$ABC$ Sum-Product Theorems for Katz-Tao Sets

Basic Information

• Paper ID: 2511.05091
• Title: ABC sum-product theorems for Katz-Tao sets
• Author: Tuomas Orponen
• Classification: math.CO (Combinatorics), math.CA (Analysis)
• Publication Date: November 10, 2025
• Paper Link: https://arxiv.org/abs/2511.05091

Abstract

This paper proves two variants of $ABC$ sum-product theorems for $\delta$-separated sets $A,B,C \subset [0,1]$ satisfying the Katz-Tao interval condition. The main innovation is that the cardinalities of sets $B,C$ need not match their non-concentration exponents. The new $ABC$ theorems are optimal under their respective assumptions and imply previous results.

Research Background and Motivation

Problem Background

The $ABC$ sum-product problem is a fundamental question in combinatorics and analysis, asking: for three sets $A,B,C \subset \mathbb{R}$, under what sufficient conditions can we guarantee that $A + cB$ (for some $c \in C$) has significantly larger "size" than $A$?

Historical Development

1. Finite Set Case: When $A,B,C$ are finite sets measured by cardinality $|\cdot|$, the Szemerédi-Trotter theorem provides necessary and sufficient conditions: $\max\{|B|, |C|\} = 1 \text{ and } |B||C| = |A|$
2. Compact Set Case: For compact sets $A,B,C \subset \mathbb{R}$ measured by Hausdorff dimension $\dim_H$, Orponen-Shmerkin 15 resolved the $ABC$ problem with necessary and sufficient conditions: $\max\{\dim_H B, \dim_H C\} > 0 \text{ and } \dim_H B + \dim_H C > \dim_H A$

Limitations of Existing Methods

The classical discrete version (Theorem 1.1) requires $B,C$ to be Frostman sets, which means:

• Set cardinality roughly matches its non-concentration exponent
• Requires $\text{diam}(B) \gtrsim \delta^{\chi/\beta}$ and $\text{diam}(C) \gtrsim \delta^{\chi/\gamma}$
• When $B$ or $C$ has small diameter, the theorem provides almost no information

Research Motivation

This paper aims to prove more flexible $ABC$ theorems where:

1. Set cardinality need not match non-concentration exponents
2. Sets $B,C$ may have small diameter
3. Natural Katz-Tao conditions replace Frostman conditions

Core Contributions

1. Theorem 1.6: Proves the first main variant where $A,B,C$ all satisfy Katz-Tao conditions, and condition (Π): $|B|^\gamma |C|^\beta \delta^{\beta\gamma} \geq \delta^{-\eta}$ is optimal.
2. Theorem 1.12: Proves a second variant removing the Katz-Tao condition on $B$, at the cost of imposing a mild Frostman condition (two-ends condition) on $C$.
3. Theoretical Completeness:
   • Shows the new theorems imply the classical Theorem 1.1
   • Provides Example 1.13 proving condition (Π) in Theorem 1.12 is optimal
   • New theorems are optimal under their respective assumptions
4. Technical Innovation: Introduces new scale decomposition techniques using "branching functions" and multiscale analysis to handle cases where cardinality and non-concentration exponents mismatch.

Detailed Methodology

Core Concepts

1. Katz-Tao Sets

Definition 1.5: A set $P \subset \mathbb{R}^d$ is called a Katz-Tao $(δ, s, C)$-set if: $|P \cap B(x,r)|_\delta \leq C\left(\frac{r}{\delta}\right)^s, \quad \forall x \in \mathbb{R}^d, r \geq \delta$

This is a non-concentration condition controlling set distribution at different scales.

2. Frostman Sets

Definition 1.3: A set $P \subset \mathbb{R}^d$ is called a Frostman $(δ, s, C)$-set if: $|P \cap B(x,r)|_\delta \leq Cr^s|P|_\delta, \quad \forall x \in \mathbb{R}^d, r \geq \delta$

3. Key Condition (Π)

In Theorem 1.6, condition (Π): $|B|^\gamma |C|^\beta \delta^{\beta\gamma} \geq \delta^{-\eta}$ has profound geometric meaning:

• It is equivalent to $\text{diam}(B) \cdot \text{diam}(C) \gtrsim \delta$
• This is a necessary condition to avoid trivial cases (Remark 1.9)

Proof Strategy for Theorem 1.6

Step One: Reduction to $C \subset [1/2, 1]$

Proposition 4.1: Through scale renormalization, we may assume $C \subset [1/2, 1]$. The key observation is the invariance of condition (Π) under renormalization (equation 4.10).

