May 2024 arXiv papers — page 69

Lena Schmidt, Kaitlyn Hair, Sergio Graziosi, Fiona Campbell

This paper describes a rapid feasibility study of using GPT-4, a large language model (LLM), to (semi)automate data extraction in systematic reviews. Despite the recent surge of interest in LLMs there is still a lack of understanding of how to design LLM-based automation tools and how to robustly evaluate their performance. During the 2023 Evidence Synthesis

Yitong Yang, Xinli Xu, Haigen Hu, Haixia Long

Despite the recent progress in medical image segmentation with scribble-based annotations, the segmentation results of most models are still not ro-bust and generalizable enough in open environments. Evidential deep learn-ing (EDL) has recently been proposed as a promising solution to model predictive uncertainty and improve the reliability of medical image

John E. Bravo, Jean C. Cortissoz

We explore Liouville's theorem and the Strong Liouville Property (SLP) for harmonic functions on Riemannian cones and surfaces. Our approach recasts the classical Liouville property in terms of the growth of radial eigenfunctions (in the case of manifolds with rotational symmetry), allowing us to recover and sharpen known results under minimal assumptions. W

Dyk Chung Nguyen, Yuriy V. Pershin

We present a fully parallel digital memcomputing solver implemented on a field-programmable gate array (FPGA) board. For this purpose, we have designed an FPGA code that solves the ordinary differential equations associated with digital memcomputing in parallel. A feature of the code is the use of only integer-type variables and integer constants to enhance

Alexandre Girard, H. Harry Asada

This paper presents a robotic system where the gear-ratio of an actuator is dynamically changed to either leverage or attenuate the natural load dynamics. Based on this principle, lightweight robotic systems can be made fast and strong; exploiting the natural load dynamics for moving at higher speeds (small reduction ratio), while also able to bear a large l

Rafael Oliveira, Dino Sejdinovic, David Howard, Edwin V. Bonilla

The process of calibrating computer models of natural phenomena is essential for applications in the physical sciences, where plenty of domain knowledge can be embedded into simulations and then calibrated against real observations. Current machine learning approaches, however, mostly rely on rerunning simulations over a fixed set of designs available in the

Gonçalo Frazao, Marco Pezzutto, Yasser Omar, Emmanuel Zambrini Cruzeiro

The metrological limits of thermometry operated in nonequilibrium dynamical regimes are analyzed. We consider a finite-dimensional quantum system, employed as a quantum thermometer, in contact with a thermal bath inducing Markovian thermalization dynamics. The quantum thermometer is initialized in a generic quantum state, possibly including quantum coherence

Dominik J. Mühlematter, Michelle Halbheer, Alexander Becker, Dominik Narnhofer

Numerous real-world decisions rely on machine learning algorithms and require calibrated uncertainty estimates. However, modern methods often yield overconfident, uncalibrated predictions. The dominant approach to quantifying the uncertainty inherent in the model is to train an ensemble of separate predictors and measure their empirical variance. In an expli

Alejo Lopez-Avila, Víctor Suárez-Paniagua

Recently, using large pretrained Transformer models for transfer learning tasks has evolved to the point where they have become one of the flagship trends in the Natural Language Processing (NLP) community, giving rise to various outlooks such as prompt-based, adapters or combinations with unsupervised approaches, among many others. This work proposes a 3 Ph

Mohamed Mejri, Chandramouli Amarnath, Abhijit Chatterjee

Human cognition excels at symbolic reasoning, deducing abstract rules from limited samples. This has been explained using symbolic and connectionist approaches, inspiring the development of a neuro-symbolic architecture that combines both paradigms. In parallel, recent studies have proposed the use of a "relational bottleneck" that separates object-level fea

Kehui Zhang, Lingfeng Li, Hao Liu, Jing Yuan

Shape compactness is a key geometrical property to describe interesting regions in many image segmentation tasks. In this paper, we propose two novel algorithms to solve the introduced image segmentation problem that incorporates a shape-compactness prior. Existing algorithms for such a problem often suffer from computational inefficiency, difficulty in reac

Bianka Bakullari, Wil M. P. van der Aalst

Process mining traditionally relies on input consisting of low-level events that capture individual activities, such as filling out a form or processing a product. However, many of the complex problems inherent in processes, such as bottlenecks and compliance issues, extend beyond the scope of individual events and process instances. Consider congestion, for

Simon R. G. Joyce, John P. Pye, Jonathan D. Nichols, Richard Alexander

Planet formation takes place in protoplanetary discs around young T-Tauri stars. PDS 70 is one of the first confirmed examples of a system where the planets are currently forming in gaps in the disc, and can be directly imaged. One of the main early influences on planet formation is the lifetime of the protoplanetary disk, which is limited by the intense ste

