Heights of Drinfeld modular polynomials and Hecke images

arXiv:2410.11132 · math.NT · Submitted 2024-10-14 · Read on arXiv

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: I'm Ada.

Emmy: And I'm Emmy, here with Kolya.

Ada: We're looking at the paper: "Heights of Drinfeld modular polynomials and Hecke images".

Emmy: Here's the gist of the paper. We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq

t: <ref:2410.11132#pg5> These polynomials vanish at pairs of j-invariants of Drinfeld Fq

t: -modules of rank 2 linked by cyclic isogenies of degree N <ref:2410.11132#pg10> The main term in both bounds is asymptotically optimal as deg N tends to infinity <ref:2410.11132#pg8> We also obtain precise estimates on the Weil height and Taguchi height of Hecke images of Drinfeld modules of rank 2 <ref:2410.11132#pg9> Estimating the Modular Polynomial Height The primary goal is to estimate the size of the coefficients of ΦN, defined by its height h(f):= log max c c, where c ranges over all coefficients of f <ref:2410.11132#pg3> The authors establish an explicit, asymptotically optimal estimate for h(ΦN), stating that the height satisfies h(ΦN) = q squared − 1/2 ψ(N) (deg N − 2λN + bq(N)) <ref:2410.11132#pg8> This result is derived by estimating the Mahler measure of ΦN (X, j(z)) and reducing the problem to estimating the sum of logarithms of “imaginary parts” of the representatives of the Hecke images <ref:2410.11132#pg4> Relating Heights via Isogenies The paper proves several height comparisons between a Drinfeld module and its images under a Hecke correspondence <ref:2410.11132#pg5> Specifically, Theorem 1.2 provides an analog of a result of Silverman

Sil90: , comparing the Weil height of a j-invariant with the height of its Hecke image <ref:2410.11132#pg9> Furthermore, Theorem 1.3 shows that the stable Taguchi height gives cleaner expressions than the Weil height when estimating heights of Hecke images <ref:2410.11132#pg10> Geometric Tools for Estimates The proof relies on several geometric tools related to Drinfeld modules and function fields <ref:2410.11132#pg7> These include the Bruhat-Tits tree, which is used to control "large" representatives from the Hecke images <ref:2410.11132#pg6> The building map λ on omega is defined, and Proposition 2.15 states that if λ(z) ∈ X(T) and λ(z) is represented by the matrix π k u 0 1, then − log zi = k <ref:2410.11132#pg8> Function field Farey sequences are employed to provide explicit approximations for the representatives of the Hecke images <ref:2410.11132#pg7> Final Bounds and Specialization The main result, Theorem 1.1, provides a lower bound of q squared − 1/2 ψ(N) (deg N − 2λN) and an upper bound involving the constant bq(N) <ref:2410.11132#pg8> The final estimate for the height is bounded by q squared − 1/2 ψ(N) (deg N − 2λN + 4 + 2q cubed + q squared − 2q − 1 / (q(q - 1) 2)) <ref:2410.11132#pg8> This is achieved by combining bounds derived from the analysis of the sum over large dγ <ref:2410.11132#pg8> The result is obtained by specializing at a single suitable value of j, as shown in Lemma 3.6 <ref:2410.11132#pg8>.

Ada: Comparison of Heights The paper concludes by proving Theorem 4.3, which establishes the equality log max(1, j(φ)v) = 1/ψ(N) X C log max(1, j(φ/C)v) <ref:2410.11132#pg9> This comparison relates the height of a Drinfeld module to the average height of its cyclic N-submodules <ref:2410.11132#pg9> The proof uses Lemma 3.7, which implies that for z ∈ F and log j(z) ≤ q, then zi = 1 and log ∆(z) = q squared <ref:2410.11132#pg8> This leads to the final bound of h(ΦN (X, Y)) ≥ q squared − 1/2 ψ(N) (deg N − 2λN) <ref:2410.11132#pg8>.

