October 2024 arXiv papers — page 105

Arvid Frydenlund

The recently introduced path-star task is a minimal task designed to exemplify limitations to the abilities of language models (Bachmann and Nagarajan, 2024). It involves a path-star graph where multiple arms radiate from a single starting node and each node is unique. Given the start node and a specified target node that ends an arm, the task is to generate

Michelangelo Olmo Nogara Notarianni, Filippo Leveni, Diego Stucchi, Luca Frittoli

We present Kernel-QuantTree Exponentially Weighted Moving Average (KQT-EWMA), a non-parametric change-detection algorithm that combines the Kernel-QuantTree (KQT) histogram and the EWMA statistic to monitor multivariate data streams online. The resulting monitoring scheme is very flexible, since histograms can be used to model any stationary distribution, an

Eliahu Horwitz, Bar Cavia, Jonathan Kahana, Yedid Hoshen

The number of publicly available models is rapidly increasing, yet most remain undocumented. Users looking for suitable models for their tasks must first determine what each model does. Training machine learning models to infer missing documentation directly from model weights is challenging, as these weights often contain significant variation unrelated to

Corentin Fierobe, Alfonso Sorrentino, Amir Vig

Symplectic billiards were introduced by Albers and Tabachnikov as billiards in strictly convex bounded domains of the plane with smooth boundary having a specific law of reflection. This paper proves a rigidity result for symplectic billiards which is similar to a previous result on classical billiards formulated by De Simoi, Kaloshin and Wei. Namely, it sta

Georgios Chochlakis, Alexandros Potamianos, Kristina Lerman, Shrikanth Narayanan

In-context Learning (ICL) has become the primary method for performing natural language tasks with Large Language Models (LLMs). The knowledge acquired during pre-training is crucial for this few-shot capability, providing the model with task priors. However, recent studies have shown that ICL predominantly relies on retrieving task priors rather than "learn

Xiao Chen, Li Ma

Recoil correction appears at $O(\frac{1}{M})$, which turns out to be very essential for the hadronic molecules with heavy flavor. In the past, we always thought that the recoil corrections were unfavorable to the formation of the molecular states, but our research reveals its importance to form the di-hadron bound states. In some cases, we are unable to find

N. Kovács, M. A. Maia, I. B. C. M. Rocha, C. Furtado

The growing use of composite materials in engineering applications has accelerated the demand for computational methods to accurately predict their complex behavior. Multiscale modeling based on computational homogenization is a potentially powerful approach for this purpose, but its widespread adoption is prevented by its excessive computational costs. A po

Priyam Ganguly, Isha Mukherjee

In retail sales forecasting, accurately predicting future sales is crucial for inventory management and strategic planning. Traditional methods like LR often fall short due to the complexity of sales data, which includes seasonality and numerous product families. Recent advancements in machine learning (ML) provide more robust alternatives. This research ben

Argyrios Gerogiannis, Yu-Han Huang, Venugopal V. Veeravalli

We study the problem of Non-Stationary Reinforcement Learning (NS-RL) without prior knowledge about the system's non-stationarity. A state-of-the-art, black-box algorithm, known as MASTER, is considered, with a focus on identifying the conditions under which it can achieve its stated goals. Specifically, we prove that MASTER's non-stationarity detection mech

Rachid El Aitouni, Miloud Mekkaoui, Abdelhadi Bahaoui, Ahmed Jellal

Photon-assisted charge transport through a double barrier laser structure, separated by a region assisted by a magnetic field, is studied. Employing Floquet theory and matrix formalism, the transmission probabilities for the central band and sidebands are calculated. The temporal periodicity of the laser fields creates an infinite number of transmission mode

Antonio Sclocchi, Alessandro Favero, Noam Itzhak Levi, Matthieu Wyart

High-dimensional data must be highly structured to be learnable. Although the compositional and hierarchical nature of data is often put forward to explain learnability, quantitative measurements establishing these properties are scarce. Likewise, accessing the latent variables underlying such a data structure remains a challenge. In this work, we show that

Pranav Rajbhandari, Prithviraj Dasgupta, Donald Sofge

We consider the problem of team selection within multiagent adversarial team games. We propose BERTeam, a novel algorithm that uses a transformer-based deep neural network with Masked Language Model training to select the best team of players from a trained population. We integrate this with coevolutionary deep reinforcement learning, which trains a diverse