Boulsane Mourad

Since the early 1960s, the fields of signal processing, data transmission, channel equalisation, filter design and others have been technologically developed and modernised as a result of the research carried out by D. Slepian and his co-authors H. J Landau and H. O Pollack on the time and band-limited wave system known as discrete and continuous spheroidal

Youssef Allouah, Rachid Guerraoui, Nirupam Gupta, Ahmed Jellouli

Robust federated learning aims to maintain reliable performance despite the presence of adversarial or misbehaving workers. While state-of-the-art (SOTA) robust distributed gradient descent (Robust-DGD) methods were proven theoretically optimal, their empirical success has often relied on pre-aggregation gradient clipping. However, existing static clipping s

Shengyu Mao, Yong Jiang, Boli Chen, Xiao Li

As Large Language Models (LLMs) and Retrieval Augmentation Generation (RAG) techniques have evolved, query rewriting has been widely incorporated into the RAG system for downstream tasks like open-domain QA. Many works have attempted to utilize small models with reinforcement learning rather than costly LLMs to improve query rewriting. However, current metho

Jiarui Fang, Jinzhe Pan, Aoyu Li, Xibo Sun

This paper presents PipeFusion, an innovative parallel methodology to tackle the high latency issues associated with generating high-resolution images using diffusion transformers (DiTs) models. PipeFusion partitions images into patches and the model layers across multiple GPUs. It employs a patch-level pipeline parallel strategy to orchestrate communication

Friedrich Philipp, Manuel Schaller, Karl Worthmann, Sebastian Peitz

We consider the Koopman operator semigroup $(K^t)_{t\ge 0}$ associated with stochastic differential equations of the form $dX_t = AX_t\,dt + B\,dW_t$ with constant matrices $A$ and $B$ and Brownian motion $W_t$. We prove that the reproducing kernel Hilbert space $\bH_C$ generated by a Gaussian kernel with a positive definite covariance matrix $C$ is invarian

Jaewoo Yang, Hayun Kim, Younghoon Kim

Modern large language models (LLMs) have established state-of-the-art performance through architectural improvements, but still require significant computational cost for inference. In an effort to reduce the inference cost, post-training quantization (PTQ) has become a popular approach, quantizing weights and activations to lower precision, such as INT8. In

Michael Schneeberger, Florian Dörfler, Silvia Mastellone

In this paper, we implement an advanced safety filter to smoothly limit the current of an inverter-based Battery Energy Storage System. The task involves finding suitable Control Barrier Function and Control Lyapunov Function via Sum-of-Squares optimization to certify the system's safety during grid transients. In contrast to the conventional safety filter,

Andrea Iannelli, Romain Postoyan

We consider the data-driven stabilization of discrete-time linear time-varying systems. The controller is defined as a linear state-feedback law whose gain is adapted to the plant changes through a data-based event-triggering rule. To do so, we monitor the evolution of a data-based Lyapunov function along the solution. When this Lyapunov function does not sa

Kabir Dabholkar, Omri Barak

Latent variable models serve as powerful tools to infer underlying dynamics from observed neural activity. Ideally, the inferred dynamics should align with true ones. However, due to the absence of ground truth data, prediction benchmarks are often employed as proxies. One widely-used method, $\textit{co-smoothing}$, involves jointly estimating latent variab

Maria Grazia Izzo, John Russo, Giorgio Pastore

The distinctive characteristics of water, evident in its thermodynamic anomalies, have implications across disciplines from biology to geophysics. Considered a valid hypothesis to rationalize its unique properties, a liquid-liquid phase transition in water's supercooled regime has nowadays been observed in several molecular dynamics simulations and is being

Athanasios Beslikas

In this note we give a new sufficient condition for the boundedness of the composition operator on the Dirichlet-type space on the disc, via a two dimensional change of variables formula. With the same formula, we characterise the bounded composition operators on the anisotropic Dirichlet-type spaces $\mathfrak{D}_{\vec{a}}(\mathbb{D}^2)$ induced by holomorp

Pouria Saidi, Gautam Dasarathy, Visar Berisha

Machine Learning (ML) is increasingly used across many disciplines with impressive reported results. However, recent studies suggest published performance of ML models are often overoptimistic. Validity concerns are underscored by findings of an inverse relationship between sample size and reported accuracy in published ML models, contrasting with the theory

