May 2023 arXiv papers — page 51

Zhuokai Zhao, Yang Yang, Wenyu Wang, Chihuang Liu

A key puzzle in search, ads, and recommendation is that the ranking model can only utilize a small portion of the vastly available user interaction data. As a result, increasing data volume, model size, or computation FLOPs will quickly suffer from diminishing returns. We examined this problem and found that one of the root causes may lie in the so-called ``

Han Zhang, Axel Ringh

Inverse Optimal Control (IOC) is a powerful framework for learning a behaviour from observations of experts. The framework aims to identify the underlying cost function that the observed optimal trajectories (the experts' behaviour) are optimal with respect to. In this work, we considered the case of identifying the cost and the feedback law from observed tr

Omid Sadeghi, Maryam Fazel

We study a generalization of the online binary prediction with expert advice framework where at each round, the learner is allowed to pick $m\geq 1$ experts from a pool of $K$ experts and the overall utility is a modular or submodular function of the chosen experts. We focus on the setting in which experts act strategically and aim to maximize their influenc

Emiliano Catonini, Nicodemo De Vito

We define notions of cautiousness and cautious belief to provide epistemic conditions for iterated admissibility in finite games. We show that iterated admissibility characterizes the behavioral implications of "cautious rationality and common cautious belief in cautious rationality" in a terminal lexicographic type structure. For arbitrary type structures,

LHCb collaboration, R. Aaij, A. S. W. Abdelmotteleb, C. Abellan Beteta

The mass difference between the $\Omega^-_b$ and $\Xi^-_b$ baryons is measured using proton-proton collision data collected by the LHCb experiment, corresponding to an integrated luminosity of $9 \, \text{fb}^{-1}$, and is found to be \begin{equation} m(\Omega^-_b)- m(\Xi^-_b) = 248.54 \pm 0.51 \text{(stat)} \pm 0.38 \text{(syst)} \, \text{MeV}/c^2. \end{equ

Jaemin Cho, Abhay Zala, Mohit Bansal

As large language models have demonstrated impressive performance in many domains, recent works have adopted language models (LMs) as controllers of visual modules for vision-and-language tasks. While existing work focuses on equipping LMs with visual understanding, we propose two novel interpretable/explainable visual programming frameworks for text-to-imag

Cary Malkiewich

We describe a point-set category of parametrized orthogonal spectra, a model structure on this category, and a separate, more geometric class of cofibrant-and-fibrant objects. The structures we describe are "convenient" in that they are preserved by the most common operations. They allow us to reduce sophisticated statements about the homotopy category to st

José Jaime Terente Díaz, Mindaugas Karčiauskas

In the context of $F(\phi)R$ models of gravity, the conformal invariance of the curvature perturbation on uniform-field slices has been already demonstrated in several publications. In this work we study the curvature perturbation $\mathcal{R}$ defined on hypersurfaces that comove with the effective fluid whose energy-momentum tensor is covariantly conserved

Sándor Baran, Mária Lakatos

To be able to produce accurate and reliable predictions of visibility has crucial importance in aviation meteorology, as well as in water- and road transportation. Nowadays, several meteorological services provide ensemble forecasts of visibility; however, the skill, and reliability of visibility predictions are far reduced compared to other variables, such

Toby Shevlane, Sebastian Farquhar, Ben Garfinkel, Mary Phuong

Current approaches to building general-purpose AI systems tend to produce systems with both beneficial and harmful capabilities. Further progress in AI development could lead to capabilities that pose extreme risks, such as offensive cyber capabilities or strong manipulation skills. We explain why model evaluation is critical for addressing extreme risks. De

Siying Liu, Saleh Alfarhan, Wenbo Wang, Shuai Feng

The emerging Internet of Things (IoTs) invokes increasing security demands that require robust encryption or anti-counterfeiting technologies. Albeit being acknowledged as efficacious solutions in processing elaborate graphical information via multiple degrees of freedom, optical data encryption and anti-counterfeiting techniques are typically inept in deliv

Liane Vogel, Benjamin Hilprecht, Carsten Binnig

Tabular representation learning has recently gained a lot of attention. However, existing approaches only learn a representation from a single table, and thus ignore the potential to learn from the full structure of relational databases, including neighboring tables that can contain important information for a contextualized representation. Moreover, current

Manami Yamagishi, Naomichi Hatano, Hideaki Obuse

The main aim of the present paper is to define an active particle in a quantum framework as a minimal model of quantum active matter and investigate the differences and similarities of quantum and classical active matter. Although the field of active matter has been expanding, most research has been conducted on classical systems. Here, we propose a truly de

