arXiv research

New findings on normal closure of maps for genus 0.

Neural model improves text normalization for non-English languages.

New test for comparing high-dimensional text data.

Enhanced text-to-speech synthesizes expressive speech from a single example.

Improves text clustering by incorporating sequential features and word embeddings.

FlowGMM uses normalizing flows for semi-supervised learning, showing promising results across various data types.

A new model uses normalizing flows for discrete sequences, improving generation speed.

Proposes SPE for robust speaker verification.

We propose a theoretical framework for thinking about score normalization, which confirms that normalization is not needed under (admittedly fragile) ideal conditions. If, however, these conditions are not met, e.g. under data-set shift between training and runtime, our theory reveals dependencies between scores that c…

In the theory of finite order knot invariants, the universal $\text{sl}_2$ weight system maps the chord diagrams to polynomials in a single variable with integer coefficients. In this paper, we define a family of polynomials that generalize the Kreweras triangle (known to refine the normalized median Genocchi numbers),…

Guided Flows enhance sample quality in conditional image generation and text-to-speech.

This is an expanded version of the lecture notes for a minicourse that I gave at a summer school called "Advanced Course on Geometry and Dynamics of Integrable Systems" at CRM Barcelona, 9--14/September/2013. In this text we study the following aspects of integrable non-Hamiltonian systems: local and semi-local normal …

jLDADMM offers Java tools for topic modeling on texts.

In this paper, we prove that the two well-known natural normalizations of Hamiltonian functions on the symplectic manifold $(M,ω)$ canonically relates the action spectra of different normalized Hamiltonians on {\it arbitrary} symplectic manifolds $(M,ω)$. The natural class of normalized Hamiltonians consists of those w…

Improved language models by incorporating statistical discriminators.

Fine-tuning normalization layers can reconstruct smaller networks.

A new test assesses text similarity between two groups of documents.

This new research explores the effects of various training methods on a Polish to English Statistical Machine Translation system for medical texts. Various elements of the EMEA parallel text corpora from the OPUS project were used as the basis for training of phrase tables and language models and for development, tunin…

Motivated by manifold learning techniques, we give an explicit lower bound for how far a smoothly embedded compact submanifold in ${\mathbb R}^N$ can move in a normal direction and remain an embedding. In addition, given a penalty function $P : \text{Emb}(M,\mathbb{R}^N) \rightarrow \mathbb{R}$ on the space of embeddi…

BBM models short texts using biterms to improve coherence.

APo-VAE generates text in hyperbolic space for better hierarchical representation.

EBMs improve text discrimination by generating negatives from auto-regressive models.

The study examines the structure of certain subgroups of quasi-isometry groups of Euclidean spaces, proving their nontriviality and properties.

We consider maximum solution $g(t)$, $t\in [0, +\infty)$, to the normalized Ricci flow. Among other things, we prove that, if $(M, ω)$ is a smooth compact symplectic 4-manifold such that $b_2^+(M)>1$ and let $g(t),t\in[0,\infty)$, be a solution to (1.3) on $M$ whose Ricci curvature satisfies that $|\text{Ric}(g(t))|\l…

Proposes a novel framework for multi-label text classification.

Non-normal subgroups of certain groups grow homologically exponentially.

SHMM models human mobility from GPS and text data, overcoming text sparsity.

We let $\varphi$ be an ageometric fully irreducible outer automorphism so that its Handel-Mosher axis bundle consists of a single unique axis. We show that the centralizer $Cen(\langle\varphi\rangle)$ of the cyclic subgroup generated by $\varphi$ equals the stabilizer $\text{Stab}(Λ^+_\varphi)$ of the attracting lamina…

The standard state-of-the-art backend for text-independent speaker recognizers that use i-vectors or x-vectors, is Gaussian PLDA (G-PLDA), assisted by a Gaussianization step involving length normalization. G-PLDA can be trained with both generative or discriminative methods. It has long been known that heavy-tailed PLD…

End-to-end clustering without labels using normalized cuts.

The paper constructs bundles and recovers Kirillov character formula.

Let $(M,g)$ be a $n-$dimensional compact Riemannian manifold with boundary. We consider the Yamabe type problem \begin{equation} \left\{ \begin{array}{ll} -Δ_{g}u+au=0 & \text{ on }M \\ \partial_νu+\frac{n-2}{2}bu= u^{{n\over n-2}\pm\varepsilon} & \text{ on }\partial M \end{array}\right. \end{equation} where $a\in C^1(…

Normal solutions found for a specific curvature equation in 4D space.

ASK-NN detects distribution drifts in LLM-generated text.

Critical volatility triggers log-normal to power-law transitions in interconnected systems.

Deep Learning predicts e-commerce activity from Italian enterprise websites.

The study examines the normal growth exponent of submanifolds in negatively curved manifolds.

We propose a family of near-metrics based on local graph diffusion to capture similarity for a wide class of data sets. These quasi-metametrics, as their names suggest, dispense with one or two standard axioms of metric spaces, specifically distinguishability and symmetry, so that similarity between data points of arbi…

New method predicts quasar continuum near Lyman-α with high precision and accuracy.

Discrete flows extend normalizing flows to discrete data, improving various applications.

There has been renewed interest in $\mathbb{S}^3$-bundles over $\mathbb{S}^4$ since K. Grove and W. Ziller constructed metrics on nonnegative curvature on the total spaces of these bundles. In this paper we write down necessary and sufficient conditions for a CW complex to be homotopy equivalent to such a bundle. We al…

Introduces TSI, a variance-based measure for persistence barcodes.

Extends graph encoder embedding to weighted graphs and matrices.

New RL algorithm for linear MDPs with nearly optimal regret.

TFiLM expands convolutional models' receptive field with minimal overhead.

We point out important problems with the common practice of using the best single model performance for comparing deep learning architectures, and we propose a method that corrects these flaws. Each time a model is trained, one gets a different result due to random factors in the training process, which include random …

Constructs currents representing Baum-Bott residues for foliations.

New method proves asymptotic normality for matrix sensing problems.