Emmy: Hecke Image Height Comparison The proof of Theorem 4.3 involves relating the quantity to the height of the specialized modular polynomial ΦN (X, j(z)) <ref:2410.11132#pg8> The upper bound derived from this specialization is h(ΦN (X, j(z))) ≤ h(ΦN) + ψ(N) log max 1, j(z) <ref:2410.11132#pg8> This ultimately leads to the final inequality involving the minimum of terms related to the growth rate <ref:2410.11132#pg8>.

Ada: The lower bound is obtained by using Lemma 3.6 and Lemma 3.7 <ref:2410.11132#pg8> This completes the proof of Theorem 4.3 in terms of the AM-GM inequality <ref:2410.11132#pg8>.

Paper summary: Ada: The paper starts by defining the height of a polynomial f as the logarithm of the maximum absolute value of its coefficients. They aim to estimate this height for ΦN, which is defined by those vanishing conditions >

Emmy: They establish an explicit, asymptotically optimal estimate for this height. It says it satisfies a bound involving q squared minus one half psi N times the degree of N minus two lambda N plus some term bq(N) >

Kolya: That explicit estimate is good, but they get there by estimating the Mahler measure of ΦN and then reducing it to summing up logarithms of imaginary parts from Hecke images >

Conclusion: Ada: So we’ve seen they provide these explicit bounds on the coefficients of Drinfeld modular polynomials ΦN and connect them to the height of Hecke images like this. The authors are essentially giving us a very precise way to measure how large these numbers can get for any given N >

Emmy: It really connects the geometry—the structure of rank two Drinfeld modules and their isogenies—with these algebraic invariants, the heights, which tells us about the complexity of those structures >

Kolya: What this suggests is that we have a much clearer picture of the growth rate for these coefficients than before. It shows how well the Hecke correspondence works as a way to average out some of that growth >

FLORIAN BREUER, FABIEN PAZUKI, ZHENLIN RAN

IRN GandA (CNRS) · ANR-20-CE40-0003 Jinvariant · Alexander von Humboldt Foundation

math.NT

Submitted: 2024-10-14

Updated: 2026-11-10

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 78/100

The gist: We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t] <ref:2410.11132#pg5> These polynomials vanish at pairs of…

Key concepts

Drinfeld modular polynomials $\Phi_N$
These are polynomials whose coefficients encode information about pairs of $j$-invariants from rank 2 Drinfeld modules linked by cyclic isogenies of degree $N$. The paper estimates the size (height) of these coefficients to understand the complexity associated with these geometric objects.
Weil height $h(f)$
The Weil height measures the arithmetic complexity of a function or polynomial. For a polynomial, it is defined using an absolute value $|x| := q^{\deg x}$ corresponding to the place at infinity. The paper uses this to bound the size of coefficients in $\Phi_N$.
Hecke images
These are specific points or invariants obtained by applying Hecke correspondences to Drinfeld modules. The authors compare the height of a Drinfeld module's $j$-invariant with the height of these Hecke images, providing comparisons using both Weil and stable Taguchi heights.

Terminology

Summary

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t] <ref:2410. p. 0

Where the paper says it

NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract. We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t].

11132#pg5> These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N <ref:2410. p. 0
Where the paper says it

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t]. These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N.

11132#pg10> The main term in both bounds is asymptotically optimal as deg N tends to infinity <ref:2410. p. 0
Where the paper says it

These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N. The main term in both bounds is asymptotically optimal as deg N tends to infinity.

11132#pg8> We also obtain precise estimates on the Weil height and Taguchi height of Hecke images of Drinfeld modules of rank 2 <ref:2410. p. 0
Where the paper says it

The main term in both bounds is asymptotically optimal as deg N tends to infinity. We also obtain precise estimates on the Weil height and Taguchi height of Hecke images of Drinfeld modules of rank 2.

11132#pg9>

Estimating the Modular Polynomial Height

The primary goal is to estimate the size of the coefficients of ΦN, defined by its height h(f):= log max c c, where c ranges over all coefficients of f <ref:2410.11132#pg3> The authors establish an explicit, asymptotically optimal estimate for h(ΦN), stating that the height satisfies h(ΦN) = q squared − 1/2 ψ(N) (deg N − 2λN + bq(N)) <ref:2410. p. 1

Where the paper says it

Hsia proved in [Hsi98] that, as deg N → ∞, the height of ΦN satisfies h(ΦN ) = q 2 − 1 2 ψ(N) (deg N − 2κN + O(1)), where κN := X P|N deg P |P| , but the bounded term O(1) was not made explicit. Explicit upper bounds on h(ΦN ) appear in [BL97] and [BPR21], but neither of these results are asymptotically optimal.