Step Two: Uniformization

Using Proposition 2.9, decompose $A,B,C$ into uniform subsets, i.e., there exists a scale sequence $\delta = \Delta_m < \cdots < \Delta_0 = 1$ such that sets have regular branching structure at these scales.

Step Three: Branching Function Analysis

For a uniform set $P$, define the branching function $f:[0,m] \to [0,dm]$: $f(j) := \frac{\log|P|_{2^{-jT}}}{T}$

Lemma 2.13 provides a decomposition of the branching function: there exist scale sequences $\{a_j\}$ and slope sequences $\{\sigma_j\}$ such that $f$ is $(\sigma_{j+1}, 0)$-superlinear on each interval $[a_j, a_{j+1}]$.

Step Four: Finding Critical Scales

The core of the proof is finding an index $j \in \{0,\ldots,n-1\}$ such that: $\gamma_{j+1} \geq \eta/2 \text{ and } |B|^\gamma |C|^\beta\left(\frac{\Delta_j}{\Delta_{j+1}}\right)^{\beta\gamma} \geq \left(\frac{\Delta_j}{\Delta_{j+1}}\right)^{\alpha\eta/(2\gamma)}$

This condition (equation 4.16) ensures that between scales $\Delta_j, \Delta_{j+1}$, we can apply the auxiliary Proposition 3.1.

Step Five: Applying Proposition 3.1

Proposition 3.1 is a key auxiliary result, a variant of Theorem 1.1 where the Frostman condition on $B$ is replaced by a Katz-Tao condition on $A$. At the identified scales, define renormalized sets: $A_I := S_I(D_\delta(A \cap I)), \quad B_J := S_J(D_{\delta/\Delta_j}(B \cap J))$

These sets satisfy the conditions of Proposition 3.1, so there exists $c \in C$ such that: $|{a + cb : (a,b) \in G}|_\Delta \geq \Delta^{-\chi}|A_I|$

Step Six: From Local to Global

By carefully tracking contributions from different pieces, we derive the global estimate from local estimates: $|{a + cb : (a,b) \in G}|_\delta \geq \delta^{-\epsilon}|A|$

Proof of Theorem 1.12

The proof of Theorem 1.12 simplifies the argument for Theorem 1.6, since the Frostman condition on $C$ automatically ensures $\gamma_{j+1} > 0$ (equation 5.4), thus the Katz-Tao condition on $B$ is not needed.

Technical Innovations

1. Scale Separation Technique: Through branching function analysis, identify "good" scales where classical results apply.
2. Optimality of Conditions: Example 1.13 constructs precise counterexamples proving condition (Π) cannot be relaxed.
3. Application of Balog-Szemerédi-Gowers Theorem (Claim 4.2): Use BSG theorem to extract structured subsets enabling application of Plünnecke-Ruzsa inequalities.
4. Multiscale Induction: The proof cleverly establishes connections between different scales, ultimately reducing the problem to known Proposition 3.1.

Experimental Setup

This is a pure mathematical theory paper with no numerical experiments or datasets. All results are rigorous mathematical proofs.

Main Theorem Statements

Theorem 1.6 (Main Result 1)

For each $\alpha \in (0,1)$, $\beta, \gamma \in [\alpha, 1]$ and $\eta > 0$, there exist $\delta_0, \epsilon > 0$ such that for all $\delta \in 2^{-\mathbb{N}} \cap (0, \delta_0]$, if $A,B,C \subset \delta\mathbb{Z} \cap [0,1]$ satisfy:

• (A) $A$ is a Katz-Tao $(δ, α)$-set
• (B) $B$ is a Katz-Tao $(δ, β)$-set
• (C) $C$ is a Katz-Tao $(δ, γ)$-set
• (Π) $|B|^\gamma |C|^\beta \delta^{\beta\gamma} \geq \delta^{-\eta}$

then there exists $c \in C$ such that: $|{a + cb : (a,b) \in G}|_\delta \geq \delta^{-\epsilon}|A|, \quad G \subset A \times B, |G| \geq \delta^\epsilon|A||B|$