Stavros Mougiakakos, Pierre Vanhove

We apply a formulation of Einstein's general relativity with only cubic interactions for deriving the metric of a Schwarzschild black hole to all orders in perturbation theory. This cubic interactions formulation coupled to effective worldline action of a massive point particle allows to derive a recursion relation for the form factors of the off-shell gravi

Vitalie Lungu, Marina-Aura Dariescu, Cristian Stelea

A new exact magnetized solution describing a Kiselev black hole immersed in a magnetic field is used for studying the dynamics of charged particles. Different types of trajectories are discussed. In the particular case of a weak magnetic field, we employ a first-order perturbative approach to analyze the perturbed circular orbits near the minimum of the effe

Malika Nisal Ratnayake, Lex Gallon, Adel N. Toosi, Alan Dorin

Field-captured video facilitates detailed studies of spatio-temporal aspects of animal locomotion, decision-making and environmental interactions including predator-prey relationships and habitat utilisation. But even though data capture is cheap with mass-produced hardware, storage, processing and transmission overheads provide a hurdle to acquisition of hi

Michail Anthropelos, Constantinos Stefanakis

We consider an Ito-financial market at which the risky assets' returns are derived endogenously through a market-clearing condition amongst heterogeneous risk-averse investors with quadratic preferences and random endowments. Investors act strategically by taking into account the impact that their orders have on the assets' drift. A frictionless market and a

C. Santamarina Ríos, P. Rodríguez Cacheda, J. J. Saborido Silva

The hydrogen atom perturbed by a constant 1-dimensional weak quadratic potential $\lambda z^2$ is solved at first-order perturbation theory using the eigenstates of the total angular momentum operator - the coupled basis. Physical applications of this result could be found, for example, in the study of a quadratic Zeeman effect weaker than fine-structure eff

P. Nieves, I. Korniienko, A. Fraile, J. M. Fernández

We present VelCrys, a web-based interactive tool, that allows to perform further post-processing of the elastic tensor in order to compute and plot the group velocity of the acoustic waves for any crystal symmetry. We also implemented the calculation of effective magnetic corrections to the elastic tensor and corresponding fractional change in group velocity

P. Korcyl, L. Motyka, T. Stebel

The most complete high-energy evolution of Wilson line operators is described by the set of equations called Balitsky-JIMWLK evolution equations. It is known from the studies of the linear - the BFKL - evolution equation that the leading corrections come from the kinematically enhanced double collinear logarithms. A method for resumming such logarithmic corr

Haiming Wang, Huajian Xin, Zhengying Liu, Wenda Li

Recent advances in automated theorem proving leverages language models to explore expanded search spaces by step-by-step proof generation. However, such approaches are usually based on short-sighted heuristics (e.g., log probability or value function scores) that potentially lead to suboptimal or even distracting subgoals, preventing us from finding longer p

Mohammadreza Malekabbasi, Tobias Pfandzelter, Trever Schirmer, David Bermbach

The massive growth of mobile and IoT devices demands geographically distributed computing systems for optimal performance, privacy, and scalability. However, existing edge-to-cloud serverless platforms lack location awareness, resulting in inefficient network usage and increased latency. In this paper, we propose GeoFaaS, a novel edge-to-cloud Function-as-a-

Swati Dahiya, Narayan Sharma, Shivani Geete, Sarveshwar Sharma

In this study, we have investigated the production of negative ions by mixing electronegative oxygen gas with electropositive argon gas in a geometrically asymmetric cylindrical capacitively coupled radio frequency plasma discharge. The plasma parameters such as density (electron, positive and negative ion), negative ion fraction, and electron temperature ar

Nan Zhang, Christian Vergara-Marcillo, Georgios Diamantopoulos, Jingran Shen

Dynamic data-driven Digital Twins (DDTs) can enable informed decision-making and provide an optimisation platform for the underlying system. By leveraging principles of Dynamic Data-Driven Applications Systems (DDDAS), DDTs can formulate computational modalities for feedback loops, model updates and decision-making, including autonomous ones. However, unders

Kunal Pal, Kuntal Pal

We obtain the complexity geometry associated with the Hamiltonian of a quantum mechanical system, specifically in cases where the Hamiltonian is explicitly time-dependent. Using Nielsen's geometric formulation of circuit complexity, we calculate the bi-invariant cost associated with these time-dependent Hamiltonians by suitably regularising their norms and o

Moises Diaz, Miguel A. Ferrer, Soodamani Ramalingam, Richard Guest

In automatic signature verification, questioned specimens are usually compared with reference signatures. In writer-dependent schemes, a number of reference signatures are required to build up the individual signer model while a writer-independent system requires a set of reference signatures from several signers to develop the model of the system. This pape