11132#pg8> This result is derived by estimating the Mahler measure of ΦN (X, j(z)) and reducing the problem to estimating the sum of logarithms of “imaginary parts” of the representatives of the Hecke images <ref:2410. p. 1
Where the paper says it

The general idea is to estimate the Mahler measure of ΦN (X, j(z)) for some chosen z. Then using the result that the geometric mean of the Hecke image of the discriminant function equals the discriminant function itself, we can reduce our proof to estimating the sum of the logarithms of “imaginary parts” (see equation (3) for the definition) of the representatives of the Hecke images of the corresponding Drinfeld modules on the upper half plane.

11132#pg4>

Relating Heights via Isogenies

The paper proves several height comparisons between a Drinfeld module and its images under a Hecke correspondence <ref:2410.11132#pg5> Specifically, Theorem 1. p. 0

Where the paper says it

They are used for constructing towers of curves with many rational points, with applications in coding theory, see for example [Elk01]. These curves also parametrise elliptic curves over function fields via a theorem in analogy with

2 provides an analog of a result of Silverman [Sil90], comparing the Weil height of a j-invariant with the height of its Hecke image <ref:2410. p. 2
Where the paper says it

1 for definitions. In our second main result, we prove the following analog of a result of Silverman [Sil90], comparing the Weil height of a j-invariant with the height of its Hecke image.

11132#pg9> Furthermore, Theorem 1. p. 7
Where the paper says it

It turns out that the functions z 7→ g(z) and z 7→ ∆(z) are Drinfeld modular forms of type 0 and weights q − 1 and q 2 − 1, respectively. Furthermore, ∆ is a cusp form which vanishes nowhere on Ω.

3 shows that the stable Taguchi height gives cleaner expressions than the Weil height when estimating heights of Hecke images <ref:2410. p. 2
Where the paper says it

Then q 2 − 1 2 2λN − deg N − bq(N) ≤ h(j(φ)) − 1 ψ(N) X C h(j(φ/C)) ≤ q + q 2 − 1 2 min 0, 2λN − deg N + log 1 q h(j(φ)) + 1 , where C ranges over all submodules C ⊂ φ[N] isomorphic to A/NA and bq(N) is the bounded term from Theorem 1. As a third result, we prove the following analog of a result of Autissier [Aut03], which shows that the stable Taguchi height gives us much cleaner expressions than the Weil height of the j-invariant when estimating the heights of Hecke images.

11132#pg10>

Geometric Tools for Estimates

The proof relies on several geometric tools related to Drinfeld modules and function fields <ref:2410.11132#pg7> These include the Bruhat-Tits tree, which is used to control large representatives from the Hecke images <ref:2410.11132#pg6> The building map λ on omega is defined, and Proposition 2. p. 3

Where the paper says it

Setting We gather in this section several crucial tools which will be used in the proofs of our main results. We start with some background in height theory, then we cover results on the Drinfeld fundamental domain, Bruhat-Tits trees, the building map, Drinfeld modular forms, and Farey sequences in the polynomial setting.

15 states that if λ(z) ∈ X(T) and λ(z) is represented by the matrix π k u 0 1, then − log zi = k <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8> Function field Farey sequences are employed to provide explicit approximations for the representatives of the Hecke images <ref:2410. p. 1
Where the paper says it

Then using the result that the geometric mean of the Hecke image of the discriminant function equals the discriminant function itself, we can reduce our proof to estimating the sum of the logarithms of “imaginary parts” (see equation (3) for the definition) of the representatives of the Hecke images of the corresponding Drinfeld modules on the upper half plane. Function fields Farey sequences are used for these estimates.

