On Hausdorff dimensions of $k$-point configuration sets and Elekes-Rónyai type theorems

This paper establishes dimension expansion results for $k$-point configuration sets generated by real analytic functions, proving that such sets attain positive Lebesgue measure or strictly larger Hausdorff dimensions under specific conditions by leveraging optimal $L^2$-based Sobolev estimates for Fourier integral operators and extending the Mattila-Sjölin and Falconer-type frameworks.

The Gist

Here is an explanation of the paper "On Hausdorff Dimensions of k-Point Configuration Sets and Elekes-Rónyai Type Theorems" by Minh-Quy Pham, translated into everyday language with creative analogies.

The Big Picture: Shaking a Bag of Marbles

Imagine you have a bag filled with marbles. These aren't just normal marbles; they are "fractal marbles." They are so complex and crinkly that they don't have a clear size like a sphere or a cube. Instead, they have a "dimension" that is a fraction (like 0.7 or 0.8). This is what mathematicians call Hausdorff dimension.

Now, imagine you have a machine (a function) that takes three marbles from this bag, mixes them together in a specific way, and spits out a single number (like a score). The big question this paper asks is: If you feed a bunch of these weird, fractal marbles into the machine, how "big" is the pile of scores that comes out?

The author proves that if the machine isn't "boring" (mathematically, if it's not a simple sum of independent parts), then the pile of scores will be significantly bigger than the pile of marbles you put in. In fact, if you put in enough marbles, the pile of scores will fill up a whole line (or even a whole area), rather than just being a thin, dusty line.

---

Part 1: The "Boring" vs. "Exciting" Machines

The paper starts by looking at a famous idea called the Elekes-Rónyai Theorem. Think of this as a rule about "boring" machines.

• The Boring Machine: Imagine a machine that takes three numbers $x, y, z$ and just adds them up: $f(x,y,z) = x + y + z$. If you put in a set of numbers that are very sparse (like just the integers 1, 100, 1000), the output is also very sparse. The machine didn't create any "new" complexity; it just shuffled what was already there.
• The Exciting Machine: Now imagine a machine that multiplies and mixes them, like $f(x,y,z) = x(y+z)$. This machine is "chaotic" in a good way. If you put in a sparse set of numbers, the machine scrambles them so thoroughly that the output becomes a dense, solid block of numbers.

The Main Discovery:
Pham proves that for almost any "real-world" machine (specifically, real analytic functions), it is either a Boring Machine (which can be broken down into simple, separate parts) or an Exciting Machine.

• If it's an Exciting Machine, and you feed it enough fractal marbles, the output isn't just a little bigger; it explodes in size. It goes from being a "dust" (dimension < 1) to being a "solid line" (dimension = 1) or even a "solid block" (positive volume).

Part 2: The Magic Tool (Fourier Integral Operators)

How did the author prove this? He didn't just count marbles. He used a high-tech tool called Fourier Integral Operators (FIOs).

The Analogy: The Sound Engineer
Imagine you have a recording of a complex sound (the fractal set). To understand it, you don't just listen to the raw noise; you run it through a sound engineer's mixer (the FIO).

• In the past, mathematicians used "discrete" methods (like counting pixels on a screen) to prove these things. It was like trying to understand a symphony by counting the number of notes on a sheet of paper. It worked, but it was messy and didn't tell you how big the sound was.
• Pham uses the "sound engineer" approach. He treats the problem as a wave. He looks at how the waves interact.
• The "Folding" Metaphor: The key insight is that the "machine" (the function) acts like a piece of paper being folded. If the fold is sharp and specific (a "Whitney fold"), the waves get compressed and amplified in a very predictable way. This compression allows the "dust" of the input to become a "solid" output.

Part 3: The "k-Point" Configuration Sets

The paper also tackles a broader problem: Configuration Sets.

The Analogy: The Detective's Map
Imagine you are a detective looking at a map of a city. You have a list of locations (points) where a suspect might be.

• 2-Point Problem: You want to know the distance between any two suspects.
• 3-Point Problem: You want to know the area of the triangle formed by three suspects.
• k-Point Problem: You want to know the shape formed by $k$ suspects.

The question is: If your list of suspects is a "fractal dust" (very scattered), does the map of all possible distances or shapes they can form become a solid, filled-in area?