Rafael Kiesel, Kilian Rückschloß, Felix Weitkämper

A ProbLog program is a logic program with facts that only hold with a specified probability. In this contribution we extend this ProbLog language by the ability to answer "What if" queries. Intuitively, a ProbLog program defines a distribution by solving a system of equations in terms of mutually independent predefined Boolean random variables. In the theory

Yongchao Zhou, Hshmat Sahak, Jimmy Ba

Acquiring high-quality data for training discriminative models is a crucial yet challenging aspect of building effective predictive systems. In this paper, we present Diffusion Inversion, a simple yet effective method that leverages the pre-trained generative model, Stable Diffusion, to generate diverse, high-quality training data for image classification. O

J. Cernicharo, B. Tercero, N. Marcelino, M. Agundez

We present a highly sensitive 2D line survey of TMC-1 obtained with the Yebes 40m radio telescope in the Q-band (31.13-49.53 GHz). These maps cover a region of 320 arcsec x 320 arcsec centred on the position of the QUIJOTE line survey with a spatial sampling of 20 arcsec. The region covering 240 arcsec x 240 arcsec, where a longer integration time was used,

Vijayanta Jain, Sepideh Ghanavati, Sai Teja Peddinti, Collin McMillan

Mobile applications are required to give privacy notices to users when they collect or share personal information. Creating consistent and concise privacy notices can be a challenging task for developers. Previous work has attempted to help developers create privacy notices through a questionnaire or predefined templates. In this paper, we propose a novel ap

Gergely Flamich

One-shot channel simulation is a fundamental data compression problem concerned with encoding a single sample from a target distribution $Q$ using a coding distribution $P$ using as few bits as possible on average. Algorithms that solve this problem find applications in neural data compression and differential privacy and can serve as a more efficient altern

Reza Rabani, Mohammad Hassan Saidi, Ali Rajabpour, Laurent Joly

Heat transfer through the interface between a metallic nanoparticle and an electrolyte solution, has great importance in a number of applications, ranging from nanoparticle-based cancer treatments to nanofluids and solar energy conversion devices. However, the impact of surface charge and the dissolved ions on heat transfer has been scarcely explored so far.

Geyu Liang, Naichen Shi, Raed Al Kontar, Salar Fattahi

We introduce a relevant yet challenging problem named Personalized Dictionary Learning (PerDL), where the goal is to learn sparse linear representations from heterogeneous datasets that share some commonality. In PerDL, we model each dataset's shared and unique features as global and local dictionaries. Challenges for PerDL not only are inherited from classi

Rafael Ceja Ayala, Isaac Harris, Andreas Kleefeld

In this paper, we consider the inverse shape problem of recovering isotropic scatterers with a conductive boundary condition. Here, we assume that the measured far-field data is known at a fixed wave number. Motivated by recent work, we study a new direct sampling indicator based on the Landweber iteration and the factorization method. Therefore, we prove th

Andreas Maier, Michael Hartung, Mark Abovsky, Klaudia Adamowicz

In recent decades, the development of new drugs has become increasingly expensive and inefficient, and the molecular mechanisms of most pharmaceuticals remain poorly understood. In response, computational systems and network medicine tools have emerged to identify potential drug repurposing candidates. However, these tools often require complex installation

Valery F. Suleimanov, Sofia V. Forsblom, Sergey S. Tsygankov, Juri Poutanen

The phase- and energy-resolved polarization measurements of accreting X-ray pulsars (XRPs) allow us to test different theoretical models of their emission, and they also provide an avenue to determine the emission region geometry. We present the results of the observations of the XRP GX 301-2 performed with the Imaging X-ray Polarimetry Explorer (IXPE). A pe

Auchère, F., Berghmans, D.

Most observations of the solar corona beyond 2 Rs consist of broadband visible light imagery from coronagraphs. The associated diagnostics mainly consist of kinematics and derivations of the electron number density. While the measurement of the properties of emission lines can provide crucial additional diagnostics of the coronal plasma (temperatures, veloci

Manali Malakar, Sudip Sinha, S. Sinha

We investigate the formation of paired states of bosons in an optical lattice, namely, pair superfluid (PSF) and pair supersolid (PSS) in the presence of pair hopping as well as the next nearest neighbor (NNN) interaction mimicking long-range forces. Both the zero and finite temperature phase diagrams are obtained using the cluster mean field theory, which i

