Explicit Extreme Values of the Argument of the Riemann Zeta-Function

Basic Information

• Paper ID: 2510.14309
• Title: Explicit extreme values of the argument of the Riemann zeta-function
• Authors: Shōta Inoue, Hirotaka Kobayashi, Yuichiro Toma
• Classification: math.NT (Number Theory)
• Publication Date: October 16, 2025
• Paper Link: https://arxiv.org/abs/2510.14309

Abstract

This paper investigates explicit extreme values of the argument of the Riemann ζ-function over short intervals. As an application, it improves upon results by Conrey and Turnage-Butterbaugh concerning r-gaps between zeros of the Riemann ζ-function.

Research Background and Motivation

Problem Background

The argument function of the Riemann ζ-function is defined as $S(t) := \frac{1}{\pi} \arg \zeta(\frac{1}{2} + it)$, and is one of the most important functions in number theory. This function is intimately related to the distribution of zeros through the Riemann-von Mangoldt formula:

$N(T) = \frac{T}{2\pi} \log \left(\frac{T}{2\pi e}\right) + \frac{7}{8} + S(T) + O\left(\frac{1}{T}\right)$

Research Motivation

1. Importance of zero distribution: Detailed information about ζ-function zeros has important applications to the distribution of primes, making the study of extreme values of $S(t+h)-S(t)$ significant
2. Limitations of existing results: Classical results by Selberg and Tsang leave room for improvement in explicit constants
3. Methodological improvements: Employing the Montgomery-Odlyzko resonance method to refine explicit bounds from the Selberg/Tsang approach

Limitations of Existing Methods

• The Selberg-Tsang method yields explicit constants c < 1/√(2eπ)
• Lack of sharp explicit bounds
• For r-gaps problems, existing explicit constants still admit improvement

Core Contributions

1. Improved explicit extreme value bounds for S(t+h)-S(t): Using the resonance method to obtain sharper bounds with leading coefficient √(h/π)log T
2. Enhanced explicit constants for r-gaps: Improving the constant from A₀ ≈ 0.906 in Conrey-Turnage-Butterbaugh results to √2 ≈ 1.414
3. Established theoretical limitations of the Montgomery-Odlyzko method: Proved the theoretical bounds of this method for zero gaps problems
4. Provided comprehensive resonance method analysis: Systematic study of extreme values of S(t) over short intervals

Detailed Methodology

Main Theorems

Theorem 1 (Extreme values of S(t)): Assuming RH, for sufficiently large T and h ∈ C/log T, c/log log T, we have $\sup_{T≤t≤2T} \{±(S(t+h) - S(t))\} ≥ (1-E)\sqrt{\frac{h}{\pi}} \log T$ where the error term E satisfies specific bounds.

Theorem 2 (Improvement for r-gaps): Assuming RH, for sufficiently large r, we have $\lambda_r ≥ 1 + \frac{\sqrt{2}}{\sqrt{r}} - C_1 \frac{(\log r)^{3/2}}{r}$$\mu_r ≤ 1 - \frac{\sqrt{2}}{\sqrt{r}} + C_2 \frac{(\log r)^{3/2}}{r}$

Technical Methods

Resonance Method Framework

Using Dirichlet polynomials as resonators: $R(t) = \sum_{n≤L} f(n)n^{-it}$

where the arithmetic function f is chosen specifically according to the target function.

Key Proposition

Proposition 2: For any non-identically zero arithmetic function f, we have $\sup_{T≤t≤2T} \{±(S(t+h) - S(t))\} ≥ ∓\left(1 + O\left(\frac{1}{T}\right)\right) \frac{2}{\pi} \text{Re} \frac{\sum_{km≤L} \frac{\Λ(k)}{\sqrt{k \log k}} \sin\left(\frac{h}{2} \log k\right) f(m)f(km)}{\sum_{n≤L} |f(n)|^2}$

Optimal Resonator Construction

Selecting multiplicative functions f±, defined on primes p by: $f_±(p) := ±\sqrt{Q} \cdot \frac{\sin(\frac{h}{2} \log p)}{p^{1/2+κh} h \log p}$ when $M < p ≤ L$, where $M = \exp(\sqrt{\log \log L}/\sqrt{h})$.

Parameter choices:

• $κ = \frac{\log(h \log L)}{yh \log L}$
• $y = \sqrt{\log(h \log L)}/(h \log L)$
• Q determined through integral formulas

Technical Innovations

1. Precise resonator design: Careful parameter selection of κ and Q to achieve optimal leading terms
2. Sharp error term control: Using Rankin's technique and partial summation formulas to obtain sharp error estimates
3. Theoretical analysis of method limitations: Proved the theoretical bounds of the Montgomery-Odlyzko method

Experimental Setup

Theoretical Verification

This is primarily a theoretical work, with results verified through rigorous mathematical proofs.

