SearcharxivSearch

arXiv subjects

Ali Taheri

Publications and source records attributed to Ali Taheri.

18 recordsLinked to original sources

Nonlinear Diffusion, and Geometric and Functional Inequalities on Smooth Metric Measure spaces

This extended abstract is based on a talk given at the workshop and summer school ``Direct and Inverse Problems with Applications" in Ghent Analysis and PDE Centre in August 2024. It focuses on nonlinear diffusion equations of slow and fast types and their links with some geometric and functional inequalities in the framework of smooth metric measure spaces. The article presents some introduction in a summer school style as well as several new results.

math.AP

The nonlinear porous medium equation for the f-Laplacian: Hamilton-Souplet-Zhang type gradient estimates and implications

This article presents new gradient estimates for positive solutions to the nonlinear porous medium equation (NPME) in the context of smooth metric measure spaces. The diffusion operator here is the f-Laplacian and the gradient estimates of interest are mainly of Hamilton-Souplet-Zhang types. These estimates are established using a variety of methods and techniques and several implications, most notably, to parabolic Liouville-type results and characterisation of ancient solutions are given. The problem is posed in the general framework where the metric and potential evolve with time and the proofs make use of natural lower bounds on the time derivative of the metric and the Bakry-\'Emery m-Ricci curvature tensors. Our results extend and improve various existing ones in the literature.

math.AP

BarrierBench: Evaluating Large Language Models for Safety Verification in Dynamical Systems

Safety verification of dynamical systems via barrier certificates is essential for ensuring correctness in autonomous applications. Synthesizing these certificates involves discovering mathematical functions with current methods suffering from poor scalability, dependence on carefully designed templates, and exhaustive or incremental function-space searches. They also demand substantial manual expertise--selecting templates, solvers, and hyperparameters, and designing sampling strategies--requiring both theoretical and practical knowledge traditionally shared through linguistic reasoning rather than formalized methods. This motivates a key question: can such expert reasoning be captured and operationalized by language models? We address this by introducing an LLM-based agentic framework for barrier certificate synthesis. The framework uses natural language reasoning to propose, refine, and validate candidate certificates, integrating LLM-driven template discovery with SMT-based verification, and supporting barrier-controller co-synthesis to ensure consistency between safety certificates and controllers. To evaluate this capability, we introduce BarrierBench, a benchmark of 100 dynamical systems spanning linear, nonlinear, discrete-time, and continuous-time settings. Our experiments assess not only the effectiveness of LLM-guided barrier synthesis but also the utility of retrieval-augmented generation and agentic coordination strategies in improving its reliability and performance. Across these tasks, the framework achieves more than 90% success in generating valid certificates. By releasing BarrierBench and the accompanying toolchain, we aim to establish a community testbed for advancing the integration of language-based reasoning with formal verification in dynamical systems. The benchmark is publicly available at https://hycodev.com/dataset/barrierbench

cs.AI

Forgetting: A New Mechanism Towards Better Large Language Model Fine-tuning

Supervised fine-tuning (SFT) plays a critical role for pretrained large language models (LLMs), notably enhancing their capacity to acquire domain-specific knowledge while preserving or potentially augmenting their general-purpose capabilities. However, the efficacy of SFT hinges on data quality as well as data volume, otherwise it may result in limited performance gains or even degradation relative to the associated baselines. To mitigate such reliance, we suggest categorizing tokens within each corpus into two parts -- positive and negative tokens -- based on whether they are useful to improve model performance. Positive tokens can be trained in common ways, whereas negative tokens, which may lack essential semantics or be misleading, should be explicitly forgotten. Overall, the token categorization facilitates the model to learn less informative messages, and the forgetting guides the model on what information to learn more precisely. We conduct experiments across diverse and well-established benchmarks using various model architectures, demonstrating that this forgetting mechanism enhances model performance.

cs.LG

Li-Yau Estimates and Harnack Inequalities for Nonlinear Slow Diffusion Equations on a Smooth Metric Measure Space

We present new gradient estimates and Harnack inequalities for positive solutions to nonlinear slow diffusion equations. The framework is that of a smooth metric measure space $(\mathscr M,g,d\mu)$ with invariant weighted measure $d\mu=e^{-\phi} dv_g$ and diffusion operator $\Delta_\phi=e^\phi {\rm div} (e^{-\phi} \nabla)$ -- the $\phi$-Laplacian. The nonlinear slow diffusion equation, then, for $x \in {\mathscr M}$ and $t>0$, and fixed exponent $p>1$, takes the form \begin{equation*} \partial_t u (x,t) - \Delta_\phi u^p (x,t) = \mathscr N (t,x,u(x,t)). \end{equation*} We assume that the metric tensor $g$ and potential $\phi$ are space-time dependent; hence the same is true of the usual metric and potential dependent differential operators and curvature tensors. The estimates are established under natural lower bounds on the Bakry-\'Emery $m$-Ricci curvature tensor and the time derivative of metric tensor. The curious interplay between geometry, nonlinearity and evolution and their influence on the estimates is at the centre of this investigation. The results here considerably extend and improve earlier results on slow diffusion equations. Several implication, special cases and corollaries are presented and discussed.