11132#pg7>

Final Bounds and Specialization

The main result, Theorem 1.1, provides a lower bound of q squared − 1/2 ψ(N) (deg N − 2λN) and an upper bound involving the constant bq(N) <ref:2410. p. 2

Where the paper says it

The upper bound in Theorem 1. 1 is probably not sharp, in particular we expect the constant 4 to be too large, but the method in place doesn’t yield better.

11132#pg8> The final estimate for the height is bounded by q squared − 1/2 ψ(N) (deg N − 2λN + 4 + 2q cubed + q squared − 2q − 1 / (q(q - 1) 2)) <ref:2410. p. 2
Where the paper says it

Then q 2 − 1 2 2λN − deg N − bq(N) ≤ h(j(φ)) − 1 ψ(N) X C h(j(φ/C)) ≤ q + q 2 − 1 2 min 0, 2λN − deg N + log 1 q h(j(φ)) + 1 , where C ranges over all submodules C ⊂ φ[N] isomorphic to A/NA and bq(N) is the bounded term from Theorem 1. As a third result, we prove the following analog of a result of Autissier [Aut03], which shows that the stable Taguchi height gives us much cleaner expressions than the Weil height of the j-invariant when estimating the heights of Hecke images.

11132#pg8> This is achieved by combining bounds derived from the analysis of the sum over large dγ <ref:2410.11132#pg8> The result is obtained by specializing at a single suitable value of j, as shown in Lemma 3. p. 8
Where the paper says it

The Drinfeld modular polynomial ΦN (X, Y ) ∈ C∞[X, Y ] is the one obtained by replacing each occurrence of j in the minimal polynomial of the function z 7→ j(N · z) over C∞(j) by an indeterminate Y . It was shown in [Bae92] that ΦN (X, Y ) has coefficients in A.

6 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>.

Comparison of Heights

The paper concludes by proving Theorem 4.3, which establishes the equality log max(1, j(φ)v) = 1/ψ(N) X C log max(1, j(φ/C)v) <ref:2410. p. 1

Where the paper says it

the height of ΦN . We define the height of a non-zero polynomial f with coefficients in C∞ to be h(f) := log max c |c|, where c ranges over all coefficients of f.

11132#pg9> This comparison relates the height of a Drinfeld module to the average height of its cyclic N-submodules <ref:2410. p. 2
Where the paper says it

Then q 2 − 1 2 2λN − deg N − bq(N) ≤ h(j(φ)) − 1 ψ(N) X C h(j(φ/C)) ≤ q + q 2 − 1 2 min 0, 2λN − deg N + log 1 q h(j(φ)) + 1 , where C ranges over all submodules C ⊂ φ[N] isomorphic to A/NA and bq(N) is the bounded term from Theorem 1. As a third result, we prove the following analog of a result of Autissier [Aut03], which shows that the stable Taguchi height gives us much cleaner expressions than the Weil height of the j-invariant when estimating the heights of Hecke images.

11132#pg9> The proof uses Lemma 3. p. 1
Where the paper says it

Then the height of the modular polynomial ΦN (X, Y ) satisfies h(ΦN ) = q 2 − 1 2 ψ(N) (deg N − 2λN + bq(N)), where 0 ≤ bq(N) ≤ 4 + 2q 3 + q 2 − 2q + 1 q(q − 1)2 . The proof is given in Section 3.

7, which implies that for z ∈ F and log j(z) ≤ q, then zi = 1 and log ∆(z) = q squared <ref:2410. p. 1
Where the paper says it

the height of ΦN . We define the height of a non-zero polynomial f with coefficients in C∞ to be h(f) := log max c |c|, where c ranges over all coefficients of f.

11132#pg8> This leads to the final bound of h(ΦN (X, Y)) ≥ q squared − 1/2 ψ(N) (deg N − 2λN) <ref:2410. p. 1
Where the paper says it

Let N ∈ A be monic. Then the height of the modular polynomial ΦN (X, Y ) satisfies h(ΦN ) = q 2 − 1 2 ψ(N) (deg N − 2λN + bq(N)), where 0 ≤ bq(N) ≤ 4 + 2q 3 + q 2 − 2q + 1 q(q − 1)2 .