Shunnosuke Ikeda, Naoki Nishimura, Shunji Umetani

This study addresses the interpretable estimation of price bounds in the context of price optimization. In recent years, price-optimization methods have become indispensable for maximizing revenue and profits. However, effective application of these methods to real-world pricing operations remains a significant challenge. It is crucial for operators responsi

Iris Rammelmüller, Gottfried Hastermann, Jana de Wiljes

Data assimilation algorithms integrate prior information from numerical model simulations with observed data. Ensemble-based filters, regarded as state-of-the-art, are widely employed for large-scale estimation tasks in disciplines such as geoscience and meteorology. Despite their inability to produce the true posterior distribution for nonlinear systems, th

Avisek Bist, Namita Behera

Consider a multivariable state space system and associated transfer function G({\lambda}). The aim of this paper is to define and analyze two vector spaces of matrix pencils associated with the matrix G({\lambda}) and show that almost all of these pencils are linearizations of G({\lambda}). We also construct symmetric/Hermitian linearizations of G({\lambda})

He Zhang, Bang Wu, Xiangwen Yang, Xingliang Yuan

Dynamic graph neural networks (DGNNs) have emerged and been widely deployed in various web applications (e.g., Reddit) to serve users (e.g., personalized content delivery) due to their remarkable ability to learn from complex and dynamic user interaction data. Despite benefiting from high-quality services, users have raised privacy concerns, such as misuse o

Federico Zocco, Wassim M. Haddad, Andrea Corti, Monica Malvezzi

The shift from a linear to a circular economy has the potential to simultaneously reduce uncertainties of material supplies and waste generation. However, to date, the development of robotic and, more generally, autonomous systems have been rarely integrated into circular economy implementation strategies despite their potential to reduce the operational cos

Supreeth Mysore Venkatesh, Antonio Macaluso, Marlon Nuske, Matthias Klusch

Quantum computing is expected to transform a range of computational tasks beyond the reach of classical algorithms. In this work, we examine the application of variational quantum algorithms (VQAs) for unsupervised image segmentation to partition images into separate semantic regions. Specifically, we formulate the task as a graph cut optimization problem an

Wen-Ying Liu, Hua-Xing Chen

We apply the extended local hidden gauge formalism to study the meson-meson interactions with the quark constituents $cc\bar c\bar c$, $cc\bar c\bar b/\bar c\bar c cb$, $cc\bar b\bar b/\bar c\bar c bb$, $bb\bar c\bar b/\bar b\bar b cb$, and $bb\bar b\bar b$, where the exchanged mesons are the fully-heavy vector mesons $J/\psi$, $B_c^*$, and $\Upsilon$. We so

Chrysanthi Papadimitriou, Jan C. Schulze, Alexander Mitsos

The increasing interest in demand-side management (DSM) as part of the energy cost optimization calls for effective methods to determine representative electricity prices for energy optimization and scheduling investigations. We propose a practical method to construct price profiles of day-ahead (DA) and intraday (ID) electricity spot markets. We construct s

Mikalai Korbit, Adeyemi D. Adeoye, Alberto Bemporad, Mario Zanon

We present Exact Gauss-Newton (EGN), a stochastic second-order optimization algorithm that combines the generalized Gauss-Newton (GN) Hessian approximation with low-rank linear algebra to compute the descent direction. Leveraging the Duncan-Guttman matrix identity, the parameter update is obtained by factorizing a matrix which has the size of the mini-batch.

Jingbo Xia, Congquan Yan, Danjun Zhao, Jingming Zhu

By now it is a well-known fact that if $f$ is a multiplier for the Drury-Arveson space $H^2_n$, and if there is a $c>0$ such that $|f(z)|\geq c$ for every $z\in B$, then the reciprocal function 1/f is also a multiplier for $H^2_n$. We show that for such an $f$ and for every $t\in \mathbb{R}$, $f^t$ is also a multiplier for $H^2_n$. We do so by deriving a dif

Anagha Athavale, Ezio Bartocci, Maria Christakis, Matteo Maffei

We present the first automated verification technique for confidence-based 2-safety properties, such as global robustness and global fairness, in deep neural networks (DNNs). Our approach combines self-composition to leverage existing reachability analysis techniques and a novel abstraction of the softmax function, which is amenable to automated verification

Shangshang Yang, Linrui Qin, Xiaoshan Yu

In the realm of intelligent education, cognitive diagnosis plays a crucial role in subsequent recommendation tasks attributed to the revealed students' proficiency in knowledge concepts. Although neural network-based neural cognitive diagnosis models (CDMs) have exhibited significantly better performance than traditional models, neural cognitive diagnosis is

