🔢 mathematics

Rational Points and Brownian Motion

This paper establishes an almost sure asymptotic expansion for the number of rational points near the graph of Brownian motion across all parameter regimes (except a critical one), thereby extending Diophantine approximation theory beyond smooth curves and bridging Number Theory with Multifractal Analysis.

Original authors: Faustin Adiceam, Volodymyr Pavlenkov, Evgeniy Zorin

Published 2026-08-25

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Faustin Adiceam, Volodymyr Pavlenkov, Evgeniy Zorin

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the vast landscape of mathematics, there is a quiet but persistent question about how numbers fit together. Specifically, mathematicians have long been fascinated by rational numbers—those simple fractions like one-half or three-quarters—and how they are scattered across the number line. When you look at a smooth, predictable curve, such as a perfect circle or a gentle hill, you can often predict exactly how many of these fractions will land very close to it. If the curve is smooth enough, the number of nearby fractions grows in a steady, predictable way, much like counting how many raindrops might hit a specific patch of ground during a storm. This relationship between the shape of a curve and the density of nearby fractions is a cornerstone of a field called Diophantine approximation, which studies how well complex numbers can be approximated by simple ones.

However, the world of mathematics is not always smooth. There are shapes that are jagged, erratic, and infinitely complex, defying the neat rules that apply to smooth curves. One of the most famous examples of such a shape is the path traced by a particle moving under the influence of random forces, a phenomenon known as Brownian motion. Imagine a speck of dust floating in a sunbeam, jostled by invisible air molecules; its path is a continuous line that never repeats itself and is so rough that it has no defined slope at any point. For decades, mathematicians have wondered how the simple fractions behave near these chaotic, jagged paths. The old rules for smooth curves seemed to break down, leaving a gap in our understanding of how numbers interact with randomness.

A team of researchers has now stepped into this gap to explore the relationship between rational numbers and the chaotic paths of Brownian motion. Their work tackles a specific problem: counting how many fractions with denominators up to a certain size fall within a narrow strip surrounding the path of a random particle. In the world of smooth curves, the answer to this question depends heavily on the curve's shape and how "thick" the strip is. But for a random path, the shape is constantly changing in unpredictable ways. The researchers set out to see if a general rule could still be found, or if the randomness made the count entirely unpredictable.

The team discovered that, surprisingly, a clear pattern does emerge, even in the chaos. They found that for almost every possible random path, the number of fractions landing near the path follows a precise mathematical law, provided the strip around the path is not too thin. The key finding is that the number of these fractions grows in a way that is directly tied to the area of the strip surrounding the path. This confirms a long-held intuition that the count of nearby fractions should be proportional to the space they occupy, a concept known as the area heuristic. The researchers proved that this rule holds true for random paths across a wide range of conditions, bridging a divide between the orderly world of number theory and the chaotic world of probability.

However, the story is not without its exceptions. The researchers identified a very specific, critical point where the rules change. If the strip around the path becomes extremely thin—specifically, if its width shrinks at a rate related to the fifth power of the denominator limit—the predictable pattern breaks down. In this narrow regime, the behavior of the fractions becomes uncertain. The team could not prove whether the count would follow a smooth curve or behave erratically in this specific zone. They suspect this point marks a genuine transition in how the numbers behave, but pinning down the exact nature of this transition remains an open question for future study.

To reach these conclusions, the team had to develop new tools, as the standard methods used for smooth curves rely on calculus and smooth slopes, neither of which exist for a random path. Instead, they used a method based on probability and statistics, treating the path as a living, breathing entity that moves through time. They calculated the likelihood of the path hitting specific small boxes in the plane and used these probabilities to estimate the total number of fractions that would land nearby. By carefully analyzing the geometry of the path and the distribution of the fractions, they were able to derive an exact formula for the average area of the strip surrounding the path and show that the count of fractions matches this area almost perfectly.

The significance of this work extends beyond just counting numbers. It connects two seemingly unrelated fields: the study of numbers and the study of how rough or smooth a shape is at a microscopic level. The researchers found that the way the fractions cluster near the path is deeply linked to the "roughness" of the path itself. In the case of Brownian motion, the path has a specific, well-known level of roughness, and this level dictates exactly how the fractions are distributed. This suggests a deeper, unifying principle that might apply to many other irregular shapes, hinting that the local texture of a curve determines the global behavior of the numbers near it.

While the paper provides a definitive answer for most scenarios, it leaves the critical, ultra-thin regime as a mystery. The authors suggest that in this tiny zone, the behavior might be different, perhaps following a different statistical law entirely. They propose that the transition between the predictable world and this chaotic edge is a fundamental feature of how numbers interact with randomness. This work completes a major chapter in the theory of Diophantine approximation on random paths, a program that began decades ago, and it opens the door to understanding how the simple, orderly world of fractions behaves when faced with the infinite complexity of the random world.

Technical Summary: Rational Points and Brownian Motion

Problem Statement
The paper addresses the problem of counting rational points with bounded denominators lying in a shrinking tubular neighborhood of a curve. Specifically, given a real-valued function $f$ on a bounded interval $I$ , and parameters $Q \ge 1$ and $\delta > 0$ , the authors study the counting function $N_f(\delta, Q)$ , which enumerates rational points $(p/q, r/q)$ with $1 \le q \le Q$ lying within a distance $\delta/Q$ (in the sup-norm) of the graph of $f$ .

