arXiv papers from May 2023

Reuben R. W. Wang, John L. Bohn

We consider collective motion and damping of dipolar Fermi gases in the hydrodynamic regime. We investigate the trajectories of collective oscillations -- here dubbed ``weltering'' motions -- in cross-dimensional rethermalization experiments via Monte Carlo simulations, where we find stark differences from the dilute regime. These observations are interprete

Robert J. Moss, Anthony Corso, Jef Caers, Mykel J. Kochenderfer

Real-world planning problems, including autonomous driving and sustainable energy applications like carbon storage and resource exploration, have recently been modeled as partially observable Markov decision processes (POMDPs) and solved using approximate methods. To solve high-dimensional POMDPs in practice, state-of-the-art methods use online planning with

Xue Xia, Pong Eksombatchai, Nikil Pancha, Dhruvil Deven Badani

Sequential models that encode user activity for next action prediction have become a popular design choice for building web-scale personalized recommendation systems. Traditional methods of sequential recommendation either utilize end-to-end learning on realtime user actions, or learn user representations separately in an offline batch-generated manner. This

Peter T. J. Bradshaw

In a recent paper, algebraic descriptions for all non-relativistic spins were derived by elementary means directly from the Lie algebra $\specialorthogonalliealgebra{3}$, and a connection between spin and the geometry of Euclidean three-space was drawn. However, the details of this relationship and the extent to which it can be developed by elementary means

Cohen Archbold, Benjamin Brodie, Aram Ansary Ogholbake, Nathan Jacobs

The monetary value of a given piece of real estate, a parcel, is often readily available from a geographic information system. However, for many applications, such as insurance and urban planning, it is useful to have estimates of property value at much higher spatial resolutions. We propose a method to estimate the distribution over property value at the pi

Peter Shaw, Mandar Joshi, James Cohan, Jonathan Berant

Much of the previous work towards digital agents for graphical user interfaces (GUIs) has relied on text-based representations (derived from HTML or other structured data sources), which are not always readily available. These input representations have been often coupled with custom, task-specific action spaces. This paper focuses on creating agents that in

Hugo Prod'homme, Philipp del Hougne

Physics-compliant channel models of RIS-parametrized radio environments require the inversion of an "interaction matrix" to capture the mutual coupling between wireless entities (transmitters, receivers, RIS, environmental scattering objects) due to proximity and reverberation. The computational cost of this matrix inversion is typically dictated by the envi

Joshua Cooper, Gabrielle Tauscheck

Graham and Pollak showed that the determinant of the distance matrix of a tree $T$ depends only on the number of vertices of $T$. Graphical distance, a function of pairs of vertices, can be generalized to ``Steiner distance'' of sets $S$ of vertices of arbitrary size, by defining it to be the fewest edges in any connected subgraph containing all of $S$. Here

Taehyun Hwang, Kyuwook Chai, Min-hwan Oh

We consider a contextual combinatorial bandit problem where in each round a learning agent selects a subset of arms and receives feedback on the selected arms according to their scores. The score of an arm is an unknown function of the arm's feature. Approximating this unknown score function with deep neural networks, we propose algorithms: Combinatorial Neu

Kai Katsumata, Duc Minh Vo, Bei Liu, Hideki Nakayama

The exploration of the latent space in StyleGANs and GAN inversion exemplify impressive real-world image editing, yet the trade-off between reconstruction quality and editing quality remains an open problem. In this study, we revisit StyleGANs' hyperspherical prior $\mathcal{Z}$ and $\mathcal{Z}^+$ and integrate them into seminal GAN inversion methods to imp

Nasif Imtiaz, Preya Shabrina, Laurie Williams

As modern software extensively uses open source packages, developers regularly pull in new upstream code through frequent updates. While a manual review of all upstream changes may not be practical, developers may rely on the authors' and reviewers' identities, among other factors, to decide what level of review the new code may require. The goal of this stu

Yoshiyuki Ohmura, Wataru Shimaya, Yasuo Kuniyoshi

The mind-brain problem is to bridge relations between in higher-level mental events and in lower-level neural events. To address this, some mathematical models have been proposed to explain how the brain can represent the discriminative structure of qualia, but they remain unresolved due to a lack of validation methods. To understand the qualia discriminatio

Maxwell Horton, Sachin Mehta, Ali Farhadi, Mohammad Rastegari

Modern deep learning approaches usually utilize modality-specific processing. For example, the most common deep learning approach to image classification involves decoding image file bytes into an RGB tensor which is passed into a neural network. Instead, we investigate modality-independent representation learning by performing classification directly on fil

