Search

Benchmarks are important tools for tracking the rapid advancements in large language model (LLM) capabilities. However, benchmarks are not keeping pace in difficulty: LLMs now achieve over 90\% accuracy on popular benchmarks like MMLU, limiting informed measurement of state-of-the-art LLM capabilities. In response, we introduce Humanity's Last Exam (HLE), a multi-modal benchmark at the frontier of human knowledge, designed to be the final closed-ended academic benchmark of its kind with broad subject coverage. HLE consists of 3,000 questions across dozens of subjects, including mathematics, humanities, and the natural sciences. HLE is developed globally by subject-matter experts and consists of multiple-choice and short-answer questions suitable for automated grading. Each question has a known solution that is unambiguous and easily verifiable, but cannot be quickly answered via internet retrieval. State-of-the-art LLMs demonstrate low accuracy and calibration on HLE, highlighting a significant gap between current LLM capabilities and the expert human frontier on closed-ended academic questions. To inform research and policymaking upon a clear understanding of model capabilities, we publicly release HLE at https://lastexam.ai., Comment: 25 pages, 6 figures

Medical image analysis suffers from a lack of labeled data due to several challenges including patient privacy and lack of experts. Although some AI models only perform well with large amounts of data, we will move to data augmentation where there is a solution to improve the performance of our models and increase the dataset size through traditional or advanced techniques. In this paper, we evaluate the effectiveness of data augmentation techniques on two different medical image datasets. In the first step, we applied some transformation techniques to the skin cancer dataset containing benign and malignant classes. Then, we trained the convolutional neural network (CNN) on the dataset before and after augmentation, which significantly improved test accuracy from 90.74% to 96.88% and decreased test loss from 0.7921 to 0.1468 after augmentation. In the second step, we used the Mixup technique by mixing two random images and their corresponding masks using the retina and blood vessels dataset, then we trained the U-net model and obtained the Dice coefficient which increased from 0 before augmentation to 0.4163 after augmentation. The result shows the effect of using data augmentation to increase the dataset size on the classification and segmentation performance.

Multi-UAV air combat is a complex task involving multiple autonomous UAVs, an evolving field in both aerospace and artificial intelligence. This paper aims to enhance adversarial performance through collaborative strategies. Previous approaches predominantly discretize the action space into predefined actions, limiting UAV maneuverability and complex strategy implementation. Others simplify the problem to 1v1 combat, neglecting the cooperative dynamics among multiple UAVs. To address the high-dimensional challenges inherent in six-degree-of-freedom space and improve cooperation, we propose a hierarchical framework utilizing the Leader-Follower Multi-Agent Proximal Policy Optimization (LFMAPPO) strategy. Specifically, the framework is structured into three levels. The top level conducts a macro-level assessment of the environment and guides execution policy. The middle level determines the angle of the desired action. The bottom level generates precise action commands for the high-dimensional action space. Moreover, we optimize the state-value functions by assigning distinct roles with the leader-follower strategy to train the top-level policy, followers estimate the leader's utility, promoting effective cooperation among agents. Additionally, the incorporation of a target selector, aligned with the UAVs' posture, assesses the threat level of targets. Finally, simulation experiments validate the effectiveness of our proposed method.

We present Fanar, a platform for Arabic-centric multimodal generative AI systems, that supports language, speech and image generation tasks. At the heart of Fanar are Fanar Star and Fanar Prime, two highly capable Arabic Large Language Models (LLMs) that are best in the class on well established benchmarks for similar sized models. Fanar Star is a 7B (billion) parameter model that was trained from scratch on nearly 1 trillion clean and deduplicated Arabic, English and Code tokens. Fanar Prime is a 9B parameter model continually trained on the Gemma-2 9B base model on the same 1 trillion token set. Both models are concurrently deployed and designed to address different types of prompts transparently routed through a custom-built orchestrator. The Fanar platform provides many other capabilities including a customized Islamic Retrieval Augmented Generation (RAG) system for handling religious prompts, a Recency RAG for summarizing information about current or recent events that have occurred after the pre-training data cut-off date. The platform provides additional cognitive capabilities including in-house bilingual speech recognition that supports multiple Arabic dialects, voice and image generation that is fine-tuned to better reflect regional characteristics. Finally, Fanar provides an attribution service that can be used to verify the authenticity of fact based generated content. The design, development, and implementation of Fanar was entirely undertaken at Hamad Bin Khalifa University's Qatar Computing Research Institute (QCRI) and was sponsored by Qatar's Ministry of Communications and Information Technology to enable sovereign AI technology development.

