November 2022 arXiv papers — page 90

Arian Hosseini, Ankit Vani, Dzmitry Bahdanau, Alessandro Sordoni

Pretrained large generative language models have shown great performance on many tasks, but exhibit low compositional generalization abilities. Scaling such models has been shown to improve their performance on various NLP tasks even just by conditioning them on a few examples to solve the task without any fine-tuning (also known as in-context learning). In

Juan Martínez Sykora, Jaime de la Cruz Rodríguez, Milan Gošić, Alberto Sainz Dalda

The heating of the chromosphere in internetwork regions remains one of the foremost open questions in solar physics. In the present study we tackle this old problem by using a very high spatial-resolution simulation of quiet-Sun conditions performed with radiative MHD numerical models and IRIS observations. We have expanded a previously existing 3D radiative

R. A. Main, A. Parthasarathy, S. Johnston, A. Karastergiou

We present the detection of 107 pulsars with interstellar scintillation arcs at 856--1712\,MHz, observed with the MeerKAT Thousand Pulsar Array Programme. Scintillation arcs appear to be ubiquitous in clean, high S/N observations, their detection mainly limited by short observing durations and coarse frequency channel resolution. This led the survey to be se

Johannes Anschütz, Ben Heuer, Arthur-César Le Bras

We describe the category of continuous semilinear representations and their cohomology for the Galois group of a $p$-adic field $K$ with coefficients in a completed algebraic closure via vector bundles on the Hodge-Tate locus of the Cartier-Witt stack. This also gives a new perspective on classical Sen theory; for example it explains the appearance of an ana

Mahmoud Kasem, Abdelrahman Abdallah, Alexander Berendeyev, Ebrahem Elkady

Tables are everywhere, from scientific journals, papers, websites, and newspapers all the way to items we buy at the supermarket. Detecting them is thus of utmost importance to automatically understanding the content of a document. The performance of table detection has substantially increased thanks to the rapid development of deep learning networks. The go

Adam Siegel, Armands Strikis, Thomas Flatters, Simon Benjamin

Whether it is at the fabrication stage or during the course of the quantum computation, e.g. because of high-energy events like cosmic rays, the qubits constituting an error correcting code may be rendered inoperable. Such defects may correspond to individual qubits or to clusters and could potentially disrupt the code sufficiently to generate logical errors

Matthias Hutsebaut-Buysse, Kevin Mets, Tom De Schepper, Steven Latré

Finding an object of a specific class in an unseen environment remains an unsolved navigation problem. Hence, we propose a hierarchical learning-based method for object navigation. The top-level is capable of high-level planning, and building a memory on a floorplan-level (e.g., which room makes the most sense for the agent to visit next, where has the agent

Saurabh Kulshreshtha, Anna Rumshisky

Multi-hop Question Generation is the task of generating questions which require the reader to reason over and combine information spread across multiple passages using several reasoning steps. Chain-of-thought rationale generation has been shown to improve performance on multi-step reasoning tasks and make model predictions more interpretable. However, few-s

Andrea Oldofredi

The Quantum Measurement Problem is arguably one of the most debated issues in the philosophy of Quantum Mechanics, since it represents not only a technical difficulty for the standard formulation of the theory, but also a source of interpretational disputes concerning the meaning of the quantum postulates. Another conundrum intimately connected with the QMP

Sicong Huang, Asli Celikyilmaz, Haoran Li

Abstractive summarization models typically generate content unfaithful to the input, thus highlighting the significance of evaluating the faithfulness of generated summaries. Most faithfulness metrics are only evaluated on news domain, can they be transferred to other summarization tasks? In this work, we first present a systematic study of faithfulness metr

S. Scaringi, M. Monguio, C. Knigge, M. Fratta

We present a sub-arcsecond cross-match of Gaia Data Release 3 (DR3) against the INT Galactic Plane Surveys (IGAPS) and the United Kingdom Infrared Deep Sky Survey (UKIDSS). The resulting cross-match of Galactic Plane Surveys (XGAPS) provides additional precise photometry ($U_{RGO}$, $g$, $r$, $i$, H$\alpha$, $J$, $H$ and $K$) to the Gaia photometry. In build

Seungwhan Moon, Satwik Kottur, Alborz Geramifard, Babak Damavandi

Recent years have seen an increasing trend in the volume of personal media captured by users, thanks to the advent of smartphones and smart glasses, resulting in large media collections. Despite conversation being an intuitive human-computer interface, current efforts focus mostly on single-shot natural language based media retrieval to aid users query their