11132#pg8>.

Hecke Image Height Comparison

The proof of Theorem 4.3 involves relating the quantity to the height of the specialized modular polynomial ΦN (X, j(z)) <ref:2410. p. 1

Where the paper says it

Let N ∈ A be monic. Then the height of the modular polynomial ΦN (X, Y ) satisfies h(ΦN ) = q 2 − 1 2 ψ(N) (deg N − 2λN + bq(N)), where 0 ≤ bq(N) ≤ 4 + 2q 3 + q 2 − 2q + 1 q(q − 1)2 .

11132#pg8> The upper bound derived from this specialization is h(ΦN (X, j(z))) ≤ h(ΦN) + ψ(N) log max 1, j(z) <ref:2410. p. 1
Where the paper says it

the height of ΦN . We define the height of a non-zero polynomial f with coefficients in C∞ to be h(f) := log max c |c|, where c ranges over all coefficients of f.

11132#pg8> This ultimately leads to the final inequality involving the minimum of terms related to the growth rate <ref:2410.11132#pg8>. The lower bound is obtained by using Lemma 3.6 and Lemma 3. p. 2
Where the paper says it

We then make use of transformation properties of the Drinfeld modular discriminant proved in Proposition 2. We conclude by an application of Lemma 3.

7 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8> This completes the proof of Theorem 4. p. 1
Where the paper says it

Then the height of the modular polynomial ΦN (X, Y ) satisfies h(ΦN ) = q 2 − 1 2 ψ(N) (deg N − 2λN + bq(N)), where 0 ≤ bq(N) ≤ 4 + 2q 3 + q 2 − 2q + 1 q(q − 1)2 . The proof is given in Section 3.

3 in terms of the AM-GM inequality <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>. The final result is h(ΦN (X, Y)) ≥ q squared − 1/2 ψ(N) (deg N − 2λN) <ref:2410. p. 2
Where the paper says it

Then q 2 − 1 2 2λN − deg N − bq(N) ≤ h(j(φ)) − 1 ψ(N) X C h(j(φ/C)) ≤ q + q 2 − 1 2 min 0, 2λN − deg N + log 1 q h(j(φ)) + 1 , where C ranges over all submodules C ⊂ φ[N] isomorphic to A/NA and bq(N) is the bounded term from Theorem 1. As a third result, we prove the following analog of a result of Autissier [Aut03], which shows that the stable Taguchi height gives us much cleaner expressions than the Weil height of the j-invariant when estimating the heights of Hecke images.

11132#pg8>. The paper's main finding is the explicit, asymptotically optimal estimate for the height of the modular polynomial and its connection to Hecke images <ref:2410.11132#pg8>. The references include [Aut03] Autissier, P. p. 2
Where the paper says it

Then q 2 − 1 2 2λN − deg N − bq(N) ≤ h(j(φ)) − 1 ψ(N) X C h(j(φ/C)) ≤ q + q 2 − 1 2 min 0, 2λN − deg N + log 1 q h(j(φ)) + 1 , where C ranges over all submodules C ⊂ φ[N] isomorphic to A/NA and bq(N) is the bounded term from Theorem 1. As a third result, we prove the following analog of a result of Autissier [Aut03], which shows that the stable Taguchi height gives us much cleaner expressions than the Weil height of the j-invariant when estimating the heights of Hecke images.

, Hauteur des correspondances de Hecke. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

Bull. Soc. Math. France 131 (2003), 421–433 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Bae92] Bae, S. p. 0
Where the paper says it

Analoguously to the case of elliptic modular polynomials, S. Bae [Bae92] defined the Drinfeld modular polynomial ΦN (X, Y ) ∈ A[X, Y ] (see also Definition 2.

, On the modular equation for Drinfeld modules of rank 2. p. 0
Where the paper says it

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t]. These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N.