math.AP

The nonlinear fast diffusion equation on smooth metric measure spaces: Hamilton-Souplet-Zhang estimates and a Ricci-Perelman super flow

This article presents new gradient estimates for positive solutions to the nonlinear fast diffusion equation on smooth metric measure spaces, involving the $f$-Laplacian. The gradient estimates of interest are mainly of Hamilton-Souplet-Zhang or elliptic type and are proved using different set of methods and techniques. Various implications notably to parabolic Liouville type results and characterisation of ancient solutions are given. The problem is considered in the general setting where the metric and potential evolve under a super flow involving the Bakry-\'Emery $m$-Ricci curvature tensor. The remarkable interplay between geometry, nonlinearity, and evolution -- and their intricate roles in the estimates and the maximum exponent range of fast diffusion -- is at the core of the investigation.

math.AP

Curvature conditions, Liouville-type theorems and Harnack inequalities for a nonlinear parabolic equation on smooth metric measure spaces

In this paper we prove gradient estimates of both elliptic and parabolic types, specifically, of Souplet-Zhang, Hamilton and Li-Yau types for positive smooth solutions to a class of nonlinear parabolic equations involving the Witten or drifting Laplacian on smooth metric measure spaces. These estimates are established under various curvature conditions and lower bounds on the generalised Bakry-Émery Ricci tensor and find utility in proving elliptic and parabolic Harnack-type inequalities as well as general elliptic and parabolic Liouville-type and other global constancy results. Several applications and consequences are presented and discussed.

math.AP

Souplet-Zhang and Hamilton type gradient estimates for nonlinear elliptic equations on smooth metric measure spaces

In this article we present new gradient estimates for positive solutions to a class of nonlinear elliptic equations involving the f-Laplacian on a smooth metric measure space. The gradient estimates of interest are of Souplet-Zhang and Hamilton types respectively and are established under natural lower bounds on the generalised Bakry-Émery Ricci curvature tensor. From these estimates we derive amongst other things Harnack inequalities and general global constancy and Liouville-type theorems. The results and approach undertaken here provide a unified treatment and extend and improve various existing results in the literature. Some implications and applications are presented and discussed.

math.AP

Gradient estimates for a nonlinear parabolic equation on smooth metric measure spaces with evolving metrics and potentials

This article presents new parabolic and elliptic type gradient estimates for positive smooth solutions to a nonlinear parabolic equation involving the Witten Laplacian in the context of smooth metric measure spaces. The metric and potential here are time dependent and evolve under a super Perelman-Ricci flow. The estimates are derived under natural lower bounds on the associated generalised Bakry-Émery Ricci curvature tensors and are utilised in establishing fairly general local and global bounds, Harnack-type inequalities and Liouville-type global constancy theorems to mention a few. Other implications and consequences of the results are also discussed.

math.AP

Gradient estimates for nonlinear elliptic equations involving the Witten Laplacian on smooth metric measure spaces and implications

This article presents new local and global gradient estimates of Li-Yau type for positive solutions to a class of nonlinear elliptic equations on smooth metric measure spaces involving the Witten Laplacian. The estimates are derived under natural lower bounds on the associated Bakry-Émery Ricci curvature tensor and find utility in proving general Harnack inequalities and Liouville-type theorems to mention a few. The results here unify, extend and improve various existing results in the literature for special nonlinearities already of huge interest and applications. Some important consequences are presented and discussed.

math.AP

On Multiple Solutions to a Family of Nonlinear Elliptic Systems in Divergence Form Coupled with an Incompressibility Constraint