Arpan Dasgupta, Niclas Boehmer, Neha Madhiwalla, Aparna Hedge

Automated voice calls are an effective method of delivering maternal and child health information to mothers in underserved communities. One method to fight dwindling listenership is through an intervention in which health workers make live service calls. Previous work has shown that we can use AI to identify beneficiaries whose listenership gets the greates

Weiyu Guo, Ying Sun, Yijie Xu, Ziyue Qiao

Surface electromyography (sEMG) based gesture recognition offers a natural and intuitive interaction modality for wearable devices. Despite significant advancements in sEMG-based gesture-recognition models, existing methods often suffer from high computational latency and increased energy consumption. Additionally, the inherent instability of sEMG signals, c

Nicholas Tan Jerome, Suren Chilingaryan, Timo Dritschler, Andreas Kopmann

Given the rapid improvement of the detectors at high-energy physics experiments, the need for real-time data monitoring systems has become imperative. The significance of these systems lies in their ability to display experiment status, steer software and hardware instrumentation, and provide alarms, thus enabling researchers to manage their experiments bett

Mengru Ma, Jiangwei Shang

The reliable characterization of quantum states as well as any potential noise in various quantum systems is crucial for advancing quantum technologies. In this work we propose the concept of corrupted sensing quantum state tomography which enables the simultaneous reconstruction of quantum states and structured noise with the aid of simple Pauli measurement

Jianhao Shen

For the Tits building B(G) of a finite group of Lie type G(Fq), we study the edge zeta function, which enumerates edge-geodesic cycles in the 1-skeleton. We show that every nonzero edge eigenvalue becomes a power of q after raising to a bounded exponent k depending on the type of G. The proof is uniform across types using a Hecke algebra approach. This exten

Zhengyan Shi, Adam X. Yang, Bin Wu, Laurence Aitchison

Instruction tuning plays a crucial role in shaping the outputs of language models (LMs) to desired styles. In this work, we propose a simple yet effective method, Instruction Modelling (IM), which trains LMs by applying a loss function to the instruction and prompt part rather than solely to the output part. Through experiments across 21 diverse benchmarks,

Cristina Martinez Montes, Fredrik Sjögren, Adam Klevfors, Birgit Penzenstadler

The well-being and productivity of IT workers are crucial for both individual success and the overall prosperity of the organisations they serve. This study proposes mindfulness to alleviate stress and improve mental well-being for IT workers. During an 8-week program, IT workers learn about mindfulness, coupled with breathing practices. This study investiga

Alberto Cabezas, Louis Sharrock, Christopher Nemeth

Continuous normalizing flows (CNFs) learn the probability path between a reference distribution and a target distribution by modeling the vector field generating said path using neural networks. Recently, Lipman et al. (2022) introduced a simple and inexpensive method for training CNFs in generative modeling, termed flow matching (FM). In this paper, we repu

Yochay Jerby

In recent works we have introduced the parameter space $\mathcal{Z}_N$ of $A$-variations of the Hardy $Z$-function, $Z(t)$, whose elements are functions of the form \begin{equation} \label{eq:Z-sections} Z_N(t ; \overline{a} ) = \cos(\theta(t))+ \sum_{k=1}^{N} \frac{a_k}{\sqrt{k+1} } \cos ( \theta (t) - \ln(k+1) t), \end{equation} where $\overline{a} = (a_1,

Haoxuan Li, Jifan Yu, Yuanxin Ouyang, Zhuang Liu

Knowledge tracing (KT), aiming to mine students' mastery of knowledge by their exercise records and predict their performance on future test questions, is a critical task in educational assessment. While researchers achieved tremendous success with the rapid development of deep learning techniques, current knowledge tracing tasks fall into the cracks from re

William Seymour, Emilee Rader

The advent of LLMs means that CUIs are cool again, but what isn't so cool is that we're doomed to use them alone. The one user, one account, one device paradigm has dominated the design of CUIs and is not going away as new conversational technologies emerge. In this provocation we explore some of the technical, legal, and design difficulties that seem to mak

Gaia Saveri, Laura Nenzi, Luca Bortolussi, Jan Křetínský

Integrating symbolic knowledge and data-driven learning algorithms is a longstanding challenge in Artificial Intelligence. Despite the recognized importance of this task, a notable gap exists due to the discreteness of symbolic representations and the continuous nature of machine-learning computations. One of the desired bridges between these two worlds woul

Luc Bryan Heitz, Joun Chamas, Christopher Scherb

The advent of Large Language Models (LLM) has revolutionized the efficiency and speed with which tasks are completed, marking a significant leap in productivity through technological innovation. As these chatbots tackle increasingly complex tasks, the challenge of assessing the quality of their outputs has become paramount. This paper critically examines the