J. Number Theory 42 (1992), 123–133 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, and [BGP23] Breuer, F. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

, Gijón Gómez, D. and Pazuki, F. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

, Explicit bounds on the coefficients of the modular polynomials and the size of X0(N). p. 0
Where the paper says it

NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract. We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t].

https://arxiv. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

org/abs/2310.14428, 2023 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [BP24] Breuer, F. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

and Pazuki, F., Explicit bounds on the coefficients of modular polynomials for the elliptic j-function. p. 0
Where the paper says it

NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract. We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t].

Proc. Amer. Math. Soc. Ser. B 11 (2024), 277–286 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [BPR21] Breuer, F. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

, Pazuki, F. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

and Razafinjatovo, M.H., Heights and isogenies of Drinfeld modules. p. 0
Where the paper says it

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t]. These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N.

Acta Arith. 197.2 (2021), 111–128 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Bre16] Breuer, F. p. 0
Where the paper says it

11132v1 [math. NT] 14 Oct 2024 HEIGHTS OF DRINFELD MODULAR POLYNOMIALS AND HECKE IMAGES FLORIAN BREUER, FABIEN PAZUKI, AND ZHENLIN RAN Abstract.

, A note on Gekeler’s h-function. p. 7
Where the paper says it

Furthermore, ∆ is a cusp form which vanishes nowhere on Ω. Another important Drinfeld modular form is Gekeler’s h-function [Gek88] (we keep the traditional notation for Gekeler’s h-function despite the conflict with our height functions, but the context will always be clear).

Arch. Math. 93(3) (1988), 667–700 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Gek88] Gekeler, E. p. 7
Where the paper says it

Furthermore, ∆ is a cusp form which vanishes nowhere on Ω. Another important Drinfeld modular form is Gekeler’s h-function [Gek88] (we keep the traditional notation for Gekeler’s h-function despite the conflict with our height functions, but the context will always be clear).

-U., On the coefficients of Drinfel’d modular forms. p. 3
Where the paper says it

Setting We gather in this section several crucial tools which will be used in the proofs of our main results. We start with some background in height theory, then we cover results on the Drinfeld fundamental domain, Bruhat-Tits trees, the building map, Drinfeld modular forms, and Farey sequences in the polynomial setting.

Invent. Math. 93(3) (1988), 667–700 <ref:2410.11132#pg8>, [Gek97] Gekeler, E. p. 8
Where the paper says it

This proposition is an immediate consequence of the following result, which appears in [Gek97, Cor 2. 11] for the case N prime, that lets us think of Gekeler’s h (and thus also ∆) as a “multiplicative Hecke eigenform”.

-U., On the Drinfeld discriminant function. p. 1
Where the paper says it

The general idea is to estimate the Mahler measure of ΦN (X, j(z)) for some chosen z. Then using the result that the geometric mean of the Hecke image of the discriminant function equals the discriminant function itself, we can reduce our proof to estimating the sum of the logarithms of “imaginary parts” (see equation (3) for the definition) of the representatives of the Hecke images of the corresponding Drinfeld modules on the upper half plane.

Compos. Math. 106 (1997), 181–202 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [GeRe96] Gekeler, E. p. 1
Where the paper says it

2 the Taniyama-Shimura-Weil (former) conjecture, see [GeRe96]. More precisely let E be a nonisotrivial elliptic curve over F, with split multiplicative reduction at the place ∞.

-U. and Reversat, M., Jacobians of Drinfeld modular curves J. p. 0
Where the paper says it

This polynomial defines an integral model of the Drinfeld modular curve X0(N). Such modular curves are interesting for several reasons.

reine Angew. Math. 476 (1996), pp. 27–93 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Gek99] Gekeler, E. p. 7
Where the paper says it

Furthermore, ∆ is a cusp form which vanishes nowhere on Ω. Another important Drinfeld modular form is Gekeler’s h-function [Gek88] (we keep the traditional notation for Gekeler’s h-function despite the conflict with our height functions, but the context will always be clear).

-U., A survey on Drinfeld modular forms Turk. p. 3
Where the paper says it

Setting We gather in this section several crucial tools which will be used in the proofs of our main results. We start with some background in height theory, then we cover results on the Drinfeld fundamental domain, Bruhat-Tits trees, the building map, Drinfeld modular forms, and Farey sequences in the polynomial setting.