Silke Husse, Andreas Spitz

The awareness and mitigation of biases are of fundamental importance for the fair and transparent use of contextual language models, yet they crucially depend on the accurate detection of biases as a precursor. Consequently, numerous bias detection methods have been proposed, which vary in their approach, the considered type of bias, and the data used for ev

Laura Nicolás-Sáenz, Agapito Ledezma, Javier Pascau, Arrate Muñoz-Barrutia

In any computer vision task involving color images, a necessary step is classifying pixels according to color and segmenting the respective areas. However, the development of methods able to successfully complete this task has proven challenging, mainly due to the gap between human color perception, linguistic color terms, and digital representation. In this

Marco Eilers, Thibault Dardinier, Peter Müller

Information flow security ensures that the secret data manipulated by a program does not influence its observable output. Proving information flow security is especially challenging for concurrent programs, where operations on secret data may influence the execution time of a thread and, thereby, the interleaving between different threads. Such internal timi

Leo Feng, Hossein Hajimirsadeghi, Yoshua Bengio, Mohamed Osama Ahmed

Neural Processes (NPs) are popular methods in meta-learning that can estimate predictive uncertainty on target datapoints by conditioning on a context dataset. Previous state-of-the-art method Transformer Neural Processes (TNPs) achieve strong performance but require quadratic computation with respect to the number of context datapoints, significantly limiti

Evangelos Vlachos, Aryan Kaushik, Muhammad Z. Shakir

The high path loss associated with millimeter wave (mmWave) frequency communication can be compensated by large scale antenna arrays such as multiple-input multiple-output (MIMO) systems. The hybrid beamforming architecture which uses fewer radio frequency chains is implemented to reduce power consumption and hardware complexity, while still supporting multi

F. E. Cornes, G. A. Frank, C. O. Dorso

This research explores the dynamics of the emergency evacuation during the "Running of the Bulls" festival (Spain, 2013). As people run to escape from danger, many pedestrians stumble and fall down, while others will try to pass over them. We carefully examined three specific recordings of the running, that show this kind of behavior. We developed a microsco

Babhrubahan Bose

We study the relationship between the point-wise symmetry of Birkhoff-James orthogonality and the geometry of the space of operators $\mathbb{B}(\ell_\infty^n,\ell_1^m)$. We show that any non-zero left-symmetric point in this space is a smooth point. We also show that for $n\geq4$, any unit norm right-symmetric point of this space is an extreme point of the

Dany Atallah, Alessandro A. Trani, Kyle Kremer, Newlin C. Weatherford

We present a novel, few-body computational framework designed to shed light on the likelihood of forming intermediate-mass (IM) and supermassive (SM) black holes (BHs) in nuclear star clusters (NSCs) through successive BH mergers, initiated with a single BH seed. Using observationally motivated NSC profiles, we find that the probability of a ${\sim}100 \, M_

Aryan Kaushik, Evangelos Vlachos, Muhammad Z. Shakir, Wonjae Shin

Joint radar-communications (JRC) benefits from multi-functionality of radar and communication operations using same hardware and radio frequency (RF) spectrum resources. Thus JRC systems possess very high potential to be employed into the sixth generation (6G) standards. This paper designs a flexible beamformer for multiple-input multiple output (MIMO) JRC w

Sahil Singla, Soheil Feizi

Training convolutional neural networks (CNNs) with a strict 1-Lipschitz constraint under the $l_{2}$ norm is useful for adversarial robustness, interpretable gradients and stable training. 1-Lipschitz CNNs are usually designed by enforcing each layer to have an orthogonal Jacobian matrix (for all inputs) to prevent the gradients from vanishing during backpro

László Mérai, Igor E. Shparlinski, Arne Winterhof

We estimate mixed character sums of polynomial values over elements of a finite field $\mathbb F_{q^r}$ with sparse representations in a fixed ordered basis over the subfield $\mathbb F_q$. First we use a combination of the inclusion-exclusion principle with bounds on character sums over linear subspaces to get nontrivial bounds for large $q$. Then we focus

Yushi Hu, Hang Hua, Zhengyuan Yang, Weijia Shi

Knowledge-based visual question answering (VQA) involves questions that require world knowledge beyond the image to yield the correct answer. Large language models (LMs) like GPT-3 are particularly helpful for this task because of their strong knowledge retrieval and reasoning capabilities. To enable LM to understand images, prior work uses a captioning mode

Mete Ismayilzada, Antoine Bosselut

In this paper, we present kogito, an open-source tool for generating commonsense inferences about situations described in text. kogito provides an intuitive and extensible interface to interact with natural language generation models that can be used for hypothesizing commonsense knowledge inference from a textual input. In particular, kogito offers several