Riccardo Muolo, Joseph D. O'Brien, Timoteo Carletti, Malbor Asllani

The emergence of order in nature manifests in different phenomena, with synchronization being one of the most representative examples. Understanding the role played by the interactions between the constituting parts of a complex system in synchronization has become a pivotal research question bridging network science and dynamical systems. Particular attenti

Yuzhu Chen, David Saintillan, Padmini Rangamani

The initiation of directional cell motion requires symmetry breaking that can happen both with or without external stimuli. During cell crawling, forces generated by the cytoskeleton and their transmission through mechanosensitive adhesions to the extracellular substrate play a crucial role. In a recently proposed 1D model (Sens, PNAS 2020), a mechanical fee

Christopher C. Green, Mohamed M. S. Nasser

This paper is concerned with the numerical computation of the harmonic-measure distribution function, or $h$-function for short, associated with a particular planar domain. This function describes the hitting probability of a Brownian walker released from some point with the boundary of the domain. We use a fast and accurate boundary integral method for the

Shampa Banik, Trapa Banik, S. M. Mostaq Hossain, Sohag Kumar Saha

The smart-grid introduces several new data-gathering, communication, and information-sharing capabilities into the electrical system, as well as additional privacy threats, vulnerabilities, and cyber-attacks. In this study, Modbus is regarded as one of the most prevalent interfaces for control systems in power plants. Modern control interfaces are vulnerable

Anurag Bhattacharyya, Jin-Young Kim, Lee R. Alacoque, Kai A. James

Traditional robotic mechanisms contain a series of rigid links connected by rotational joints that provide powered motion, all of which is controlled by a central processor. By contrast, analogous mechanisms found in nature, such as octopus tentacles, contain sensors, actuators, and even neurons distributed throughout the appendage, thereby allowing for moti

Aditya Kumar

In this paper we study the singular limit for critical points of boundary reactions \begin{equation*} (-\Delta)^{\frac{1}{2}}u = \frac{1}{\varepsilon}(u-u^3) \quad \text{in } U \subset \textbf{R}^n . \end{equation*} We show the existence of a $(n-1)$-rectifiable energy concentration set. Furthermore, we show that the limit of the energy measures can be assoc

Andre Brasil Vieira Wyzykowski, Anil K. Jain

Forensic science heavily relies on analyzing latent fingerprints, which are crucial for criminal investigations. However, various challenges, such as background noise, overlapping prints, and contamination, make the identification process difficult. Moreover, limited access to real crime scene and laboratory-generated databases hinders the development of eff

Pi-Yueh Chuang, Lorena A. Barba

The recent surge of interest in physics-informed neural network (PINN) methods has led to a wave of studies that attest to their potential for solving partial differential equations (PDEs) and predicting the dynamics of physical systems. However, the predictive limitations of PINNs have not been thoroughly investigated. We look at the flow around a 2D cylind

Zhengyang Liu, Stefan Mada, John Regehr

A superoptimizing compiler--one that performs a meaningful search of the program space as part of the optimization process--can find optimization opportunities that are missed by even the best existing optimizing compilers. We created Minotaur: a superoptimizer for LLVM that uses program synthesis to improve its code generation, focusing on integer and float

Jiarui Zhang, Mahyar Khayatkhoei, Prateek Chhikara, Filip Ilievski

Visual Question Answering is a challenging task, as it requires seamless interaction between perceptual, linguistic, and background knowledge systems. While the recent progress of visual and natural language models like BLIP has led to improved performance on this task, we lack understanding of the ability of such models to perform on different kinds of ques

Skyler Wang, Ned Cooper, Margaret Eby

Large language models (LLMs) and dialogue agents represent a significant shift in artificial intelligence (AI) research, particularly with the recent release of the GPT family of models. ChatGPT's generative capabilities and versatility across technical and creative domains led to its widespread adoption, marking a departure from more limited deployments of

Kelly Blincoe, Markus Luczak-Roesch, Tim Miller, Matthias Galster

This article summarizes the literature on trust of digital technologies from a human-centric perspective. We summarize literature on trust in face-to-face interactions from other fields, followed by a discussion of organizational trust, technology-mediated trust, trust of software products, trust of AI, and blockchain. This report was created for the Science

Johnatan Costa, Rafael Diógenes, Neilha Pinheiro, Ernani Ribeiro

In this article, we investigate the geometry of static perfect fluid space-time on compact manifolds with boundary. We use the generalized Reilly's formula to establish a geometric inequality for a static perfect fluid space-time involving the area of the boundary and its volume. Moreover, we obtain new boundary estimates for static perfect fluid space-time.