Daniel S. Levin, Peter S. Friedman, Claudio Ferretti, Nicholas Ristow

FLASH Radiotherapy (RT) is a potentially new cancer radiotherapy technique where an entire therapeutic dose is delivered in about 0.1 s and at ~1000 times higher dose rate than in conventional RT. For clinical trials to be conducted safely, precise and fast beam monitoring that can generate an out-of-tolerance beam interrupt is required. A FLASH Beam Scintil

Tobias Dornheim, Tilo Döppner, Andrew D. Baczewski, Panagiotis Tolias

We present a formally exact and simulation-free approach for the normalization of X-ray Thomson scattering (XRTS) spectra based on the f-sum rule of the imaginary-time correlation function (ITCF). Our method works for any degree of collectivity, over a broad range of temperatures, and is applicable even in nonequilibrium situations. In addition to giving us

Susheela Sharma, Yuewan Sun, Sarah Go, Jordan P. Amadio

Towards reducing the failure rate of spinal fixation surgical procedures in osteoporotic patients, we propose a unique biomechanically-aware framework for the design of a novel concentric tube steerable drilling robot (CT-SDR). The proposed framework leverages a patient-specific finite element (FE) biomechanics model developed based on Quantitative Computed

S. Mahakal, Diptasikha Das, Pintu Singha, Aritra Banerjee

TiCoSb1+x (x=0.0, 0.01, 0.02, 0.03, 0.04, 0.06) samples have been synthesized, employing solid state reaction method followed by arc menting. Theoretical calculations, using Density Functional Theory (DFT) have been performed to estimate band structure and density of states (DOS). Further, energitic calculations, using first principle have been carried out t

Chang Liu, Henghui Ding, Yulun Zhang, Xudong Jiang

We address the problem of referring image segmentation that aims to generate a mask for the object specified by a natural language expression. Many recent works utilize Transformer to extract features for the target object by aggregating the attended visual regions. However, the generic attention mechanism in Transformer only uses the language input for atte

Martje Rave, Göran Kauermann

With the beginning of the COVID-19 pandemic, we became aware of the need for comprehensive data collection and its provision to scientists and experts for proper data analyses. In Germany, the Robert Koch Institute (RKI) has tried to keep up with this demand for data on COVID-19, but there were (and still are) relevant data missing that are needed to underst

Siarhei Finski

From the work of Phong and Sturm in 2007, for a polarised projective manifold and an ample test configuration, one can associate the geodesic ray of plurisubharmonic metrics on the polarising line bundle using the solution of the Monge-Amp\`ere equation on an equivariant resolution of singularities of the test configuration. We prove that the Mabuchi chordal

Evangelos Pournaras

Large language models of artificial intelligence (AI), such as ChatGPT, find remarkable but controversial applicability in science and research. This paper reviews epistemological challenges, ethical and integrity risks in science conduct in the advent of generative AI. This is with the aim to lay new timely foundations for a high-quality research ethics rev

Setareh Sharifi Panah, Robert Großmann, Valentino Lepro, Carsten Beta

The integration of motile cells into biohybrid microrobots offers unique properties such as sensitive responses to external stimuli, resilience, and intrinsic energy supply. Here we study biohybrid microtransporters that are driven by amoeboid Dictyostelium discoideum cells and explore how the speed of transport and the resulting viscous drag force scales wi

Noga Alon, Anurag Bishnoi, Shagnik Das, Alessandro Neri

A strong blocking set in a finite projective space is a set of points that intersects each hyperplane in a spanning set. We provide a new graph theoretic construction of such sets: combining constant-degree expanders with asymptotically good codes, we explicitly construct strong blocking sets in the $(k-1)$-dimensional projective space over $\mathbb{F}_q$ th

Marco Bellagente, Manuel Brack, Hannah Teufel, Felix Friedrich

The recent popularity of text-to-image diffusion models (DM) can largely be attributed to the intuitive interface they provide to users. The intended generation can be expressed in natural language, with the model producing faithful interpretations of text prompts. However, expressing complex or nuanced ideas in text alone can be difficult. To ease image gen

Marcelo P. Becker, Marco A. P. Idiart

The construction of transfer functions in theoretical neuroscience plays an important role in determining the spiking rate behavior of neurons in networks. These functions can be obtained through various fitting methods, but the biological relevance of the parameters is not always clear. However, for stationary inputs, such functions can be obtained without

Zhihong Shao, Yeyun Gong, Yelong Shen, Minlie Huang

Large language models are powerful text processors and reasoners, but are still subject to limitations including outdated knowledge and hallucinations, which necessitates connecting them to the world. Retrieval-augmented large language models have raised extensive attention for grounding model generation on external knowledge. However, retrievers struggle to

Annika Mauro, Jack B. Miller, Steven J. Miller

The Katz-Sarnak philosophy predicts that the behavior of zeros near the central point in families of $L$-functions agrees with that of eigenvalues near 1 of random matrix ensembles. Under GRH, Iwaniec, Luo and Sarnak showed agreement in the one-level densities for cuspidal newforms with the support of the Fourier transform of the test function in $(-2, 2)$.