Pham shows that if the rules for measuring these shapes (the function $\Phi$) are "non-degenerate" (meaning they don't have weird symmetries that cancel things out), then yes. If you have enough suspects, the map of their relationships will fill up the space.

Part 4: Why This Matters

1. Breaking the "Special Form" Barrier:
Previous work could only handle simple cases or required the sets to be huge. Pham's method works for a much wider range of functions and smaller sets. He found a "tipping point." For example, if you have sets with a combined dimension greater than $5/3$ (in the 2-variable case), the output is guaranteed to be a solid line.

2. The "Folding" Secret:
The paper reveals that the secret to making these sets grow is the geometry of the fold. If the function folds the space in a specific, non-degenerate way, it acts like a magnifying glass, turning a thin line of data into a thick, solid block.

3. Real-World Applications:
While this sounds abstract, it connects to:

• Distance Problems: How many distinct distances can you find between points in a cloud?
• Signal Processing: Understanding how complex signals interact.
• Geometry: Understanding the structure of shapes in high-dimensional spaces.

Summary in One Sentence

This paper uses advanced wave-mixing techniques (Fourier Integral Operators) to prove that if you mix "fractal dust" through a sufficiently complex machine, the result isn't just a little dust—it becomes a solid, measurable object, provided the machine isn't doing something trivially simple.

The Takeaway: Complexity begets complexity. If you mix things in a non-trivial way, the result is always "larger" than the sum of its parts.

Technical Summary

Here is a detailed technical summary of the paper "On Hausdorff Dimensions of k-Point Configuration Sets and Elekes-Rónyai Type Theorems" by Minh-Quy Pham.

1. Problem Statement

The paper addresses two interconnected problems in geometric measure theory and harmonic analysis:

1. Dimension Expansion (Elekes-Rónyai Type Theorems): Determining conditions under which the image of a product of fractal sets under a real analytic function $f$ has a larger Hausdorff dimension than the original sets, or even positive Lebesgue measure. Specifically, it investigates when $f(A \times \dots \times A)$ expands dimensionally unless $f$ belongs to a "special form" (e.g., $f(x,y) = g(h(x)+k(y))$).
2. Falconer-Type and Mattila-Sjölin Type Problems: Establishing dimensional thresholds for $k$-point configuration sets $\Delta_\Phi(A_1, \dots, A_k) = \{\Phi(x_1, \dots, x_k) : x_i \in A_i\}$ to have positive Lebesgue measure or non-empty interior.

The core challenge is to extend existing results (which often rely on discretized combinatorial arguments or specific curvature conditions) to a broader class of smooth and analytic functions, including those with asymmetric variable dimensions ($d_X \neq d_Y$) and those that do not satisfy standard non-vanishing curvature conditions globally.

2. Methodology

The author employs a microlocal analysis framework centered on Fourier Integral Operators (FIOs), diverging from the discretized sum-product and energy dispersion methods used in recent works by Raz and Zahl.

• Configuration Measure and FIOs: The problem is reformulated by studying the $L^2$ norm of a configuration measure $\nu$ supported on the image set. The squared $L^2$ norm is expressed as an integral involving a kernel $K$, which is identified as the Schwartz kernel of an FIO associated with a canonical relation $\Lambda$.
• Canonical Relations and Singularities: The geometry of the canonical relation $\Lambda$ (specifically the twisted conormal bundle of the incidence relation $\Phi(x) = \Phi(y)$) dictates the mapping properties of the FIO.
  • If $\Lambda$ is nondegenerate, standard $L^2$-Sobolev estimates apply.
  • If $\Lambda$ is degenerate, the author analyzes the specific type of singularity. A key innovation is identifying cases where the projections of $\Lambda$ exhibit Whitney fold singularities.
• Optimal Sobolev Estimates: By proving that the relevant FIOs are associated with folding canonical relations, the author utilizes the sharp $L^2$-Sobolev estimates established by Melrose and Taylor, which incur a loss of only $1/6$derivatives (compared to the$1/2$ loss for general degeneracies).
• Partition Optimization: The variables are partitioned into two groups ($I$ and $J$) to define the FIO. The author optimizes over all possible partitions microlocally to minimize the dimensional threshold required for the configuration set to have positive measure.
• Auxiliary Functions: To characterize "special forms" (degenerate cases), the author introduces auxiliary functions (e.g., $\kappa(x,y)$ for bivariate functions and $\{G_i\}$ for trivariate functions). The vanishing of these functions identically characterizes the special forms; their non-vanishing ensures the non-degeneracy (or specific folding structure) of the canonical relation.