Aaveg Aggarwal, Eleftherios Kirkinis, Monica Olvera de la Cruz

A droplet of a classical liquid surrounded by a cold gas placed on a hot substrate is accompanied by unremitting internal circulations, while the droplet remains immobile. Two identical cells with opposite sense of circulation form in the interior due to the thermocapillary effect induced by the gas and substrate temperature difference. Under the same condit

Mustafa Ridvan Cantas, Levent Guvenc

Increasing the implemented SAE level of autonomy in road vehicles requires extensive simulations and verifications in a realistic simulation environment before proving ground and public road testing. The level of detail in the simulation environment helps ensure the safety of a real-world implementation and reduces algorithm development cost by allowing deve

Kaushal Kumar

Nonlinear systems play a significant role in numerous scientific and engineering disciplines, and comprehending their behavior is crucial for the development of effective control and prediction strategies. This paper introduces a novel data-driven approach for accurately modeling and estimating parameters of nonlinear systems utilizing trust region optimizat

Jakob Bossek, Christian Grimme

We contribute to the efficient approximation of the Pareto-set for the classical $\mathcal{NP}$-hard multi-objective minimum spanning tree problem (moMST) adopting evolutionary computation. More precisely, by building upon preliminary work, we analyse the neighborhood structure of Pareto-optimal spanning trees and design several highly biased sub-graph-based

Dmitry Manning-Coe, Barry Bradlyn

Recently, models with long-range interactions -- known as Hatsugai-Kohmoto (HK) models -- have emerged as a promising tool to study the emergence of superconductivity and topology in strongly correlated systems. Two obstacles, however, have made it difficult to understand the applicability of these models, especially to topological features: they have thermo

Alexander Clow

In this paper we consider the cop number of graphs with no, or few, short cycles. We show that when $G$ is graph of girth $g$ and the minimum degree $\delta \geq 2$, then $c(G) = O(n\log(n)(\delta-1)^{-\lfloor \frac{g+1}{4} \rfloor})$ as a function of $n$. This extends work of Frankl and implies that if $G$ is large and dense in the sense that $\delta \geq n

Peyman Gholami, Robert Xiao

Text-to-image generative models have made remarkable advancements in generating high-quality images. However, generated images often contain undesirable artifacts or other errors due to model limitations. Existing techniques to fine-tune generated images are time-consuming (manual editing), produce poorly-integrated results (inpainting), or result in unexpec

Daniel P. Arnold, Aakanksha Gubbala, Sho C. Takatori

Phase separation of multicomponent lipid membranes is characterized by the nucleation and coarsening of circular membrane domains that grow slowly in time as $\sim t^{1/3}$, following classical theories of coalescence and Ostwald ripening. In this work, we study the coarsening kinetics of phase-separating lipid membranes subjected to nonequilibrium forces an

Patrick Lee, Iyanuoluwa Shode, Alain Chirino Trujillo, Yuan Zhao

Transformers have been shown to work well for the task of English euphemism disambiguation, in which a potentially euphemistic term (PET) is classified as euphemistic or non-euphemistic in a particular context. In this study, we expand on the task in two ways. First, we annotate PETs for vagueness, a linguistic property associated with euphemisms, and find t

Dimitri Jordan Kenne

Let $K \subset \mathbb{C}^n$ be a compact set satisfying the following Bernstein inequality: for any $m \in \{ 1,..., n\}$ and for any $n$-variate polynomial $P$ of degree $\mbox{deg}(P)$ we have \begin{align*} \max_{z\in K}\left|\frac{\partial P}{\partial z_m}(z)\right| \le M\ \mbox{deg}(P) \max_{z\in K}|P(z)| \ \mbox{ for } z = (z_1, \dots, z_n). \end{alig

S. Arthamonov, Sh. Shakirov

We construct an algebra that is an elliptic generalization of $A_1$ spherical DAHA acting on its finite-dimensional module at $t=-q^{-K/2}$ with $K=2$. We prove that $PSL(2,\mathbb Z)$ acts by automorphisms of the algebra we constructed, and provide an explicit representation of automorphisms and algebra operators alike by $3\times 3$ matrices of elliptic fu