Random number generation plays a vital role in cryptographic systems and computational applications, where uniformity, unpredictability, and robustness are essential. This paper presents the Entropy Mixing Network (EMN), a novel hybrid random number generator designed to enhance randomness quality by combining deterministic pseudo-random generation with periodic entropy injection. To evaluate its effectiveness, we propose a comprehensive assessment framework that integrates statistical tests, advanced metrics, and visual analyses, providing a holistic view of randomness quality, predictability, and computational efficiency. The results demonstrate that EMN outperforms Python's SystemRandom and MersenneTwister in critical metrics, achieving the highest Chi-squared p-value (0.9430), entropy (7.9840), and lowest predictability (-0.0286). These improvements come with a trade-off in computational performance, as EMN incurs a higher generation time (0.2602 seconds). Despite this, its superior randomness quality makes it particularly suitable for cryptographic applications where security is prioritized over speed.

This paper focuses on the analysis of a stochastic SAIRS-type epidemic model that explicitly incorporates the roles of asymptomatic and symptomatic infectious individuals in disease transmission dynamics. Asymptomatic carriers, often undetected due to the lack of symptoms, play a crucial role in the spread of many communicable diseases, including COVID-19. Our model also accounts for vaccination and considers the stochastic effects of environmental and population-level randomness using L\'evy processes. We begin by demonstrating the existence and uniqueness of a global positive solution to the proposed stochastic system, ensuring the model's mathematical validity. Subsequently, we derive sufficient conditions under which the disease either becomes extinct or persists over time, depending on the parameters and initial conditions. The analysis highlights the influence of random perturbations, asymptomatic transmission, and vaccination strategies on disease dynamics. Finally, we conduct comprehensive numerical simulations to validate the theoretical findings and illustrate the behavior of the model under various scenarios of randomness and parameter settings. These results provide valuable insights into the stochastic dynamics of epidemic outbreaks and inform strategies for disease management and control.

Understanding how large language models (LLMs) grasp the historical context of concepts and their semantic evolution is essential in advancing artificial intelligence and linguistic studies. This study aims to evaluate the capabilities of various LLMs in capturing temporal dynamics of meaning, specifically how they interpret terms across different time periods. We analyze a diverse set of terms from multiple domains, using tailored prompts and measuring responses through both objective metrics (e.g., perplexity and word count) and subjective human expert evaluations. Our comparative analysis includes prominent models like ChatGPT, GPT-4, Claude, Bard, Gemini, and Llama. Findings reveal marked differences in each model's handling of historical context and semantic shifts, highlighting both strengths and limitations in temporal semantic understanding. These insights offer a foundation for refining LLMs to better address the evolving nature of language, with implications for historical text analysis, AI design, and applications in digital humanities.

Modular invariance has recently paved new promising directions in flavor model building. Motivated by this development, we present in this work the first implementation of the $S_3$ modular symmetry within the Pati-Salam unification framework, addressing the flavor structure of quarks and leptons. Assigning left- and right-handed matter fields as $S_3$ doublets or singlets, we propose three benchmark models that achieve compelling fits to sixteen observables including charged fermion mass ratios and flavor mixing parameters. Light neutrino masses arise via the type-I seesaw mechanism, and neutrino oscillation parameters are explored in light of the latest NuFIT-6.0 results. All models favor a normal neutrino mass ordering, with the atmospheric mixing angle lie in the lower octant. For models I and III, the effective Majorana mass $m_{\beta\beta}$ is within the reach of upcoming neutrinoless double-beta decay experiments, while it is too small to be detected in model II. Predicted leptonic CP-violating phases, the sum of active neutrino masses, and Majorana phases span wide but distinctive ranges, enabling future experiments to test and differentiate the proposed models., Comment: 17 pages, 3 figures. Comments are welcome

Machine Unlearning has emerged as a critical area in artificial intelligence, addressing the need to selectively remove learned data from machine learning models in response to data privacy regulations. This paper provides a comprehensive comparative analysis of six state-of-theart unlearning techniques applied to image and text classification tasks. We evaluate their performance, efficiency, and compliance with regulatory requirements, highlighting their strengths and limitations in practical scenarios. By systematically analyzing these methods, we aim to provide insights into their applicability, challenges,and tradeoffs, fostering advancements in the field of ethical and adaptable machine learning.