Axel Ringh, Isabel Haasler, Yongxin Chen, Johan Karlsson

In this work we consider mean field type control problems with multiple species that have different dynamics. We formulate the discretized problem using a new type of entropy-regularized multimarginal optimal transport problems where the cost is a decomposable structured tensor. A novel algorithm for solving such problems is derived, using this structure and

Yulle G. F. Borges, Rafael C. S. Schouery, Flávio K. Miyazawa

This paper focuses on exact approaches for the Colored Bin Packing Problem (CBPP), a generalization of the classical one-dimensional Bin Packing Problem in which each item has, in addition to its length, a color, and no two items of the same color can appear consecutively in the same bin. To simplify modeling, we present a characterization of any feasible pa

Heena Khanchandani, Stefan Zeiler, Lucas Strobel, Mathias Goeken

High strength steels are susceptible to H-induced failure, which is typically caused by the presence of diffusible H in the microstructure. The diffusivity of H in austenitic steels with fcc crystal structure is slow. The austenitic steels are hence preferred for applications in the hydrogen-containing atmospheres. However, the fcc structure of austenitic st

Igor Pavlov

For the classical N-body problem, an approach is proposed based on the introduction of some natural in the physical sense optimization problems of mathematical programming for finding a conditional minimum for the characteristics of the system on the set of its possible states. The solution of these problems then makes it possible to construct families of fl

T V Anoop, Ujjal Das, Subhajit Roy

Let $\Omega$ be an open subset of $\mathbb{R}^N$ with $N\geq 2.$ We identify various classes of Young functions $\Phi$ and $\Psi$, and function spaces for a weight function $g$ so that the following weighted Orlicz-Sobolev inequality holds: \begin{equation*}\label{ineq:Orlicz} \Psi^{-1}\left(\int_{\Omega}|g(x)|\,\Psi(|u(x)| )dx \right)\leq C\Phi^{-1}\left(\i

Sukriti Singh, Anusha Srikanthan, Vivek Mallampati, Harish Ravichandar

Task allocation can enable effective coordination of multi-robot teams to accomplish tasks that are intractable for individual robots. However, existing approaches to task allocation often assume that task requirements or reward functions are known and explicitly specified by the user. In this work, we consider the challenge of forming effective coalitions f

Yan Dai, Kwangjun Ahn, Suvrit Sra

Sharpness-Aware Minimization (SAM) is a recently proposed gradient-based optimizer (Foret et al., ICLR 2021) that greatly improves the prediction performance of deep neural networks. Consequently, there has been a surge of interest in explaining its empirical success. We focus, in particular, on understanding the role played by normalization, a key component

Hans Johnson, Tianyang Fang, Alejandro Perez-Vicente, Jafar Saniie

We propose a distributed system based on lowpower embedded FPGAs designed for edge computing applications focused on exploring distributing scheduling optimizations for Deep Learning (DL) workloads to obtain the best performance regarding latency and power efficiency. Our cluster was modular throughout the experiment, and we have implementations that consist

August Geelmuyden, Vitor S. Barroso, Sreelekshmi C. Ajithkumar, Anthony J. Kent

Digital holography (DH) has been widely used for imaging and characterization of micro and nanostructures in materials science and biology and has the potential to provide high-resolution, non-destructive measurements of fluid surfaces as well. Digital holographic setups capture the complex wavefronts of light scattered by an object or reflected from a surfa

Ansgar Jüngel, Annamaria Massimini

A modified Poisson-Nernst-Planck system in a bounded domain with mixed Dirichlet-Neumann boundary conditions is analyzed. It describes the concentrations of ions immersed in a polar solvent and the correlated electric potential due to the ion--solvent interaction. The concentrations solve cross-diffusion equations, which are thermodynamically consistent. The