Deion Elzie, Samir Fridhi, Blake Mellor, Daniel Silva

The {\em topological symmetry group} of an embedding $\Gamma$ of an abstract graph $\gamma$ in $S^3$ is the group of automorphisms of $\gamma$ which can be realized by homeomorphisms of the pair $(S^3, \Gamma)$. These groups are motivated by questions about the symmetries of molecules in space. The Petersen family of graphs is an important family of graphs f

Valentina Marin, Carlos Molinero, Elsa Arcaute

Urban systems are primarily relational. The uneven intensities and distribution of flows between systems of cities results in hierarchically organised complex networks of urban exchange. Distinct urban spatial structures reflect the diversity of functional and social patterns which vary or remain constant across multiple scales. In this work, we examine the

Dongsheng Ding, Xiaohan Wei, Zhuoran Yang, Zhaoran Wang

We examine online safe multi-agent reinforcement learning using constrained Markov games in which agents compete by maximizing their expected total rewards under a constraint on expected total utilities. Our focus is confined to an episodic two-player zero-sum constrained Markov game with independent transition functions that are unknown to agents, adversari

Wenqian Chen, Yucheng Fu, Panos Stinis

In this paper, we present a physics-informed neural network (PINN) approach for predicting the performance of an all-vanadium redox flow battery, with its physics constraints enforced by a two-dimensional (2D) mathematical model. The 2D model, which includes 6 governing equations and 24 boundary conditions, provides a detailed representation of the electroch

Guillermo F. Quispe Peña, Andrei V. Frolov

To create high-fidelity cosmic microwave background maps, current component separation methods rely on availability of information on different foreground components, usually through multi-band frequency coverage of the instrument. Internal linear combination (ILC) methods provide an unbiased estimators for CMB which are easy to implement, but component sepa

Ali TehraniJamsaz, Quazi Ishtiaque Mahmud, Le Chen, Nesreen K. Ahmed

The remarkable growth and significant success of machine learning have expanded its applications into programming languages and program analysis. However, a key challenge in adopting the latest machine learning methods is the representation of programming languages, which directly impacts the ability of machine learning methods to reason about programs. The

Ali Barki, Sergey G. Bobkov, Esther Bou Dagher, Cyril Roberto

We revisit several results on exponential integrability in probability spaces and derive some new ones. In particular, we give a quantitative form of recent results by Cianchi-Musil and Pick in the framework of Moser-Trudinger-type inequalities, and recover Ivanisvili-Russell's inequality for the Gaussian measure. One key ingredient is the use of a dual argu

Santosh Kesiraju, Marek Sarvas, Tomas Pavlicek, Cecile Macaire

This paper presents techniques and findings for improving the performance of low-resource speech to text translation (ST). We conducted experiments on both simulated and real-low resource setups, on language pairs English - Portuguese, and Tamasheq - French respectively. Using the encoder-decoder framework for ST, our results show that a multilingual automat

Carolina Araujo, Alessio Corti, Alex Massarenti

We develop a framework that allows one to describe the birational geometry of Calabi-Yau pairs $(X,D)$. After establishing some general results for Calabi-Yau pairs $(X,D)$ with mild singularities, we focus on the special case when $X=\mathbb{P}^3$ and $D\subset \mathbb{P}^3$ is a quartic surface. We investigate how the appearance of increasingly worse singu

Young-Jin Park, Hao Wang, Shervin Ardeshir, Navid Azizan

Self-supervised learning models extract general-purpose representations from data. Quantifying the reliability of these representations is crucial, as many downstream models rely on them as input for their own tasks. To this end, we introduce a formal definition of representation reliability: the representation for a given test point is considered to be reli

Jin Wang, Ming Ma, Erio Tosatti

The ultra-low kinetic friction F_k of 2D structurally superlubric interfaces, connected with the fast motion of the incommensurate moir\'e pattern, is often invoked for its linear increase with velocity v_0 and area A, but never seriously addressed and calculated so far. Here we do that, exemplifying with a twisted graphene layer sliding on top of bulk graph

Yan Pan, Yuanzhi Li

While stochastic gradient descent (SGD) is still the most popular optimization algorithm in deep learning, adaptive algorithms such as Adam have established empirical advantages over SGD in some deep learning applications such as training transformers. However, it remains a question that why Adam converges significantly faster than SGD in these scenarios. In