Theorem 1.12 (Main Result 2)

For each $\alpha \in (0,1)$, $\gamma \in [\alpha, 1]$ and $\eta > 0$, there exist $\delta_0, \epsilon > 0$ such that for all $\delta \in 2^{-\mathbb{N}} \cap (0, \delta_0]$, if $A,B,C \subset \delta\mathbb{Z} \cap [0,1]$ satisfy:

• (A) $A$ is a Katz-Tao $(δ, α)$-set
• (C) $C$ is a Katz-Tao $(δ, γ)$-set and a Frostman $(δ, η, δ^{-\epsilon})$-set
• (Π) $|B|^\gamma |C|^\alpha \delta^{\alpha\gamma} \geq \delta^{-\eta}$

then there exists $c \in C$ such that the conclusion holds as in Theorem 1.6.

Related Work

Historical Context

1. Szemerédi-Trotter Theorem 23: Provides incidence bounds for finite sets, foundational for the $ABC$ problem.
2. Bourgain's Work 1,2: First studied discretized sum-product problems, introducing key techniques.
3. Orponen-Shmerkin 15: Resolved the Hausdorff dimension version of the $ABC$ problem; Theorem 1.1 originates from this work.
4. Furstenberg Set Conjecture: Finally resolved by Ren-Wang 18, building on 15.

Relation to Recent Work

Demeter-Wang 4 and Wang-Wu 25,26: Recently proved very strong incidence theorems under Katz-Tao assumptions. Their main result (equation 1.8) requires exponent $\gamma$ to satisfy: $\gamma = \min\{\alpha + \beta, 2 - \alpha - \beta\}$

In contrast, Theorem 1.6 requires more flexible exponent relationships, though 26 yields stronger results in certain parameter ranges.

Advantages of This Work

1. Broader Applicability: Does not require cardinality to match exponents
2. Allows Small Diameter Sets: Overcomes the main limitation of Theorem 1.1
3. Optimality: Achieves optimal bounds under respective assumptions

Conclusions and Discussion

Main Conclusions

1. Proves two new $ABC$ sum-product theorems (Theorems 1.6 and 1.12) working under Katz-Tao conditions without requiring cardinality to match non-concentration exponents.
2. New theorems imply the classical Theorem 1.1 (Section 6), providing a more unified framework.
3. Condition (Π) is optimal under respective assumptions, as shown by Example 1.13.

Limitations

1. Exponent Range Restrictions: Theorem 1.6 requires $\beta, \gamma \in [\alpha, 1]$; while Remark 1.7 indicates this is not very restrictive, limitations remain.
2. Constant Dependence: The constant $\epsilon$ may be very small, depending on parameters $\alpha, \beta, \gamma, \eta$ in complex ways.
3. Comparison with 26: In certain parameter ranges (e.g., $\beta < \alpha$), stronger results might exist if $B$ simultaneously satisfies Katz-Tao conditions (see Section 1.1 discussion).

Future Directions

1. Optimal Constants: Determine optimal dependence of $\epsilon$.
2. Unification with Incidence Theory: Explore deeper connections with Demeter-Wang and Wang-Wu work.
3. Higher-Dimensional Generalizations: Extend results to $\mathbb{R}^d$ ($d \geq 2$).
4. Applications: Apply new theorems to Furstenberg sets, Kakeya problems, and other geometric measure theory questions.