Alireza Ghafarollahi, Markus J. Buehler

A multi-agent AI model is used to automate the discovery of new metallic alloys, integrating multimodal data and external knowledge including insights from physics via atomistic simulations. Our multi-agent system features three key components: (a) a suite of LLMs responsible for tasks such as reasoning and planning, (b) a group of AI agents with distinct ro

Gabriel Conant

We show that if $G$ is an amenable group and $A\subseteq G$ has positive upper Banach density, then there is an identity neighborhood $B$ in the Bohr topology on $G$ that is almost contained in $AA^{-1}$ in the sense that $B\backslash AA^{-1}$ has upper Banach density $0$. This generalizes the abelian case (due to F{\o}lner) and the countable case (due to Be

Yu Xia, Junda Wu, Sungchul Kim, Tong Yu

Large language models (LLMs) have been used to generate query expansions augmenting original queries for improving information search. Recent studies also explore providing LLMs with initial retrieval results to generate query expansions more grounded to document corpus. However, these methods mostly focus on enhancing textual similarities between search que

Matthew Kirk, Bastian Kubis, Méril Reboud, Danny van Dyk

We discuss a novel and simple parametrisation of the pion vector form factor that transparently connects spacelike and timelike regions of the momentum transfer $q^2$. Our parametrisation employs the framework of conformal mapping and respects the known analyticity properties of the form factor, accounting explicitly for the $\rho(770)$-meson pole. The param

Junyu Luo, Xiao Luo, Xiusi Chen, Zhiping Xiao

Supervised fine-tuning (SFT) is crucial in adapting large language model (LLMs) to a specific domain or task. However, only a limited amount of labeled data is available in practical applications, which poses a severe challenge for SFT in yielding satisfactory results. Therefore, a data-efficient framework that can fully exploit labeled and unlabeled data fo

Arthur Brigatto, Alexandre Street, Cristiano Fernandes, Davi Valladao

Hydroelectricity accounted for roughly 61.4% of Brazil's total generation in 2024 and addressed most of the intermittency of wind and solar generation. Thus, inflow forecasting plays a critical role in the operation, planning, and market in this country, as well as in any other hydro-dependent power system. These forecasts influence generation schedules, res

Raisa Bentay Hossain, Farid Ahmed, Kazuma Kobayashi, Seid Koric

Effective real-time monitoring is a foundation of digital twin technology, crucial for detecting material degradation and maintaining the structural integrity of nuclear systems to ensure both safety and operational efficiency. Traditional physical sensor systems face limitations such as installation challenges, high costs, and difficulty measuring critical

Guibin Zhang, Haonan Dong, Yuchen Zhang, Zhixun Li

Training high-quality deep models necessitates vast amounts of data, resulting in overwhelming computational and memory demands. Recently, data pruning, distillation, and coreset selection have been developed to streamline data volume by retaining, synthesizing, or selecting a small yet informative subset from the full set. Among these methods, data pruning

Lohit Petikam, Charlie Hewitt, Fatemeh Saleh, Tadas Baltrušaitis

Eyelid shape is integral to identity and likeness in human facial modeling. Human eyelids are diverse in appearance with varied skin fold and epicanthal fold morphology between individuals. Existing parametric face models express eyelid shape variation to an extent, but do not preserve sufficient likeness across a diverse range of individuals. We propose a n

Joao F. Doriguello, George Giapitzakis, Alessandro Luongo, Aditya Morolia

One of the main candidates of post-quantum cryptography is lattice-based cryptography. Its cryptographic security against quantum attackers is based on the worst-case hardness of lattice problems like the shortest vector problem (SVP), which asks to find the shortest non-zero vector in an integer lattice. Asymptotic quantum speedups for solving SVP are known

Dingding Dong, Nitya Mani, Huy Tuan Pham, Jonathan Tidor

We study the number of monochromatic solutions to linear equations in a $2$-coloring of $\{1,\ldots,n\}$. We show that any nontrivial linear equation has a constant fraction of solutions that are monochromatic in any $2$-coloring of $\{1,\ldots,n\}$. We further study commonness of four-term equations and disprove a conjecture of Costello and Elvin by showing

Zichen Zhu, Hao Tang, Yansi Li, Dingye Liu

Existing Multimodal Large Language Model (MLLM)-based agents face significant challenges in handling complex GUI (Graphical User Interface) interactions on devices. These challenges arise from the dynamic and structured nature of GUI environments, which integrate text, images, and spatial relationships, as well as the variability in action spaces across diff