Yashish M. Siriwardena, Carol Espy-Wilson, Suzanne Boyce, Mark K. Tiede

The velopharyngeal (VP) valve regulates the opening between the nasal and oral cavities. This valve opens and closes through a coordinated motion of the velum and pharyngeal walls. Nasalance is an objective measure derived from the oral and nasal acoustic signals that correlate with nasality. In this work, we evaluate the degree to which the nasalance measur

Adrian Shuai Li, Elisa Bertino, Rih-Teng Wu, Ting-Yan Wu

Deep learning (DL) techniques are highly effective for defect detection from images. Training DL classification models, however, requires vast amounts of labeled data which is often expensive to collect. In many cases, not only the available training data is limited but may also imbalanced. In this paper, we propose a novel domain adaptation (DA) approach to

Keyi Chen, Francesco Orabona

We propose a new class of online learning algorithms, generalized implicit Follow-The-Regularized-Leader (FTRL), that expands the scope of FTRL framework. Generalized implicit FTRL can recover known algorithms, as FTRL with linearized losses and implicit FTRL, and it allows the design of new update rules, as extensions of aProx and Mirror-Prox to FTRL. Our t

Jiashun Wang, Xueting Li, Sifei Liu, Shalini De Mello

Transferring the pose of a reference avatar to stylized 3D characters of various shapes is a fundamental task in computer graphics. Existing methods either require the stylized characters to be rigged, or they use the stylized character in the desired pose as ground truth at training. We present a zero-shot approach that requires only the widely available de

Matthias Christandl, Bergfinnur Durhuus, Lasse Harboe Wolff

Relations among von Neumann entropies of different parts of an $N$-partite quantum system have direct impact on our understanding of diverse situations ranging from spin systems to quantum coding theory and black holes. Best formulated in terms of the set $\Sigma^*_N$ of possible vectors comprising the entropies of the whole and its parts, the famous strong

Carolina Zheng, Claudia Shi, Keyon Vafa, Amir Feder

Controlled generation refers to the problem of creating text that contains stylistic or semantic attributes of interest. Many approaches reduce this problem to training a predictor of the desired attribute. For example, researchers hoping to deploy a large language model to produce non-toxic content may use a toxicity classifier to filter generated text. In

Peter Belcak, Luca A. Lanzendörfer, Roger Wattenhofer

We conduct a preliminary inquiry into the ability of generative transformer models to deductively reason from premises provided. We observe notable differences in the performance of models coming from different training setups and find that the deductive reasoning ability increases with scale. Further, we discover that the performance generally does not decr

Ziang Xu, Jens Rittscher, Sharib Ali

Data-driven methods have shown tremendous progress in medical image analysis. In this context, deep learning-based supervised methods are widely popular. However, they require a large amount of training data and face issues in generalisability to unseen datasets that hinder clinical translation. Endoscopic imaging data incorporates large inter- and intra-pat

Yige Hong, Qiaomin Xie, Yudong Chen, Weina Wang

We study the infinite-horizon restless bandit problem with the average reward criterion, in both discrete-time and continuous-time settings. A fundamental goal is to efficiently compute policies that achieve a diminishing optimality gap as the number of arms, $N$, grows large. Existing results on asymptotic optimality all rely on the uniform global attractor

Ranjani G Sundaram, Himanshu Gupta

Scalability is currently one of the most sought-after objectives in the field of quantum computing. Distributing a quantum circuit across a quantum network is one way to facilitate large computations using current quantum computers. In this paper, we consider the problem of distributing a quantum circuit across a network of heterogeneous quantum computers, w

Joao Marcos do O, Guozhen Lu, Raoni Ponciano

We establish embeddings on a class of Sobolev spaces with potential weights on unbounded domains. Our results provide embeddings into weighted Lebesgue spaces $L^q_\theta$ with radial power weights and establish the existence and non-existence of the maximizers for their Trudinger-Moser type inequalities. We also sharpen the maximal integrability by ``removi

Graham Harper, Ray Tuminaro

Patch-based relaxation refers to a family of methods for solving linear systems which partitions the matrix into smaller pieces often corresponding to groups of adjacent degrees of freedom residing within patches of the computational domain. The two most common families of patch-based methods are block-Jacobi and Schwarz methods, where the former typically c

Daniel Margineda, Alessandro Crippa, Elia Strambini, Yuri Fukaya

Supercurrent diodes are nonreciprocal electronic elements whose switching current depends on their flow direction. Recently, a variety of composite systems combining different materials and engineered asymmetric superconducting devices have been proposed. Yet, ease of fabrication and tunable sign of supercurrent rectification joined to large efficiency have