Thomas Gorordo, Simon Knapen, Benjamin Nachman, Dean J. Robinson

The proposed designs of many auxiliary long-lived particle (LLP) detectors at the LHC call for the instrumentation of a large surface area inside the detector volume, in order to reliably reconstruct tracks and LLP decay vertices. Taking the CODEX-b detector as an example, we provide a proof-of-concept optimization analysis that demonstrates the required ins

Kang Yang, Ipsita Mandal

Exceptional points (EPs) are degeneracy of non-Hermitian Hamiltonians, at which the eigenvalues, along with their eigenvectors, coalesce. Their orders are given by the Jordan decomposition. Here, we focus on higher-order EPs arising in fermionic systems with a sublattice symmetry, which restricts the eigenvalues of the Hamitlonian to appear in pairs of $\lbr

ChunJun Cao, Gong Cheng, Brian Swingle

In this paper, we explore the possibility of building a quantum memory that is robust to thermal noise using large $N$ matrix quantum mechanics models. First, we investigate the gauged $SU(N)$ matrix harmonic oscillator and different ways to encode quantum information in it. By calculating the mutual information between the system and a reference which purif

Aiqi Jiang, Arkaitz Zubiaga

The goal of sexism detection is to mitigate negative online content targeting certain gender groups of people. However, the limited availability of labeled sexism-related datasets makes it problematic to identify online sexism for low-resource languages. In this paper, we address the task of automatic sexism detection in social media for one low-resource lan

Petr Jakubčík, Matteo Marcoli, Giovanni Stagnitto

We investigate the quantum chromodynamics (QCD) corrections to hadronic final states in electron-positron collisions at $\mathcal{O}(\alpha_s^3)$ in the strong coupling constant $\alpha_s$. Namely, we analytically compute the total cross section for this process by separately integrating the tree-level five-parton, the one-loop four-parton, the two-loop thre

Siddhartha Karmakar, Jai More, Akhila Kumar Pradhan, S. Uma Sankar

We study the constraints on the doublet left-right symmetric model (DLRSM) arising due to the Higgs data. The $SU(2)_L$ symmetry of this model is broken by three vacuum expectation values, $\kappa_1$, $\kappa_2$, and $v_L$. Most studies of this model assume that the ratios $r = \kappa_2/\kappa_1$ and $w = v_L/\kappa_1$ are very small. In this work, we study

Siyao Xu, Alex Lazarian

The ubiquitous turbulence in astrophysical plasmas is important for both magnetic reconnection and reconnection acceleration. We study the particle acceleration during fast 3D turbulent reconnection with reconnection-driven turbulence. Particles bounce back and forth between the reconnection-driven inflows due to the mirror reflection and convergence of stro

Bhavesh Chauhan

The strange quark contribution to proton's spin ($\Delta s$) is a fundamental quantity that is poorly determined from current experiments. Neutrino-proton elastic scattering (pES) is a promising channel to measure this quantity, and requires an intense source of low-energy neutrinos and a low-threshold detector with excellent resolution. In this paper, we pr

Josh Borrow, Matthieu Schaller, Yannick M. Bahe, Joop Schaye

All modern galaxy formation models employ stochastic elements in their sub-grid prescriptions to discretise continuous equations across the time domain. In this paper, we investigate how the stochastic nature of these models, notably star formation, black hole accretion, and their associated feedback, that act on small ($<$ kpc) scales, can back-react on mac

Myrann Baker-Rasooli, Wei Liu, Tangui Aladjidi, Alberto Bramati

Turbulence in quantum fluids has, surprisingly, a lot in common with its classical counterpart. Recently, cold atomic gases has emerged as a well controlled experimental platform to study turbulent dynamics. In this work, we introduce a novel system to study quantum turbulence in optics, with the major advantage of having access to a wide range of characteri

S. G. Parsons, M. S. Hernandez, O. Toloza, M. Zorotovic

Virtually all binaries consisting of a white dwarf with a non-degenerate companion can be classified as either close post-interaction systems (with orbital periods of a few days or less), or wide systems (with periods longer than decades), in which both components have effectively evolved as single stars. Binaries with periods between these two extremes can

Chiara Brandenstein, Stefan Stelzl, Erwin Gutsmiedl, Wolfgang Schott

We describe a new table-top electrostatic storage ring concept for $30$ keV polarized ions at frozen spin condition. The device will ultimately be capable of measuring magnetic fields with a resolution of 10$^{-21}$ T with sub-mHz bandwidth. With the possibility to store different kinds of ions or ionic molecules and access to prepare and probe states of the