Brian N. Granzow, D. Thomas Seidl, Stephen D. Bond

This paper is concerned with goal-oriented a posteriori error estimation for nonlinear functionals in the context of nonlinear variational problems solved with continuous Galerkin finite element discretizations. A two-level, or discrete, adjoint-based approach for error estimation is considered. The traditional method to derive an error estimate in this cont

Eric Eaton, Marcel Hussing, Michael Kearns, Jessica Sorrell

The replicability crisis in the social, behavioral, and data sciences has led to the formulation of algorithm frameworks for replicability -- i.e., a requirement that an algorithm produce identical outputs (with high probability) when run on two different samples from the same underlying distribution. While still in its infancy, provably replicable algorithm

Enrico Picco, Piotr Antonik, Serge Massar

The recognition of human actions in videos is one of the most active research fields in computer vision. The canonical approach consists in a more or less complex preprocessing stages of the raw video data, followed by a relatively simple classification algorithm. Here we address recognition of human actions using the reservoir computing algorithm, which all

Rohan Bhambhoria, Lei Chen, Xiaodan Zhu

In recent years, large language models (LLMs) have achieved strong performance on benchmark tasks, especially in zero or few-shot settings. However, these benchmarks often do not adequately address the challenges posed in the real-world, such as that of hierarchical classification. In order to address this challenge, we propose refactoring conventional tasks

Markus Fellner, Ansgar Jüngel

A one-dimensional cross-diffusion system modeling the transport of vesicles in neurites is analyzed. The equations are coupled via nonlinear Robin boundary conditions to ordinary differential equations for the number of vesicles in the reservoirs in the cell body and the growth cone at the end of the neurite. The existence of bounded weak solutions is proved

Somnath Das, P K Suresh

Inflationary cosmology has made significant strides in understanding the physics driving the rapid expansion of the early universe. However, many inflation models with diverse potential shapes present analysis, comparison, and classification challenges. In this paper, we propose a novel approach to tackle this issue. We introduce a general potential formula

Ryuzo Hirota, Hayato Saigo, Shigeru Taguchi

Life continuously changes its own components and states at each moment through interaction with the external world, while maintaining its own individuality in a cyclical manner. Such a property, known as "autonomy," has been formulated using the mathematical concept of "closure." We introduce a branch of mathematics called "category theory" as an "arrow-firs

Changmin Yu, Neil Burgess, Maneesh Sahani, Samuel J. Gershman

Exploration is essential in reinforcement learning, particularly in environments where external rewards are sparse. Here we focus on exploration with intrinsic rewards, where the agent transiently augments the external rewards with self-generated intrinsic rewards. Although the study of intrinsic rewards has a long history, existing methods focus on composin

Jianhao Ma, Rui Ray Chen, Yinghui He, Salar Fattahi

In this paper, we study the problem of sparse mean estimation under adversarial corruptions, where the goal is to estimate the $k$-sparse mean of a heavy-tailed distribution from samples contaminated by adversarial noise. Existing methods face two key limitations: they require prior knowledge of the sparsity level $k$ and scale poorly to high-dimensional set

Qihuang Zhong, Liang Ding, Juhua Liu, Bo Du

Masked language modeling, widely used in discriminative language model (e.g., BERT) pretraining, commonly adopts a random masking strategy. However, random masking does not consider the importance of the different words in the sentence meaning, where some of them are more worthy to be predicted. Therefore, various masking strategies (e.g., entity-level maski

Daniele Cassani, Zhisu Liu, Giulio Romani

We study the nonlinear Schr\"odinger equation for the $s-$fractional $p-$Laplacian strongly coupled with the Poisson equation in dimension two and with $p=\frac2s$, which is the limiting case for the embedding of the fractional Sobolev space $W^{s,p}(\mathbb{R}^2)$. We prove existence of solutions by means of a variational approximating procedure for an auxi

Qihuang Zhong, Liang Ding, Juhua Liu, Xuebo Liu

Token dropping is a recently-proposed strategy to speed up the pretraining of masked language models, such as BERT, by skipping the computation of a subset of the input tokens at several middle layers. It can effectively reduce the training time without degrading much performance on downstream tasks. However, we empirically find that token dropping is prone

Jingfeng Yao, Xinggang Wang, Shusheng Yang, Baoyuan Wang

Recently, plain vision Transformers (ViTs) have shown impressive performance on various computer vision tasks, thanks to their strong modeling capacity and large-scale pretraining. However, they have not yet conquered the problem of image matting. We hypothesize that image matting could also be boosted by ViTs and present a new efficient and robust ViT-based