Xabier Legaspi, Markus Steenbock

An open question asks whether every group acting acylindrically on a hyperbolic space has uniform exponential growth. We prove that the class of groups of uniform uniform exponential growth acting acylindrically on a hyperbolic space is closed under taking certain geometric small cancellation quotients. There are two consequences: firstly, there is a finitel

Miles Everett, Aiden Durrant, Mingjun Zhong, Georgios Leontidis

Learning invariant representations has been the long-standing approach to self-supervised learning. However, recently progress has been made in preserving equivariant properties in representations, yet do so with highly prescribed architectures. In this work, we propose an invariant-equivariant self-supervised architecture that employs Capsule Networks (Caps

Aline Étienne, Delphine Battistelli, Gwénolé Lecorvé

The objective of this paper is to predict (A) whether a sentence in a written text expresses an emotion, (B) the mode(s) in which it is expressed, (C) whether it is basic or complex, and (D) its emotional category. One of our major contributions, through a dataset and a model, is to integrate the fact that an emotion can be expressed in different modes: from

Marion Neumeier, Sebastian Dorn, Michael Botsch, Wolfgang Utschick

This work introduces the conditioned Vehicle Motion Diffusion (cVMD) model, a novel network architecture for highway trajectory prediction using diffusion models. The proposed model ensures the drivability of the predicted trajectory by integrating non-holonomic motion constraints and physical constraints into the generative prediction module. Central to the

Zhihua Wen, Zhiliang Tian, Zexin Jian, Zhen Huang

Large Language Models (LLMs) are widely used for knowledge-seeking yet suffer from hallucinations. The knowledge boundary (KB) of an LLM limits its factual understanding, beyond which it may begin to hallucinate. Investigating the perception of LLMs' KB is crucial for detecting hallucinations and LLMs' reliable generation. Current studies perceive LLMs' KB o

Julien Feiguel, Mouhamed NDiaye, Pascal Chambaud, Adrien Chambellan

Rehabilitating cast iron pipes through lining offers several advantages, including increased durability, reduced water leaks, and minimal disruption.This approach presents a cost effective and environmentally friendly solution by sealing cracks and joints, extending the pipeline's lifespan, and reducing water wastage, all while avoiding the need for trench e

Martin Ekerå, Joel Gärtner

We provide a high-level cost comparison between Regev's quantum algorithm with Eker{\aa}-G\"artner's extensions on the one hand, and existing state-of-the-art quantum algorithms for factoring and computing discrete logarithms on the other. This when targeting cryptographically relevant problem instances, and when accounting for the space-saving optimizations

Miguel A. Benitez-Rathgeb, André H. Hoang, Vicent Mateu, Iain W. Stewart

We update and extend a previous N$^3$LL$^\prime$+${\cal O}(\alpha_s^3)$ strong coupling determination from thrust data. In particular, we carry out a fit with data fully restricted to the dijet region seeking to minimize the potential impact of power corrections that go beyond dijet configurations. In addition, we parametrize deviations from the dijet power

Thomas Greatrix, Roger Whitaker, Liam Turner, Walter Colombo

The potential for Large Language Models (LLMs) to generate new information offers a potential step change for research and innovation. This is challenging to assert as it can be difficult to determine what an LLM has previously seen during training, making "newness" difficult to substantiate. In this paper we observe that LLMs are able to perform sophisticat

Marios Georgiou, Ioannis Rousochatzakis, Damian J. J. Farnell, Johannes Richter

We study the spin-$S$ Kitaev-Heisenberg model on the honeycomb lattice for $S\!=\!1/2$, $1$ and $3/2$, by using the coupled cluster method (CCM) of microscopic quantum many-body theory. This system is one of the earliest extensions of the Kitaev model and is believed to contain two extended spin liquid phases for any value of the spin quantum number $S$. We

Zi Yang, Ziyue Liu, Samridhi Choudhary, Xinfeng Xie

Training large AI models such as LLMs and DLRMs costs massive GPUs and computing time. The high training cost has become only affordable to big tech companies, meanwhile also causing increasing concerns about the environmental impact. This paper presents CoMERA, a Computing- and Memory-Efficient training method via Rank-Adaptive tensor optimization. CoMERA a

W. J. T. Mollema

Is it sensical to ascribe psychological predicates to AI systems like chatbots based on large language models (LLMs)? People have intuitively started ascribing emotions or consciousness to social AI ('affective artificial agents'), with consequences that range from love to suicide. The philosophical question of whether such ascriptions are warranted is thus