Jacopo Gliozzi, Mao Lin, Taylor L. Hughes

We study the orbital magnetic quadrupole moment (MQM) in three dimensional higher-order topological phases. Much like electric quadrupole moment, which is associated with a charge response on the boundaries of a finite sample, the diagonal components of the MQM manifest as surface-localized magnetization and hinge currents. The hinge current is generally not

Robert J. J. Grand, Rüdiger Pakmor, Francesca Fragkoudi, Facundo A. Gómez

We utilize a novel numerical technique to model star formation in cosmological simulations of galaxy formation - called Superstars - to simulate a Milky Way-like galaxy with $\gtrsim10^8$ star particles to study the formation and evolution of out-of-equilibrium stellar disc structures in a full cosmological setting. In the plane defined by the coordinate and

Thibault D. Décoppet, Matthew Yu

We generalize the notion of an anomaly for a symmetry to a noninvertible symmetry enacted by surface operators using the framework of condensation in 2-categories. Given a multifusion 2-category, potentially with some additional levels of monoidality, we prove theorems about the structure of the 2-category obtained by condensing a suitable algebra object. We

Run Zhou, Ryan J. Marshman, Sougato Bose, Anupam Mazumdar

Creating a massive spatial quantum superposition, such as the Schr\"odinger cat state, where the mass and the superposition size within the range $10^{-19}-10^{-14}$ kg and $\Delta x \sim 10~{\rm nm}-100~\mu {\rm m}$, is a challenging task. The methods employed so far rely either on wavepacket expansion or on a quantum ancilla, e.g. single spin dependent for

David Villaseñor, Saúl Pilatowsky-Cameo, Miguel A. Bastarrachea-Magnani, Sergio Lerma-Hernández

We present a detailed analysis of the connection between chaos and the onset of thermalization in the spin-boson Dicke model. This system has a well-defined classical limit with two degrees of freedom, and it presents both regular and chaotic regions. Our studies of the eigenstate expectation values and the distributions of the off-diagonal elements of the n

Mudit Jain, Mustafa A. Amin

We provide an algorithm and a publicly available code to numerically evolve multicomponent Schr\"{o}dinger-Poisson (SP) systems with a SO($n$) symmetry, including attractive or repulsive self-interactions in addition to gravity. Focusing on the case where the SP system represents the non-relativistic limit of a massive vector field, non-gravitational self-in

Clara Murgui, Yikun Wang, Kathryn M. Zurek

We propose a novel technique to search for axions with an optomechanical cavity filled with a material such as superfluid helium. Axion absorption converts a pump laser photon to a photon plus a phonon. The axion absorption rate is enhanced by the high occupation number of coherent photons or phonons in the cavity, allowing our proposal to largely overcome t

Thomas C. Gehrman, Barmak Shams Es Haghi, Kuver Sinha, Tao Xu

Gravitational waves (GWs) in the MHz - GHz frequency range are motivated by a host of early Universe phenomena such as oscillons, preheating, and cosmic strings. We point out that baryogenesis too serves as a motivation to probe GWs in this frequency range. The connection is through primordial black holes (PBHs): on the one hand, PBHs induce baryogenesis by

L. Bassini, R. Feldmann, J. Gensior, C. C. Hayward

Recent observations indicate that galactic outflows are ubiquitous in high redshift galaxies, including normal star forming galaxies, quasar hosts, and dusty star forming galaxies (DSFGs). However, the impact of outflows on the evolution of their hosts is still an open question. Here, we analyse the star formation histories (SFH) and galactic outflow propert

Ekdeep Singh Lubana, Eric J. Bigelow, Robert P. Dick, David Krueger

We study neural network loss landscapes through the lens of mode connectivity, the observation that minimizers of neural networks retrieved via training on a dataset are connected via simple paths of low loss. Specifically, we ask the following question: are minimizers that rely on different mechanisms for making their predictions connected via simple paths

Raffaele Mazzilli, Alex Levchenko, Elio J. König

Recent advances in material synthesis made it possible to realize two-dimensional monolayers of candidate materials for a quantum spin liquid (QSL) such as $\alpha$-RuCl$_3$,1T-TaSe$_2$ and 1T-TaS$_2$. In this work, we propose an experimental setup that exploits nonlocal electrical probes to gain information on the transport properties of a gapless QSL. The

Zhongtian Dong, Kyoungchul Kong, Konstantin T. Matchev, Katia Matcheva

We demonstrate the use of symbolic regression in deriving analytical formulas, which are needed at various stages of a typical experimental analysis in collider phenomenology. As a first application, we consider kinematic variables like the stransverse mass, $M_{T2}$, which are defined algorithmically through an optimization procedure and not in terms of an