Serena Dipierro, Ovidiu Savin, Enrico Valdinoci

Differently from their classical counterpart, nonlocal minimal surfaces are known to present boundary discontinuities, by sticking at the boundary of smooth domains. It has been observed numerically by J. P. Borthagaray, W. Li, and R. H. Nochetto ``that stickiness is larger near the concave portions of the boundary than near the convex ones, and that it is a

Tong Xu, Micol Spitale, Hao Tang, Lu Liu

Generating facial reactions in a human-human dyadic interaction is complex and highly dependent on the context since more than one facial reactions can be appropriate for the speaker's behaviour. This has challenged existing machine learning (ML) methods, whose training strategies enforce models to reproduce a specific (not multiple) facial reaction from eac

Abulhair Saparov, Richard Yuanzhe Pang, Vishakh Padmakumar, Nitish Joshi

Given the intractably large size of the space of proofs, any model that is capable of general deductive reasoning must generalize to proofs of greater complexity. Recent studies have shown that large language models (LLMs) possess some abstract deductive reasoning ability given chain-of-thought prompts. However, they have primarily been tested on proofs usin

Zhengwei Tao, Zhi Jin, Xiaoying Bai, Haiyan Zhao

Events serve as fundamental units of occurrence within various contexts. The processing of event semantics in textual information forms the basis of numerous natural language processing (NLP) applications. Recent studies have begun leveraging large language models (LLMs) to address event semantic processing. However, the extent that LLMs can effectively tack

Chen-Hao Chao, Wei-Fang Sun, Yen-Chang Hsu, Zsolt Kira

In this paper, we establish a connection between the parameterization of flow-based and energy-based generative models, and present a new flow-based modeling approach called energy-based normalizing flow (EBFlow). We demonstrate that by optimizing EBFlow with score-matching objectives, the computation of Jacobian determinants for linear transformations can b

Eloi Moliner, Vesa Välimäki

Audio inpainting aims to reconstruct missing segments in corrupted recordings. Most of existing methods produce plausible reconstructions when the gap lengths are short, but struggle to reconstruct gaps larger than about 100 ms. This paper explores recent advancements in deep learning and, particularly, diffusion models, for the task of audio inpainting. The

Zirui Liu, Guanchu Wang, Shaochen Zhong, Zhaozhuo Xu

With the rapid growth in model size, fine-tuning the large pre-trained language model has become increasingly difficult due to its extensive memory usage. Previous works usually focus on reducing the number of trainable parameters in the network. While the model parameters do contribute to memory usage, the primary memory bottleneck during training arises fr

Peter Richtárik, Elnur Gasanov, Konstantin Burlachenko

We provide the first proof that gradient descent $\left({\color{green}\sf GD}\right)$ with greedy sparsification $\left({\color{green}\sf TopK}\right)$ and error feedback $\left({\color{green}\sf EF}\right)$ can obtain better communication complexity than vanilla ${\color{green}\sf GD}$ when solving the distributed optimization problem $\min_{x\in \mathbb{R}

Michael Hahsler

The R arules package implements a comprehensive infrastructure for representing, manipulating, and analyzing transaction data and patterns using frequent itemsets and association rules. The package also provides a wide range of interest measures and mining algorithms, including the code of Christian Borgelt's popular and efficient C implementations of the as

Kejuan Yang, Xiao Liu, Kaiwen Men, Aohan Zeng

We identify two crucial limitations in the evaluation of recent parallel-integrated method Parallel Context Windows (PCW), which extends the maximum context lengths of language models, e.g., 2048 for LLaMA, by harnessing window-wise attention and positional embedding techniques. We first show that a simple yet strong baseline, weighted sum ensemble, is missi

Valentina Rein, Hao Gao, Hendrik H. Heenen, Wissal Sghaier

In recent years, liquid metal catalysts have emerged as a compelling choice for the controllable, large-scale, and high-quality synthesis of two-dimensional materials. At present, there is little mechanistic understanding of the intricate catalytic process, though, of its governing factors or what renders it superior to growth at the corresponding solid cata

Jorge Antezana, Diana Carbajal, José Luis Romero

We consider multi-variate signals spanned by the integer shifts of a set of generating functions with distinct frequency profiles and the problem of reconstructing them from samples taken on a random periodic set. We show that such a sampling strategy succeeds with high probability provided that the density of the sampling pattern exceeds the number of frequ

Qi Wang, Junming Yang, Yunbo Wang, Xin Jin

Training offline RL models using visual inputs poses two significant challenges, i.e., the overfitting problem in representation learning and the overestimation bias for expected future rewards. Recent work has attempted to alleviate the overestimation bias by encouraging conservative behaviors. This paper, in contrast, tries to build more flexible constrain