Walker Byrnes, Miroslav Bogdanovic, Avi Balakirsky, Stephen Balakirsky

Intelligent and reliable task planning is a core capability for generalized robotics, requiring a descriptive domain representation that sufficiently models all object and state information for the scene. We present CLIMB, a continual learning framework for robot task planning that leverages foundation models and execution feedback to guide domain model cons

Xiaoxiao Cheng, Jonathan Eden, Bastien Berret, Atsushi Takagi

To manipulate objects or dance together, humans and robots exchange energy and haptic information. While the exchange of energy in human-robot interaction has been extensively investigated, the underlying exchange of haptic information is not well understood. Here, we develop a computational model of the mechanical and sensory interactions between agents tha

Jinjie Ni, Yifan Song, Deepanway Ghosal, Bo Li

Perceiving and generating diverse modalities are crucial for AI models to effectively learn from and engage with real-world signals, necessitating reliable evaluations for their development. We identify two major issues in current evaluations: (1) inconsistent standards, shaped by different communities with varying protocols and maturity levels; and (2) sign

Peilin He, Chenkai Lin, Isabella Montoya

In recent years, the financial sector has faced growing pressure to adopt advanced machine learning models to derive valuable insights while preserving data privacy. However, the highly sensitive nature of financial data presents significant challenges to sharing and collaboration. This paper presents DPFedBank, an innovative framework enabling financial ins

Dayeol Lee, Jorge António, Hisham Khan

This paper addresses privacy protection in decentralized Artificial Intelligence (AI) using Confidential Computing (CC) within the Atoma Network, a decentralized AI platform designed for the Web3 domain. Decentralized AI distributes AI services among multiple entities without centralized oversight, fostering transparency and robustness. However, this structu

B. Barbuy, A. C. S. Friaça, H. Ernandes, P. da Silva

Stars presently identified in the bulge spheroid are probably very old, and their abundances can be interpreted as due to the fast chemical enrichment of the early Galactic bulge. The abundances of the iron-peak elements are important tracers of nucleosynthesis processes, in particular oxygen burning, silicon burning, the weak s-process, and alpha-rich freez

Shan Tai Chan

We study holomorphic isometries between bounded symmetric domains with respect to the Bergman metrics up to a normalizing constant. In particular, we first consider a holomorphic isometry from the complex unit ball into an irreducible bounded symmetric domain with respect to the Bergman metrics. In this direction, we show that images of (nonempty) affine-lin

Albert Gong, Kyuseong Choi, Raaz Dwivedi

The kernel thinning algorithm of Dwivedi & Mackey (2024) provides a better-than-i.i.d. compression of a generic set of points. By generating high-fidelity coresets of size significantly smaller than the input points, KT is known to speed up unsupervised tasks like Monte Carlo integration, uncertainty quantification, and non-parametric hypothesis testing, wit

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

Lepton flavour universality in rare $b\rightarrow s$ transitions is tested for the first time using $B_s^0$ meson decays. The measurements are performed using $pp$ collision data collected by the LHCb experiment between 2011 and 2018, corresponding to a total integrated luminosity of 9$\,{\rm fb}^{-1}$. Branching fraction ratios between the $B_s^0 \rightarro

Matthew Grayling, Brodie Popovic

In recent years there has been significant debate around the impact of dust on SNe Ia, a major source of uncertainty in cosmological analyses. We perform the first validation of the probabilistic hierarchical SN Ia SED model BayeSN on the conventional SALT model, an important test given the history of conflicting conclusions regarding the distributions of ho

Yuchen Liang, Peizhong Ju, Yingbin Liang, Ness Shroff

The denoising diffusion model has recently emerged as a powerful generative technique, capable of transforming noise into meaningful data. While theoretical convergence guarantees for diffusion models are well established when the target distribution aligns with the training distribution, practical scenarios often present mismatches. One common case is in th

D. A. Baiko

Thermal evolution of the central region of a $0.9 \, M_\odot$ C/O white dwarf at the initial stage of the ion mixture crystallization is studied by numerically solving the heat equation on a fine spatial and temporal grid and by including a detailed treatment of the latent heat release. Formation of two spherical shells is observed. The outer one surrounds a

Matteo Framba, Veronica Vinciotti, Ernst C. Wit

In the modelling of stochastic phenomena, such as quasi-reaction systems, parameter estimation of kinetic rates can be challenging, particularly when the time gap between consecutive measurements is large. Local linear approximation approaches account for the stochasticity in the system but fail to capture the nonlinear nature of the underlying process. At t