Jorg Bornschein, Alexandre Galashov, Ross Hemsley, Amal Rannen-Triki

A shared goal of several machine learning communities like continual learning, meta-learning and transfer learning, is to design algorithms and models that efficiently and robustly adapt to unseen tasks. An even more ambitious goal is to build models that never stop adapting, and that become increasingly more efficient through time by suitably transferring t

Sean McGregor

Emerging artificial intelligence (AI) applications often balance the preferences and impacts among diverse and contentious stakeholder groups. Accommodating these stakeholder groups during system design, development, and deployment requires tools for the elicitation of disparate system interests and collaboration interfaces supporting negotiation balancing t

David Ruhe, Kaze Wong, Miles Cranmer, Patrick Forré

We propose parameterizing the population distribution of the gravitational wave population modeling framework (Hierarchical Bayesian Analysis) with a normalizing flow. We first demonstrate the merit of this method on illustrative experiments and then analyze four parameters of the latest LIGO/Virgo data release: primary mass, secondary mass, redshift, and ef

Tarek. M. Elgindi, Ryan. W. Murray, Ayman. R. Said

We study the long-time behavior of scale-invariant solutions of the 2d Euler equation satisfying a discrete symmetry. We show that all scale-invariant solutions with bounded variation on $\mathbb{S}^1$ relax to states that are piece-wise constant with finitely many jumps. All continuous scale-invariant solutions become singular and homogenize in infinite tim

Quentin Chuet, Johanne Cohen, François Pirot

Given a graph $G$, a colouring of $G$ is \emph{acyclic} if it is a proper colouring of $G$ and every cycle contains at least three colours. Its acyclic chromatic number $\chi_a(G)$ is the minimum~$k$ such that an acyclic $k$-colouring of $G$ exists. When $G$ has maximum degree $\Delta$, it is known that $\chi_a(G) = \mathcal {O}(\Delta^{4/3})$ as $\Delta \to

Huihan Liu, Soroush Nasiriany, Lance Zhang, Zhiyao Bao

With the rapid growth of computing powers and recent advances in deep learning, we have witnessed impressive demonstrations of novel robot capabilities in research settings. Nonetheless, these learning systems exhibit brittle generalization and require excessive training data for practical tasks. To harness the capabilities of state-of-the-art robot learning

Naofumi Hama, Masayoshi Mase, Art B. Owen

A model-agnostic variable importance method can be used with arbitrary prediction functions. Here we present some model-free methods that do not require access to the prediction function. This is useful when that function is proprietary and not available, or just extremely expensive. It is also useful when studying residuals from a model. The cohort Shapley

Enrique Tomás Martínez Beltrán, Mario Quiles Pérez, Pedro Miguel Sánchez Sánchez, Sergio López Bernal

In recent years, Federated Learning (FL) has gained relevance in training collaborative models without sharing sensitive data. Since its birth, Centralized FL (CFL) has been the most common approach in the literature, where a central entity creates a global model. However, a centralized approach leads to increased latency due to bottlenecks, heightened vulne

Derek Tam, Anisha Mascarenhas, Shiyue Zhang, Sarah Kwan

While large language models (LLMs) have proven to be effective on a large variety of tasks, they are also known to hallucinate information. To measure whether an LLM prefers factually consistent continuations of its input, we propose a new benchmark called FIB(Factual Inconsistency Benchmark) that focuses on the task of summarization. Specifically, our bench

Nikhil Kandpal, Haikang Deng, Adam Roberts, Eric Wallace

The Internet contains a wealth of knowledge -- from the birthdays of historical figures to tutorials on how to code -- all of which may be learned by language models. However, while certain pieces of information are ubiquitous on the web, others appear extremely rarely. In this paper, we study the relationship between the knowledge memorized by large languag

Joshua Wang

We provide the first complete computations of colored sl(N) homology for a nontrivial knot. In doing so, we show that the colored sl(N) homology of the trefoil labeled by an exterior power of the defining representation is isomorphic to the cohomology of a closed manifold naturally associated to the trefoil. This manifold is the set of homomorphisms from the

Sourya Dey, Eric Davis

We present DLKoopman -- a software package for Koopman theory that uses deep learning to learn an encoding of a nonlinear dynamical system into a linear space, while simultaneously learning the linear dynamics. While several previous efforts have either restricted the ability to learn encodings, or been bespoke efforts designed for specific systems, DLKoopma

Keller Jordan, Hanie Sedghi, Olga Saukh, Rahim Entezari

In this paper we look into the conjecture of Entezari et al. (2021) which states that if the permutation invariance of neural networks is taken into account, then there is likely no loss barrier to the linear interpolation between SGD solutions. First, we observe that neuron alignment methods alone are insufficient to establish low-barrier linear connectivit

