der Primzahlen 2 B¨ande Chelsea Publishing Co., New York 1953
[10] H Nakaya The generalized divisor problem and the Riemann hypothesis in Nagoya Math. J. 122 (1991) pp 149–159
[12] B Randol The Riemann Hypothesis for Selberg’s Zeta-Function and the Asymptotic Behavior of Eigenvalues of the Laplace Operator Trans. Amer. Math. Soc. 236 (1978) pp 209–223
[13] Z Rudnick Closed geodesics in short intervals for random hyperbolic surfaces arXiv:2605.22059 May 2026
[14] P Sarnak Determinants of Laplacians in Comm. Math. Phys. 110 (1987) pp 113–120
[15] L G Sathe On a problem of Hardy on the distribution of integers having a given number of prime factors I in J. Indian Math. p. 0 In this paper, we consider the following counting function
P
m(X, x) = {Γ ∈ GX,m : ℓΓ ≤ x},
1
arXiv:2611. 08165v1 [math. In this paper, we consider the following counting function
P
m(X, x) = {Γ ∈ GX,m : ℓΓ ≤ x},
1
arXiv:2611. 08165v1 [math. In this paper, we consider the following counting function
P
m(X, x) = {Γ ∈ GX,m : ℓΓ ≤ x},
1
arXiv:2611. 08165v1 [math.Where the paper says it
Where the paper says it
Where the paper says it
[16] L G Sathe On a problem of Hardy on the distribution of integers having a given number of prime factors II in J. Indian Math. Soc. (N.S.) 17 (1953) pp 83–141
[17] A Selberg Note on a paper by L G Sathe in J. Indian Math. Soc. (N.S.) 18 (2004) pp 83–87
[18] G Tenenbaum Introduction to analytic and probabilistic number theory Third edition Grad. Stud. in Math volume 163 Amer. p. 0 In this paper, we consider the following counting function
P
m(X, x) = {Γ ∈ GX,m : ℓΓ ≤ x},
1
arXiv:2611. 08165v1 [math.Where the paper says it
[20] Y Wu and Y Xue Prime geodesic theorem and closed geodesics for large genus in J Eur Math Soc (JEMS) (2025)
Page 4, five or more words copied word for word from that page
The primitive length spectrum of X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity. p. 0 1 Introduction
Consider a closed, oriented, connected, hyperbolic surface X of genus g. The primitive length spectrum of
X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity. 1 Introduction
Consider a closed, oriented, connected, hyperbolic surface X of genus g. The primitive length spectrum of
X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity. 1 Introduction
Consider a closed, oriented, connected, hyperbolic surface X of genus g. The primitive length spectrum of
X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity.Where the paper says it
Where the paper says it
Where the paper says it
The Prime Geodesic Theorem by Huber [6] states that P(X, x) = Li(x) +X sj Li(x sj) + OX x 3/4 log x. p. 0 The primitive length spectrum
shares many similarities with prime numbers. For example, the Prime Geodesic Theorem by Huber [6] states
that
P(X, x) = Li(x) +X
sj
Li(x
sj
) + OX
x
3/4
log x
,
where P(X, x) = {γ ∈ P(X) : ℓγ ≤ log x}, P(X) is the set of primitive oriented closed geodesics on
X, ℓγ denotes the length of the geodesic γ, and sj =
1
2 +
q
1
4 − λj are parameters corresponding to the
eigenvalues λj of the Laplace-Beltrami operator ∆X in the interval (0,
1
4
) (note that this set may be empty).Where the paper says it
The Prime Number Theorem in its simplest form (see e. p. 0 For example, the Prime Geodesic Theorem by Huber [6] states
that
P(X, x) = Li(x) +X
sj
Li(x
sj
) + OX
x
3/4
log x
,
where P(X, x) = {γ ∈ P(X) : ℓγ ≤ log x}, P(X) is the set of primitive oriented closed geodesics on
X, ℓγ denotes the length of the geodesic γ, and sj =
1
2 +
q
1
4 − λj are parameters corresponding to the
eigenvalues λj of the Laplace-Beltrami operator ∆X in the interval (0,
1
4
) (note that this set may be empty). In comparison, the Prime Number Theorem in its simplest form (see, e. The primitive length spectrum
shares many similarities with prime numbers. For example, the Prime Geodesic Theorem by Huber [6] states
that
P(X, x) = Li(x) +X
sj
Li(x
sj
) + OX
x
3/4
log x
,
where P(X, x) = {γ ∈ P(X) : ℓγ ≤ log x}, P(X) is the set of primitive oriented closed geodesics on
X, ℓγ denotes the length of the geodesic γ, and sj =
1
2 +
q
1
4 − λj are parameters corresponding to the
eigenvalues λj of the Laplace-Beltrami operator ∆X in the interval (0,
1
4
) (note that this set may be empty).Where the paper says it