The aim of this paper is to prove the existence of multiple solutions for a family of nonlinear elliptic systems in divergence form coupled with a pointwise gradient constraint: \begin{align*} \left\{ \begin{array}{ll} \dive\{\A(|x|,|u|^2,|\nabla u|^2) \nabla u\} + \B(|x|,|u|^2,|\nabla u|^2) u = \dive \{ \mcP(x) [{\rm cof}\,\nabla u] \} \quad &\text{ in} \ Ω, \\ \text{det}\, \nabla u = 1 \ &\text{ in} \ Ω, \\ u =φ\ &\text{ on} \ \partial Ω, \end{array} \right. \end{align*} where $Ω\subset \mathbb{R}^n$ ($n \ge 2$) is a bounded domain, $u=(u_1, \dots, u_n)$ is a vector-map and $φ$ is a prescribed boundary condition. Moreover $\mathscr{P}$ is a hydrostatic pressure associated with the constraint $\det \nabla u \equiv 1$ and $\A = \A(|x|,|u|^2,|\nabla u|^2)$, $\B = \B(|x|,|u|^2,|\nabla u|^2)$ are sufficiently regular scalar-valued functions satisfying suitable growths at infinity. The system arises in diverse areas, e.g., in continuum mechanics and nonlinear elasticity, as well as geometric function theory to name a few and a clear understanding of the form and structure of the solutions set is of great significance. The geometric type of solutions constructed here draws upon intimate links with the Lie group ${\bf SO}(n)$, its Lie exponential and the multi-dimensional curl operator acting on certain vector fields. Most notably a discriminant type quantity $Δ=Δ(\A,\B)$, prompting from the PDE, will be shown to have a decisive role on the structure and multiplicity of these solutions.

math.AP

RC-RNN: Reconfigurable Cache Architecture for Storage Systems Using Recurrent Neural Networks

Solid-State Drives (SSDs) have significant performance advantages over traditional Hard Disk Drives (HDDs) such as lower latency and higher throughput. Significantly higher price per capacity and limited lifetime, however, prevents designers to completely substitute HDDs by SSDs in enterprise storage systems. SSD-based caching has recently been suggested for storage systems to benefit from higher performance of SSDs while minimizing the overall cost. While conventional caching algorithms such as Least Recently Used (LRU) provide high hit ratio in processors, due to the highly random behavior of Input/Output (I/O) workloads, they hardly provide the required performance level for storage systems. In addition to poor performance, inefficient algorithms also shorten SSD lifetime with unnecessary cache replacements. Such shortcomings motivate us to benefit from more complex non-linear algorithms to achieve higher cache performance and extend SSD lifetime. In this paper, we propose RC-RNN, the first reconfigurable SSD-based cache architecture for storage systems that utilizes machine learning to identify performance-critical data pages for I/O caching. The proposed architecture uses Recurrent Neural Networks (RNN) to characterize ongoing workloads and optimize itself towards higher cache performance while improving SSD lifetime. RC-RNN attempts to learn characteristics of the running workload to predict its behavior and then uses the collected information to identify performance-critical data pages to fetch into the cache. Experimental results show that RC-RNN characterizes workloads with an accuracy up to 94.6% for SNIA I/O workloads. RC-RNN can perform similarly to the optimal cache algorithm by an accuracy of 95% on average, and outperforms previous SSD caching architectures by providing up to 7x higher hit ratio and decreasing cache replacements by up to 2x.

cs.AR

Machine Learning Interpretability Meets TLS Fingerprinting

Protecting users' privacy over the Internet is of great importance; however, it becomes harder and harder to maintain due to the increasing complexity of network protocols and components. Therefore, investigating and understanding how data is leaked from the information transmission platforms and protocols can lead us to a more secure environment. In this paper, we propose a framework to systematically find the most vulnerable information fields in a network protocol. To this end, focusing on the transport layer security (TLS) protocol, we perform different machine-learning-based fingerprinting attacks on the collected data from more than 70 domains (websites) to understand how and where this information leakage occurs in the TLS protocol. Then, by employing the interpretation techniques developed in the machine learning community and applying our framework, we find the most vulnerable information fields in the TLS protocol. Our findings demonstrate that the TLS handshake (which is mainly unencrypted), the TLS record length appearing in the TLS application data header, and the initialization vector (IV) field are among the most critical leaker parts in this protocol, respectively.

cs.NI

A Hybrid Deep Learning Architecture for Privacy-Preserving Mobile Analytics

Internet of Things (IoT) devices and applications are being deployed in our homes and workplaces. These devices often rely on continuous data collection to feed machine learning models. However, this approach introduces several privacy and efficiency challenges, as the service operator can perform unwanted inferences on the available data. Recently, advances in edge processing have paved the way for more efficient, and private, data processing at the source for simple tasks and lighter models, though they remain a challenge for larger, and more complicated models. In this paper, we present a hybrid approach for breaking down large, complex deep neural networks for cooperative, privacy-preserving analytics. To this end, instead of performing the whole operation on the cloud, we let an IoT device to run the initial layers of the neural network, and then send the output to the cloud to feed the remaining layers and produce the final result. In order to ensure that the user's device contains no extra information except what is necessary for the main task and preventing any secondary inference on the data, we introduce Siamese fine-tuning. We evaluate the privacy benefits of this approach based on the information exposed to the cloud service. We also assess the local inference cost of different layers on a modern handset. Our evaluations show that by using Siamese fine-tuning and at a small processing cost, we can greatly reduce the level of unnecessary, potentially sensitive information in the personal data, and thus achieving the desired trade-off between utility, privacy, and performance.

