REVIEW
Large value estimates for Dirichlet polynomials and trilinear sums over short intervals
Number Theory · Submitted 12 Oct 2026 · 80 references
Pith's verdict: A careful, incremental extension of Matomaki et al.'s method to a new range, but the entire construction depends on two unverified recent large-value estimates.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
This paper improves the Baker–Harman–Pintz mean-value estimate for products of three Dirichlet polynomials, obtaining a larger exponent range and a 5.5% improvement at θ = 0.52, with direct consequences for the existence of primes in…
The reading
What carries the argument
The proof reduces the mean-value estimate to showing that a certain piecewise-linear function $P₁(t)+P₂(t)+P₃(t)-t-ε^{10}$ is positive for all $t∈[ε⁶/2,1-θ]$ (Lemma 2.1). The $P_j$ are defined via a lower bound $R_α(t)$, whose explicit formula (9) comes from the large-value estimates of Guth–Maynard and Chen–Gupta–Li applied to powers of the Dirichlet polynomials. The core of the paper is a careful case analysis—captured in Lemmas 2.2 and 2.3—that verifies positivity of these combinations under the exponent conditions stated in Theorems 1 and 2. The new $γ₂$ and $γ₃$ terms arise from choosing integer parameters $g₂,g₃$ that optimise the resulting inequalities.
Extended reading notes
Core claim
The central claim, stated as Theorem 1, is that for $\frac12<θ<\frac47$ the mean-value estimate (1) holds whenever $R(s)$ is prime-factored and the exponents satisfy $|α₁-α₂|<2θ-1$ and $1-α₁-α₂ < γ(θ)$, where $γ(θ)$ is the maximum of three families of explicit piecewise-linear functions $γ₁_{g₁}(θ)$, $γ₂_{g₂}(θ)$ (for $g₂\ge5$ with $2θ-1$ in a specific interval $I_{g₂}$), and $γ₃_{g₃}(θ)$ (with $g₃$ depending on $θ$). The term $γ₁$ reproduces the 2001 Baker–Harman–Pintz bound; the new terms $γ₂$ and $γ₃$ enlarge the admissible region. At $θ=0.52$, $γ$ increases from $0.08$ to $\frac{19}{225}-ε≈0.0844$, a $5.5\%$ improvement. Theorem 2 provides additional regimes (i)–(iii) covering all $\frac12<θ<1$, and Theorem 3 asserts the existence of primes in $[x-x^{0.52-c},x]$ for some $c>0$.
Load-bearing premise
The entire proof assumes the correctness and applicability of the Guth–Maynard and Chen–Gupta–Li large-value estimates for the specific parameter ranges (e.g., k up to $ε^{{-7}}$, ν∈[ε⁶,$ε^{{-1}}$]) and for Dirichlet polynomials that may not be prime-factored; if either estimate fails or cannot be applied in these ranges, the derived functions R_α(t) and thus the advertised γ(θ) would not be guaranteed.
Editorial extensions
If this is right
- The improved $γ(θ)$ directly sharpens the numerical exponent in the author's earlier prime-in-short-intervals result, giving a concrete $c>0$ such that $[x-x^{0.52-c},x]$ always contains a prime (Theorem 3).
- Because the mean-value estimate (1) is a basic ingredient in many sieve arguments, the new ranges for $α₁,α₂$ can be applied to related problems such as Goldbach representations in short intervals and the distribution of almost-primes.
- The method of checking the sign of a piecewise-linear function via $R_α(t)$ provides a template that may be reused whenever future large-value estimates become available, automatically extending the admissible exponent range.
- The improvement is quantitative: at $θ=0.52$ the gain is $5.5\%$; at other $θ<4/7$ the relative gain varies but is always positive, so the theorem strengthens the BHP estimate across its entire range of applicability.
Reading between the lines
- The condition that $R(s)$ be prime-factored is inherited from the BHP framework and is not proved for arbitrary divisor-bounded polynomials; applications that require the estimate for non-prime-factored $R$ will need an alternative argument or a separate case-by-case verification.
- The explicit formulas for $R_α(t)$ depend on the constants $\frac85$ and $\frac25$ that arise from the slope of the Guth–Maynard estimate. If future large-value estimates improve these slopes, the same algebraic procedure would produce smaller $R_α(t)$ and thus larger $γ(θ)$, potentially pushing the admissible $θ$ below $1/2$.
- The paper does not numerically optimise $g₂$ and $g₃$ beyond the stated ranges; a computational search over integer parameters could reveal whether the piecewise-linear expression for $γ(θ)$ can be marginally improved without new analytic input.
Original abstract
Recently, Matom\"{a}ki, Pandey, Pliego, Ter\"{a}v\"{a}inen and Wang improved the Baker--Harman--Pintz estimate of trilinear sums over short intervals of length $x^{\theta}$, which naturally reduces to estimating mean values of products of three Dirichlet polynomials. Based on recent works of Guth and Maynard and Chen, Gupta and Li on new large value estimates for Dirichlet polynomials, they improved the Baker--Harman--Pintz estimate for $\theta > \frac{3}{5}$. In this paper, we use their method to improve the Baker--Harman--Pintz estimate for $\frac{1}{2} < \theta < \frac{4}{7}$ and get new results for all $\frac{1}{2} < \theta < 1$. Our result yields improvements on many problems, including the work of Li on primes in short intervals of length $x^{0.52}$.
The compiled record
The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.
The abstract, sentence by sentence
Recently, Matom\"{a}ki, Pandey, Pliego, Ter\"{a}v\"{a}inen and Wang improved the Baker--Harman--Pintz estimate of trilinear sums over short intervals of length $x^{\theta}$, which naturally reduces to estimating mean values of products of three Dirichlet polynomials. Based on recent works of Guth and Maynard and Chen, Gupta and Li on new large value estimates for Dirichlet polynomials, they improved the Baker--Harman--Pintz estimate for $\theta > \frac{3}{5}$. In this paper, we use their method to improve the Baker--Harman--Pintz estimate for $\frac{1}{2} < \theta < \frac{4}{7}$ and get new results for all $\frac{1}{2} < \theta < 1$. Our result yields improvements on many problems, including the work of Li on primes in short intervals of length $x^{0.52}$.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | Recently, Matom\"{a}ki, Pandey, Pliego, Ter\"{a}v\"{a}inen and Wang improved the Baker--Harman--Pintz estimate of trilinear sums over short intervals of length $x^{\theta}$, which naturally reduces to estimating mean values of products of three Dirichlet polynomials. | background 1.00 | 0.04 | 0.04 | asserted 1.00 |
| s1 | Based on recent works of Guth and Maynard and Chen, Gupta and Li on new large value estimates for Dirichlet polynomials, they improved the Baker--Harman--Pintz estimate for $\theta > \frac{3}{5}$. | background 0.94 | 0.13 | 0.07 | asserted 1.00 |
| s2 | In this paper, we use their method to improve the Baker--Harman--Pintz estimate for $\frac{1}{2} < \theta < \frac{4}{7}$ and get new results for all $\frac{1}{2} < \theta < 1$. | contribution 0.71 | 0.95 | 0.89 | asserted 1.00 |
| s3 | Our result yields improvements on many problems, including the work of Li on primes in short intervals of length $x^{0.52}$. | interpretation 0.83 | 0.81 | 0.15 | asserted 1.00 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
References
Discussion
Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.