Yue Huang, Zhaoxian Wu, Shiqian Ma, Qing Ling

Stochastic approximation (SA) that involves multiple coupled sequences, known as multiple-sequence SA (MSSA), finds diverse applications in the fields of signal processing and machine learning. However, existing theoretical understandings {of} MSSA are limited: the multi-timescale analysis implies a slow convergence rate, whereas the single-timescale analysi

En Chen, Xi Chen, Xuepeng Chen, Min Fang

We report that the gas components in the N59 bubble suffered from sequential multiple cloud-cloud collision (CCC) processes. The molecular gas in the N59 bubble can be decomposed into four velocity components, namely Cloud A [95, 108] km s$^{-1}$, Cloud B [86, 95] km s$^{-1}$, Cloud C [79, 86] km s$^{-1}$ and Cloud D [65, 79] km s$^{-1}$. Four CCC processes

Cedric Luger

We prove that the weak Hilbert property ascends along a morphism of varieties over an arbitrary field of characteristic zero, under suitable assumptions.

Arnaud Rémi, François Damanet, Christophe Geuzaine

Solving Helmholtz problems using finite elements leads to the resolution of a linear system which is challenging to solve for classical computers. In this paper, we investigate how quantum annealers could address this challenge. We first express the linear system arising from the Helmholtz problem as a generalized eigenvalue problem (gEVP). The obtained gEVP

Yasmín Navarrete, Carlos Femenías, Sergio Davis, Claudia Loyola

The formation of large social groups having uniform opinions influenced by mass media is currently an important topic in the social sciences. In this work, we explore and extend an off-lattice, two-dimensional Potts model (Eur. Phys. J. B 87, 78 [2014]) that describes the formation and dynamics of opinions in social groups according to individual consequence

Yuchen Jiao, Gen Li

Score-based diffusion models have demonstrated outstanding empirical performance in machine learning and artificial intelligence, particularly in generating high-quality new samples from complex probability distributions. Improving the theoretical understanding of diffusion models, with a particular focus on the convergence analysis, has attracted significan

Daniel Cortild, Claire Delplancke, Nadia Oudjane, Juan Peypouquet

We present an optimization algorithm that can identify a global minimum of a potentially nonconvex smooth function with high probability, assuming the Gibbs measure of the potential satisfies a logarithmic Sobolev inequality. Our contribution is twofold: on the one hand we propose a global optimization method, which is built on an oracle sampling algorithm p

Megan Fairchild

Whitehead doubles provide a plethora of examples of knots that are topologically slice but not smoothly slice. We discuss the problem of the Whitehead double of the Figure 8 knot and survey commonly used techniques to obstructing sliceness. Additionally, we improve bounds in general for the non-orientable 4 genus of $t$-twisted Whitehead doubles and provide

Minxing Zheng, Shixiang Zhu

Conformal prediction (CP) provides model-agnostic uncertainty quantification with guaranteed coverage, but conventional methods often produce overly conservative uncertainty sets, especially in multi-dimensional settings. This limitation arises from simplistic non-conformity scores that rely solely on prediction error, failing to capture the prediction error

Yuchen Guo, Shuo Yang

Recent studies have unveiled new possibilities for discovering intrinsic quantum phases that are unique to open systems, including phases with average symmetry-protected topological (ASPT) order and strong-to-weak spontaneous symmetry breaking (SWSSB) order in systems with global symmetry. In this work, we propose a new class of phases, termed the double ASP

Yanpeng Sun, Huaxin Zhang, Qiang Chen, Xinyu Zhang

We focus on improving the visual understanding capability for boosting the vision-language models. We propose \textbf{Arcana}, a multiModal language model, which introduces two crucial techniques. First, we present Multimodal LoRA (MM-LoRA), a module designed to enhance the decoder. Unlike traditional language-driven decoders, MM-LoRA consists of two paralle

Bernhard Bermeitinger, Tomas Hrycej, Massimo Pavone, Julianus Kath

Transformers are a widespread and successful model architecture, particularly in Natural Language Processing (NLP) and Computer Vision (CV). The essential innovation of this architecture is the Attention Mechanism, which solves the problem of extracting relevant context information from long sequences in NLP and realistic scenes in CV. A classical neural net

Manika Bag, Sheetal Dharmatti, Manil T Mohan

This paper presents a mathematical analysis of the evolution of a mixture of two incompressible, isothermal fluids flowing through a porous medium in a three dimensional bounded domain. The model is governed by a coupled system of convective Brinkman Forchheimer equations and the Cahn Hilliard equation, considering a regular potential and non degenerate mobi

Helmut Gfrerer, Simon Hubmer, Ronny Ramlau

We consider the efficient numerical minimization of Tikhonov functionals with nonlinear operators and non-smooth and non-convex penalty terms, which appear for example in variational regularization. For this, we consider a new class of SCD semismooth$^*$ Newton methods, which are based on a novel concept of graphical derivatives, and exhibit locally superlin

Pedro A. Garcia-Sanchez

Every numerical semigroup can be expressed as an intersection of irreducible numerical semigroups. We show that the unions of sets of lengths of factorizations of numerical semigroups into irreducible numerical semigroups are all equal to $\mathbb{N}_{\ge 2}$.