Derek Xu, Shuyan Dong, Changhan Wang, Suyoun Kim

Recent studies find existing self-supervised speech encoders contain primarily acoustic rather than semantic information. As a result, pipelined supervised automatic speech recognition (ASR) to large language model (LLM) systems achieve state-of-the-art results on semantic spoken language tasks by utilizing rich semantic representations from the LLM. These s

Yawen Zhang, Michael Hannigan, Qin Lv

Air pollution is a major global environmental health threat, in particular for people who live or work near pollution sources. Areas adjacent to pollution sources often have high ambient pollution concentrations, and those areas are commonly referred to as air pollution hotspots. Detecting and characterizing pollution hotspots are of great importance for air

Shibo Liu, Li-Feng Yin

In this paper, we consider the following quasilinear Schr\"{o}dinger equation \begin{align*} -\Delta u-u\Delta(u^{2})=k(x)\left\vert u\right\vert ^{q-2}u-h(x)\left\vert u\right\vert ^{s-2}u\text{, }u\in D^{1,2}(\mathbb{R}^{N})\text{,} \end{align*} where $1<q<2<s<+\infty$. Unlike most results in the literature, the exponent $s$ here is allowed to be supercrit

Yadi Wei, Roni Khardon

Direct Loss Minimization (DLM) has been proposed as a pseudo-Bayesian method motivated as regularized loss minimization. Compared to variational inference, it replaces the loss term in the evidence lower bound (ELBO) with the predictive log loss, which is the same loss function used in evaluation. A number of theoretical and empirical results in prior work s

Zhaoyi Li, Paolo Glorioso, Joaquin F. Rodriguez-Nieva

We study the far-from-equilibrium dynamics of isolated two-dimensional Heisenberg antiferromagnets. We consider spin spiral initial conditions which imprint a position-dependent staggered-magnetization (or Neel order) in the two-dimensional lattice. Remarkably, we find a long-lived prethermal regime characterized by self-similar behavior of staggered magneti

Emmy Lewis

Given a Noetherian ring $A$, the collection of all integrally closed ideals in $A$ which contain a nonzerodivisor, denoted $ic(A)$, forms a cancellative monoid under the operation $I*J=\overline{IJ}$, the integral closure of the product. The monoid is torsion-free and atomic -- every integrally closed ideal in $A$ containing a nonzerodivisor can be factored

Hartmut Führ, Irina Shafkulovska

We study the mapping properties of metaplectic operators $\widehat{S}\in \mathrm{Mp}(2d,\mathbb{R})$ on modulation spaces of the type $\mathrm{M}^{p,q}_m(\mathbb{R}^d)$. Our main result is a full characterisation of the pairs $(\widehat{S},\mathrm{M}^{p,q}(\mathbb{R}^d))$ for which the operator $\widehat{S}:\mathrm{M}^{p,q}(\mathbb{R}^d) \to \mathrm{M}^{p,q}

Siddharth Chaini, Atharva Bagul, Anish Deshpande, Rishi Gondkar

We present MargNet, a deep learning-based classifier for identifying stars, quasars and compact galaxies using photometric parameters and images from the Sloan Digital Sky Survey (SDSS) Data Release 16 (DR16) catalogue. MargNet consists of a combination of Convolutional Neural Network (CNN) and Artificial Neural Network (ANN) architectures. Using a carefully

Hayate Iso

Lexically constrained text generation is one of the constrained text generation tasks, which aims to generate text that covers all the given constraint lexicons. While the existing approaches tackle this problem using a lexically constrained beam search algorithm or dedicated model using non-autoregressive decoding, there is a trade-off between the generated

Dan Su

In this thesis, we investigated the relevance, faithfulness, and succinctness aspects of Long Form Question Answering (LFQA). LFQA aims to generate an in-depth, paragraph-length answer for a given question, to help bridge the gap between real scenarios and the existing open-domain QA models which can only extract short-span answers. LFQA is quite challenging

Tushar Agarwal, Emre Ertin

We present CardiacGen, a Deep Learning framework for generating synthetic but physiologically plausible cardiac signals like ECG. Based on the physiology of cardiovascular system function, we propose a modular hierarchical generative model and impose explicit regularizing constraints for training each module using multi-objective loss functions. The model co

Yiran Huang, Yexu Zhou, Michael Hefenbrock, Till Riedel

The vulnerability of the high-performance machine learning models implies a security risk in applications with real-world consequences. Research on adversarial attacks is beneficial in guiding the development of machine learning models on the one hand and finding targeted defenses on the other. However, most of the adversarial attacks today leverage the grad