3. Key Contributions and Results

A. Elekes-Rónyai Type Theorems for Analytic Functions

The paper provides alternative proofs and significant improvements to recent results by Raz and Zahl (2024) and Koh, Pham, and Shen (2024).

• Bivariate Case (Theorem 1.4):

  • For a real analytic function $f: \Omega \subset \mathbb{R}^2 \to \mathbb{R}$, if $f$ is not an analytic special form, and $A, B \subset [0,1]$ have Hausdorff dimensions $\alpha, \beta$:
    • If $\alpha + \beta > 5/3$, then $f(A \times B)$ has positive Lebesgue measure.
    • If $\alpha + \beta \le 5/3$, then $\dim_H f(A \times B) \ge \alpha + \beta - 2/3$.
  • Significance: This improves the previous bound $\alpha + \beta > 2$ (for positive measure) to $5/3$and provides explicit quantitative gains for dimensions$> 2/3$. It also establishes that the image has positive measure for$\alpha > 5/6$(when$\alpha=\beta$).

• Trivariate Case (Theorem 1.8):

  • For a trivariate real analytic function $f(x,y,z)$ not of the form $g(h(x)+k(y)+l(z))$:
    • If $\dim_H A + \dim_H B + \dim_H C > 2$, then $f(A \times B \times C)$ has positive Lebesgue measure.
  • Significance: This extends the result of Koh et al. from quadratic polynomials to general real analytic functions, removing the need for specialized combinatorial arguments.

B. General k-Point Configuration Sets

The paper generalizes Falconer-type results to asymmetric settings and functions with lower rank mixed Hessians.

• General Thresholds (Theorem 1.14 & 3.8):

  • For a smooth function $\Phi: X \times Y \to \mathbb{R}$ with mixed Hessian rank $r \ge 1$:
    • If $\dim_H A + \dim_H B > d_X + d_Y + 1 - r$, then $\Delta_\Phi(A,B)$ has positive Lebesgue measure.
  • This recovers the classic result for distance sets (where rank is maximal) but applies to functions where the rank $r$ is lower, adjusting the threshold accordingly.

• Asymmetric Configurations: The framework handles cases where the dimensions of the domain variables differ ($d_X \neq d_Y$), which previous Mattila-Sjölin type results could not easily address.

C. Application to Distance Sets on Hypersurfaces

• Theorem 6.1: For a smooth hypersurface $S \subset \mathbb{R}^d$ and sets $A \subset \mathbb{R}^d, B \subset S$:
  • If $\dim_H A + \dim_H B > d$, the distance set $\Delta(A,B)$ has positive Lebesgue measure.
  • This generalizes previous results to cases where one set lies on a curved surface and the other is in the ambient space, provided the set $A$ is not contained in the union of tangent spaces of $S$.

4. Significance and Impact

1. Unification of Methods: The paper successfully bridges the gap between combinatorial Elekes-Rónyai theorems and harmonic analytic Falconer-type problems. It demonstrates that FIO techniques, previously used primarily for non-degenerate curvature conditions, can be adapted to handle "folding" singularities arising from analytic special forms.
2. Quantitative Improvements: By exploiting the specific geometry of Whitney folds (Melrose-Taylor estimates), the author achieves explicit dimensional thresholds (e.g., $5/3$instead of non-explicit$\epsilon$ gains) and proves the existence of positive Lebesgue measure (a stronger condition than just dimension expansion) for sets with relatively low dimensions.
3. Characterization of Degeneracy: The introduction of auxiliary functions ($\kappa$ and $G_i$) provides a rigorous analytic criterion to distinguish between "expanding" functions and "special forms" in the continuous setting, mirroring the algebraic characterization in the discrete setting.
4. Extension to Asymmetry: The results significantly broaden the scope of configuration set problems to include asymmetric domains and functions with vanishing mixed Hessians, areas where previous methods were often vacuous.

In summary, Minh-Quy Pham's work offers a powerful microlocal toolkit that refines our understanding of how analytic functions map fractal sets, providing sharper bounds and broader applicability than previous combinatorial or discretized approaches.