This study examines the phytochemical characteristics of Ayurvedic products. An analysis was performed on Kottakkal Ayurveda Triphala (T), Kottakkal Ayurveda Hinguvachadi Churnam (H), and Kottakkal Ayurveda Jirakadyarishtam (J) using GC-MS and LC-MS techniques to determine their bioactive constituents, while also assessing their antimicrobial, docking, anticancer, and anti-diabetic activities. The GC-MS analysis identified 30, 45, and 8 chemical components in Kottakkal Ayurveda Triphala (T), Kottakkal Ayurveda Hinguvachadi Churnam (H), and Kottakkal Ayurveda Jirakadyarishtam (J), respectively. The LC-MS analysis produced 15, 20, and 16 peaks for Kottakkal Ayurveda Triphala (T), Kottakkal Ayurveda Hinguvachadi Churnam (H), and Kottakkal Ayurveda Jirakadyarishtam (J), with m/z values of 982, 981, 972, and 933; 987, 985, 974, and 945; and 969, 965, 951, and 941, respectively, confirming their precision. Moreover, characterization of the Ayurvedic products was carried out using FT-IR, UV-vis, and 1H-NMR spectroscopy to identify significant functional groups and chemical substances. Kottakkal Ayurveda Triphala (T) was evaluated for antibacterial activity against Gram-positive bacteria (Streptococcus pneumoniae and Staphylococcus aureus) along with Gram-negative bacteria (Escherichia coli and Klebsiella pneumoniae), yielding a P value of 0.0650 (P < 0.0001). Both Kottakkal Ayurveda Hinguvachadi Churnam (H) and Kottakkal Ayurveda Jirakadyarishtam (J) were subjected to analysis for their effectiveness against Aspergillus niger and Aspergillus fumigatus, also revealing a P value within the acceptable range of 0.0650 (P < 0.0001). The anti-diabetic properties of Kottakkal Ayurveda Triphala (T) were assessed using the {\alpha}-glucosidase inhibitory method, which exhibited a significant inhibitory effect on {\alpha}-glucosidase, resulting in an average P value of 0.001 (P < 0.0001)., Comment: for associated mpeg file, see http://myhost.domain/file.mpg

Understanding the deep meanings of the Qur'an and bridging the language gap between modern standard Arabic and classical Arabic is essential to improve the question-and-answer system for the Holy Qur'an. The Qur'an QA 2023 shared task dataset had a limited number of questions with weak model retrieval. To address this challenge, this work updated the original dataset and improved the model accuracy. The original dataset, which contains 251 questions, was reviewed and expanded to 629 questions with question diversification and reformulation, leading to a comprehensive set of 1895 categorized into single-answer, multi-answer, and zero-answer types. Extensive experiments fine-tuned transformer models, including AraBERT, RoBERTa, CAMeLBERT, AraELECTRA, and BERT. The best model, AraBERT-base, achieved a MAP@10 of 0.36 and MRR of 0.59, representing improvements of 63% and 59%, respectively, compared to the baseline scores (MAP@10: 0.22, MRR: 0.37). Additionally, the dataset expansion led to improvements in handling "no answer" cases, with the proposed approach achieving a 75% success rate for such instances, compared to the baseline's 25%. These results demonstrate the effect of dataset improvement and model architecture optimization in increasing the performance of QA systems for the Holy Qur'an, with higher accuracy, recall, and precision.

Large language models (LLMs) have significantly advanced natural language processing, excelling in areas like text generation, summarization, and question-answering. Despite their capabilities, these models face challenges when fine-tuned on small, domain-specific datasets, often struggling to generalize and deliver accurate results with unfamiliar inputs. To tackle this issue, we introduce RIRO, a novel two-layer architecture designed to improve performance in data-scarce environments. The first layer leverages advanced prompt engineering to reformulate inputs, ensuring better alignment with training data, while the second layer focuses on refining outputs to minimize inconsistencies. Through fine-tuning models like Phi-2, Falcon 7B, and Falcon 1B, with Phi-2 outperforming the others. Additionally, we introduce a benchmark using evaluation metrics such as cosine similarity, Levenshtein distance, BLEU score, ROUGE-1, ROUGE-2, and ROUGE-L. While these advancements improve performance, challenges like computational demands and overfitting persist, limiting the potential of LLMs in data-scarce, high-stakes environments such as healthcare, legal documentation, and software testing.

