Cyclotomic integral points for affine dynamics

Basic Information

• Paper ID: 2511.13443
• Title: Cyclotomic integral points for affine dynamics
• Authors: Zhuchao Ji, Junyi Xie, Geng-Rui Zhang
• Classification: math.DS (Dynamical Systems), math.AG (Algebraic Geometry), math.NT (Number Theory)
• Submission Date: November 17, 2025
• Paper Link: https://arxiv.org/abs/2511.13443

Abstract

This paper studies regular endomorphisms $f:\mathbb{A}^N\to\mathbb{A}^N$ of affine space defined over number fields (algebraic degree $d\geq2$). The main result proves that if the set of $f$-preperiodic cyclotomic points is Zariski dense in $\mathbb{A}^N$, then some iterate $f^{\circ l}$ is a quotient of an algebraic group endomorphism $g:\mathbb{G}_m^N\to\mathbb{G}_m^N$. This is a higher-dimensional generalization of the Dvornicich-Zannier theorem on cyclotomic preperiodic points for univariate polynomials. More generally, for dominant endomorphisms of affine varieties defined over number fields, the paper proves rigidity results concerning Zariski-dense subsets of cyclotomic integral points that are "almost invariant." As applications, the results are applied to backward orbits of regular endomorphisms and periodic points of Hénon-type automorphisms.

Research Background and Motivation

1. Research Questions

This paper investigates the distribution of special points (cyclotomic points) in algebraic dynamical systems, particularly addressing:

• When preperiodic points contain a Zariski-dense set of cyclotomic points, does the dynamical system possess special algebraic structure?
• Can such phenomena characterize the "monomial type" property of the system?

2. Importance of the Problem

This problem connects multiple mathematical fields:

• Arithmetic Dynamics: Understanding the distribution of special points (such as cyclotomic points, torsion points) in dynamical systems
• Unlikely Intersection Problems: Continuing the philosophy of Zannier and others—special points should not be dense in non-special varieties
• Diophantine Geometry: Arithmetic properties of cyclotomic integral points

3. Limitations of Existing Work

Dvornicich-Zannier Theorem (2007) only addresses the one-dimensional case:

• For polynomials $f\in K[z]$ of degree $d\geq2$, if there are infinitely many preperiodic cyclotomic points, then $f$ is affinely conjugate to $z^d$ or $\pm T_d(z)$ (Chebyshev polynomial)

Limitations:

• Restricted to polynomial maps on $\mathbb{A}^1$
• Does not handle higher-dimensional affine spaces
• Cannot be applied to more general dynamical systems (such as Hénon maps)

4. Research Motivation

The core motivation of this paper is:

• Higher-dimensional Generalization: Extend the Dvornicich-Zannier result to $\mathbb{A}^N$
• Unified Framework: Establish rigidity theory applicable to general dominant endomorphisms
• Extended Applications: Resolve concrete problems concerning backward orbits and Hénon-type automorphisms

Core Contributions

1. Main Rigidity Theorem (Theorem 1.2): For a dominant endomorphism $f$ on an affine variety $X$, proves that Zariski density of cyclotomic integral points satisfying three conditions (DCI, BH, AI) implies $(X,f)$ is of monomial type.
2. Equivalence Theorem (Theorem 1.3): For cohomologically hyperbolic systems, monomial type is equivalent to strongly monomial type.
3. Regular Endomorphism Application (Theorem 1.4): Proves that for regular endomorphisms on $\mathbb{A}^N$ of algebraic degree $d\geq2$, if preperiodic cyclotomic points are Zariski dense, then the system is strongly monomial type—a direct higher-dimensional generalization of the Dvornicich-Zannier theorem.
4. Backward Orbit Characterization (Theorem 1.5): Provides necessary and sufficient conditions for cyclotomic points in backward orbits to satisfy conditions (DCI, BH, AI).
5. Hénon-type Automorphisms (Theorem 1.7): Proves that periodic cyclotomic points of Hénon-type automorphisms cannot be Zariski dense, providing new evidence for the "unlikely intersection" philosophy.

