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

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
WordPress大学
WordPress大学
T
Tailwind CSS Blog
V
Visual Studio Blog
月光博客
月光博客
Hugging Face - Blog
Hugging Face - Blog
小众软件
小众软件
奇客Solidot–传递最新科技情报
奇客Solidot–传递最新科技情报
博客园 - Franky
freeCodeCamp Programming Tutorials: Python, JavaScript, Git & More
Last Week in AI
Last Week in AI
阮一峰的网络日志
阮一峰的网络日志
量子位
有赞技术团队
有赞技术团队
酷 壳 – CoolShell
酷 壳 – CoolShell
Apple Machine Learning Research
Apple Machine Learning Research
博客园_首页
Jina AI
Jina AI
雷峰网
雷峰网
博客园 - 【当耐特】
博客园 - 叶小钗
美团技术团队
宝玉的分享
宝玉的分享
IT之家
IT之家

math.PR updates on arXiv.org

Visibility in the Boolean Model on Harmonic Manifolds Global estimates on the Brenier map Geodesics and Wandering Exponents in Brochette First-Passage Percolation State-dependent inverse-subordinator time changes of regenerative processes: Excursion structure and multiscale occupation-time limits Randomly twisted transfer operators and singular values statistics Generalized Bessel-Dunkl diffusions An almost sure invariance principle for the Takagi-van der Waerden class functions Central limit theorems for high dimensional lattice polytopes: cosmological polytopes Convergence rate estimates for semigroups and heat kernels associated with resistance forms Second-order Poincaré inequalities and localization on the Poisson space Maximum Probability of Independence in Transitive Matroids On global solutions to the semidiscrete stochastic heat equation The Poisson Tail Conjecture for primes in short intervals A Complete Spectral Analysis of the CEV Operator with Applications to Arbitrage Holographic functions and neural networks From Betting to Empirical Bernstein LIL Concentration of General Stochastic Approximation Under Heavy-Tailed Markovian Noise Pointwise Generalization in Deep Neural Networks Bayesian Latent Space Models for Graphs Are Misspecified: Toward Robust Inference via Generalized Posteriors Wasserstein bounds for denoising diffusion probabilistic models via the Föllmer process A note on connections between the Föllmer process and the denoising diffusion probabilistic model Simple Approximation and Derivative Free Inference-Time Scaling for Diffusion Models via Sequential Monte Carlo on Path Measures Diffusion-Based Stochastic Operator Networks for Uncertainty Quantification in Stochastic Partial Differential Equations A Fourier perspective on the learning dynamics of neural networks: from sample complexities to mechanistic insights Propagation of Chaos in Contextual Flow Maps Dimension-Uniform Discretization Analysis of Preconditioned Annealed Langevin Dynamics for Multimodal Gaussian Mixtures $α$-TCAV: A Unified Framework for Testing with Concept Activation Vectors Scaling Laws from Sequential Feature Recovery: A Solvable Hierarchical Model On the Limits of Latent Reuse in Diffusion Models State-of-art minibatches via novel DPP kernels: discretization, wavelets, and rough objectives
A $Γ$-convergence of level-two large deviation for metast...
Kyuhyeon Choi · 2024-05-17 · via math.PR updates on arXiv.org

This study explores the relationship between the precise asymptotics of the level-two large deviation rate function and the behavior of metastable stochastic systems. Initially identified for overdamped Langevin dynamics (Ges{ù} et al., SIAM J Math Anal 49(4), 3048-3072, 2017), this connection has been validated across various models, including random walks in a potential field. We extend this connection to condensing zero-range processes, a complex interacting particle system. Specifically, we investigate a certain class of zero-range processes on a fixed graph $G$ with $N > 0$ particles and interaction parameter $α> 1$. On the time scale $N^2$, this process behaves like an absorbing-type diffusion and converges to a condensed state where all particles occupy a single vertex of $G$ as $N$ approaches infinity. Once condensed, on the time scale $N^{1+α}$, the condensed site moves according to a Markov chain on $G$, showing metastable behavior among condensed states. The time scales $N^2$ and $N^{1+α}$ are called the pre-metastable and metastable time scales. It is conjectured that this behavior is encapsulated in the level-two large deviation rate function $\mathcal{I}_N$ of the zero-range process. Specifically, it is expected that the $Γ$-expansion of $\mathcal{I}_N$ can be expressed as:$$\mathcal{I}_N = \frac{1}{N^2} \mathcal{K} + \frac{1}{N^{1+α}} \mathcal{J},$$ where $\mathcal{K}$ and $\mathcal{J}$ are the level-two large deviation rate functions of the absorbing diffusion processes and the Markov chain on $G$. We rigorously prove this $Γ$-expansion by developing a methodology for $Γ$-convergence in the pre-metastable time scale and establishing a link between the resolvent approach to metastability (Landim et al., J Eur Math Soc, 2023. arXiv:2102.00998) and the $Γ$-expansion in the metastable time scale.