Alessio Cipriani, Jon Woolf

Highest weight categories are an abstraction of the representation theory of semisimple Lie algebras introduced by Cline, Parshall and Scott in the late 1980s. There are by now many characterisations of when an abelian category is highest weight, but most are hard to verify in practice. We present two new criteria - one numerical in terms of the Grothendieck

Rajkumar Pujari, Dan Goldwasser

Conversations often adhere to well-understood social norms that vary across cultures. For example, while "addressing parents by name" is commonplace in the West, it is rare in most Asian cultures. Adherence or violation of such norms often dictates the tenor of conversations. Humans are able to navigate social situations requiring cultural awareness quite ad

Hanbo Cheng, Limin Lin, Chenyu Liu, Pengcheng Xia

Talking head generation intends to produce vivid and realistic talking head videos from a single portrait and speech audio clip. Although significant progress has been made in diffusion-based talking head generation, almost all methods rely on autoregressive strategies, which suffer from limited context utilization beyond the current generation step, error a

E. Gerard, W. van Driel, L. D. Matthews, T. Le Bertre

We present a compendium of HI 21-cm line observations of circumstellar envelopes (CSEs) of 290 evolved stars, mostly (~84%) on the asymptotic giant branch (AGB), made with the 100m-class, single-dish Nancay Radio Telescope. The observational and data reduction procedures were optimised for separating genuine CSE HI emission from surrounding Galactic line fea

Shirin Panahi, Li-Li Ye, Ying-Cheng Lai

Exceptional points, a remarkable phenomenon in physical systems, have been exploited for sensing applications. It has been demonstrated recently that it can also utilize as sensory threshold in which the interplay between exceptional-point dynamics and noise can lead to enhanced performance. Most existing works focused on second-order exceptional points. We

Sixtus Dakurah, Jessi Cisewski-Kehe

A time-delay embedding (TDE), grounded in the framework of Takens's Theorem, provides a mechanism to represent and analyze the inherent dynamics of time-series data. Recently, topological data analysis (TDA) methods have been applied to study this time series representation mainly through the lens of persistent homology. Current literature on the fusion of T

Yiming Zhang, Javier Rando, Ivan Evtimov, Jianfeng Chi

Large language models are pre-trained on uncurated text datasets consisting of trillions of tokens scraped from the Web. Prior work has shown that: (1) web-scraped pre-training datasets can be practically poisoned by malicious actors; and (2) adversaries can compromise language models after poisoning fine-tuning datasets. Our work evaluates for the first tim

M. A. Wolfe, Thomas McJunkin, Daniel R. Ward, DeAnna Campbell

The challenges of operating qubits in a cryogenic environment point to a looming bottleneck for large-scale quantum processors, limited by the number of input-output connections. Classical processors solve this problem via multiplexing; however, on-chip multiplexing circuits have not been shown to have similar benefits for cryogenic quantum devices. In this

Adam Polyak, Amit Zohar, Andrew Brown, Andros Tjandra

We present Movie Gen, a cast of foundation models that generates high-quality, 1080p HD videos with different aspect ratios and synchronized audio. We also show additional capabilities such as precise instruction-based video editing and generation of personalized videos based on a user's image. Our models set a new state-of-the-art on multiple tasks: text-to

Cristián Erices, Luis Guajardo, Kristiansen Lara

We study a class of solutions within the context of modified gravity theories, characterized by a non-trivial field that does not generate any back-reaction on the metric. These stealth configurations are effectively defined by the stealth conditions, which correspond to a vanishing stress-energy tensor. In this work, we introduce a novel approach to constru

Alain R. Ndjiongue, Octavia A. Dobre, Hyundong Shin

Given the importance of reconfigurable intelligent surfaces (RISs) in next-generation mobile systems, several RIS variants have been proposed in recent years. Omni-digital-RIS (omni-DRIS) is one of the newly introduced variants of optical RIS that can successfully be driven by bit sequences to control lights emerging from simultaneous reflection and refracti