Detailed Methodology

Task Definition

Input:

• Affine variety $X\subseteq \mathbb{A}^N_K$ ($K$ a number field)
• Dominant endomorphism $f:X\to X$
• Point set $P\subseteq X(K)$

Task: Determine whether $(X,f)$ is of monomial type, i.e., whether there exist:

• Integers $l\geq1, n\geq\dim(X)$
• Group endomorphism $g:\mathbb{G}_m^n\to\mathbb{G}_m^n$
• Dominant morphism $\phi:\mathbb{G}_m^n\to X$

such that $f^{\circ l}\circ\phi = \phi\circ g$

Conditions: $P$ satisfies

• (DCI) Dense Cyclotomic Integers: $P$ is Zariski dense with coordinates in $\frac{1}{M}\mathcal{O}_{K^c}$
• (BH) Bounded House: $\mathcal{C}(y)\leq c$ for all $y\in P$
• (AI) Almost Invariant: $P\setminus f^{-1}(P)$ is not Zariski dense

Proof Architecture

The proof consists of four key steps:

Step 1: Application of Loxton's Theorem

Core Tool: Loxton's theorem as generalized by Dvornicich-Zannier (Theorem 2.1)

Conclusion: There exist a finite set $E\subset K$ and integer $b$ such that $P\subseteq \left(\sum_{i=1}^b E\cdot U(\mathbb{C})\right)^N$

Construction: For each $a=(a_{ij})\in E^{bN}$, define

• Morphism $\phi_a:\mathbb{G}_m^{bN}\to\mathbb{A}^N$, $(z_{ij})\mapsto\left(\sum_j a_{ij}z_{ij}\right)_i$
• Torsion point set $\Lambda_a = \phi_a^{-1}(P)\cap\mathbb{G}_m^{bN}(K)_{tors}$
• Closed subset $Z_a = \overline{\Lambda_a}^{Zar}\subseteq \mathbb{G}_m^{bN}$

Key Property: By removing exceptional sets, one can assume $\phi_a(\Lambda_a)^{Zar}=X$ for all $a\in M$.

Step 2: Construction of Correspondence

Define the correspondence $\Gamma = \bigsqcup_{(a_1,a_2)\in M^2}\Gamma_{a_1,a_2}\subseteq Z\times Z$ where $\Gamma_{a_1,a_2} = \overline{\{(\xi_1,\xi_2)\in\Lambda_{a_1}\times\Lambda_{a_2}: f(\phi_{a_1}(\xi_1))=\phi_{a_2}(\xi_2)\}}^{Zar}$

Key Properties:

• $\pi_1(\Gamma) = Z$ (surjectivity)
• $(\phi\times\phi)(\Gamma)^{Zar} = \Gamma_f$ (related to the graph of $f$)
• By the torsion point theorem (Theorem 2.3), each $\Gamma_{a_1,a_2}$ is a finite union of torsion cosets

Step 3: Irreducibility and Surjectivity

Through a series of reductions:

1. Decompose $Z$ into irreducible components $Z=\bigsqcup_{\alpha\in I}Y_\alpha$, each $Y_\alpha\cong\mathbb{G}_m^{\gamma_\alpha}$
2. Select a periodic point $\alpha_0\in J$ (where $J=\{\alpha:\phi(Y_\alpha)^{Zar}=X\}$)
3. Replace by iterates $f^{\circ n}$ and compositions of corresponding $\psi$ to obtain:
   • $Y$ irreducible
   • $\psi\subseteq Y\times Y$ irreducible
   • $\psi(Y)=Y$ (surjectivity)

Step 4: Induction of Group Endomorphism

Stabilizer Technique:

• For each $y\in Y$, define the stabilizer $T_y=\text{Stab}_Y(F_y)$ of the fiber $F_y=\phi^{-1}(\phi(y))$
• Let $T=\bigcap_{y\in Y}T_y$
• After quotienting by $T$, one can assume $T=1$

Key Observation:

• For any $y\in Y$ and $z\in\psi(y)$, we have $\psi(y)=z\cdot V$, where $V$ is an algebraic subgroup
• By invariance arguments, prove $V=1$
• Therefore $\psi$ is the graph of a morphism $g:Y\to Y$

Group Structure Adjustment:

• Since $\psi$ is a torsion coset, $g(y)=\tau_0\cdot g_0(y)$ ($g_0$ is a group endomorphism, $\tau_0$ is a torsion point)
• By iteration, make $g(1)=1$, so $g$ becomes a group endomorphism