M. Järvinen, E. Kiritsis, F. Nitti, E. Préau

A (toy) model for cold and luke-warm strongly-coupled nuclear matter at finite baryon density is used to study neutrino transport. The complete charged current two-point correlators are computed in the strongly-coupled medium and their impact on neutrino transport is analyzed. The full result is compared with various approximations for the current correlator

Hao Wu, Shu Liu, Xiaojing Ye, Haomin Zhou

In this work, we propose a numerical method to compute the Wasserstein Hamiltonian flow (WHF), which is a Hamiltonian system on the probability density manifold. Many well-known PDE systems can be reformulated as WHFs. We use parameterized function as push-forward map to characterize the solution of WHF, and convert the PDE to a finite-dimensional ODE system

Gautam Yadav, Ying-Jui Tseng, Xiaolin Ni

Contextualizing problems to align with student interests can significantly improve learning outcomes. However, this task often presents scalability challenges due to resource and time constraints. Recent advancements in Large Language Models (LLMs) like GPT-4 offer potential solutions to these issues. This study explores the ability of GPT-4 in the contextua

Giancarlo Rossi

This is the second of two companion papers in which we continue developing the construction of an elementary particle model with no Higgs. Here we show that the recently identified non-perturbative field-theoretical feature, alternative to the Higgs mechanism and capable of giving masses to quarks, Tera-quarks and $W$, can also provide mass to leptons and Te

Guangyao Zheng, Shuhao Lai, Vladimir Braverman, Michael A. Jacobs

Deep reinforcement learning(DRL) is increasingly being explored in medical imaging. However, the environments for medical imaging tasks are constantly evolving in terms of imaging orientations, imaging sequences, and pathologies. To that end, we developed a Lifelong DRL framework, SERIL to continually learn new tasks in changing imaging environments without

Kailash Gogineni, Yongsheng Mei, Peng Wei, Tian Lan

Multi-Agent Experience Replay (MER) is a key component of off-policy reinforcement learning~(RL) algorithms. By remembering and reusing experiences from the past, experience replay significantly improves the stability of RL algorithms and their learning efficiency. In many scenarios, multiple agents interact in a shared environment during online training und

Paul Roit, Johan Ferret, Lior Shani, Roee Aharoni

Despite the seeming success of contemporary grounded text generation systems, they often tend to generate factually inconsistent text with respect to their input. This phenomenon is emphasized in tasks like summarization, in which the generated summaries should be corroborated by their source article. In this work, we leverage recent progress on textual enta

B. Posselt, G. G. Pavlov, O. Kargaltsev, J. Hare

Previous observations of the middle-aged $\gamma$-ray, X-ray, and radio pulsar B1055-52 indicated some peculiarities, such as a suspected changing of the X-ray flux and spectral parameters, a large excess of the alleged thermal component of the ultraviolet (UV) spectrum over the Rayleigh-Jeans extension of the X-ray thermal spectrum, and a possible double br

Siu Wun Cheung, Youngsoo Choi, H. Keo Springer, Teeratorn Kadeethum

Understanding the mechanisms of shock-induced pore collapse is of great interest in various disciplines in sciences and engineering, including materials science, biological sciences, and geophysics. However, numerical modeling of the complex pore collapse processes can be costly. To this end, a strong need exists to develop surrogate models for generating ec

Vedant Nanda, Till Speicher, John P. Dickerson, Soheil Feizi

Representations learned by pre-training a neural network on a large dataset are increasingly used successfully to perform a variety of downstream tasks. In this work, we take a closer look at how features are encoded in such pre-trained representations. We find that learned representations in a given layer exhibit a degree of diffuse redundancy, ie, any rand

Gabriel Rioux, Ziv Goldfeld, Kengo Kato

The Gromov-Wasserstein (GW) distance quantifies discrepancy between metric measure spaces and provides a natural framework for aligning heterogeneous datasets. Alas, as exact computation of GW alignment is NP hard, entropic regularization provides an avenue towards a computationally tractable proxy. Leveraging a recently derived variational representation fo

Florentin Guth, Etienne Lempereur, Joan Bruna, Stéphane Mallat

There is a growing gap between the impressive results of deep image generative models and classical algorithms that offer theoretical guarantees. The former suffer from mode collapse or memorization issues, limiting their application to scientific data. The latter require restrictive assumptions such as log-concavity to escape the curse of dimensionality. We