Advancements in quantum optics and squeezed light generation have revolutionized various fields of quantum science over the past three decades, with notable applications such as gravitational wave detection. Here, we extend the use of squeezed light to the realm of ultrafast quantum science. We demonstrate the generation of ultrafast, broadband quantum light pulses spanning 0.33 to 0.73 PHz using light field synthesizer and a four-wave mixing nonlinear process. Experimental results confirm that these pulses exhibit amplitude squeezing, which is consistent with theoretical predictions. This work lays the groundwork for a new field of ultrafast quantum science, enabling real-time studies of quantum light-matter interaction dynamics, which expect to reveal new physics. We also demonstrate the encoding of binary digital data onto these quantum light waveforms, synthesized with attosecond resolution, showcasing potential applications in secure quantum communication. This work paves the way for ultrafast quantum optoelectronics, quantum computing, and next-generation encrypted quantum communication networks, capable of achieving petahertz-scale data transmission speeds., Comment: 15 pages, 5 figures

While numerical simulations are presented in most papers introducing new methods to enhance the VQE performance, comprehensive, comparative, and applied studies remain relatively rare. We present a comprehensive, yet concise guide for the implementation of the VQE for molecular problems on NISQ devices, specifically applied to estimate the ground state energy of the BeH2 molecule using hardware-efficient and chemically informed ans\"atze. This work clarifies several under-documented aspects in the literature, such as the construction of the electronic Hamiltonian, the transformation of fermionic operators into qubit operators via second quantization, and the mathematical framework's details for the unitary coupled cluster single and double (UCCSD) ansatz. Our methodology, implemented using Qiskit 1.2, the latest release as of the date of this writing, is demonstrated on a noiseless simulator and further tested with noisy quantum circuits. The resilience of the VQE to quantum noise remains an open question. This study compares the computational accuracy of ground state energy estimations for molecules using the VQE across three different current quantum hardware noise models. Furthermore, our experiment on IBM's 156-qubit actual quantum computer revealed valuable insights on the real performance of the VQE on current quantum hardware.

Piezo1 ion channels are stretch-activated ion channels involved in a variety of important physiological and pathophysiological processes, as for example cardiovascular development and homeostasis. Since its discovery, it is known that this type of ion channel desensitize when exposed to stretch. However, recent experiments on Piezo1 ion channels have uncovered that their stretch response is qualitatively different when exposed to positive electrochemical driving forces, where the desensitization is reset. In this work, we propose a novel voltage-modulated mathematical model of Piezo1 based on a continuous-time Markov chain. We show that our Piezo1 model is able to quantitatively reproduce a wide range of experimental observations. Furthermore, we integrate our new ion channel model into the Mahajan-Shiferaw ventricular cardiomyocyte model to study the effect of electromechanical pacing at the cellular scale. This integrated cell model is able to qualitatively reproduce some aspects of the experimental observations regarding the rate-dependence of electromechanical pacing protocols. Our studies suggest that the Piezo1 ion channel is an important component that significantly contributes to the electromechanical coupled response of cardiomyocytes.

Careful parametrization of networking protocols is crucial to maximize the performance of low-power wireless systems and ensure that stringent application requirements can be met. This is a non-trivial task involving thorough characterization on testbeds and requiring expert knowledge. Unfortunately, the community still lacks a tool to facilitate parameter exploration while minimizing the necessary experimentation time on testbeds. Such a tool would be invaluable, as exhaustive parameter searches can be time-prohibitive or unfeasible given the limited availability of testbeds, whereas non-exhaustive unguided searches rarely deliver satisfactory results. In this paper, we present APEX, a framework enabling an automated and informed parameter exploration for low-power wireless protocols and allowing to converge to an optimal parameter set within a limited number of testbed trials. We design APEX using Gaussian processes to effectively handle noisy experimental data and estimate the optimality of a certain parameter combination. After developing a prototype of APEX, we demonstrate its effectiveness by parametrizing two IEEE 802.15.4 protocols for a wide range of application requirements. Our results show that APEX can return the best parameter set with up to 10.6x, 4.5x and 3.25x less testbed trials than traditional solutions based on exhaustive search, greedy approaches, and reinforcement learning, respectively., Comment: 28 pages, 13 figures

Magnetic Resonance Imaging (MRI) offers critical insights into microstructural details, however, the spatial resolution of standard 1.5T imaging systems is often limited. In contrast, 7T MRI provides significantly enhanced spatial resolution, enabling finer visualization of anatomical structures. Though this, the high cost and limited availability of 7T MRI hinder its widespread use in clinical settings. To address this challenge, a novel Super-Resolution (SR) model is proposed to generate 7T-like MRI from standard 1.5T MRI scans. Our approach leverages a diffusion-based architecture, incorporating gradient nonlinearity correction and bias field correction data from 7T imaging as guidance. Moreover, to improve deployability, a progressive distillation strategy is introduced. Specifically, the student model refines the 7T SR task with steps, leveraging feature maps from the inference phase of the teacher model as guidance, aiming to allow the student model to achieve progressively 7T SR performance with a smaller, deployable model size. Experimental results demonstrate that our baseline teacher model achieves state-of-the-art SR performance. The student model, while lightweight, sacrifices minimal performance. Furthermore, the student model is capable of accepting MRI inputs at varying resolutions without the need for retraining, significantly further enhancing deployment flexibility. The clinical relevance of our proposed method is validated using clinical data from Massachusetts General Hospital. Our code is available at https://github.com/ZWang78/SR.