J. Math 23 (1999), 485–518 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Hsi98] Hsia, L. p. 1
Where the paper says it

There exist some results in this direction. Hsia proved in [Hsi98] that, as deg N → ∞, the height of ΦN satisfies h(ΦN ) = q 2 − 1 2 ψ(N) (deg N − 2κN + O(1)), where κN := X P|N deg P |P| , but the bounded term O(1) was not made explicit.

-C., On the coefficients of modular polynomials for Drinfeld modules. p. 0
Where the paper says it

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t]. These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N.

J. Number Theory 72 (1998), 236–256 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Ked01] Kedlaya, K.S., The algebraic closure of the power series field in positive characteristic. p. 0
Where the paper says it

Introduction Let A = Fq[t] and F = Fq(t), equipped with the absolute value |x| := q deg x corresponding to the place ∞. We denote by F∞ = Fq(( 1 t )) the completion of F at ∞ and by C∞ the completion of an algebraic closure of F∞.

Proc. Am. Math. Soc. 129 (2001), 3461–3470 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Ked17] Kedlaya, K. S., On the algebraicity of generalized power series. Beitr. Algebra Geom 58 (2017), 499–527 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Pap23] Papikian, M., Drinfeld Modules, Springer International Publishing, 2023 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Pet22] Petit, V. p. 1
Where the paper says it

Then there exists a modular parametrisation X0(N) → E. See [Pet22] for recent explicit computations.

, Explicit computation of the modular parametrization of elliptic curves over function fields by Drinfeld modular curves (2022), https://arxiv. p. 0
Where the paper says it

They are used for constructing towers of curves with many rational points, with applications in coding theory, see for example [Elk01]. These curves also parametrise elliptic curves over function fields via a theorem in analogy with

org/pdf/2206.00896 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Paz19] Pazuki, F. p. 2
Where the paper says it

Heights of Drinfeld modular polynomials and Hecke images 3 papers (for instance in [Paz19]) one needed to specialize at a number of Y -values and then use Lagrange interpolation to estimate h(ΦN ). This trick relies on the non-Archimedean metric on C∞.

, Modular invariants and isogenies. p. 0
Where the paper says it

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials ΦN for any monic N ∈ Fq[t]. These polynomials vanish at pairs of j-invariants of Drinfeld Fq[t]-modules of rank 2 linked by cyclic isogenies of degree N.

Inter. J. Number Theory 15. p. 2
Where the paper says it

Heights of Drinfeld modular polynomials and Hecke images 3 papers (for instance in [Paz19]) one needed to specialize at a number of Y -values and then use Lagrange interpolation to estimate h(ΦN ). This trick relies on the non-Archimedean metric on C∞.

3 (2019), 569–584 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Sil90] Silverman, J. p. 2
Where the paper says it

1 for definitions. In our second main result, we prove the following analog of a result of Silverman [Sil90], comparing the Weil height of a j-invariant with the height of its Hecke image.

H., Hecke points on modular curves. p. 0
Where the paper says it

Such modular curves are interesting for several reasons. They are used for constructing towers of curves with many rational points, with applications in coding theory, see for example [Elk01].

Duke Math. J. 60.2 (1990), 401–423 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Tag93] Taguchi, Y. p. 2
Where the paper says it

We denote by hTag(φ) the stable Taguchi height of φ (see Section 2. 1, or [Tag93], for definitions).

, Semi-simplicity of the Galois representations attached to Drinfeld modules over fields of “infinite characteristics”. p. 1
Where the paper says it

Then using the result that the geometric mean of the Hecke image of the discriminant function equals the discriminant function itself, we can reduce our proof to estimating the sum of the logarithms of “imaginary parts” (see equation (3) for the definition) of the representatives of the Hecke images of the corresponding Drinfeld modules on the upper half plane. Function fields Farey sequences are used for these estimates.

J. Number Theory 44 (1993), 292–314 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>, [Wei20] Wei, F.-T., On Kronecker terms over global function fields. p. 0
Where the paper says it

They are used for constructing towers of curves with many rational points, with applications in coding theory, see for example [Elk01]. These curves also parametrise elliptic curves over function fields via a theorem in analogy with

Invent. Math. 220.3 (2020), 847–907 <ref:2410. p. 0
Where the paper says it

arXiv:2410. 11132v1 [math.

