Source author record

Kewei Zhang

Explore this author’s publications, collaborators and research topics. This catalog record has not been claimed by its author.

ResearcherUnclaimed source record

math.AG math.DG math.CV math.MG Artificial Intelligence Computer Vision math.NA math.OC Numerical Analysis

Catalog footprint

Research overview

10works

9topics

4close collaborators

Actions

Is this your work?

Open graph Browse papers

preprint2026arXiv

MHLA: Restoring Expressivity of Linear Attention via Token-Level Multi-Head

While the Transformer architecture dominates many fields, its quadratic self-attention complexity hinders its use in large-scale applications. Linear attention offers an efficient alternative, but its direct application often degrades performance, with existing fixes typically re-introducing computational overhead through extra modules (e.g., depthwise separable convolution) that defeat the original purpose. In this work, we identify a key failure mode in these methods: global context collapse, where the model loses representational diversity. To address this, we propose Multi-Head Linear Attention (MHLA), which preserves this diversity by computing attention within divided heads along the token dimension. We prove that MHLA maintains linear complexity while recovering much of the expressive power of softmax attention, and verify its effectiveness across multiple domains, achieving a 3.6\% improvement on ImageNet classification, a 6.3\% gain on NLP, a 12.6\% improvement on image generation, and a 41\% enhancement on video generation under the same time complexity.

preprint2026arXiv

Transcendental Okounkov bodies

We show that the volume of transcendental big $(1,1)$ -classes on compact Kähler manifolds can be realized by convex bodies, thus answering questions of Lazarsfeld-Mustaţă and Deng. In our approach we use an approximation process by partial Okounkov bodies together with properties of the restricted volume, and we study the extension of Kähler currents, as well as the bimeromorphic behavior of currents with analytic singularities. We also establish a connection between transcendental Okounkov bodies and toric degenerations.

preprint2024arXiv

A transcendental approach to non-Archimedean metrics of pseudoeffective classes

We introduce the concept of non-Archimedean metrics attached to a transcendental pseudoeffective cohomology class on a compact Kähler manifold. This is obtained via extending the Ross-Witt Nyström correspondence to the relative case, and we point out that our construction agrees with that of Boucksom-Jonsson when the class is induced by a pseudoeffective $\mathbb Q$ -line bundle. We introduce the notion of a flag configuration attached to a transcendental big class, recovering the notion of a test configuration in the ample case. We show that non-Archimedean finite energy metrics are approximable by flag configurations, and very general versions of the radial Ding energy are continuous, a novel result even in the ample case. As applications, we characterize the delta invariant as the Ding semistability threshold of flag configurations and filtrations and prove a YTD type existence theorem in terms of flag configurations.

preprint2024arXiv

Twisted Kähler-Einstein metrics in big classes

We prove existence of twisted Kähler-Einstein metrics in big cohomology classes, using a divisorial stability condition. In particular, when $-K_X$ is big, we obtain a uniform Yau-Tian-Donaldson existence theorem for Kähler-Einstein metrics. To achieve this, we build up from scratch the theory of Fujita-Odaka type delta invariants in the transcendental big setting, using pluripotential theory. We do not use the K-energy in our arguments, and our techniques provide a simple roadmap to prove Yau-Tian-Donaldson existence theorems for Kähler-Einstein type metrics, that only needs convexity of the appropriate Ding energy. As an application, we give a simplified proof of Li-Tian-Wang's existence theorem in the log Fano setting.

preprint2022arXiv

Continuity of delta invariants and twisted Kähler--Einstein metrics

We show that delta invariant is a continuous function on the big cone. We will also introduce an analytic delta invariant and show its continuity in the Kähler cone, from which we deduce the continuity of the greatest Ricci lower bound. Then building on the work Berman-Boucksom-Jonsson, we obtain a uniform Yau-Tian-Donaldson theorem for twisted Kähler-Einstein metrics in general Kähler classes.

preprint2021arXiv

Angle deformation of Kähler-Einstein edge metrics on Hirzebruch surfaces

We construct a family of Kähler-Einstein edge metrics on all Hirzebruch surfaces using the Calabi ansatz and study their angle deformation. This allows us to verify in some special cases a conjecture of Cheltsov-Rubinstein that predicts convergence towards a non-compact Calabi-Yau fibration in the small angle limit. We also give an example of a Kähler-Einstein edge metric whose edge singularity is rigid, answering a question posed by Cheltsov.

preprint2021arXiv

Compensated Convexity on Bounded Domains, Mixed Moreau Envelopes and Computational Methods

Compensated convex transforms have been introduced for extended real-valued functions defined over $\mathbb{R}^n$ . In their application to image processing, interpolation, and shape interrogation, where one deals with functions defined over a bounded domain, one was making the implicit assumption that the function coincides with its transform at the boundary of the data domain. In this paper, we introduce local compensated convex transforms for functions defined in bounded open convex subsets $Ω$ of $\mathbb{R}^n$ by making specific extensions of the function to the whole space, and establish their relations to globally defined compensated convex transforms via the mixed critical Moreau envelopes. We find that the compensated convex transforms of such extensions coincide with the local compensated convex transforms in the closure of $Ω$ . We also propose a numerical scheme for computing Moreau envelopes, establishing convergence of the scheme with the rate of convergence depending on the regularity of the original function. We give an estimate of the number of iterations needed for computing the discrete Moreau envelope. We then apply the local compensated convex transforms to image processing and shape interrogation. Our results are compared with those obtained by using schemes based on computing the convex envelope from the original definition of compensated convex transforms.