This paper introduces GLLM, an innovative tool that leverages Large Language Models (LLMs) to automatically generate G-code from natural language instructions for Computer Numerical Control (CNC) machining. GLLM addresses the challenges of manual G-code writing by bridging the gap between human-readable task descriptions and machine-executable code. The system incorporates a fine-tuned StarCoder-3B model, enhanced with domain-specific training data and a Retrieval-Augmented Generation (RAG) mechanism. GLLM employs advanced prompting strategies and a novel self-corrective code generation approach to ensure both syntactic and semantic correctness of the generated G-code. The architecture includes robust validation mechanisms, including syntax checks, G-code-specific verifications, and functional correctness evaluations using Hausdorff distance. By combining these techniques, GLLM aims to democratize CNC programming, making it more accessible to users without extensive programming experience while maintaining high accuracy and reliability in G-code generation.

Left and right circularly polarized transverse electromagnetic waves propagate at slightly different speeds in a magnetic material leading to a polarization rotation by an amount proportional to the projection of the magnetic field along the direction of the wave propagation. We show how this magneto-optical effect can serve as a vectorial polarization shaper if the input mode is either a radially-polarised or an azimuthally-polarised Laguerre-Gaussian (LG) mode. The specific polarization map of the output field can be achieved by choosing appropriately the magnetic material and/or its geometry. We show further that when the LG beam waist is comparable to the wavelength the fields are no longer purely transverse but acquire an additional longitudinal (axial) component. We demonstrate how this modifies the polarisation texturing.

Question answering systems face critical limitations in languages with limited resources and scarce data, making the development of robust models especially challenging. The Quranic QA system holds significant importance as it facilitates a deeper understanding of the Quran, a Holy text for over a billion people worldwide. However, these systems face unique challenges, including the linguistic disparity between questions written in Modern Standard Arabic and answers found in Quranic verses written in Classical Arabic, and the small size of existing datasets, which further restricts model performance. To address these challenges, we adopt a cross-language approach by (1) Dataset Augmentation: expanding and enriching the dataset through machine translation to convert Arabic questions into English, paraphrasing questions to create linguistic diversity, and retrieving answers from an English translation of the Quran to align with multilingual training requirements; and (2) Language Model Fine-Tuning: utilizing pre-trained models such as BERT-Medium, RoBERTa-Base, DeBERTa-v3-Base, ELECTRA-Large, Flan-T5, Bloom, and Falcon to address the specific requirements of Quranic QA. Experimental results demonstrate that this cross-language approach significantly improves model performance, with RoBERTa-Base achieving the highest MAP@10 (0.34) and MRR (0.52), while DeBERTa-v3-Base excels in Recall@10 (0.50) and Precision@10 (0.24). These findings underscore the effectiveness of cross-language strategies in overcoming linguistic barriers and advancing Quranic QA systems

This paper introduces the Precomputed Numerical Green Function (PNGF) method, a new approach for rapid inverse design of electromagnetic devices. The static components of the design are incorporated into a numerical Green function obtained from a single fully parallelized precomputation step, reducing the cost of evaluating candidate designs during optimization to only being proportional to the size of the region under modification. When used with the direct binary search optimization algorithm, a low-rank update technique is leveraged to further decrease the iteration time to seconds without approximations or compromises in accuracy. The total runtime for an inverse design is reduced by several orders of magnitude compared to using conventional Maxwell solvers due to the linear time complexity of the method, attaining speedups of up to 700x for the design examples considered and lowering the process from multiple days to weeks down to less than an hour. The performance and flexibility of the approach are highlighted with design studies, including experimental results, on an ultrawideband 30GHz substrate antenna with 50% fractional bandwidth, a 6GHz switched beam antenna steerable between angles 90{\deg} apart, and a broadband, ultra-short-length microstrip to substrate-integrated waveguide transition. The approach stands to reshape inverse design in electromagnetics.

