How many points contain homothetic copies in their Hurwitz continued fraction expansion?

Technical Summary: Homothetic Copies in Hurwitz Continued Fraction Expansions

Problem Statement
The paper addresses the existence of multidimensional pattern structures within number-theoretic expansions. While Szemerédi's theorem and its multidimensional extensions by Furstenberg and Katznelson establish that subsets of integers with positive upper density contain arithmetic progressions and homothetic copies of finite sets, the question remains whether such patterns emerge in the partial quotients of continued fraction expansions. Specifically, the authors investigate the Hurwitz continued fraction expansion for complex irrationals. They seek to determine the Hausdorff dimension of the set of complex numbers whose partial quotients (viewed as a subset of $\mathbb{Z}^2$ via the identification of Gaussian integers) contain infinitely many homothetic copies of any finite subset of $\mathbb{Z}^2$.

Methodology
The authors employ a constructive approach using Iterated Function Systems (IFS) derived from the Hurwitz continued fraction. The methodology proceeds in three main stages:

1. Framework Construction: The authors restrict their analysis to the set $E$ of complex irrationals in a fundamental domain $U$ whose partial quotients are distinct. They utilize the canonical infinite IFS associated with the Hurwitz expansion, defined by maps $\phi_i(x) = (x + k + i\ell)^{-1}$ for Gaussian integers $(k, \ell)$ with sufficiently large norms. This system satisfies the open set condition, conformality, and uniform contraction properties.

2. Seed Set Construction (Lower Bound): To establish a lower bound for the dimension, the authors construct a "seed set" $S \subset E$. This set consists of points where the partial quotients $c_n(x)$ grow rapidly and strictly ($3^n \leq \|c_n(x)\|_\infty < 2 \cdot 3^n$). Using a mass distribution principle, they prove that this seed set has a Hausdor dimension of at least 1. The proof involves defining a probability measure on the space of symbolic sequences and estimating the measure of balls in the Euclidean space relative to their radii, utilizing distortion theorems for univalent functions.

3. Pattern Insertion (Modification): To ensure the set contains the required homothetic copies, the authors modify the seed set. They insert "integer squares" (finite blocks of Gaussian integers forming square grids) into the sequence of partial quotients at specific, widely spaced indices. These inserted blocks are chosen to be large enough to contain homothetic copies of any finite subset of $\mathbb{Z}^2$. The insertion is performed carefully to maintain the distinctness of partial quotients (keeping the set within $E$) while ensuring the Hausdorff dimension remains close to 1.

4. Dimension Estimation: The authors prove that the mapping from the modified set $\tilde{S}$ back to the seed set $S$ (which removes the inserted blocks) is Hölder continuous with an exponent arbitrarily close to 1. This allows them to transfer the dimension lower bound from $S$ to the final set $H$.

Key Contributions and Results

• Main Theorem (Theorem 1.1): The authors prove that the Hausdorff dimension of the set $H$ (complex irrationals in $E$ containing infinitely many homothetic copies of any finite subset of $\mathbb{Z}^2$) is exactly 1. Since $H \subset E$ and the dimension of $E$ is known to be at most 1, the result is $\dim_H H = \dim_H E = 1$.
• Explicit Construction: Unlike previous results on regular continued fractions which often yielded dimensions strictly less than 1 (e.g., $1/2$ for strictly increasing partial quotients), this work provides a construction where the dimension is maximal (1) within the constrained set $E$.
• Multidimensional Pattern Emergence: The paper demonstrates that the combinatorial richness of multidimensional patterns (homothetic copies) is compatible with the fractal complexity of the Hurwitz expansion.

Significance
The paper claims to provide a "clear and concrete example of multidimensional pattern emergence in number-theoretic expansions." The significance lies in showing that the set of points exhibiting these complex combinatorial structures is not a negligible set (in terms of dimension) but rather a maximal dimensional subset of the naturally constrained space. The authors note that their method, which involves constructing a seed set and modifying it with large integer squares, extends to other expansions, suggesting a broader relevance for the study of multidimensional pattern formation in dynamical systems. The work bridges combinatorial number theory (Szemerédi-type theorems) with fractal geometry (Hausdorff dimension) in the context of complex continued fractions.