Numerical Computation

• Numerical calculations verify the limitations of the Montgomery-Odlyzko method
• Computed bounds: λ₁ ≥ 3.022, μ₁ ≤ 0.508 as theoretical limits of the method

Parameter Selection

• L ≤ T/(log T)²
• h ∈ C/log T, c/log log T, where C is large and c is small
• W = 22.6 (for λ₁ computation), W = 4.9 (for μ₁ computation)

Experimental Results

Main Results

Improvement in S(t) Extreme Values

Compared to the Selberg-Tsang method's constant c < 1/√(2eπ) ≈ 0.225, this paper achieves a leading coefficient of √(1/π) ≈ 0.564, a significant improvement.

Enhancement of r-gaps Constants

• Previous results: Conrey-Turnage-Butterbaugh obtained Θ = A₀ ≈ 0.906
• This paper's results: Improved to Θ = √2 ≈ 1.414
• Improvement magnitude: Approximately 56% enhancement

Method Limitations

Proved the theoretical limits of the Montgomery-Odlyzko method:

• λ₁ ≥ 3.022 (slight improvement over previous 3.74)
• μ₁ ≤ 0.508
• General case: λᵣ ≥ 1 + √2/√r - O(1/r)

Technical Verification

Error Term Analysis

The error term E in Theorem 1 satisfies: $E \ll \sqrt{h \log \log T} + \min\left\{\sqrt{\frac{\log^3(h \log T)}{h \log T}}, \frac{(\log \log T)^{3/2}}{h^{3/2} \log T}\right\}$

Optimality Proof

Theorem 3 proves the optimality of the √2 constant, demonstrating that the results are optimal within the Montgomery-Odlyzko framework.

Related Work

Historical Development

1. Selberg (unpublished): First proved lower bounds for S(t+h)-S(t)
2. Tsang (1986): Provided complete proof of Selberg's results
3. Montgomery-Odlyzko (1984): Developed the resonance method for studying zero gaps
4. Conrey-Turnage-Butterbaugh (2018): Provided explicit bounds for r-gaps

Method Comparison

• Selberg-Tsang method: Direct but with non-sharp constants
• Montgomery-Odlyzko resonance method: More refined but technically demanding
• Hall method: Can break through resonance method limitations in some cases

Contribution of This Paper

This paper cleverly applies the resonance method to the study of S(t), establishes connections with zero gaps, and achieves significant improvements in explicit constants.

Conclusions and Discussion

Main Conclusions

1. Under RH, the extreme values of S(t+h)-S(t) achieve the sharp bound (1-E)√(h/π)log T
2. The normalized bounds for r-gaps improve to leading terms 1 ± √2/√r
3. Proved these results are optimal within the Montgomery-Odlyzko framework

Limitations

1. RH dependence: Main results require the Riemann Hypothesis
2. Parameter range restrictions: h must lie within specific ranges
3. Method limitations: Proved theoretical bounds exist for the resonance method

Future Directions

1. Seek new techniques to break through Montgomery-Odlyzko method limitations
2. Study extreme value problems over more general intervals
3. Explore possibilities for unconditional results

In-Depth Evaluation

Strengths

1. Technical innovation: Sophisticated application of the resonance method to S(t) research
2. Significant results: Large constant improvements, from 0.906 to 1.414
3. Theoretical completeness: Not only provides improved results but also proves method limitations
4. Rigorous proofs: Precise mathematical reasoning with careful technical details

Weaknesses

1. Strong dependence: Main results depend on the RH assumption
2. Limited application scope: Parameter h restricted to specific ranges
3. Computational complexity: Resonator construction and analysis are quite involved

Impact

1. Theoretical contribution: Important contribution to zero distribution theory in analytic number theory
2. Methodological value: Demonstrates the power of the resonance method on new problems
3. Future research: Provides theoretical foundation and technical framework for further improvements

Applicable Scenarios

1. Study of extreme value problems in analytic number theory
2. Related problems on zero distribution of L-functions
3. Further applications and development of the resonance method

References

This paper cites important literature in the field, including:

1. Selberg's pioneering work
2. Montgomery-Odlyzko's resonance method
3. Conrey-Turnage-Butterbaugh's r-gaps research
4. Developments by Soundararajan and Bondarenko-Seip on the resonance method

Overall Assessment: This is a high-quality analytic number theory paper with significant technical innovations and important improvements in results. It contributes substantially to the development of zero distribution theory. The paper combines deep theoretical insights with refined technical handling, making it an important contribution to the field.