One of the significant challenges to generating value-aligned behavior is to not only account for the specified user objectives but also any implicit or unspecified user requirements. The existence of such implicit requirements could be particularly common in settings where the user's understanding of the task model may differ from the agent's estimate of the model. Under this scenario, the user may incorrectly expect some agent behavior to be inevitable or guaranteed. This paper addresses such expectation mismatch in the presence of differing models by capturing the possibility of unspecified user subgoal in the context of a task captured as a Markov Decision Process (MDP) and querying for it as required. Our method identifies bottleneck states and uses them as candidates for potential implicit subgoals. We then introduce a querying strategy that will generate the minimal number of queries required to identify a policy guaranteed to achieve the underlying goal. Our empirical evaluations demonstrate the effectiveness of our approach in inferring and achieving unstated goals across various tasks.

High-dimensional multiplex graphs are characterized by their high number of complementary and divergent dimensions. The existence of multiple hierarchical latent relations between the graph dimensions poses significant challenges to embedding methods. In particular, the geometric distortions that might occur in the representational space have been overlooked in the literature. This work studies the problem of high-dimensional multiplex graph embedding from a geometric perspective. We find that the node representations reside on highly curved manifolds, thus rendering their exploitation more challenging for downstream tasks. Moreover, our study reveals that increasing the number of graph dimensions can cause further distortions to the highly curved manifolds. To address this problem, we propose a novel multiplex graph embedding method that harnesses hierarchical dimension embedding and Hyperbolic Graph Neural Networks. The proposed approach hierarchically extracts hyperbolic node representations that reside on Riemannian manifolds while gradually learning fewer and more expressive latent dimensions of the multiplex graph. Experimental results on real-world high-dimensional multiplex graphs show that the synergy between hierarchical and hyperbolic embeddings incurs much fewer geometric distortions and brings notable improvements over state-of-the-art approaches on downstream tasks., Comment: Published in Proceedings of the ACM Conference on Information and Knowledge Management (CIKM) 2024, DOI: 10.1145/3627673.3679541

We consider the problem of finding the asymptotic capacity of symmetric private information retrieval (SPIR) with $B$ Byzantine servers. Prior to finding the capacity, a definition for the Byzantine servers is needed since in the literature there are two different definitions. In \cite{byzantine_tpir}, where it was first defined, the Byzantine servers can send any symbol from the storage, their received queries and some independent random symbols. In \cite{unresponsive_byzantine_1}, Byzantine servers send any random symbol independently of their storage and queries. It is clear that these definitions are not identical, especially when \emph{symmetric} privacy is required. To that end, we define Byzantine servers, inspired by \cite{byzantine_tpir}, as the servers that can share everything, before and after the scheme initiation. In this setting, we find an upper bound, for an infinite number of messages case, that should be satisfied for all schemes that protect against this setting and develop a scheme that achieves this upper bound. Hence, we identify the capacity of the problem.

We present the results from the first axion-like particle search conducted using a dish antenna. The experiment was conducted at room temperature and sensitive to axion-like particles in the $44-52\,\mu\mathrm{eV}$ range ($10.7 - 12.5\,\mathrm{GHz}$). The novel dish antenna geometry was proposed by the BREAD collaboration and previously used to conduct a dark photon search in the same mass range. To allow for axion-like particle sensitivity, the BREAD dish antenna was placed in a $3.9\,\mathrm{T}$ solenoid magnet at Argonne National Laboratory. In the presence of a magnetic field, axion-like dark matter converts to photons at the conductive surface of the reflector. The signal is focused onto a custom coaxial horn antenna and read out with a low-noise radio-frequency receiver. No evidence of axion-like dark matter was observed in this mass range and we place the most stringent laboratory constraints on the axion-photon coupling strength, $g_{a\gamma\gamma}$, in this mass range at 90\% confidence., Comment: 7 pages, 3 figures

The double Heston model is one of the most popular option pricing models in financial theory. It is applied to several issues such that risk management and volatility surface calibration. This paper deals with the problem of global parameter estimations in this model. Our main stochastic results are about the stationarity and the ergodicity of the double Heston process. The statistical part of this paper is about the maximum likelihood and the conditional least squares estimations based on continuous-time observations; then for each estimation method, we study the asymptotic properties of the resulted estimators in the ergodic case.

