On the Number of Small Points for Rational Maps

Basic Information

• Paper ID: 2510.12039
• Title: On the Number of Small Points for Rational Maps
• Author: Jit Wu Yap
• Classification: math.NT (Number Theory), math.DS (Dynamical Systems)
• Publication Date: October 14, 2025 (arXiv preprint)
• Paper Link: https://arxiv.org/abs/2510.12039

Abstract

Let $K$ be a number field and $f: \mathbb{P}^1 \to \mathbb{P}^1$ be a rational map of degree $d \geq 2$ with at most $s$ places of bad reduction (including all archimedean places). This paper proves that there exist constants $c_1, c_2 > 0$ depending only on $d$ and independent of $f$ or $K$, such that $\# \left\{ x \in \mathbb{P}^1(K) \mid \widehat{h}_f(x) \leq \frac{c_1}{s} h_{\operatorname{rat}_d}(\langle f \rangle) \right\} \leq c_2 s \log(s)$ where $\operatorname{rat}_d$ is the moduli space of rational maps under conjugacy action, $h_{\operatorname{rat}_d}$ is the ample height, and $\langle f \rangle$ is the equivalence class associated with $f$.

Research Background and Motivation

Problem Background

This research belongs to the field of arithmetic dynamics, focusing on the small points problem for rational maps. Specifically:

1. Core Problem: For a given rational map $f$, how many points $x$ possess small canonical height $\widehat{h}_f(x)$? This problem is closely related to the distribution of preperiodic points, since preperiodic points are precisely those with canonical height zero.
2. Significance:
   • This is a fundamental problem in arithmetic geometry, related to important conjectures such as Silverman's conjecture
   • It is crucial for understanding the arithmetic properties of rational maps
   • It occupies a central position in the intersection of algebraic number theory and dynamical systems
3. Existing Limitations:
   • Baker's (2006) results depend on the specific map $f$, with non-uniform constants
   • Results by Benedetto (2007) and Looper (2019) apply only to polynomial cases
   • Lack of uniform bounds for general rational maps

Research Motivation

This paper aims to generalize existing polynomial results to general rational maps and provide uniform bounds that depend only on the degree $d$ and the number of bad reduction places $s$.

Core Contributions

1. Uniformization of Baker's Theorem: Proves a uniform version of Baker's theorem where constants depend only on degree $d$, independent of the specific map or number field
2. Generalization of Benedetto-Looper Results: Extends the results of Benedetto and Looper on polynomials to general rational maps
3. Optimal Dependence: Achieves $O(s \log s)$ dependence on the number of bad reduction places, significantly improving Troncoso's exponential dependence
4. Technical Innovation: First systematic application of Luo's degeneration theory for rational map sequences and Favre-Gong's formalization of Berkovich spaces to the small points problem
5. Lower Bound Results: As a corollary, obtains uniform lower bounds for the canonical height of non-preperiodic points

Detailed Methodology

Problem Setup

Given a rational map $f: \mathbb{P}^1 \to \mathbb{P}^1$ of degree $d$ over a number field $K$ with at most $s$ places of bad reduction, the goal is to estimate the size of the set $\left\{ x \in \mathbb{P}^1(K) \mid \widehat{h}_f(x) \leq \frac{c_1}{s} h_{\operatorname{rat}_d}(\langle f \rangle) \right\}$

Core Technical Framework

1. Arakelov-Green Function Method

For each place $v \in M_K$, define the dynamical Arakelov-Green function: $g_{f,v}(x,y) = -\log |x̃ \wedge ỹ| + \widehat{H}_F(x̃) + \widehat{H}_F(ỹ) - \frac{1}{d(d-1)}\log|\text{Res}(F)|$

where $F$ is a homogeneous lift of $f$, and $x̃, ỹ$ are lifts of $x, y$ to $\mathbb{A}^2$.

2. Key Inequality

Baker proved that for $z_1, \ldots, z_N \in \mathbb{P}^1(K)$: $\sum_{i \neq j} g_{f,v}(z_i, z_j) \geq -O_f(N \log N)$

3. Achievement of Uniformity

The core technical challenge is to uniformize the inequality $g_{f,w}(x,y) > \delta$ for $x,y$ in the same open set, such that $\delta$ is proportional to $h_{\operatorname{rat}_d}(\langle f \rangle)$.

Application of Degeneration Theory

1. Berkovich Space Construction

For a sequence of rational maps $(f_n)$, construct the Banach ring: $A^ε = \left\{(x_n) \in \prod_{n=1}^∞ K_n \mid |x_n|^{ε_n} \text{ is bounded}\right\}$

where $ε_n = (-\log|\text{Res}(f_n)|)^{-1}$.

2. Properties of Limit Maps

Via an ultrafilter $ω$, obtain the limit map $f_ω$ on the field $H(ω)$, satisfying:

• If $\lim_ω |x_n|^{ε_n} = 0$, then $H(ω)$ is a non-archimedean field
• $f_ω$ has bad reduction properties

3. Continuity Argument

Utilize density of model functions and compactness arguments to transfer results from the non-archimedean case to the archimedean case.