Ezgi Kantarcı Oğuz, Cem Yalım Özel, Mohan Ravichandran

We introduce a class of polytopes that we call chainlink polytopes and which allow us to construct infinite families of pairs of non isomorphic rational polytopes with the same Ehrhart quasi-polynomial. Our construction is based on circular fence posets, which admit a non-obvious and non-trivial symmetry in their rank sequences that turns out to be reflected

Sungjin Im, Benjamin Moseley, Hung Q. Ngo, Kirk Pruhs

We consider a class of optimization problems that involve determining the maximum value that a function in a particular class can attain subject to a collection of difference constraints. We show that a particular linear programming technique, based on duality and projections, can be used to rederive some structural results that were previously established u

Ziniu Hu, Yichong Xu, Wenhao Yu, Shuohang Wang

Answering open-domain questions requires world knowledge about in-context entities. As pre-trained Language Models (LMs) lack the power to store all required knowledge, external knowledge sources, such as knowledge graphs, are often used to augment LMs. In this work, we propose knOwledge REasOning empowered Language Model (OREO-LM), which consists of a novel

Hsin-Hung Chen, Alexander Lerch

The performance of approaches to Music Instrument Classification, a popular task in Music Information Retrieval, is often impacted and limited by the lack of availability of annotated data for training. We propose to address this issue with "reprogramming," a technique that utilizes pre-trained deep and complex neural networks originally targeting a differen

Ali Behrouz, Margo Seltzer

The problem of identifying anomalies in dynamic networks is a fundamental task with a wide range of applications. However, it raises critical challenges due to the complex nature of anomalies, lack of ground truth knowledge, and complex and dynamic interactions in the network. Most existing approaches usually study networks with a single type of connection b

Varinder Singh, Vahid Shaghaghi, Özgür E. Müstecaplıoğlu, Dario Rosa

We investigate the thermodynamic uncertainty relation (TUR), \textit{i.e.} a trade-off between entropy production rate and relative power fluctuations, for nondegenerate three-level and degenerate four-level maser heat engines. In the nondegenerate case, we consider two slightly different configurations of three-level maser heat engine and contrast their deg

Matias Salibian-Barrera

Nonparametric regression models offer a way to understand and quantify relationships between variables without having to identify an appropriate family of possible regression functions. Although many estimation methods for these models have been proposed in the literature, most of them can be highly sensitive to the presence of a small proportion of atypical

Binjie Li, Xiaoping Xie

For the Euler scheme of the stochastic linear evolution equations, discrete stochastic maximal $ L^p $-regularity estimate is established, and a sharp error estimate in the norm $ \|\cdot\|_{L^p((0,T)\times\Omega;L^q(\mathcal O))} $, $ p,q \in [2,\infty) $, is derived via a duality argument.

Zachary Chase, Mayank Pandey

For a given positive integer $n$, how long can the process $x \mapsto n\text{ }(\text{mod } x)$ last before reaching $0$? We improve Erd\H{o}s and Shallit's upper bound of $O(n^{\frac{1}{3}+\varepsilon})$ to $O(n^{\frac{1}{3}-\frac{2}{177}+\varepsilon})$ for any $\varepsilon > 0$.

Joshua Brakensiek, Venkatesan Guruswami, Sai Sandeep

For a constraint satisfaction problem (CSP), a robust satisfaction algorithm is one that outputs an assignment satisfying most of the constraints on instances that are near-satisfiable. It is known that the CSPs that admit efficient robust satisfaction algorithms are precisely those of bounded width, i.e., CSPs whose satisfiability can be checked by a simple

Giulia Gianfagna, Elena Rasia, Weiguang Cui, Marco De Petris

We use a set of about 300 simulated clusters from The Three Hundred Project to calculate their hydrostatic masses and evaluate the associated bias by comparing them with the true cluster mass. Over a redshift range from 0.07 to 1.3, we study the dependence of the hydrostatic bias on redshift, concentration, mass growth, dynamical state, mass, and halo shapes

Daniel Fried, Nicholas Tomlin, Jennifer Hu, Roma Patel

People rely heavily on context to enrich meaning beyond what is literally said, enabling concise but effective communication. To interact successfully and naturally with people, user-facing artificial intelligence systems will require similar skills in pragmatics: relying on various types of context -- from shared linguistic goals and conventions, to the vis

Damián Quijano, Richard Gil-Herrera

We disclose a methodology to determine the participants in discussions and their contributions in social networks with a local relationship (e.g., nationality), providing certain levels of trust and efficiency in the process. The dynamic is a challenge that has demanded studies and some approximations to recent solutions. The study addressed the problem of i

