arXiv subjects

Uma Roy

Publications and source records attributed to Uma Roy.

8 recordsLinked to original sources

arXiv2025

vApps: Verifiable Applications at Internet Scale

Blockchain technology promises a decentralized, trustless, and interoperable infrastructure. However, widespread adoption remains hindered by issues such as limited scalability, high transaction costs, and the complexity of maintaining coherent verification logic across different blockchain layers. This paper introduces Verifiable Applications (vApps), a novel development framework designed to streamline the creation and deployment of verifiable blockchain computing applications. vApps offer a unified Rust-based Domain-Specific Language (DSL) within a comprehensive SDK, featuring modular abstractions for verification, proof generation, and inter-chain connectivity. This eases the developer's burden in securing diverse software components, allowing them to focus on application logic. The DSL also ensures that applications can automatically take advantage of specialized precompiles and hardware acceleration to achieve consistently high performance with minimal developer effort, as demonstrated by benchmark results for zero-knowledge virtual machines (zkVMs). Experiments show that native Rust execution eliminates interpretation overhead, delivering up to an 197x cycle count improvement compared to EVM-based approaches. Precompiled circuits can accelerate the proof by more than 95%, while GPU acceleration increases throughput by up to 30x and recursion compresses the proof size by up to 230x, enabling succinct and efficient verification. The framework also supports seamless integration with the Web2 and Web3 systems, enabling developers to focus solely on their application logic. Through modular architecture, robust security guarantees, and composability, vApps pave the way toward a trust-minimized and verifiable Internet-scale application environment.

cs.CR↗

arXiv2020

Covariance Matrix Estimation under Total Positivity for Portfolio Selection

Selecting the optimal Markowitz porfolio depends on estimating the covariance matrix of the returns of $N$ assets from $T$ periods of historical data. Problematically, $N$ is typically of the same order as $T$, which makes the sample covariance matrix estimator perform poorly, both empirically and theoretically. While various other general purpose covariance matrix estimators have been introduced in the financial economics and statistics literature for dealing with the high dimensionality of this problem, we here propose an estimator that exploits the fact that assets are typically positively dependent. This is achieved by imposing that the joint distribution of returns be multivariate totally positive of order 2 ($\text{MTP}_2$). This constraint on the covariance matrix not only enforces positive dependence among the assets, but also regularizes the covariance matrix, leading to desirable statistical properties such as sparsity. Based on stock-market data spanning over thirty years, we show that estimating the covariance matrix under $\text{MTP}_2$ outperforms previous state-of-the-art methods including shrinkage estimators and factor models.

stat.AP↗

arXiv2020

LAReQA: Language-agnostic answer retrieval from a multilingual pool

We present LAReQA, a challenging new benchmark for language-agnostic answer retrieval from a multilingual candidate pool. Unlike previous cross-lingual tasks, LAReQA tests for "strong" cross-lingual alignment, requiring semantically related cross-language pairs to be closer in representation space than unrelated same-language pairs. Building on multilingual BERT (mBERT), we study different strategies for achieving strong alignment. We find that augmenting training data via machine translation is effective, and improves significantly over using mBERT out-of-the-box. Interestingly, the embedding baseline that performs the best on LAReQA falls short of competing baselines on zero-shot variants of our task that only target "weak" alignment. This finding underscores our claim that languageagnostic retrieval is a substantively new kind of cross-lingual evaluation.

cs.CL↗

arXiv2020

Learning High-dimensional Gaussian Graphical Models under Total Positivity without Adjustment of Tuning Parameters

We consider the problem of estimating an undirected Gaussian graphical model when the underlying distribution is multivariate totally positive of order 2 (MTP2), a strong form of positive dependence. Such distributions are relevant for example for portfolio selection, since assets are usually positively dependent. A large body of methods have been proposed for learning undirected graphical models without the MTP2 constraint. A major limitation of these methods is that their structure recovery guarantees in the high-dimensional setting usually require a particular choice of a tuning parameter, which is unknown a priori in real world applications. We here propose a new method to estimate the underlying undirected graphical model under MTP2 and show that it is provably consistent in structure recovery without adjusting the tuning parameters. This is achieved by a constraint-based estimator that infers the structure of the underlying graphical model by testing the signs of the empirical partial correlation coefficients. We evaluate the performance of our estimator in simulations and on financial data.

stat.ME↗

arXiv2017

Online Multi-Armed Bandit

We introduce a novel variant of the multi-armed bandit problem, in which bandits are streamed one at a time to the player, and at each point, the player can either choose to pull the current bandit or move on to the next bandit. Once a player has moved on from a bandit, they may never visit it again, which is a crucial difference between our problem and classic multi-armed bandit problems. In this online context, we study Bernoulli bandits (bandits with payout Ber($p_i$) for some underlying mean $p_i$) with underlying means drawn i.i.d. from various distributions, including the uniform distribution, and in general, all distributions that have a CDF satisfying certain differentiability conditions near zero. In all cases, we suggest several strategies and investigate their expected performance. Furthermore, we bound the performance of any optimal strategy and show that the strategies we have suggested are indeed optimal up to a constant factor. We also investigate the case where the distribution from which the underlying means are drawn is not known ahead of time. We again, are able to suggest algorithms that are optimal up to a constant factor for this case, given certain mild conditions on the universe of distributions.

cs.AI↗

arXiv2015

An Explicit Description of the B(\infty) Crystal For Generalized Quantum Groups of a Family of Comet Quivers

Tristan Bozec gave a definition of generalized quantum groups that extends the usual definition of quantum groups to finite quivers with loops at vertices, and he introduced a theory of generalized crystals for this new family of Hopf algebras. We explicitly characterize the generalized crystal $\mathcal{B}(\infty)$ associated to a certain family of quivers with multiple loops by providing a complete set of relations among the Kashiwara operators themselves.

math.RT↗

arXiv2015

Diagrammatic Calculus of Coxeter and Braid Groups

We investigate a novel diagrammatic approach to examining strict actions of a Coxeter group or a braid group on a category. This diagrammatic language, which was developed in a series of papers by Elias, Khovanov and Williamson, provides new tools and methods to attack many problems of current interest in representation theory. In our research we considered a particular problem which arises in this context. To a Coxeter group $W$ one can associate a real hyperplane arrangement, and can consider the complement of these hyperplanes in the complexification $Y_W$. The celebrated $K(π,1)$ conjecture states that $Y_W$ should be a classifying space for the pure braid group, and thus a natural quotient ${Y_W}/{W}$ should be a classifying space for the braid group. Salvetti provided a cell complex realization of the quotient, which we refer to as the Salvetti complex. In this paper we investigate a part of the $K(π,1)$ conjecture, which we call the $K(π,1)$ conjecturette, that states that the second homotopy group of the Salvetti complex is trivial. In this paper we present a diagrammatic proof of the $K(π,1)$ conjecturette for a family of braid groups as well as an analogous result for several families of Coxeter groups.

math.GR↗

arXiv2014

A Generalization of Tokuyama's Formula to the Hall-Littlewood Polynomials

A theorem due to Tokuyama expresses Schur polynomials in terms of Gelfand-Tsetlin patterns, providing a deformation of the Weyl character formula and two other classical results, Stanley's formula for the Schur $q$-polynomials and Gelfand's parametrization for the Schur polynomial. We generalize Tokuyama's formula to the Hall-Littlewood polynomials by extending Tokuyama's statistics. Our result, in addition to specializing to Tokuyama's result and the aforementioned classical results, also yields connections to the monomial symmetric function and a new deformation of Stanley's formula.

math.CO↗