Where the paper says it
This analogy is further amplified when one considers the Selberg zeta function defined for s ∈ C, R(s) > 1 as ZX(s) = Y γ∈P(X) Y∞ n=0 (1 − e -(s+n)lγ), where the outer product is over primitive oriented closed geodesics on X. p. 0 [18]) states that
π(x) ∼ Li(x),
where π(x) denotes the number of prime numbers up to x. This analogy is further amplified when one considers the Selberg zeta function defined for s ∈ C, ℜ(s) > 1
as
ZX(s) = Y
γ∈P(X)
Y∞
n=0
(1 − e
−(s+n)ℓγ
),
where the outer product is over primitive oriented closed geodesics on X.Where the paper says it
This is an entire function whose non-trivial zeros lie on the set 1/2 + it: t ∈ R ∪ (0, 1) and are given by 1/2 + q i /4 − λ j where λ j are the nontrivial eigenvalues of ∆X. p. 0 This analogy is further amplified when one considers the Selberg zeta function defined for s ∈ C, ℜ(s) > 1
as
ZX(s) = Y
γ∈P(X)
Y∞
n=0
(1 − e
−(s+n)ℓγ
),
where the outer product is over primitive oriented closed geodesics on X. This is an entire function whose
non-trivial zeros lie on the set {
1
2 + it : t ∈ R} ∪ (0, 1) and are given by 1
2 +
q
1
4 − λj where λj are the nontrivial eigenvalues of ∆X.Where the paper says it
Page 5, five or more words copied word for word from that page
The counting function P m(X, x) = GX,m = Y Γ = Y distinct primitive oriented closed geodesics on X and lΓ = lγ1 + … + lγm is the set of unordered distinct primitive m-multigeodesics on X and lΓ = lγ1 + … + lγm. The number theoretic analogue of P m(X, x) is the counting function πm(x) which denotes the number of integers ≤ x with exactly m distinct prime factors. p. 1 Since we take the
geodesics to be oriented, γ and γ
−1 are considered distinct. The number theoretic analogue of P
m(X, x) is the counting function πm(x) which denotes the number of
integers ≤ x with exactly m distinct prime factors. for m fixed. A later result of Sathe [15], [16] and Selberg [17] gives the asymptotic for m ≤ B log log x for
any B > 0 fixed
πm(x) = x
log x
(log log x)
m−1
(m − 1)! A later result of Sathe [15], [16] and Selberg [17] gives the asymptotic for m ≤ B log log x for
any B > 0 fixed
πm(x) = x
log x
(log log x)
m−1
(m − 1)! λ
m − 1
log log x
+ OB
m
(log log x)
2
, (1)
where the leading coefficient is given by the following function evaluated at m−1
log log x
λ (z) = 1
Γ(z + 1)
Y
p prime
1 +
z
p − 1
−1
1 −
1
p
z
,
and Γ denotes the gamma function.Where the paper says it
Where the paper says it
Where the paper says it
Page 6, five or more words copied word for word from that page
We adapt the Selberg-Delange method to the context of the Selberg zeta function and study the asymptotics of P m(X, x). For the statement, let s1 = 1/2 + q i /4 − λ i be the largest non-trivial zero of ZX(s), and let E = 2ζ'(-1) − 1/4 + 1/2 log 2π ≈ 0.3381 be a fixed constant. p. 1 For the statement, let s1 =
1
2 +
q
1
4 − λ1 be the largest non-trivial zero of ZX(s), and
let
E = 2ζ
′
(−1) −
1
4
+
1
2
log 2π ≈ 0. 3381
be a fixed constant. Recall also the zeta-regularised determinant det ∆ (see [14]), which can be related to
ZX(s) by Z
′
X(1) = det ∆ e
−(2g−2)E. Our main theorem about P
m(X, x) then reads as follows.Where the paper says it
Where the paper says it
Page 7, five or more words copied word for word from that page
The distribution of the function ω1, defined on multigeodesics Γ as ω1(Γ) = k − 1 for Γ = (γ1, …, γk) a k-multigeodesic is close to that of a Poisson random variable with parameter log x as x → ∞. Both of these theorems rely on studying the generating function of the random variable ω1 which we relate to the Selberg zeta function. p. 0 We show that as x → ∞, the
distribution of ω1 in the set of multigeodesics of length ≤ x is close to that of a Poisson random variable
with parameter log x. Both of these theorems rely on studying the generating function of the random
variable ω1 which we relate to the Selberg zeta function.Where the paper says it
Page 8, five or more words copied word for word from that page
The proofs of the above theorems are based on the study of the function FX,x(z) = X Γ∈GX(x) z ω(Γ), where the sum is over all primitive multigeodesics of length at most x on the surface X, z ∈ C, and ω(Γ) = ω1 + 1.
