arXiv research

The paper studies orbifold splice quotients and log covers of surface pairs.

The study finds effective lower bounds for spectra of random surfaces and bundles.

Study geometric properties of log Calabi-Yau manifolds, focusing on Fano manifolds with smooth or two proportional components.

Existence of Kähler-Einstein metrics on toric varieties proven.

We study how the systole of principal congruence coverings of a Hilbert modular variety grows when the degree of the covering goes to infinity. We prove that given a Hilbert modular variety $M$ of real dimension $2n$, the sequence of principal congruence coverings $M_{I}$ eventually satisfies $$sysπ_{1}(M_{I})\geq \fra…

We show relationships between uniform K-stability and plt blowups of log Fano pairs. We see that it is enough to evaluate certain invariants defined by volume functions for all plt blowups in order to test uniform K-stability of log Fano pairs. We also discuss the uniform K-stability of two log Fano pairs under crepant…

New SGD covering technique yields dimension-independent generalization bounds.

Using stable log maps, we introduce log twisted differentials extending the notion of abelian differentials to the Deligne-Mumford boundary of stable curves. The moduli stack of log twisted differentials provides a compactification of the strata of abelian differentials. The open strata can have up to three connected c…

New algorithm for online portfolio selection with reduced runtime.

Equivalence proven between divisorial stability and quotient log divisorial stability.

The paper improves transformer generalization bounds using rank-dependent covering number bounds.

The logarithmic Riemann surface Sigma_{log} is a classical holomorphic 1-manifold. It lives into R^4 and induces a covering space of C - 0 defined by exp. This paper suggests a geometric construction of it, derived as the limit of a sequence of vector fields extending exp suitably to embeddings of C into R^3, which tur…

This paper examines Bachelier implied volatility at extreme strikes.

Let T(x,r) denote the first hitting time of the disc of radius r centered at x for Brownian motion on the two dimensional torus. We prove that sup_{x} T(x,r)/|log r|^2 --> 2/pi as r --> 0. The same applies to Brownian motion on any smooth, compact connected, two-dimensional, Riemannian manifold with unit area and no bo…

A log symplectic manifold is a Poisson manifold which is generically nondegenerate. We develop two methods for constructing the symplectic groupoids of log symplectic manifolds. The first is a blow-up construction, corresponding to the notion of an elementary modification of a Lie algebroid along a subalgebroid. The se…

The paper explores the shape of filling-systole subspace in surface moduli space and critical points of systole function.

Optimal SGD rates achieved with shuffling, covering non-convex and convex cases.

We prove that on a closed, orientable surface of genus $g$, a set of simple loops with the property that no two are homotopic or intersect in more than $k$ points has cardinality $\lesssim_k g^{k+1} \log g$. The bound matches the size of the largest known construction to within a factor of $\sim_k \log g$. It generaliz…

A translation structure equips a Riemann surface with a singular flat metric. Not much is known about the shape of a random translation surface. We compute an upper bound on the expected value of the covering radius of a translation surface in any stratum H_1(kappa). The covering radius of a translation surface is the …

A new method using mod n covering improves systolic inequalities.

The paper tightens bounds on covering numbers for deep ReLU networks.

The paper uses Tannakian reconstruction to understand hyperbolic log-orbi curves.

The paper proposes a new method for clustering survival data using smoothed log-hazard trajectories.

The paper classifies certain singular projective varieties with specific properties.

Cover's celebrated theorem states that the long run yield of a properly chosen "universal" portfolio is as good as the long run yield of the best retrospectively chosen constant rebalanced portfolio. The "universality" pertains to the fact that this result is model-free, i.e., not dependent on an underlying stochastic …

3-manifold groups' word problem solved in nearly linear time.

Improved DDPMs achieve high log-likelihoods and sample quality with fewer passes.

In this paper we prove that, for any arithmetic hyperbolic $n$-manifold $M$ of the first type, the systole of most of the principal congruence coverings $M_{I}$ satisfy $sys_{1}(M_{I})\geq \frac{8}{n(n+1)}\log(vol(M_{I}))-c,$ where $c$ is a constant independent of $I$. This generalizes previous work of Buser and Sarn…

We provide upper bounds of the expected Wasserstein distance between a probability measure and its empirical version, generalizing recent results for finite dimensional Euclidean spaces and bounded functional spaces. Such a generalization can cover Euclidean spaces with large dimensionality, with the optimal dependence…

We investigate the geometry of $π_1$-injective surfaces in closed hyperbolic 3-manifolds. First we prove that for any $e>0$, if the manifold $M$ has sufficiently large systole $\sys_1(M)$, the genus of any such surface in $M$ is bounded below by $\exp((1/2-e)\sys_1(M))$. Using this result we show, in particular, that f…

New algorithms verify and search causal graphs with minimal interventions.

The paper analyzes log-optimal portfolios in markets with random time events.

Study shows offline RL with partial coverage and weak function classes is possible.

In this paper, a statistical analysis of log-return fluctuations of the IPC, the Mexican Stock Market Index is presented. A sample of daily data covering the period from $04/09/2000-04/09/2010$ was analyzed, and fitted to different distributions. Tests of the goodness of fit were performed in order to quantitatively as…

Improved Langevin algorithms with prior diffusion achieve dimension-independent convergence for non-log-concave distributions.

We provide a differentially private algorithm for hypothesis selection. Given samples from an unknown probability distribution $P$ and a set of $m$ probability distributions $\mathcal{H}$, the goal is to output, in a $\varepsilon$-differentially private manner, a distribution from $\mathcal{H}$ whose total variation di…

Generative Adversarial Networks (GANs) can achieve state-of-the-art sample quality in generative modelling tasks but suffer from the mode collapse problem. Variational Autoencoders (VAE) on the other hand explicitly maximize a reconstruction-based data log-likelihood forcing it to cover all modes, but suffer from poore…

Discrete time analogues of ergodic stochastic differential equations (SDEs) are one of the most popular and flexible tools for sampling high-dimensional probability measures. Non-asymptotic analysis in the $L^2$ Wasserstein distance of sampling algorithms based on Euler discretisations of SDEs has been recently develop…

This paper focuses on numéraire portfolio and log-optimal portfolio (portfolio with finite expected utility that maximizes the expected logarithm utility from terminal wealth), when a market model $(S,\mathbb F)$ -specified by its assets' price $S$ and its flow of information $\mathbb F$- is stopped at a random time $τ…

We discuss the asymptotic lower bound on the inner radius of nodal domains that arise from Laplacian eigenfunctions $φ_λ$ on a closed Riemannian manifold $(M,g)$. First, in the real-analytic case we present an improvement of the currently best known bounds, due to Mangoubi (\cite{Man1}). Furthermore, using recent re…

New method uses higher-order Langevin dynamics for efficient parallel sampling.

We study the problem of reconstructing an unknown matrix M of rank r and dimension d using O(rd poly log d) Pauli measurements. This has applications in quantum state tomography, and is a non-commutative analogue of a well-known problem in compressed sensing: recovering a sparse vector from a few of its Fourier coeffic…

New insights show coverage conditions are crucial for efficient online reinforcement learning.

Study on geodesics on high genus expander surfaces, proving filling and non-simple properties.

Onflow optimizes portfolio allocation with gradient flows, robust to transaction fees.

Study minimax regret in sequential probability assignment with and without side information.

The paper studies positivity properties of cotangent bundles in complex hyperbolic manifolds with cusps.

Unified framework for data-free sampling using Wasserstein gradient flows.