Marcel Moosbrugger, Julian Müllner, Laura Kovács

We present an exact approach to analyze and quantify the sensitivity of higher moments of probabilistic loops with symbolic parameters, polynomial arithmetic and potentially uncountable state spaces. Our approach integrates methods from symbolic computation, probability theory, and static analysis in order to automatically capture sensitivity information abo

Arturo Salmi, Szabolcs Cséfalvay, James Imber

Application of realism enhancement methods, particularly in real-time and resource-constrained settings, has been frustrated by the expense of existing methods. These achieve high quality results only at the cost of long runtimes and high bandwidth, memory, and power requirements. We present an efficient alternative: a high-performance, generative shader-bas

Ning-Chen Bai, Lei Li, Jun Tao

We observe a superfluid $\lambda$ transition in the $P-V$ criticality of charged AdS black holes, within a holographic extended thermodynamics that considers the variation of Newton's constant $G$. We calculate the critical exponents and find that they coincide with those of a superfluid transition in liquid $^4\text{He}$ and the Bose-Einstein condensation o

Min Liu, Zhiqiang Cai, Karthik Ramani

In this paper, we study the deep Ritz method for solving the linear elasticity equation from a numerical analysis perspective. A modified Ritz formulation using the $H^{1/2}(\Gamma_D)$ norm is introduced and analyzed for linear elasticity equation in order to deal with the (essential) Dirichlet boundary condition. We show that the resulting deep Ritz method

Munyque Mittelmann, Aniello Murano, Laurent Perrussel

Discounting is an important dimension in multi-agent systems as long as we want to reason about strategies and time. It is a key aspect in economics as it captures the intuition that the far-away future is not as important as the near future. Traditional verification techniques allow to check whether there is a winning strategy for a group of agents but they

Eliya Nachmani, Alon Levkovitch, Roy Hirsch, Julian Salazar

We present Spectron, a novel approach to adapting pre-trained large language models (LLMs) to perform spoken question answering (QA) and speech continuation. By endowing the LLM with a pre-trained speech encoder, our model becomes able to take speech inputs and generate speech outputs. The entire system is trained end-to-end and operates directly on spectrog

Benno Käch, Isabell Melzer-Pellmann

The generation of collider data using machine learning has emerged as a prominent research topic in particle physics due to the increasing computational challenges associated with traditional Monte Carlo simulation methods, particularly for future colliders with higher luminosity. Although generating particle clouds is analogous to generating point clouds, a

Han Yu, Xingxuan Zhang, Renzhe Xu, Jiashuo Liu

Domain generalization aims to solve the challenge of Out-of-Distribution (OOD) generalization by leveraging common knowledge learned from multiple training domains to generalize to unseen test domains. To accurately evaluate the OOD generalization ability, it is required that test data information is unavailable. However, the current domain generalization pr

Sadaf Ul Zuhra, Prasanna Chaporkar, Abhay Karandikar, H. Vincent Poor

In video streaming applications especially during live streaming events (such as the Super Bowl), video traffic can account for a significant portion of network traffic and can lead to severe network congestion. During such events, multicast transmission can be used to avoid network congestion since the same video content is being streamed to multiple users

Taiping Zhang

In this thesis, we propose to tackle this important issue by designing and realizing a novel nano-optical device based on the use of a photonic crystal (PC) structure to generate an efficient coupling between the external source and a NA. In this dissertation, the content is arranged into three charpters. Chapter 1 introduces the theoritical background of th

Sonia Cambiaso, Fabio Rasera, Antonio Tinti, Davide Bochicchio

Hydrophobic nanoporous materials can be intruded by water only by exerting an external action, typically increasing pressure. For some materials, water extrudes when the pressure is lowered again. Controlling intrusion/extrusion hysteresis is central in a number of technological applications, including materials for energy applications and for high performan

Sharan Vaswani, Amirreza Kazemi, Reza Babanezhad, Nicolas Le Roux

Actor-critic (AC) methods are widely used in reinforcement learning (RL) and benefit from the flexibility of using any policy gradient method as the actor and value-based method as the critic. The critic is usually trained by minimizing the TD error, an objective that is potentially decorrelated with the true goal of achieving a high reward with the actor. W