Page 9, five or more words copied word for word from that page
The key observation is that GX(s, z) behaves like ZX(s) z in the region R(s) > 1/2. Thus, we can use the Perron formula to derive an asymptotic for FX,x(z), which relies on having good bounds for ZX(s) in the region R(s) > 1/2. p. 2 Thus, we can use the Perron formula to derive an asymptotic for FX,x(z). This relies
on having good bounds for ZX(s) in the region ℜ(s) >
1
2
, and a major component of the paper is to derive
bounds that are explicit in terms of the geometry of X.Where the paper says it
Page 10, five or more words copied word for word from that page
The spirit of the proof is inspired by the classical Selberg-Delange method in number theory. The new challenges arise from the fact that ZX(s) grows much faster than ζ(s) on the critical strip. p. 2 The spirit of the proof is inspired by the classical Selberg-Delange method in number theory. The new
challenges arise from the fact that ZX(s) grows much faster than ζ(s) on the critical strip. The new
challenges arise from the fact that ZX(s) grows much faster than ζ(s) on the critical strip. In particular, even
without assuming the Riemann hypothesis, we have the bound ζ(s) = o(|t|) for s = σ + it with 1
2 ≤ |σ| ≤ 1
(see [2] for a more precise result).Where the paper says it
Where the paper says it
Page 3, five or more words copied word for word from that page
Consider a closed, oriented, connected, hyperbolic surface X of genus g. The primitive length spectrum of X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity. A geodesic is primitive if it cannot be written as γ k for some k ≥ 2 and closed geodesic γ. p. 0 The primitive length spectrum of
X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity. A geodesic is
primitive if it cannot be written as γ
k
for some k ≥ 2 and closed geodesic γ.Where the paper says it
The Prime Geodesic Theorem by Huber [6] states that P(X, x) = Li(x) +X sj Li(x sj) + OX x 3/4 log x, where P(X, x) = the number of primitive oriented closed geodesics on X of length at most x.
The counting function P m(X, x) = GX,m = Y Γ = Y distinct primitive oriented closed geodesics on X and lΓ = lγ1 + … + lγm is the set of unordered distinct primitive m-multigeodesics on X and lΓ = lγ1 + … + lγm.
Page 2, five or more words copied word for word from that page
Consider a closed, oriented, connected, hyperbolic surface X of genus g. The primitive length spectrum of X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity.
Consider a closed, oriented, connected, hyperbolic surface X of genus g. The primitive length spectrum of X is defined as the set of lengths of primitive oriented closed geodesics on X with multiplicity.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved asymptotic analysis for counting multigeodesics allows systems to determine whether an asymptotic is surface-independent or depends on an infinite Euler product over primitive closed geodesics based on the value of m relative to log x.
-
Enhanced ability to predict the distribution of component counts in random multigeodesics, as Theorem 1.2 shows that the normalized variable omega x,1 converges in probability to 1 and is approximable by Poisson random variables with parameter log x.
-
Increased capacity for computational number theory applications by using the Selberg-Delange method to study the distribution of sums of multiplicative functions, as mentioned in the abstract and introduction.
-
Superior handling of spectral data integration by providing explicit upper bounds for the error terms in asymptotic expansions, which are given in terms of spectral-geometric data of X.
-
Precision in approximating the leading order term for counting multigeodesics, allowing systems to calculate P m(X, x) using a polynomial of degree m minus 1 evaluated at log n as shown in Theorem 4.6.
Abstract
We study the counting function P m(X,x) for m-multigeodesics of length at most x on a compact hyperbolic surface X where m is a positive integer. Building on the Selberg-Delange method, originally developed for counting integers with exactly m prime factors in the interval [1,x], we prove a large x asymptotic for P m(X,x). We show that for fixed m as x to infinity the asymptotic is independent of the surface, whereas for m over x bounded away from zero, the asymptotic depends on an infinite Euler product over the lengths of primitive closed geodesics. We also study the distribution of the function ω 1, defined on multigeodesics Γ as ω 1(Γ)=k-1 for Γ=(γ 1,...,γ k) a k-multigeodesic. We show that as x to infinity, the distribution of ω 1 in the set of multigeodesics of length at most x is close to that of a Poisson random variable with parameter x. Both of these theorems rely on studying the generating function of the random variable ω 1 which we relate to the Selberg zeta function.
Sources
Related papers
- Uncomputable trisection invariants of 4-manifolds
- One internal stabilization is enough for resolving self-referential tubes
- The proper geometric dimension of the mapping class groups of non-orientable surfaces with punctures
- 2-Linearizability of Geometric 3-Manifold Groups Over Commutative Rings
- A Faithful Action of the Mapping Class Group Mod(S g) on the (2g-1) -Sphere
- Profinite Non-Rigidity of Arithmetic Lattices and the K"ahler Property