11132#pg8>.

Pruning redundant reasoning tokens

The paper introduces two notions of height for a Drinfeld module: the Weil height of the j-invariant and the stable Taguchi height <ref:2410.11132#pg6> The Weil height h(x) is defined using an absolute value x:= q deg x corresponding to the place ∞ <ref:2410. p. 0

Where the paper says it

——— 1. Introduction Let A = Fq[t] and F = Fq(t), equipped with the absolute value |x| := q deg x corresponding to the place ∞.

11132#pg7> The stable Taguchi height hTag(φ) is defined as a sum of finite and infinite parts <ref:2410. p. 2
Where the paper says it

As a third result, we prove the following analog of a result of Autissier [Aut03], which shows that the stable Taguchi height gives us much cleaner expressions than the Weil height of the j-invariant when estimating the heights of Hecke images. We denote by hTag(φ) the stable Taguchi height of φ (see Section 2.

11132#pg6>

How the training signal is built

The proof of Theorem 3.3 utilizes Lemma 3. p. 2

Where the paper says it

We then make use of transformation properties of the Drinfeld modular discriminant proved in Proposition 2. We conclude by an application of Lemma 3.

6, which asserts that for any non-zero polynomial f(X, Y) in C∞[X, Y], there exists y ∈ Fq such that h(f(X, Y)) = h(f(X, y)) <ref:2410. p. 1
Where the paper says it

the height of ΦN . We define the height of a non-zero polynomial f with coefficients in C∞ to be h(f) := log max c |c|, where c ranges over all coefficients of f.

11132#pg8> This allows the authors to compute the height of ΦN (X, Y) by specializing at a single suitable value of j <ref:2410.11132#pg8> Lemma 3. p. 2
Where the paper says it

We then make use of transformation properties of the Drinfeld modular discriminant proved in Proposition 2. We conclude by an application of Lemma 3.

7 further specifies that if z ∈ F and log j(z) ≤ q, then zi = 1 and log ∆(z) = q squared <ref:2410. p. 1
Where the paper says it

the height of ΦN . We define the height of a non-zero polynomial f with coefficients in C∞ to be h(f) := log max c |c|, where c ranges over all coefficients of f.

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

  1. Improved bound estimation for Drinfeld modular polynomials: The AI system can now derive explicit, asymptotically optimal upper and lower bounds for the size of coefficients of Drinfeld modular polynomials using Theorem 1.1, which provides a precise formula involving λN and bq(N).

  2. Enhanced comparison between module heights: The system can perform an analog to Silverman's result (Theorem 1.2) by comparing the Weil height of a j-invariant with the height of its Hecke image, providing explicit bounds for these differences over global function fields.

  3. Precise calculation of stable Taguchi height differences: The AI system can compute the difference between the stable Taguchi height and the average over submodules (Theorem 1.3) by utilizing Lemma 3.2.3 to equate terms involving log D(Λv) and log D(ΛC,v).

  4. Control over fundamental representatives via building maps: The system can determine the fundamental representative of a point v(k, u) in the Bruhat-Tits tree using Theorem 2.13, enabling accurate estimation of log zeγi by relating it to the structure of the tree's vertices.

  5. Numerical approximation using Farey sequences: The system can approximate representatives zeγ by explicit Farey representatives zbγ (Section 2.7) when d is large, allowing for precise calculation of sums involving log zeγi up to a certain bound, as shown in Proposition 3.5.

Abstract

We obtain explicit upper and lower bounds on the size of the coefficients of the Drinfeld modular polynomials Φ N for any monic N in F q[t]. These polynomials vanish at pairs of j-invariants of Drinfeld F q[t] -modules of rank 2 linked by cyclic isogenies of degree N. The main term in both bounds is asymptotically optimal as deg(N) tends to infinity. We also obtain precise estimates on the Weil height and Taguchi height of Hecke images of Drinfeld modules of rank 2.

Sources

Related papers