惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
Apple Machine Learning Research
Apple Machine Learning Research
aimingoo的专栏
aimingoo的专栏
H
Help Net Security
腾讯CDC
T
Tailwind CSS Blog
Hugging Face - Blog
Hugging Face - Blog
人人都是产品经理
人人都是产品经理
酷 壳 – CoolShell
酷 壳 – CoolShell
MongoDB | Blog
MongoDB | Blog
宝玉的分享
宝玉的分享
有赞技术团队
有赞技术团队
美团技术团队
雷峰网
雷峰网
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
博客园 - 司徒正美
博客园_首页
Recent Announcements
Recent Announcements
云风的 BLOG
云风的 BLOG
B
Blog RSS Feed
钛媒体:引领未来商业与生活新知
钛媒体:引领未来商业与生活新知
D
Docker
博客园 - Franky
Jina AI
Jina AI

math updates on arXiv.org

Coupling-Robust Accuracy in Multiphysics Physics Informed Neural Networks via Kronecker-Preconditioned Optimization Non-normal spectral signatures of instability in neural network training dynamics Optimization of randomized neural networks for transfer operator approximation Selective Ambulance Dispatch Under Contextual Travel-Time Uncertainty LLAMA LIMA: A Living Meta-Analysis on the Effects of Generative AI on Learning Mathematics Neural Flow Operators can Approximate any Operator: Abstract Frameworks and Universal Approximations LLMs as Noisy Channels: A Shannon Perspective on Model Capacity and Scaling Laws On the Stability of Spherical Hellinger-Kantorovich Flows and Their Implications for Differential Privacy Training-Free Looped Transformers Move on Muon : A Hamiltonian probability gradient flow perspective of Muon optimizer Entrywise Error Bounds for Spectral Ranking with Semi-Random Adversaries Asymmetric Scaling Laws from Sparse Features Is Dimensionality a Barrier for Retrieval Models? RA-DCA: A Randomized Active-Set DCA for Directional Stationarity in Max-Structured DC Programs Commutator-Induced Uncertainty in VAEs Weisfeiler-Leman Is Incomplete on Simple Spectrum Graphs, so Canonicalize Them Sparse In-Network Learning via Shortest-Path Backpropagation and Finite-Rate Gating Instance-Optimal Estimation with Multiple LLM Judges on a Budget Entropy Equivalence Testing Expand More, Shrink Less: Shaping Effective-Rank Dynamics for Dense Scaling in Recommendation Any-Dimensional Invariant Universality Operationalizing Individual Fairness via Gradient Descent and Bradley-Terry Models Anytime Training with Schedule-Free Spectral Optimization Diffusion-based Denoising Beats Vanilla Score Matching in Parameter Estimation: A Theoretical Explanation Resilience Characterization of AI-Native Wireless Receivers via Persistent Homology The General Theory of Localization Methods Group-Algebraic Tensors: Provably-optimal Equivariant Learning and Physical Symmetry Discovery General Lower Bounds for Differentially Private Federated Learning with Arbitrary Public-Transcript Interactions PilotWiMAE: Pilot-Native Representation Learning for Wireless Channels Proximal basin hopping: global optimization with guarantees
The Fyodorov--Hiary--Keating Conjecture on Mesoscopic Int...
Louis-Pierre Arguin, Jad Hamdan · 2024-05-10 · via math updates on arXiv.org

We derive precise upper bounds for the maximum of the Riemann zeta function on a typical short interval of the critical line. We show that for fixed $θ\in(-1,0]$, large $T$, and $y\geq 2$ satisfying $y=O(\log\log T/\log\log\log T)$, the proportion of points $t\in [T,2T]$ for which \begin{align*} \max_{|h|\leq \log^θT}\big|ζ(&\tfrac{1}{2}+it+ih)\big|>e^{y} \cdot e^{S\sqrt{(\log\log T)|θ|/2}}\frac{(\log T)^{(1+θ)}}{(\log\log T)^{3/4}} \end{align*} is bounded above by a constant times $y\exp({-2y-y^2/((1+θ)\log\log T)})$, where $S=S(t)$ is a quantity whose value distribution is approximately that of a standard Gaussian. Up to a multiplicative constant, this settles the upper bound of a conjecture of Fyodorov--Hiary--Keating which was only known in the leading order for $θ\in(-1,0)$. Using similar techniques, we also derive upper bounds for the second moment of the zeta function on such intervals. We show that for large $T$, the proportion of $t\in [T,2T]$ for which \begin{align*} \frac{1}{\log^θT}\int_{-\log^θT}^{\log^θT} \big|ζ(&\tfrac{1}{2}+it+ih)\big|^2\mathrm{d}h > A e^{S\sqrt{2|θ|\log\log T}} \frac{(\log T)^{(1+θ)}}{\sqrt{\log\log T}} \end{align*} tends to zero as $A\to\infty$, for the same $S$ as above. This proves a weak form of another conjecture of Fyodorov--Keating and generalizes a result of Harper, which is recovered at $θ= 0$ (in which case $S$ is defined to be zero). Our proofs use an adaptation of the recursive scheme introduced by one of the authors, Bourgade and Radziwiłł.