Josip Jukić, Martin Tutek, Jan Šnajder

A popular approach to unveiling the black box of neural NLP models is to leverage saliency methods, which assign scalar importance scores to each input component. A common practice for evaluating whether an interpretability method is faithful has been to use evaluation-by-agreement -- if multiple methods agree on an explanation, its credibility increases. Ho

Zheng-Meng Zhai, Ling-Wei Kong, Ying-Cheng Lai

Can noise be beneficial to machine-learning prediction of chaotic systems? Utilizing reservoir computers as a paradigm, we find that injecting noise to the training data can induce a stochastic resonance with significant benefits to both short-term prediction of the state variables and long-term prediction of the attractor of the system. A key to inducing th

Troy McMahon, Aravind Sivaramakrishnan, Edgar Granados, Kostas E. Bekris

Sampling-based methods are widely adopted solutions for robot motion planning. The methods are straightforward to implement, effective in practice for many robotic systems. It is often possible to prove that they have desirable properties, such as probabilistic completeness and asymptotic optimality. Nevertheless, they still face challenges as the complexity

Rajsuryan Singh, Pablo Zinemanas, Xavier Serra, Juan Pablo Bello

Most recent work in visual sound source localization relies on semantic audio-visual representations learned in a self-supervised manner, and by design excludes temporal information present in videos. While it proves to be effective for widely used benchmark datasets, the method falls short for challenging scenarios like urban traffic. This work introduces t

Miguel G. Silva, Rui Henriques, Sara C. Madeira

Collaborative Filtering (CF), the most common approach to build Recommender Systems, became pervasive in our daily lives as consumers of products and services. However, challenges limit the effectiveness of Collaborative Filtering approaches when dealing with recommendation data, mainly due to the diversity and locality of user preferences, structural sparsi

Chaitanya Chadha, Vandit Gupta, Deepak Gupta, Ashish Khanna

Text classification helps analyse texts for semantic meaning and relevance, by mapping the words against this hierarchy. An analysis of various types of texts is invaluable to understanding both their semantic meaning, as well as their relevance. Text classification is a method of categorising documents. It combines computer text classification and natural l

Rebecca Robinson, Mats Carlsson, Guillaume Aulanier

Magnetic reconnection in the quiet Sun is a phenomenon that is consistently observed, and it has recently become feasible to simulate via 3D numerical models of realistically stratified and convection-driven reconnection. We aim to illustrate ways by which quiet Sun fields may contribute to solar atmospheric heating via magnetic reconnection that is driven b

Sourav Kesharee Sahoo, Radhika Vathsan, Tabish Qureshi

Emergence of classicality from quantum mechanics, a hotly debated topic, has had no satisfactory resolution so far. Various approaches including decoherence and gravitational interactions have been suggested. In the present work, the Schr\"odinger-Newton model is used to study the role of semi-classical self-gravity in the evolution of massive spin-1/2 parti

Christopher Henney-Turner

We introduce two schemes of quantifiers analogous to $I$ and $Q^\text{e.c.}$, which tell us about regular cardinals of small Cantor-Bendixson rank. We examine how the L\"owenheim-Skolem-Tarski numbers of these quantifiers interact with one another, and with those of $I$ and $Q^{\text{e.c.}}$. We then find the exact lower bound for each of the LST numbers, as

Philipp Scharpf, Moritz Schubotz, Andreas Spitz, Andre Greiner-Petter

Since the COVID-19 outbreak, the use of digital learning or education platforms has significantly increased. Teachers now digitally distribute homework and provide exercise questions. In both cases, teachers need to continuously develop novel and individual questions. This process can be very time-consuming and should be facilitated and accelerated both thro

Daniel Menges, Adil Rasheed

This paper proposes a robust disturbance observer framework for maritime autonomous surface vessels considering model and measurement uncertainties. The core contribution lies in a nonlinear disturbance observer, reconstructing the forces on a vessel impacted by the environment. For this purpose, mappings are found leading to synchronized global exponentiall

Alessandro Strumia

We consider generic freeze-in processes for generation of Dark Matter, together with the consequent re-thermalization of the Standard Model fluid. We find that Dark Matter inherits the Standard Model adiabatic inhomogeneities on the cosmological scales probed by current observations, that were super-horizon during freeze-in. Thereby, freeze-in satisfies the

Abdullatif Köksal, Timo Schick, Hinrich Schütze

Few-shot classification has made great strides due to foundation models that, through priming and prompting, are highly effective few-shot learners. However, this approach has high variance both across different sets of few shots (data selection) and across different finetuning runs (run variability). This is problematic not only because it impedes the fair