Lanke Fu, E. Litvinova Mitra, R. Nies, A. H. Reiman

Theoretical studies have suggested that nonlinear effects can lead to "radio frequency condensation", which coalesces RF power deposition and driven current near the center of a magnetic island. It is predicted that an initially broad current profile can coalesce in islands when they reach sufficient width, providing automatic stabilization. Experimental val

Nandan Thakur, Suleman Kazi, Ge Luo, Jimmy Lin

Traditional retrieval-augmented generation (RAG) benchmarks evaluate systems using heuristic-based metrics, but these require human preferences as the ground truth for reference. In contrast, arena-based benchmarks, where systems compete against each other, require an expensive large language model (LLM) as a judge for a reliable evaluation. We present a sim

Hidemaro Suwa, Shang-Shun Zhang, Cristian D. Batista

The excitonic insulator, a quantum mechanical state arising from exciton condensation, was proposed theoretically many years ago but has yet to be experimentally confirmed. The discovery of correlated transition metal oxides based on $4d$ and $5d$ elements, where the on-site Coulomb repulsion is comparable to the dominant hopping amplitude, presents a unique

Jiaxun Li, Vinod Raman, Ambuj Tewari

We study generation through the lens of statistical learning theory. First, we abstract and formalize the results of Gold [1967], Angluin [1979], Angluin [1980] and Kleinberg and Mullainathan [2024] in terms of a binary hypothesis class defined over an abstract example space. Then, we extend the notion of "generation" from Kleinberg and Mullainathan [2024] t

Kaipeng Zheng, Weiran Huang, Wanli Ouyang, Han-Sen Zhong

Atomic structure analysis of crystalline materials is a paramount endeavor in both chemical and material sciences. This sophisticated technique necessitates not only a solid foundation in crystallography but also a profound comprehension of the intricacies of the accompanying software, posing a significant challenge in meeting the rigorous daily demands. For

Aurora Ireland, Gordan Krnjaic, Takuya Okawa

The Standard Model of particle physics does not explain the origin of the universe's baryon asymmetry or its primordial fluctuations. The Affleck-Dine mechanism is a well-motivated scenario for generating the baryon asymmetry through the post-inflationary dynamics of a complex scalar field with baryon number. The curvaton mechanism is a popular approach for

Angéline Lafleur, Soo-hyon Phark

Leveraging scanning tunneling microscopy (STM) for atomic-scale fabrication has led to many advancements such as the creation of atomic electron-spin qubit structures on surfaces. However, the time-consuming and tedious nature of this process calls for improvements, and this study explores the use of machine learning (ML) to automate certain steps, notably i

Michael McCreesh, Erfan Nozari, Jorge Cortes

Over the past two decades, an increasing array of control-theoretic methods have been used to study the brain as a complex dynamical system and better understand its structure-function relationship. This article provides an overview on one such family of methods, based on the linear-threshold rate (LTR) dynamics, which arises when modeling the spiking activi

Mustofa Ahmed, Abdul Muntakim, Nawrin Tabassum, Mohammad Asifur Rahim

Depression detection using deep learning models has been widely explored in previous studies, especially due to the large amounts of data available from social media posts. These posts provide valuable information about individuals' mental health conditions and can be leveraged to train models and identify patterns in the data. However, distributed learning

Zhenhong Zhou, Haiyang Yu, Xinghua Zhang, Rongwu Xu

Large language models (LLMs) achieve state-of-the-art performance on multiple language tasks, yet their safety guardrails can be circumvented, leading to harmful generations. In light of this, recent research on safety mechanisms has emerged, revealing that when safety representations or component are suppressed, the safety capability of LLMs are compromised

Ege Atacan Doğan, Peter F. Patel-Schneider

Disjointness checks are among the most important constraint checks in a knowledge base and can be used to help detect and correct incorrect statements and internal contradictions. Wikidata is a very large, community-managed knowledge base. Because of both its size and construction, Wikidata contains many incorrect statements and internal contradictions. We a

Ronja Bärwolf, Alexander Sushchyev, Francesco Parisen Toldin, Stefan Wessel

Using a combination of unbiased quantum Monte Carlo simulations and a decoupled dimer mean-field theory, we investigate the thermal and quantum phase transitions of the spin-1/2 Heisenberg model on the dimerized diamond lattice. We find that at sufficiently strong dimerization the system exhibits a quantum disordered ground state, in contrast to the antiferr