Large-scale deployment of Internet of Things (IoT) networks in the industrial, scientific, and medical (ISM) band leads to spectrum congestion and requires multiple gateways to cover wide areas. This will increase cost, complexity, and energy consumption. TV White Spaces (TVWS) provides an abundant spectrum that is sufficient for low data rate IoT applications. This low-frequency band offers coverage over larger areas due to the ability of wireless signals to penetrate obstacles and terrain. In this paper, we examine the performance of narrowband data communications in TVWS through an outdoor experiment in a suburban area with line-of-sight (LOS) and non-line-of-sight (NLOS) propagation scenarios. We implement a software-defined radio (SDR) testbed and develop a GNU radio benchmark tool to perform outdoor experiments for TVWS narrowband data communication between a gateway and wireless nodes at various locations. The results reveal that the system can achieve a throughput of up to 97 Kbps with a packet error rate (PER) and packet loss rate (PLR) under 1% over NLOS paths, making it suitable for low-data rate applications. This work offers valuable insights for designing the physical layer of narrowband white space devices (WSDs). The developed benchmark tool will also greatly assist other researchers in evaluating the performance of SDR-based communication systems.

Maintaining high data quality is crucial for reliable data analysis and machine learning (ML). However, existing data quality management tools often lack automation, interactivity, and integration with ML workflows. This demonstration paper introduces DataLens, a novel interactive dashboard designed to streamline and automate the data quality management process for tabular data. DataLens integrates a suite of data profiling, error detection, and repair tools, including statistical, rule-based, and ML-based methods. It features a user-in-the-loop module for interactive rule validation, data labeling, and custom rule definition, enabling domain experts to guide the cleaning process. Furthermore, DataLens implements an iterative cleaning module that automatically selects optimal cleaning tools based on downstream ML model performance. To ensure reproducibility, DataLens generates DataSheets capturing essential metadata and integrates with MLflow and Delta Lake for experiment tracking and data version control. This demonstration showcases DataLens's capabilities in effectively identifying and correcting data errors, improving data quality for downstream tasks, and promoting reproducibility in data cleaning pipelines.

This paper introduces EdgeMLOps, a framework leveraging Cumulocity IoT and thin-edge.io for deploying and managing machine learning models on resource-constrained edge devices. We address the challenges of model optimization, deployment, and lifecycle management in edge environments. The framework's efficacy is demonstrated through a visual quality inspection (VQI) use case where images of assets are processed on edge devices, enabling real-time condition updates within an asset management system. Furthermore, we evaluate the performance benefits of different quantization methods, specifically static and dynamic signed-int8, on a Raspberry Pi 4, demonstrating significant inference time reductions compared to FP32 precision. Our results highlight the potential of EdgeMLOps to enable efficient and scalable AI deployments at the edge for industrial applications.

Recommender systems often rely on graph-based filters, such as normalized item-item adjacency matrices and low-pass filters. While effective, the centralized computation of these components raises concerns about privacy, security, and the ethical use of user data. This work proposes two decentralized frameworks for securely computing these critical graph components without centralizing sensitive information. The first approach leverages lightweight Multi-Party Computation and distributed singular vector computations to privately compute key graph filters. The second extends this framework by incorporating low-rank approximations, enabling a trade-off between communication efficiency and predictive performance. Empirical evaluations on benchmark datasets demonstrate that the proposed methods achieve comparable accuracy to centralized state-of-the-art systems while ensuring data confidentiality and maintaining low communication costs. Our results highlight the potential for privacy-preserving decentralized architectures to bridge the gap between utility and user data protection in modern recommender systems.

Wireless mouse rings offer subtle, reliable pointing interactions for wearable computing platforms, but the small battery below 27 mAh in the miniature rings restricts the ring's continuous lifespan to just 1-2 hours due to the power consumption of current low-powered wireless communication like BLE. However, the picoRing mouse addresses this by enabling continuous ring-based mouse interaction with ultra-low-powered ring-to-wristband wireless communication through a coil-based impedance sensing method called semi-passive inductive telemetry. This allows a wristband coil to capture a unique frequency response of a nearby ring coil via sensitive inductive coupling, converting the user's mouse input into the unique frequency response via an 820 uW mouse-driven modulation module. Thus, the picoRing mouse can operate continuously for over 92 hours on a single charge of a 20 mAh battery while supporting subtle scrolling and pressing interactions.