Cameron Smith, Yilun Du, Ayush Tewari, Vincent Sitzmann

Reconstruction of 3D neural fields from posed images has emerged as a promising method for self-supervised representation learning. The key challenge preventing the deployment of these 3D scene learners on large-scale video data is their dependence on precise camera poses from structure-from-motion, which is prohibitively expensive to run at scale. We propos

Chenghao Wang

The main contribution of this MS Thesis is centered around taking steps towards successful multi-modal demonstrations using Northeastern's legged-aerial robot, Husky Carbon. This work discusses the challenges involved in achieving multi-modal locomotion such as trotting-hovering and thruster-assisted incline walking and reports progress made towards overcomi

Konstantin Wernli

These are the lecture notes for a short course on geometric quantization given by the author at the XVIII Modave Summer School on Mathematical Physics, Sep 5 - Sep 9.

Haopeng Zhang, Xiao Liu, Jiawei Zhang

The extended structural context has made scientific paper summarization a challenging task. This paper proposes CHANGES, a contrastive hierarchical graph neural network for extractive scientific paper summarization. CHANGES represents a scientific paper with a hierarchical discourse graph and learns effective sentence representations with dedicated designed

Nicholas Pangakis, Samuel Wolken, Neil Fasching

Generative large language models (LLMs) can be a powerful tool for augmenting text annotation procedures, but their performance varies across annotation tasks due to prompt quality, text data idiosyncrasies, and conceptual difficulty. Because these challenges will persist even as LLM technology improves, we argue that any automated annotation process using a

Alex Altair

When formulated using Bayesian networks, two standard decision algorithms (Evidential Decision Theory and Causal Decision Theory) can be shown to fail systematically when faced with aspects of the prisoner's dilemma and so-called "Newcomblike" problems. We describe a new form of decision algorithm, called Timeless Decision Theory, which consistently wins on

Usman Nazir, Muhammad Ahmad Waseem, Falak Sher Khan, Rabia Saeed

A reliable yet inexpensive tool for the estimation of flood water spread is conducive for efficient disaster management. The application of optical and SAR imagery in tandem provides a means of extended availability and enhanced reliability of flood mapping. We propose a methodology to merge these two types of imagery into a common data space and demonstrate

Rajgowrav Cheenikundil, Massimiliano d'Aquino, Riccardo Hertel

Three-dimensional magnetic nanostructures have recently emerged as artificial magnetic material types with unique properties bearing potential for applications, including magnonic devices. Interconnected magnetic nanowires are a sub-category within this class of materials that is attracting particular interest. We investigate the high-frequency magnetization

Shani Avitan, Ram Brustein, Yotam Sherf

General Relativity predicts that black holes do not possess an internal structure and consequently cannot be excited. This leads to a specific prediction about the waveform of gravitational waves, which they emit during a binary black hole inspiral and to the vanishing of their Love numbers. However, if astrophysical black holes do possess an internal struct

Pengfei Li, Jianyi Yang, Shaolei Ren

Many problems, such as online ad display, can be formulated as online bipartite matching. The crucial challenge lies in the nature of sequentially-revealed online item information, based on which we make irreversible matching decisions at each step. While numerous expert online algorithms have been proposed with bounded worst-case competitive ratios, they ma

Jeff Kahn, Charles Kenney

It is shown that the following holds for each $\varepsilon>0$. For $G$ an $n$-vertex graph of maximum degree $D$ and "lists" $L_v$ ($v \in V(G)$) chosen independently and uniformly from the ($(1+\varepsilon)\ln n$)-subsets of $\{1, ..., D+1\}$, \[ G \text{ admits a proper coloring } \sigma \text{ with } \sigma_v \in L_v \forall v \] with probability tending

Edison M. Murairi, Michael J. Cervia

A variety of quantum algorithms employ Pauli operators as a convenient basis for studying the spectrum or evolution of Hamiltonians or measuring multi-body observables. One strategy to reduce circuit depth in such algorithms involves simultaneous diagonalization of Pauli operators generating unitary evolution operators or observables of interest. We propose

Rie Johnson, Tong Zhang

As deep neural networks are highly expressive, it is important to find solutions with small generalization gap (the difference between the performance on the training data and unseen data). Focusing on the stochastic nature of training, we first present a theoretical analysis in which the bound of generalization gap depends on what we call inconsistency and