Technical Innovations

1. Cross-Characteristic Degeneration: First systematic realization of degeneration from archimedean to non-archimedean fields, breaking through limitations of traditional methods
2. Uniform Constants: Through refined analysis, obtains uniform constants depending only on degree $d$
3. Minimal Resultant Height: Introduces the minimal resultant "height" $h_{\text{res}}(f)$ and proves its comparability with the moduli height $h_{\operatorname{rat}_d}(\langle f \rangle)$

Main Theorems

Theorem 1.1 (Main Result)

Let $K$ be a number field and $f: \mathbb{P}^1_K \to \mathbb{P}^1_K$ be a rational map of degree $d \geq 2$ with $s$ places of bad reduction. There exist constants $c_1, c_2 > 0$ depending only on $d$ such that $\# \left\{ x \in \mathbb{P}^1(K) \mid \widehat{h}_f(x) \leq \frac{c_1}{s} h_{\operatorname{rat}_d}(\langle f \rangle) \right\} \leq c_2 s \log(s)$

Theorem 1.2 (Height Lower Bound)

Under the same conditions, there exists a constant $c_3 > 0$ such that for $x \in \mathbb{P}^1(K)$: $\widehat{h}_f(x) = 0 \quad \text{or} \quad \widehat{h}_f(x) \geq \frac{1}{dc_3 s \log s} h_{\operatorname{rat}_d}(\langle f \rangle)$

Theorem 1.3 (Uniform Open Cover)

There exist constants $\delta > 0$ and positive integer $N$ depending only on $d$ such that for any rational map $f$ of degree $d$ over a complete algebraically closed field $(K, |\cdot|)$, one can cover $\mathbb{P}^1(K)$ with $N$ open sets $U_1, \ldots, U_N$ satisfying $g_f(x,y) > \delta(-\log|\text{res}(f)|)$ for any $x, y \in U_i$.

Proof Strategy

Step One: Reduction to Uniform Cover

By contradiction, assume the existence of too many small height points and use the pigeonhole principle to distribute them among open sets in the cover.

Step Two: Apply Arakelov-Green Function Inequality

For points in the same open set, use Theorem 1.3 to obtain lower bounds for the Green function.

Step Three: Global Summation and Contradiction

Sum contributions over all places, utilizing the global formula for canonical height: $\sum_{v \in M_K} \sum_{i \neq j} N_v g_{f,v}(z_i, z_j) = 2(M-1) \sum_{i=1}^M \widehat{h}_f(z_i)$

Step Four: Degeneration Argument

The proof of Theorem 1.3 relies on degeneration theory:

• Assume no uniform cover exists
• Construct a sequence of rational maps $(f_n)$
• Apply Favre-Gong degeneration to obtain limit maps
• Utilize Baker's results in the non-archimedean case
• Transfer back to the original problem via continuity arguments

Related Work

Historical Development

1. Baker (2006): First proved small points bounds in finite extensions, but constants depend on the map
2. Benedetto (2007): Provided uniform bounds for polynomial cases
3. Looper (2019): Improved Benedetto's results and gave height lower bounds
4. Troncoso (2017): Gave uniform bounds for preperiodic points, but with exponential dependence

Technical Background

1. Kiwi (2006, 2014): First used non-archimedean fields to understand degeneration of rational map families
2. DeMarco-Faber (2014, 2016): Developed degeneration theory for equilibrium measures
3. Luo (2021, 2022): Established degeneration theory for rational map sequences
4. Favre-Gong (2024): Formalized Luo's construction via Berkovich spaces

Conclusions and Discussion

Main Conclusions

This paper successfully generalizes small points bounds from the polynomial case to general rational maps, achieving optimal $O(s \log s)$ dependence and providing uniform constants. This resolves a weakened version of Silverman's conjecture.

Limitations

1. Effectiveness of Constants: Although uniform constants are obtained, they are ineffective (non-computable)
2. Technical Complexity: The proof relies on complex degeneration theory, making explicit estimates difficult
3. Scope of Application: Results apply only to rational maps on $\mathbb{P}^1$

Future Directions

1. Higher-Dimensional Generalization: Extend results to injective maps on $\mathbb{P}^n$
2. Effective Constants: Seek methods for obtaining effective constants
3. Function Field Case: Further investigate analogous problems over function fields
4. Moduli Space Compactification: Develop satisfactory compactifications of moduli spaces of rational maps

In-Depth Evaluation

Strengths

1. Theoretical Depth: Cleverly combines arithmetic geometry, dynamical systems, and Berkovich space theory
2. Technical Innovation: First systematic application of degeneration theory to the small points problem
3. Optimal Results: Achieves optimal dependence, improving existing results
4. Uniformity: Provides truly uniform treatment independent of specific maps

Weaknesses

1. Non-Effectiveness of Constants: Key constants are non-computable, limiting practical applications
2. Technical Threshold: Requires deep background in algebraic geometry and Berkovich spaces
3. Difficulty of Generalization: Methods are difficult to directly extend to more general settings

Impact

1. Theoretical Contribution: Provides new technical tools for arithmetic dynamics
2. Methodological Significance: Demonstrates the power of degeneration theory in number-theoretic problems
3. Subsequent Research: Provides new ideas and methods for related problem investigations

Applicable Scenarios

1. Theoretical Research: Theoretical development of arithmetic dynamics
2. Related Problems: Other small points problems and height bound problems
3. Method Borrowing: Application of degeneration methods to other number-theoretic problems

References

The paper cites major works in the field, including:

• Baker's foundational work
• Results by Benedetto and Looper on polynomials
• Luo's degeneration theory
• Favre-Gong's formalization via Berkovich spaces
• Work by DeMarco and others on degeneration of dynamical systems

This paper represents significant progress in arithmetic dynamics, solving long-standing problems through innovative technical methods and making important contributions to the development of the field.