In-Depth Evaluation

Strengths

1. Theoretical Depth:
   • Proof techniques are sophisticated, particularly the multiscale analysis and branching function usage
   • Results are optimal under respective assumptions with precise counterexamples
   • Unifies previous results (Theorem 1.1)
2. Technical Innovation:
   • Scale renormalization technique (Proposition 4.1) handles difficulties when $C$ may lie on small intervals
   • Claim 4.2 cleverly applies BSG theorem to extract structure
   • Proposition 3.1 serves as key bridge connecting different assumptions
3. Clarity of Exposition:
   • Well-organized paper progressing from simple to complex
   • Abundant remarks explaining necessity of conditions and result significance
   • Detailed counterexamples (Example 1.13)
4. Completeness:
   • Not only proves new results but shows how they imply old ones
   • Thorough comparison with related work (Section 1.1)

Weaknesses

1. Readability Challenges:
   • Proof is extremely technical, requiring reader familiarity with substantial background
   • Constant choices (e.g., equations 4.2, 4.11, 5.1) involve complex dependencies, difficult to track
2. Computational Complexity:
   • Specific bounds for constant $\epsilon$ are unclear, potentially very small
   • Threshold $\delta_0$ dependence is complex
3. Limited Application Guidance:
   • As pure theory, lacks discussion of concrete applications
   • Unclear guidance on when to use Theorem 1.6 vs. Theorem 1.12
4. Unresolved Questions:
   • Section 1.1 end mentions: if $B$ simultaneously satisfies Katz-Tao $(δ, β)$ condition ($\beta < \alpha$), what is the optimal result?

Impact

1. Contribution to Field:
   • Advances understanding of discretized sum-product problems
   • Provides new tools for Katz-Tao set research
   • May have further applications to Furstenberg sets, Kakeya problems
2. Theoretical Value:
   • Proof techniques (particularly multiscale analysis) may apply to other problems
   • Optimality of condition (Π) provides deep geometric insight
3. Reproducibility:
   • As mathematical proof, completely reproducible
   • All cited lemmas have clear references

Applicable Scenarios

1. Theoretical Research:
   • Sum-product problems in combinatorics
   • Projection and incidence problems in geometric measure theory
   • Restriction estimates in harmonic analysis
2. Specific Problems:
   • When sets satisfy Katz-Tao conditions but cardinality mismatches exponents
   • When handling small diameter sets
   • Studying Furstenberg sets and related conjectures

Technical Details Supplement

Role of Branching Functions

The branching function $f:[0,m] \to [0,dm]$ encodes uniform set structure at different scales:

• $f(j) = \frac{\log|P|_{2^{-jT}}}{T}$
• Function slope $\sigma_j = s_f(a_{j-1}, a_j)$ reflects set "dimension" in that scale range
• Lemmas 2.15 and 2.16 convert branching function properties to Frostman/Katz-Tao conditions

Derivation of Key Inequalities

In the proof, derivation of equation (4.16) is critical: $|B|^{\alpha/\beta}_{\delta/\Delta_{j+1} \to \delta/\Delta_j} \left(\frac{\Delta_j}{\Delta_{j+1}}\right)^{(\alpha/\gamma)\gamma_{j+1}} \geq \left(\frac{\Delta_j}{\Delta_{j+1}}\right)^{\alpha(1+\eta/(2\gamma))}$

This is proved by contradiction: if it fails for all $j$, taking products contradicts condition (Π).

Constant Selection Strategy

Constant choices in the paper follow a "reverse engineering" principle:

1. First determine final needed $\epsilon$ (from Proposition 3.1)
2. Then choose $\xi$ so that $\xi \leq \epsilon_0 \tau$ (equation 4.11)
3. Next choose $\zeta$ enabling BSG theorem application
4. Finally choose $\epsilon$ satisfying all constraints

While technically demanding, this strategy ensures proof rigor.

Key References

2 J. Bourgain. The discretized sum-product and projection theorems. J. Anal. Math., 2010.

15 T. Orponen and P. Shmerkin. Projections, Furstenberg sets, and the ABC sum-product problem. arXiv:2301.10199, 2023.

18 K. Ren and H. Wang. Furstenberg sets estimate in the plane. arXiv:2308.08819, 2023.

26 H. Wang and S. Wu. Two-ends Furstenberg estimates in the plane. arXiv:2509.21869, 2025.

Overall Assessment: This is a high-quality theoretical mathematics paper making important progress in discretized sum-product theory. The proof techniques are sophisticated, results are optimal with precise counterexamples, and the work unifies previous theory. While highly technical, it represents a significant contribution for researchers in this field. The paper not only proves new results but demonstrates how they unify and improve existing theory, reflecting deep mathematical insight.