Usman Nazir, Momin Uppal, Muhammad Tahir, Zubair Khalid

This paper proposes a multi-spectral random forest classifier with suitable feature selection and masking for tree cover estimation in urban areas. The key feature of the proposed classifier is filtering out the built-up region using spectral indices followed by random forest classification on the remaining mask with carefully selected features. Using Sentin

Nitay Calderon, Naveh Porat, Eyal Ben-David, Alexander Chapanin

Existing research on Domain Robustness (DR) suffers from disparate setups, limited task variety, and scarce research on recent capabilities such as in-context learning. Furthermore, the common practice of measuring DR might not be fully accurate. Current research focuses on challenge sets and relies solely on the Source Drop (SD): Using the source in-domain

Ziyu Ji, Julian Wolfson

Increasingly during the past decade, researchers have sought to leverage auxiliary data for enhancing individualized inference. Many existing methods, such as multisource exchangeability models (MEM), have been developed to borrow information from multiple supplemental sources to support parameter inference in a primary source. MEM and its alternatives decid

C. Augier, G. Baulieu, V. Belov, L. Bergé

The future Ricochet experiment aims to search for new physics in the electroweak sector by measuring the Coherent Elastic Neutrino-Nucleus Scattering process from reactor antineutrinos with high precision down to the sub-100 eV nuclear recoil energy range. While the Ricochet collaboration is currently building the experimental setup at the reactor site, it i

Omid Ashtari, Tobias M. Schneider

Invariant solutions of the Navier-Stokes equations play an important role in the spatiotemporally chaotic dynamics of turbulent shear flows. Despite the significance of these solutions, their identification remains a computational challenge, rendering many solutions inaccessible and thus hindering progress towards a dynamical description of turbulence in ter

Isaac B. W. Harris, Cathryn P. Michaels, Kevin C. Chen, Ryan A. Parker

A quantum register coupled to a spin-photon interface is a key component in quantum communication and information processing. Group-IV color centers in diamond (SiV, GeV, and SnV) are promising candidates for this application, comprising an electronic spin with optical transitions coupled to a nuclear spin as the quantum register. However, the creation of a

F. Kahil, A. Gandorfer, J. Hirzberger, D. Calchetti

We use wavefront sensing to characterise the image quality of the the High Resolution Telescope (HRT) of the Polarimetric and Helioseismic Imager (SO/PHI) data products during the second remote sensing window of the Solar Orbiter (SO) nominal mission phase. Our ultimate aims are to reconstruct the HRT data by deconvolving with the HRT point spread function (

Conrad Kosowsky

In this paper, I prove that existence of pure-strategy Nash equilibrium in games with infinitely many players is equivalent to the axiom of choice.

Mohammed Sayyad, Ying Qin, Jan Kopaczek, Adway Gupta

Two-dimensional Janus transition metal dichalcogenides (TMDs) have attracted attention due to their emergent properties arising from broken mirror symmetry and self-driven polarisation fields. While it has been proposed that their vdW superlattices hold the key to achieving superior properties in piezoelectricity and photovoltiacs, available synthesis has ul

Arturo Ortiz-Tapia

The Paul Erd\H{o}s-Tur\'an inequality is used as a quantitative form of Weyl' s criterion, together with other criteria to asses equidistribution properties on some patterns of sequences that arise from indexation of prime numbers, Jumping Champions (called here and in previous work, "meta-distances" or even md, for short). A statistical meta-analysis is als

Héctor Martel, Julius Richter, Kai Li, Xiaolin Hu

We propose Audio-Visual Lightweight ITerative model (AVLIT), an effective and lightweight neural network that uses Progressive Learning (PL) to perform audio-visual speech separation in noisy environments. To this end, we adopt the Asynchronous Fully Recurrent Convolutional Neural Network (A-FRCNN), which has shown successful results in audio-only speech sep

Philippe Charron, Dan Mangoubi

We prove that every nodal domain of an eigenfunction of the Laplacian of eigenvalue $\lambda$ on a $d$-dimensional closed Riemannian manifold contains a ball of radius $c\lambda^{-1/2}(\log\lambda)^{-(d-2)/2}$. This ball is centered at a point at which the eigenfunction attains its maximum in absolute value within the nodal domain.

Shangjia Zhang, Zhaohuan Zhu, Takahiro Ueda, Akimasa Kataoka

Dust particle sizes constrained from dust continuum and polarization observations by radio interferometry are inconsistent by at least an order of magnitude. Motivated by porous dust observed in small Solar System bodies (e.g., from the Rosetta mission), we explore how the dust particle's porosity affects the estimated particle sizes from these two methods.