Technical Innovations

1. Correspondence Method: Using $\Gamma$ to encode dynamical information is a key technical innovation for handling higher-dimensional problems
2. Systematic Application of Torsion Point Theorem: Laurent's torsion point theorem (multiplicative Manin-Mumford) serves as the core tool
3. Cohomological Hyperbolicity Theory: Theorem 1.3 utilizes log-concavity of dynamical degrees and cohomological Lyapunov exponents
4. Unified Framework: Three conditions (DCI, BH, AI) provide a unified framework for treating different problems

Experimental Setup

Note: This is a pure mathematics theory paper with no numerical experiments. All results are rigorous mathematical theorems.

Application Verification

The paper verifies the applicability of the theory by proving the following theorems:

Application 1: Regular Endomorphisms (Theorem 1.4)

• Setup: $f:\mathbb{A}^N\to\mathbb{A}^N$ is a regular endomorphism of algebraic degree $d\geq2$
• Verification: Check that $P=\text{PrePer}(f,\mathbb{A}^N(K^c))$ satisfies (DCI), (BH), (AI)
• Key: Utilize properties of the Green function $G(z)=\lim_{n\to\infty}\frac{1}{d^n}\log\max\{1,\|f^{\circ n}(z)\|\}$

Application 2: Backward Orbits (Theorem 1.5)

• Setup: $P=\{z\in\mathbb{A}^N(K^c):\exists n\geq1, f^{\circ n}(z)=x\}$
• Necessary and Sufficient Condition: $P$ satisfies (DCI), (BH), (AI) $\Leftrightarrow$ $P$ is Zariski dense

Application 3: Hénon-type Automorphisms (Theorem 1.7)

• Setup: $f:\mathbb{A}^N\to\mathbb{A}^N$ is Hénon-type ($\deg_1(f)\geq2$, $I(f)\cap I(f^{-1})=\emptyset$)
• Conclusion: $\text{Per}(f,\mathbb{A}^N(K^c))$ is not Zariski dense (by contradiction)
• Key Tool: Kawaguchi's Green function theory

Experimental Results

Verification of Main Theorems

Verification of Theorem 1.2

Input Condition Checking:

• For regular endomorphisms, verify (BH) through Green function compactness arguments
• Construct integer $M$ using the condition $f^{-1}_h(0)=\{0\}$ to verify (DCI)
• (AI) is automatically satisfied by $P\subseteq f^{-1}(P)$

Output Result: Successfully construct $f^{\circ l}\circ\phi = \phi\circ g$ where $g:\mathbb{G}_m^N\to\mathbb{G}_m^N$ is a surjective group endomorphism

Proof Strategy for Theorem 1.3

For cohomologically hyperbolic systems:

• Utilize growth rate estimates from Lemma 2.7
• Decompose via Lemma 2.6 into root-of-unity eigenvalue part and positive part
• Derive contradiction using projection formula, proving $n=\dim(X)$ is necessary

Key Case Analysis

Case 1: One-dimensional case recovering Dvornicich-Zannier theorem

For $f\in K[z]$ of degree $d\geq2$:

• By Theorem 1.4, if there are infinitely many preperiodic cyclotomic points, then $f$ is strongly monomial type
• There exist $n\in\mathbb{Z}, h\in\mathbb{Q}(z)\setminus\mathbb{Q}$ such that $z^n\circ h=f^{\circ l}\circ h$
• By $n=\pm d^l$ and classical results, $f$ is conjugate to $z^d$ or $\pm T_d(z)$

Case 2: Periodic points of Hénon maps

For $f(x,y)=(p(x)-ay,x)$ ($a\neq0, \deg p\geq2$):

• Compute dynamical degrees: $\lambda_1(f)=d, \lambda_2(f)=1$ (where $d=\deg p$)
• If periodic cyclotomic points are dense, by contrapositive of Theorem 1.7:
  • There must exist $A\in\text{GL}_2(\mathbb{Z})$ such that $\lambda_i(\phi_A)=\lambda_i(f)$
  • But $|\det(A)|=1$ contradicts $\lambda_2(f)=d^2\geq4$
• Conclusion: Periodic cyclotomic points are not dense