Cheng-Ze Lu, Xiaojie Jin, Qibin Hou, Jun Hao Liew

Understanding whether self-supervised learning methods can scale with unlimited data is crucial for training large-scale models. In this work, we conduct an empirical study on the scaling capability of masked image modeling (MIM) methods (e.g., MAE) for visual recognition. Unlike most previous works that depend on the widely-used ImageNet dataset, which is m

Anna N. Morozovska, Eugene A. Eliseev, Yongtao Liu, Kyle P. Kelley

Using Landau-Ginzburg-Devonshire (LGD) phenomenological approach we analyze the bending-induced re-distribution of electric polarization and field, elastic stresses and strains inside ultrathin layers of van der Waals ferrielectrics. We consider a CuInP2S6 (CIPS) thin layer with fixed edges and suspended central part, the bending of which is induced by exter

Bálint Vető, Bálint Virág

Coalescing simple random walks in the plane form an infinite tree. A natural directed distance on this tree is given by the number of jumps between branches when one is only allowed to move in one direction. The Brownian web distance is the scale-invariant limit of this directed metric. It is integer-valued and has scaling exponents 0:1:2 as compared to 1:2:

Fu Xing Long, Diederick Vermetten, Anna V. Kononova, Roman Kalkreuth

Within the optimization community, the question of how to generate new optimization problems has been gaining traction in recent years. Within topics such as instance space analysis (ISA), the generation of new problems can provide new benchmarks which are not yet explored in existing research. Beyond that, this function generation can also be exploited for

Daniel Layeghi, Steve Tonneau, Michael Mistry

Despite impressive results, reinforcement learning (RL) suffers from slow convergence and requires a large variety of tuning strategies. In this paper, we investigate the ability of RL algorithms on simple continuous control tasks. We show that without reward and environment tuning, RL suffers from poor convergence. In turn, we introduce an optimal control (

Jacob B Khurgin

Photonic Time crystals (PTC) arise in time-modulated media when the frequency of modulation of permittivity is on the order of twice the frequency of light and are manifested by the generation and amplification of so-called time reversed waves propagating in the direction opposite to the incoming light. Superficially, the observed phenomenon bears resemblanc

Luciano Floridi

The article explores the cultural shift from recording to deleting information in the digital age and its implications on privacy, intellectual property (IP), and Large Language Models like ChatGPT. It begins by defining a delete culture where information, in principle legal, is made unavailable or inaccessible because unacceptable or undesirable, especially

Huanran Chen, Yinpeng Dong, Zhengyi Wang, Xiao Yang

Diffusion models have been applied to improve adversarial robustness of image classifiers by purifying the adversarial noises or generating realistic data for adversarial training. However, diffusion-based purification can be evaded by stronger adaptive attacks while adversarial training does not perform well under unseen threats, exhibiting inevitable limit

Tomas Lazna, Ludek Zalud

We discuss the localization of radiation sources whose number and other relevant parameters are not known in advance. The data collection is ensured by an autonomous mobile robot that performs a survey in a defined region of interest populated with static obstacles. The measurement trajectory is information-driven rather than pre-planned, and the localizatio

Travis LaCroix, Simon J. D. Prince

This article appears as chapter 21 of Prince (2023, Understanding Deep Learning); a complete draft of the textbook is available here: http://udlbook.com. This chapter considers potential harms arising from the design and use of AI systems. These include algorithmic bias, lack of explainability, data privacy violations, militarization, fraud, and environmenta

Alessandra Fumagalli, Yodovina Piškur, Anže Slosar

We present a new class of models that have potential to alleviate tensions present in the cosmological data today. We postulate about variation in the sound horizon scale on superhorizon scales, i.e., on scales that are larger than that of the present observable low-redshift Universe ($\gtrsim 1\,$Gpc) while at the same time smaller than the largest scales p

Ramy Taki Eldin, Patrick Sole

In this study, we consider the Euclidean and Galois hulls of multi-twisted (MT) codes over a finite field $\mathbb{F}_{p^e}$ of characteristic $p$. Let $\mathbf{G}$ be a generator polynomial matrix (GPM) of a MT code $\mathcal{C}$. For any $0\le \kappa<e$, the $\kappa$-Galois hull of $\mathcal{C}$, denoted by $h_\kappa\left(\mathcal{C}\right)$, is the inters

Claudio Bonati, Andrea Pelissetto, Ettore Vicari

We study the critical behavior of three-dimensional (3D) lattice Abelian-Higgs (AH) gauge models with noncompact gauge variables and multicomponent complex scalar fields, along the transition line between the Coulomb and Higgs phases. Previous works that focused on gauge-invariant correlations provided evidence that, for a sufficiently large number of scalar