András Grabarits, Kasturi Ranjan Swain, Mahsa Seyed Heydari, Pranav Chandarana

We report a systematic investigation of universal quantum chaotic signatures in the transverse field Ising model on an Erd\H{o}s-R\'enyi network. This is achieved by studying local spectral measures such as the level spacing and the level velocity statistics. A spectral form factor analysis is also performed as a global measure, probing energy level correlat

Anandam Choudhary, Laxman Prasad Goswami, C. Aparajit, Amit D. Lad

The interaction of intense linearly polarized femtosecond laser pulses with solids is known to generate azimuthal magnetic fields, while circularly polarized light has been shown to create axial fields. We demonstrate through experiments and particle-in-cell simulations that circularly polarized light can generate both axial and azimuthal fields of comparabl

Charles A. Hepburn, Yue Jin, Giovanni Montana

Traditional offline reinforcement learning (RL) methods predominantly operate in a batch-constrained setting. This confines the algorithms to a specific state-action distribution present in the dataset, reducing the effects of distributional shift but restricting the policy to seen actions. In this paper, we alleviate this limitation by introducing state-con

Yuga Iguchi, Samuel Livingstone, Nikolas Nüsken, Giorgos Vasdekis

We propose a new simple and explicit numerical scheme for time-homogeneous stochastic differential equations. The scheme is based on sampling increments at each time step from a skew-symmetric probability distribution, with the level of skewness determined by the drift and volatility of the underlying process. We show that as the step-size decreases the sche

Francesco Emanuele Stradi, Anna Lunghi, Matteo Castiglioni, Alberto Marchesi

In constrained Markov decision processes (CMDPs) with adversarial rewards and constraints, a well-known impossibility result prevents any algorithm from attaining both sublinear regret and sublinear constraint violation, when competing against a best-in-hindsight policy that satisfies constraints on average. In this paper, we show that this negative result c

Mingjin Zhang, Jiannong Cao, Xiaoming Shen, Zeyang Cui

Large language models (LLMs) have shown great potential in natural language processing and content generation. However, current LLMs heavily rely on cloud computing, leading to prolonged latency, high bandwidth cost, and privacy concerns. Edge computing is promising to address such concerns by deploying LLMs on edge devices, closer to data sources. Some work

Xiaoxi Lin, Yunpeng Zi, Fengling Li, Jingyan Li

This is a summary of mathematical tools we used in research of analyzing the structure of proteins with amyloid form \cite{xi2024Top}. We defined several geometry indicators on the discrete curve namely the hop distance, the discrete curvature and the discrete torsion. Then, we used these indicators to analyze the structure of amyloid fibrils by regarding it

Haixu Wu, Huakun Luo, Yuezhou Ma, Jianmin Wang

Physics-informed neural networks (PINNs) have been widely applied to solve partial differential equations (PDEs) by enforcing outputs and gradients of deep models to satisfy target equations. Due to the limitation of numerical computation, PINNs are conventionally optimized on finite selected points. However, since PDEs are usually defined on continuous doma

W. J. T. Mollema

The development and deployment of machine learning and AI engender 'AI colonialism', a term that conceptually overlaps with 'data colonialism', as a form of injustice. AI colonialism is in need of decolonization for three reasons. Politically, because it enforces digital capitalism's hegemony. Ecologically, as it negatively impacts the environment and intens

Anna N. Morozovska, Sergei V. Kalinin, Eugene. A. Eliseev, Svitlana Kopyl

We consider nanoflakes of van der Waals ferrielectric CuInP$_2$S$_6$ covered by an ionic surface charge and reveal the appearance of polar states with relatively high polarization ~5 microC/cm$^2$ and stored free charge ~10 microC/cm$%2$, which can mimic "mid-gap" states associated with a surface field-induced transfer of Cu and/or In ions in the van der Waa

Ce Ge, Zhijian Ma, Daoyuan Chen, Yaliang Li

Large language models have demonstrated remarkable capabilities across various tasks, primarily attributed to the utilization of diversely sourced data. However, the impact of pretraining data composition on model performance remains poorly understood. This paper introduces $\textbf{BiMix}$, a novel bivariate data mixing law that models the joint scaling beh

Akide Liu, Jing Liu, Zizheng Pan, Yefei He

A critical approach for efficiently deploying computationally demanding large language models (LLMs) is Key-Value (KV) caching. The KV cache stores key-value states of previously generated tokens, significantly reducing the need for repetitive computations and thereby lowering latency in autoregressive generation. However, the size of the KV cache grows line