Theoretical Findings

1. Rigidity Phenomenon: Zariski density of cyclotomic points forces the system to have algebraic group structure
2. Dimensional Obstruction: For cohomologically hyperbolic systems, there is no "redundant dimension" ($n=\dim X$)
3. Unlikely Intersection Verification: Results on Hénon-type automorphisms align with Zannier's philosophy

Related Work

Historical Development

1. Loxton's Theorem (1972)

• Content: Cyclotomic integers can be expressed as sums of a bounded number of roots of unity
• Generalization: Dvornicich-Zannier (2007) generalized to arbitrary number fields

2. Dvornicich-Zannier Theorem (2007)

• Content: Rigidity of cyclotomic preperiodic points for one-dimensional polynomials
• Method: Loxton's theorem + classification of dynamical systems
• Limitation: Only handles $\mathbb{P}^1$

3. Subsequent Work

• Ostafe (2017): Backward orbits of rational functions with periodic critical points
• Chen (2018): Backward orbits in cyclotomic field closures
• Ferraguti-Ostafe-Zannier (2024): Backward orbits of abelian points

Related Theories

Torsion Point Theorem (Laurent 1984)

• Manin-Mumford Conjecture: The Zariski closure of torsion points in an abelian variety is a finite union of cosets of abelian subvarieties
• Multiplicative Version: The Zariski closure of torsion points in $\mathbb{G}_m^n$ is a finite union of torsion cosets
• Application in This Paper: Theorem 2.3 is the core tool in the proof

Dynamical Degree Theory

• Definition: $\lambda_i(f)=\lim_{n\to\infty}\deg_{i,L}(f^{\circ n})^{1/n}$
• Property: Log-concavity (Truong 2020)
• Cohomological Hyperbolicity: $\mu_i(f)=\lambda_i(f)/\lambda_{i-1}(f)\neq1$ for all $i$

Unique Contributions of This Paper

Compared to existing work:

1. Dimension: First to handle affine varieties of arbitrary dimension
2. Generality: Applicable to all dominant endomorphisms (not just polynomials)
3. Uniformity: (DCI)+(BH)+(AI) provides a unified framework
4. Breadth of Applications: Covers regular endomorphisms, backward orbits, Hénon maps

Conclusions and Discussion

Main Conclusions

1. Higher-dimensional Rigidity Theorem: Zariski density of cyclotomic points combined with "almost invariance" implies monomial type structure
2. Higher-dimensional Generalization of Dvornicich-Zannier Theorem: For regular endomorphisms on $\mathbb{A}^N$, if preperiodic cyclotomic points are dense, then the system is strongly monomial type
3. Negative Result for Hénon-type Automorphisms: Periodic cyclotomic points cannot be Zariski dense, verifying the "unlikely intersection" philosophy
4. Equivalence Characterization: For cohomologically hyperbolic systems, monomial type $\Leftrightarrow$ strongly monomial type

Limitations

1. Number Field Restriction: Theorems require definition over number fields; cannot be directly applied to function fields
2. Cyclotomic Point Restriction: Only addresses cyclotomic points (roots of unity), not more general algebraic points
3. Almost Invariance: Condition (AI) may be difficult to verify in some cases
4. Non-constructive: Theorems are existence results without providing algorithms to explicitly construct $\phi$ and $g$
5. Cohomological Hyperbolicity: Theorem 1.3 requires additional assumptions; the relationship between monomial type and strongly monomial type in general cases remains unclear

Future Directions

Directions suggested by the paper:

1. Function Field Generalization: Can results be extended to function fields of positive characteristic?
2. Abelian Points: For $K=\mathbb{Q}$, distribution problems for points in abelian extensions (partial results already exist)
3. Algorithmic Questions: Given $f$, can one algorithmically determine whether it is of monomial type?
4. Quantitative Results: Can one provide upper bounds on the number of preperiodic cyclotomic points?
5. Other Dynamical Systems: Can results be generalized to rational maps, birational maps, etc.?