Weizhen Tang, Yating Zheng, Amir Shee, Guozheng Lin

We study the emerging collective states in a simple mechanical model of a dense group of self-propelled polar disks with off-centered rotation, confined within a circular arena. Each disk presents self-alignment towards the sum of contact forces acting on it, resulting from disk-substrate interactions, while also displaying mutual alignment with neighbors du

Baptiste Alperin, Ilya Golokolenov, Gwénaëlle Julié, Bruno Fernandez

Mechanical objects have been widely used at low temperatures for decades, for various applications; from quantum fluids sensing with vibrating wires or tuning forks, to torsional oscillators for the study of mechanical properties of glasses, and finally micro and nano-mechanical objects with the advent of clean room technologies. These small structures opene

Toan T. Nguyen

In plasma physics, collisionless charged particles are transported following the dynamics of a meanfield Vlasov equation with a self-consistent electric field generated by the charge density. Due to the long range interaction between particles, the generating electric field oscillates and disperses like a Klein-Gordon dispersive wave, known in the physical l

Adnan A. E. Hajomer, Florian Kanitschar, Nitin Jain, Michael Hentschel

Establishing secure data communication necessitates secure key exchange over a public channel. Quantum key distribution (QKD), which leverages the principles of quantum physics, can achieve this with information-theoretic security. The discrete modulated (DM) continuous variable (CV) QKD protocol, in particular, is a suitable candidate for large-scale deploy

Edward McDonald

We prove that order zero operators in the pseudodifferential calculus associated to a filtration defined by Androulidakis, Mohsen and Yuncken are bounded on $L_p$ spaces for $1<p<\infty.$

Aditi Saxena, Twinkle Tripathy, Rajasekhar Anguluri

In this paper, we explore the property of eventual exponential positivity (EEP) in complex matrices. We show that this property holds for the real part of the matrix exponential for a certain class of complex matrices. Next, we present the relation between the spectral properties of the Laplacian matrix of an unsigned digraph with complex edge-weights and th

Yiming Zhang, Baoyi He, Shengyu Zhang, Yuhao Fu

Recent advancements in building domain-specific large language models (LLMs) have shown remarkable success, especially in tasks requiring reasoning abilities like logical inference over complex relationships and multi-step problem solving. However, creating a powerful all-in-one LLM remains challenging due to the need for proprietary data and vast computatio

Oliver Boine-Frankenheim, Thilo Egenolf

For a single hadron bunch affected by longitudinal space charge in a stationary rf bucket we analyze the frequency spectrum close to the expected loss of Landau damping for the lowest order dipole mode. For different bunch intensity parameters we obtain the bunch oscillation spectrum from a conventional longitudinal particle tracking code with a grid-based s

Davor Dragicevic, Cesar M. Silva, Helder Vilarinho

In this paper, we introduce generalized dichotomies for nonautonomous random linear dynamical systems acting on arbitrary Banach spaces, and obtain their complete characterization in terms of an appropriate admissibility property. These generalized dichotomies are associated to growth rates satisfying mild conditions and they include the standard exponential

Wei Huang, Richard Combes, Andrea Araldo, Hind Castel-Taleb

We study the placement of virtual functions aimed at minimizing the cost. We propose a novel algorithm, using ideas based on multi-armed bandits. We prove that these algorithms learn the optimal placement policy rapidly, and their regret grows at a rate at most $O( N M \sqrt{T\ln T} )$ while respecting the feasibility constraints with high probability, where

Mervyn Tong

Since K\H{o}v\'ari, S\'os, and Tur\'an proved upper bounds for the Zarankiewicz problem in 1954, much work has been undertaken to improve these bounds, and some have done so by restricting to particular classes of graphs. In 2017, Fox, Pach, Sheffer, Suk, and Zahl proved better bounds for semialgebraic binary relations, and this work was extended by Do in th

Yifan Du, Yuqi Huo, Kun Zhou, Zijia Zhao

Video Multimodal Large Language Models (MLLMs) have shown remarkable capability of understanding the video semantics on various downstream tasks. Despite the advancements, there is still a lack of systematic research on visual context representation, which refers to the scheme to select frames from a video and further select the tokens from a frame. In this

Dingjia Cao, Marina I. Knight, Guy P. Nason

Data collected over networks can be modelled as noisy observations of an unknown function over the nodes of a graph or network structure, fully described by its nodes and their connections, the edges. In this context, function estimation has been proposed in the literature and typically makes use of the network topology such as relative node arrangement, oft