preprint2016arXiv

Compensated Convex Transforms and Geometric Singularity Extraction from Semiconvex Functions

We apply upper and lower compensated convex transforms, which are `tight' one-sided approximations of a given function, to the extraction of fine geometric singularities from semiconvex/semiconcave functions and DC-functions in $\mathbb{R}^n$ (difference of convex functions). Well-known examples of (locally) semiconcave functions include the Euclidean distance and squared distance functions. For a locally semiconvex function $f$ with general modulus, we show that `locally' a point is a singular (non-differentiable) point if and only if it is a scale $1$ -valley point, and if $x$ is a singular point, then locally the limit of the scaled valley transform exists at every point $x$ and $\lim_{λ\to \infty}λV_λ(f)(x)=r_x^2/4$ , where $r_x$ is the radius of the minimal bounding sphere of the (Fréchet) subdifferential $\partial_- f(x)$ and $V_λ(f)(x)$ is the valley transform at $x$ . Thus the limit function $\mathcal{V}_\infty(f)(x):=\lim_{λ\to+\infty}λV_λ(f)(x)=r_x^2/4$ gives a `scale $1$ -valley landscape function' of the singular set for a locally semiconvex function $f$ , and also provides an asymptotic expansion of the upper transform $C^u_λ(f)(x)$ when $λ\to \infty$ . For a locally semiconvex function $f$ with linear modulus we show that the limit of the gradient of the upper compensated convex transform $\lim_{λ\to+\infty}\nabla C^u_λ(f)(x)$ exists and equals the centre of the minimal bounding sphere of $\partial_- f(x$ , and that for a DC-function $f=g-h$ , the scale $1$ -edge transform satisfies $\liminf_{λ\to+\infty}λE_λ(f)(x)\geq (r_{g,x}-r_{h,x})^2/4$ , where $r_{g,x}$ and $r_{h,x}$ are the radii of the minimal bounding spheres of the subdifferentials $\partial_- g$ and $\partial_- h$ of the convex functions $g$ and $h$ at $x$ respectively.

preprint2016arXiv

Compensated Convexity Methods for Approximations and Interpolations of Sampled Functions in Euclidean Spaces: Theoretical Foundations

We introduce Lipschitz continuous and $C^{1,1}$ geometric approximation and interpolation methods for sampled bounded uniformly continuous functions over compact sets and over complements of bounded open sets in $\mathbb{R}^n$ by using compensated convex transforms. Error estimates are provided for the approximations of bounded uniformly continuous functions, of Lipschitz functions, and of $C^{1,1}$ functions. We also prove that our approximation methods, which are differentiation and integration free and not sensitive to sample type, are stable with respect to the Hausdorff distance between samples.

preprint2015arXiv

Compensated Convexity, Multiscale Medial Axis Maps and Sharp Regularity of the Squared Distance Function

We introduce a new stable mathematical model for locating and measuring the medial axis of geometric objects, called the quadratic multiscale medial axis map of scale $λ$ , and prove a sharp regularity result for the squared-distance function to any closed non-empty subset $K$ of $\mathbb{R}^n$ . Our results exploit properties of the function $C^l_λ(dist^2(\cdot;\, K))$ obtained by applying the quadratic lower compensated convex transform of parameter $λ$ to $dist^2(\cdot;\, K)$ , the Euclidean squared-distance function to $K$ . Using an estimate for the tight approximation of $dist^2(\cdot;\, K)$ by $C^l_λ(dist^2(\cdot;\, K))$ , we prove $C^{1,1}$ -regularity of $dist^2(\cdot;\, K)$ outside a neighbourhood of the closure of the medial axis $M_K$ of $K$ , and give an asymptotic formula for $C^l_λ(dist^2(\cdot;\, K))$ in terms of the scaled squared distance to $K$ and to the convex hull of the set of points that realize the minimum distance to $K$ . The multiscale medial axis map, $M_λ(\cdot;\, K)$ , is a family of non-negative functions whose limit as $λ\to \infty$ exists and is called the multiscale medial axis landscape map, $M_{\infty}(\cdot;\, K)$ . We show $M_{\infty}(\cdot;\, K)$ is strictly positive on the medial axis $M_K$ and zero elsewhere. We give conditions to ensure $M_λ(\cdot;\, K)$ keeps a constant height along parts of $M_K$ generated by two-point subsets with the height dependent on the distance between the generating points, so giving a hierarchy between different parts of $M_K$ that enables subsets of $M_K$ to be selected by thresholding. Given a compact subset $K$ of $\mathbb{R}^n$ , while it is well known that $M_K$ is not Hausdorff stable, we prove $M_λ(\cdot;\, K)$ is stable under Hausdorff distance, and deduce implications for localization of the stable parts of $M_K$ . Examples are included.

Inspect adjacent papers, topics, institutions and collaborators without losing the researcher page.

Building this map preview

BZPEER is loading the nearby papers, people, topics and institutions for this page.