In-Depth Evaluation

Strengths

1. Mathematical Depth

• Theoretical Breakthrough: First rigidity theorem for cyclotomic points in higher dimensions
• Technical Innovation: Correspondence method elegantly encodes dynamical information
• Tool Integration: Skillfully combines Loxton's theorem, torsion point theorem, and dynamical degree theory

2. Proof Structure

• Clarity: Four-step proof is logically clear with well-defined objectives at each step
• Completeness: Forms a complete chain from general theorems to concrete applications
• Technical Sophistication: Stabilizer technique and group structure adjustment demonstrate high-level expertise

3. Application Value

• Breadth: Unified treatment of regular endomorphisms, backward orbits, Hénon maps
• Extensibility: Framework potentially applicable to other dynamical system problems
• Theoretical Significance: Provides new evidence for "unlikely intersection" philosophy

4. Writing Quality

• Organization: Detailed introduction, clear motivation, step-by-step proof exposition
• Readability: Despite technical difficulty, logic is easy to follow
• Literature Review: Comprehensive coverage of related work with clear positioning of contributions

Weaknesses

1. Technical Limitations

• Cohomological Hyperbolicity: Additional assumption in Theorem 1.3 limits generality
• Number Field Assumption: Function field cases completely unaddressed
• Constructive Deficiency: No effective algorithms provided

2. Application Scope

• Point Type Restriction: Limited to cyclotomic points; other special points (CM points, etc.) not included
• Dynamical System Types: Primarily focused on polynomial endomorphisms; rational function cases unclear

3. Quantitative Issues

• Constant Dependence: Specific dependence of constants $M,c$ not provided
• Complexity Analysis: Complexity of decision algorithms not discussed

4. Missing Examples

• Concrete Examples: Beyond Chebyshev polynomials, few non-trivial examples provided
• Boundary Cases: No discussion of cases where conditions fail

Impact Assessment

Contribution to the Field

• Arithmetic Dynamics: Opens new directions for higher-dimensional cyclotomic point research
• Diophantine Geometry: Provides new techniques for handling cyclotomic integers
• Algebraic Geometry: Correspondence method may have broader applications

Practical Value

• Theoretical Guidance: Provides theoretical framework for studying special point distributions
• Problem Resolution: Resolves long-standing questions about Hénon map periodic points

Reproducibility

• Proof Completeness: All theorems have complete proofs
• Standard Tools: Uses only standard mathematical tools
• Logical Clarity: Proofs can be verified step-by-step

Applicable Scenarios

Theoretical Research

• Studying structure of higher-dimensional algebraic dynamical systems
• Exploring distribution of special points (torsion points, CM points, etc.)
• Developing "unlikely intersection" theory

Concrete Problems

• Determining whether a given map is of monomial type
• Studying arithmetic properties of preperiodic points
• Analyzing backward orbit structures

Method Borrowing

• Correspondence technique applicable to other geometric problems
• Stabilizer method useful for other quotient constructions
• Green function techniques applicable to other height problems

Selected References

Core References

1. Dvornicich-Zannier (2007): Cyclotomic Diophantine problems, Duke Math. J.
   • Original one-dimensional results
2. Laurent (1984): Equations diophantiennes exponentielles, Invent. Math.
   • Proof of torsion point theorem
3. Kawaguchi (2013): Local and global canonical height functions, Algebra Number Theory
   • Green function theory for Hénon-type automorphisms
4. Truong (2020): Relative dynamical degrees of correspondences, J. Reine Angew. Math.
   • Log-concavity of dynamical degrees
5. Zannier (2012): Some problems of unlikely intersections in arithmetic and geometry
   • Systematic exposition of "unlikely intersection" philosophy

Related Work

6. Ostafe (2017): On roots of unity in orbits of rational functions, Proc. AMS
7. Ferraguti-Ostafe-Zannier (2024): Cyclotomic and abelian points in backward orbits, Adv. Math.
8. Pink-Roessler (2004): On ψ-invariant subvarieties, J. Algebraic Geom.

Overall Assessment: This is a high-quality pure mathematics paper making important contributions to arithmetic dynamics. Through elegant techniques and deep insights, it successfully generalizes the classical one-dimensional result to higher dimensions and establishes a unified theoretical framework. Despite some technical limitations, its methods and results provide important guidance for future research in the field. The paper's rigor, completeness, and originality meet the standards of top-tier mathematics journals.