Kun Zhou, Beichen Zhang, Jiapeng Wang, Zhipeng Chen

Mathematical reasoning is an important capability of large language models~(LLMs) for real-world applications. To enhance this capability, existing work either collects large-scale math-related texts for pre-training, or relies on stronger LLMs (\eg GPT-4) to synthesize massive math problems. Both types of work generally lead to large costs in training or sy

Mancho Manev

Almost contact complex Riemannian manifolds, also known as almost contact B-metric manifolds, are equipped with a pair of pseudo-Riemannian metrics that are mutually associated with each other using the tensor structure. Here we consider a special class of these manifolds, those of the Sasaki-like type. They have an interesting geometric interpretation: the

Federica Storiale, Enrico Ferrentino, Federico Salvioli, Konstantinos Kapellos

Space robotics poses unique challenges arising from the limitation of energy and computational resources, and the complexity of the environment and employed platforms. At the control center, offline motion planning is fundamental in the computation of optimized trajectories accounting for the system's constraints. Smooth movements, collision and forbidden ar

Changze Lv, Dongqi Han, Yansen Wang, Xiaoqing Zheng

Spiking neural networks (SNNs) represent a promising approach to developing artificial neural networks that are both energy-efficient and biologically plausible. However, applying SNNs to sequential tasks, such as text classification and time-series forecasting, has been hindered by the challenge of creating an effective and hardware-friendly spike-form posi

Lukas Schwenkel, Daniel Briem, Matthias A. Müller, Frank Allgöwer

In this paper, we investigate discounted economic model predictive control (E-MPC) schemes without terminal conditions in scenarios where the optimal operating behavior is a periodic orbit. For such a setting, it is known that a linearly discounted stage cost guarantees asymptotic stability of any arbitrarily small neighborhood of the optimal orbit if the pr

Ben Samia, Mohamed Lazhar Tayeb, Nicolas Vauchelet

In this article, we investigate a competitive reaction-diffusion system modelling the interaction between several species of mosquitoes. In particular, it has been observed that in tropical regions, Aedes aegypti mosquitoes are well established in urban area whereas Aedes albopictus mosquitoes spread widely in forest region. The aim of this paper is to propo

Lei Zheng, Ning Li, Yanhuan Huang, Ruiwen Xu

Sequential recommendation aims to estimate how a user's interests evolve over time via uncovering valuable patterns from user behavior history. Many previous sequential models have solely relied on users' historical information to model the evolution of their interests, neglecting the crucial role that future information plays in accurately capturing these d

Chen Wang, Yan Song, Shuai Wu, Sa Wu

Between 2021 and 2023, AI-Olympics, a series of online AI competitions was hosted by the online evaluation platform Jidi in collaboration with the IJCAI committee. In these competitions, an agent is required to accomplish diverse sports tasks in a two-dimensional continuous world, while competing against an opponent. This paper provides a brief overview of t

Huifeng Wang, Nan Liang

In this paper, we compile a \emph{Fermi} sample of the \emph{long} GRB observations from 15 years of GBM catalogue with identified redshift, in which the GOLD sample contains 123 long GRBs at $z\le5.6$ and the FULL sample contains 151 long GRBs with redshifts at $z\le8.2$. The Amati relation (the $E_{\rm p,i}$-$E_{\rm iso}$ correlation) are calibrated at $z<

Leon Bollmann, Peter Müller

We prove a logarithmically enhanced area law for all R\'enyi entanglement entropies of the ground state of a free gas of relativistic Dirac fermions. Such asymptotics occur in any dimension if the modulus of the Fermi energy is larger than the mass of the particles and in the massless case at Fermi energy zero in one space dimension. In all other cases of ma

Gaia Saveri, Luca Bortolussi

The integration of cyber-physical systems (CPS) into everyday life raises the critical necessity of ensuring their safety and reliability. An important step in this direction is requirement mining, i.e. inferring formally specified system properties from observed behaviors, in order to discover knowledge about the system. Signal Temporal Logic (STL) offers a

Jarnail Singh, Pankaj Bhardwaj

Transparent conducting oxides (TCOs) represent a remarkable class of materials that possess both excellent electrical conductivity and high optical transparency, which are typically considered mutually exclusive in traditional materials. In conventional materials, achieving both high electrical conductivity and optical transparency is difficult. Materials wi

Milena Röhrs, Alexey Bochkarev, Arcesio C. Medina

This work presents a detailed empirical analysis of Bayesian optimisation with information sharing (BOIS) for the variational quantum eigensolver (VQE). The method is applied to computing the potential energy surfaces (PES) of the hydrogen and water molecules. We performed noise-free simulations and investigated the algorithms' performance under the influenc