cs.LG

Deep Private-Feature Extraction

We present and evaluate Deep Private-Feature Extractor (DPFE), a deep model which is trained and evaluated based on information theoretic constraints. Using the selective exchange of information between a user's device and a service provider, DPFE enables the user to prevent certain sensitive information from being shared with a service provider, while allowing them to extract approved information using their model. We introduce and utilize the log-rank privacy, a novel measure to assess the effectiveness of DPFE in removing sensitive information and compare different models based on their accuracy-privacy tradeoff. We then implement and evaluate the performance of DPFE on smartphones to understand its complexity, resource demands, and efficiency tradeoffs. Our results on benchmark image datasets demonstrate that under moderate resource utilization, DPFE can achieve high accuracy for primary tasks while preserving the privacy of sensitive features.

stat.ML

Privacy-Preserving Deep Inference for Rich User Data on The Cloud

Deep neural networks are increasingly being used in a variety of machine learning applications applied to rich user data on the cloud. However, this approach introduces a number of privacy and efficiency challenges, as the cloud operator can perform secondary inferences on the available data. Recently, advances in edge processing have paved the way for more efficient, and private, data processing at the source for simple tasks and lighter models, though they remain a challenge for larger, and more complicated models. In this paper, we present a hybrid approach for breaking down large, complex deep models for cooperative, privacy-preserving analytics. We do this by breaking down the popular deep architectures and fine-tune them in a particular way. We then evaluate the privacy benefits of this approach based on the information exposed to the cloud service. We also asses the local inference cost of different layers on a modern handset for mobile applications. Our evaluations show that by using certain kind of fine-tuning and embedding techniques and at a small processing costs, we can greatly reduce the level of information available to unintended tasks applied to the data feature on the cloud, and hence achieving the desired tradeoff between privacy and performance.

cs.CV

On Weyl's asymptotics and remainder term for the orthogonal and unitary groups

We examine the asymptotics of the spectral counting function of a compact Riemannian manifold by V.G.~Avakumovic \cite{Avakumovic} and L.~Hörmander \cite{Hormander-eigen} and show that for the scale of orthogonal and unitary groups ${\bf SO}(N)$, ${\bf SU}(N)$, ${\bf U}(N)$ and ${\bf Spin}(N)$ it is not sharp. While for negative sectional curvature improvements are possible and known, {\it cf.} e.g., J.J.~Duistermaat $\&$ V.~Guillemin \cite{Duist-Guill}, here, we give sharp and contrasting examples in the positive Ricci curvature case [non-negative for ${\bf U}(N)$]. Furthermore here the improvements are sharp and quantitative relating to the dimension and {\it rank} of the group. We discuss the implications of these results on the closely related problem of closed geodesics and the length spectrum.

math.AP

Twist maps as energy minimisers in homotopy classes: symmetrisation and the coarea formula

Let $\X = \X[a, b] = \{x: a<|x|<b\}\subset \R^n$ with $0<a<b<\infty$ fixed be an open annulus and consider the energy functional, \begin{equation*} {\mathbb F} [u; \X] = \frac{1}{2} \int_\X \frac{|\nabla u|^2}{|u|^2} \, dx, \end{equation*} over the space of admissible incompressible Sobolev maps \begin{equation*} {\mathcal A}_ϕ(\X) = \bigg\{ u \in W^{1,2}(\X, \R^n) : \det \nabla u = 1 \text{ {\it a.e.} in $\X$ and $u|_{\partial \X} = ϕ$} \bigg\}, \end{equation*} where $ϕ$ is the identity map of $\overline \X$. Motivated by the earlier works \cite{TA2, TA3} in this paper we examine the {\it twist} maps as extremisers of ${\mathbb F}$ over ${\mathcal A}_ϕ(\X)$ and investigate their minimality properties by invoking the coarea formula and a symmetrisation argument. In the case $n=2$ where ${\mathcal A}_ϕ(\X)$ is a union of infinitely many disjoint homotopy classes we establish the minimality of these extremising twists in their respective homotopy classes a result that then leads to the latter twists being $L^1$-local minimisers of ${\mathbb F}$ in ${\mathcal A}_ϕ(\X)$. We discuss variants and extensions to higher dimensions as well as to related energy functionals.

math.AP