Cristian Maria-Moreno, Ignacio Mateos, Guillermo Pacheco-Ramos, Francisco Rivas

In recent years, nanosatellites have revolutionized the space sector due to their significant economic and time-saving advantages. As a result, they have fostered the testing of advanced instruments intended for larger space science missions. The case of MELISA is presented in this work. MELISA is a magnetic measurement instrument which aims at demonstrating

Alexander Robey, Zachary Ravichandran, Vijay Kumar, Hamed Hassani

The recent introduction of large language models (LLMs) has revolutionized the field of robotics by enabling contextual reasoning and intuitive human-robot interaction in domains as varied as manipulation, locomotion, and self-driving vehicles. When viewed as a stand-alone technology, LLMs are known to be vulnerable to jailbreaking attacks, wherein malicious

Inês Aniceto, Samuel Crew

We investigate geometric aspects of co-equational parametric resurgence, by studying physical problems whose formal asymptotic solutions give rise to Borel transforms lying on an algebraic curve. This perspective allows us to elucidate concepts unique to parametric resurgence such as singularity structures, (virtual) turning points and the higher-order Stoke

Henryk Fukś, José Manuel Gómez Soto

We propose a simple cellular automaton model of a self-healing system and investigate its properties. In the model, the substrate is a two-dimensional checkerboard configuration which can be damaged by changing values of a finite number of sites. The cellular automaton we consider is a checkerboard voting rule, a binary rule with Moore neighbourhood which is

Wenbo Fan, Xiaotian Yan, Zhanbo Sun, Xiaohui Yang

Ride-pooling (RP) service, as a form of shared mobility, enables multiple riders with similar itineraries to share the same vehicle and split the fee. This makes RP a promising on-demand feeder service for patrons with a common trip end in urban transportation. We propose the RP as Feeder (RPaF) services with tailored operational algorithms. Specifically, we

ZiDong Wang, Zeyu Lu, Di Huang, Cai Zhou

\textit{Nature is infinitely resolution-free}. In the context of this reality, existing diffusion models, such as Diffusion Transformers, often face challenges when processing image resolutions outside of their trained domain. To address this limitation, we conceptualize images as sequences of tokens with dynamic sizes, rather than traditional methods that p

Abeynaya Gnanasekaran, Amit Surana, Hongyu Zhu

We present a novel variational quantum framework for nonlinear partial differential equation (PDE) constrained optimization problems. The proposed work extends the recently introduced bi-level variational quantum PDE constrained optimization (BVQPCO) framework for linear PDE to a nonlinear setting by leveraging Carleman linearization (CL). CL framework allow

Antonio Alarcon

We survey the recent history of the conformal Calabi-Yau problem consisting in determining the complex structures admitted by complete bounded minimal surfaces in $\mathbb{R}^3$. Moreover, we prove that for any minimally convex domain $\Omega$ in $\mathbb{R}^3$ and any compact Riemann surface $R$ there is a Cantor set $C$ in $R$ whose complement $R\setminus

Adam Kanigowski, Davide Ravotti

The famous Rokhlin Problem asks whether mixing implies higher order mixing. So far, all the known examples of zero entropy, mixing dynamical systems enjoy a variant of the mixing via shearing mechanism. In this paper we introduce the notion of locally uniformly shearing systems (LUS) which is a rigorous way of describing the mixing via shearing mechanism. We

Sania Sinha, Aarham Wasit, Won Seob Kim, Jongkyoo Kim

Assessing the quality of bovine satellite cells (BSCs) is essential for the cultivated meat industry, which aims to address global food sustainability challenges. This study aims to develop a label-free method for predicting fluorescence markers in isolated BSCs using deep learning. We employed a U-Net-based CNN model to predict multiple fluorescence signals

Francesco Toschi, Axel Brunold, Lea Burmeister, Klaus Eitel

Superfluid ${}^4$He is an ideal candidate for the direct detection of light dark matter via nuclear recoils thanks to its low nuclear mass and the possibility to reach a low detection energy threshold by exploiting the generated quasiparticles. The design of future detectors based on this target, such as the DELight experiment, requires a proper understandin

Joseph H. P. Jackson, Hooshyar Assadullahi, Andrew D. Gow, Kazuya Koyama

We simulate the distribution of very rare, large excursions in the primordial density field produced in models of inflation in the very early universe which include a strong enhancement of the power spectrum. The stochastic $\delta \mathcal{N}$ formalism is used to identify the probability distribution for the primordial curvature perturbation with the first