The central question is the validity of the "Area Heuristic," which predicts that $N_f(\delta, Q)$ should be asymptotically proportional to the area of the tubular neighborhood multiplied by $Q^3$ , provided the neighborhood is sufficiently thick relative to $Q$ . While this heuristic is well-established for sufficiently regular curves (e.g., twice continuously differentiable with non-vanishing Lipschitz second derivative), it remains unexplored for curves with minimal regularity. The paper focuses on the case where $f$ is an almost sure realization of the graph of standard linear Brownian motion, a continuous but nowhere differentiable function.

Methodology
The authors employ probabilistic methods, specifically the Second Moment Method, adapted to the stochastic setting of Brownian motion. The core of the proof involves:

Geometric Decomposition: The problem is reduced to estimating the probability that a Brownian trajectory hits specific rectangular regions in the plane. These rectangles are centered at rational points with side lengths determined by the approximation function.
Probability Estimates: A significant portion of the work (Section 4) is dedicated to deriving sharp, closed-form estimates and asymptotic bounds for the probability that Brownian motion hits a rectangle. This involves decomposing the hitting probability into boundary contributions and using properties of Gaussian densities, the complementary error function ( $\text{erfc}$ ), and Owen's $T$ -function.
Moment Analysis: The counting function is treated as a sum of indicator random variables. The authors establish asymptotic expansions for the first moment (expectation) and the second moment (variance) of this sum.
- Diagonal Terms: These correspond to the expectation of the count and are estimated using the hitting probabilities of single rectangles.
- Off-Diagonal Terms: These correspond to the covariance between counts at different rational points. The authors decompose the summation domain into four distinct regimes based on the relative positions of the rational points (horizontally distant, vertically distant, close with distinct centers, and same centers) to handle the correlations.
Area Heuristic Derivation: A crucial intermediate step is the derivation of the almost sure asymptotic behavior of the area of the tubular neighborhood of the Brownian graph. This is achieved using a slicing technique and Fubini's Theorem to relate the area to the local range of the Brownian motion.

Key Results

Tubular Neighborhood Area Theorem: The authors prove that for a Brownian motion restricted to $[T_1, T_2]$ , the area $\hat{A}(T; \eta)$ of its $\eta$ -tubular neighborhood satisfies the almost sure asymptotic:
$\hat{A}(T; \eta) \sim \frac{4}{\sqrt{\pi}} (T_2 - T_1) \sqrt{\eta} \quad \text{as } \eta \to 0^+.$
Notably, this leading term lacks the logarithmic corrections often found in the theory of Wiener sausages in the plane.
First Main Theorem (Asymptotic Counting): For a general approximation function $\xi(q)$ (satisfying monotonicity conditions), the paper establishes an almost sure asymptotic expansion for the counting function $\hat{N}_\xi(T; Q)$ . The main term is proportional to the partial sum $\Xi(Q) = \sum_{q=1}^Q q^{3/2}\sqrt{\xi(q)}$ .
Second Main Theorem (Uniform Counting): For the specific case of a uniform tubular neighborhood of width $\delta/Q$ (where $\delta = \delta(Q)$ ), the paper establishes:
- If $\delta(Q) \cdot Q^5 \gg (\log Q)^{2+\eta}$ , then almost surely:
  $N^\flat(T; (\delta, Q)) \sim \frac{4}{3\sqrt{\pi}} (T_2 - T_1) \sqrt{\delta Q^5}.$
- If $\delta(Q) \cdot Q^5 \ll (\log Q)^{-2-\eta}$ , then almost surely $N^\flat(T; (\delta, Q)) = 0$ for all large $Q$ .
- The critical regime where $\delta(Q) \cdot Q^5$ is of order $(\log Q)^k$ remains undetermined regarding an almost sure limit, though a dichotomy in probability is established.
Auxiliary Theorem: In the critical regime where $\delta Q^5 \to \infty$ or $\to 0$ , the counting function converges in probability to the expected value or zero, respectively.

Significance and Claims
The paper claims to complete the counting aspect of the theory of Diophantine approximation on the graph of Brownian motion, a program initiated by Sprindžuk. The primary significance lies in extending the Area Heuristic to a regime of minimal regularity (nowhere differentiable curves), a case previously inaccessible to methods relying on Taylor expansions.

The authors posit that their results hint at a unifying theory connecting the analysis of rational points near curves with the local Hölder regularity and fine-scale oscillations of the curve. Specifically, they propose an "Oscillation Principle" (Conjecture 1.2), suggesting that for monofractal functions with Hölder exponent $\gamma$ , the counting function scales as $(\delta/Q)^\gamma Q^3$ . The Brownian motion case ( $\gamma = 1/2$ ) serves as a probabilistic verification of this principle for $\alpha < 5$ (where $\delta \gg Q^{-\alpha}$ ).

The work is presented as building a bridge between Number Theory and Multifractal Analysis. The authors note that while the probabilistic method works for Brownian motion, the deterministic case for monofractal functions (like Weierstrass functions) remains open, as neither probabilistic nor Taylor-based methods are directly available.

Limitations and Open Questions
The paper explicitly identifies the regime where $\delta(Q) \cdot Q^5$ is of order $(\log Q)^k$ as a critical transition point where the almost sure behavior is not fully determined. The authors conjecture this corresponds to a genuine transition in asymptotic behavior but leave the complete determination of the probability distribution in this regime open. Additionally, the extension of the Oscillation Principle to multifractal maps (e.g., Riemann's non-differentiable function) is noted as a challenging open problem.

More like this