We study a linear computation problem over a quantum multiple access channel (LC-QMAC), where $S$ servers share an entangled state and separately store classical data streams $W_1,\cdots, W_S$ over a finite field $\mathbb{F}_d$. A user aims to compute $K$ linear combinations of these data streams, represented as $Y = \mathbf{V}_1 W_1 + \mathbf{V}_2 W_2 + \cdots + \mathbf{V}_S W_S \in \mathbb{F}_d^{K \times 1}$. To this end, each server encodes its classical information into its local quantum subsystem and transmits it to the user, who retrieves the desired computations via quantum measurements. In this work, we propose an achievable scheme for LC-QMAC based on the stabilizer formalism and the ideas from entanglement-assisted quantum error-correcting codes (EAQECC). Specifically, given any linear computation matrix, we construct a self-orthogonal matrix that can be implemented using the stabilizer formalism. Also, we apply precoding matrices to minimize the number of auxiliary qudits required. Our scheme achieves more computations per qudit, i.e., a higher computation rate, compared to the best-known methods in the literature, and attains the capacity in certain cases.

The dominance of large decoder-only language models has overshadowed encoder-decoder architectures, despite their fundamental efficiency advantages in sequence processing. For small language models (SLMs) - those with 1 billion parameters or fewer - our systematic analysis across GPU, CPU, and NPU platforms reveals that encoder-decoder architectures achieve 47% lower first-token latency and 4.7x higher throughput compared to decoder-only models on edge devices. These gains may be attributed to encoder-decoder's one-time input processing and efficient separation of understanding and generation phases. We introduce a novel knowledge distillation framework that enables encoder-decoder models to leverage capabilities from large scalable decoder-only teachers while preserving their architectural advantages, achieving up to 6 average performance points improvement across diverse tasks, with significant gains in asymmetric sequence tasks where input and output distributions can benefit from different processing approaches. When combined with modern advances like Rotary Positional Embeddings (RoPE) and Vision encoders, our systematic investigation demonstrates that encoder-decoder architectures provide a more practical path toward deploying capable language models in resource-constrained environments. Our findings challenge the prevailing trend toward decoder-only scaling, showing that architectural choices become increasingly crucial as parameter budgets decrease, particularly for on-device and edge deployments where computational efficiency is paramount., Comment: 13 pages, 5 figures. LLMs/SLMs, encoder-decoder and decoder-only

We study multidimensional generalized backward stochastic differential equations (GBSDEs) within a general filtration that supports a Brownian motion under weak assumptions on the associated data. We establish the existence and uniqueness of solutions in $\mathbb{L}^p$ for $p \in (1,2]$. Our results apply to generators that are stochastic monotone in the $y$-variable, stochastic Lipschitz in the $z$-variable, and satisfy a general stochastic linear growth condition.

In this paper, we calculate the unit groups and the $2$-class numbers of the fields $ \mathbb{K}= \mathbb{Q}(\sqrt{2}, \sqrt{p_1}, \sqrt{p_2})$ and $ \mathbb{L}= \mathbb{Q}( \sqrt{-1},\sqrt{2}, \sqrt{p_1}, \sqrt{p_2})$, where $p_1$ and $p_2$ are two prime numbers satisfying $p_1\equiv p_2 \equiv 1 \pmod{4}$., Comment: 36 pages

The progress and integration of intelligent transport systems (ITS) have therefore been central to creating safer and more efficient transport networks. The Internet of Vehicles (IoV) has the potential to improve road safety and provide comfort to travelers. However, this technology is exposed to a variety of security vulnerabilities that malicious actors could exploit. One of the most serious threats to IoV is the Distributed Denial of Service (DDoS) attack, which could be used to disrupt traffic flow, disable communication between vehicles, or even cause accidents. In this paper, we propose a novel Deep Multimodal Learning (DML) approach for detecting DDoS attacks in IoV, addressing a critical aspect of cybersecurity in intelligent transport systems. Our proposed DML model integrates Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU), enhanced by Attention and Gating mechanisms, and Multi-Layer Perceptron (MLP) with a multimodal intermediate fusion architecture. This innovative method effectively identifies and mitigates DDoS attacks in real-time by utilizing the Framework for Misbehavior Detection (F2MD) to generate a synthetic dataset, thereby overcoming the limitations of the existing Vehicular Reference Misbehavior (VeReMi) extension dataset. The proposed approach is evaluated in real-time across different simulated real-world scenario with 10\%, $30\%$, and $50\%$ attacker densities. The proposed DML model achieves an average accuracy of 96.63\%, outperforming the classical Machine Learning (ML) approaches and state-of-the-art methods which demonstrate significant efficacy and reliability in protecting vehicular networks from malicious cyber-attacks.

Age-of-information (AoI) and transmission power are crucial performance metrics in low energy wireless networks, where information freshness is of paramount importance. This study examines a power-limited internet of things (IoT) network supported by a flying unmanned aerial vehicle(UAV) that collects data. Our aim is to optimize the UAV flight trajectory and scheduling policy to minimize a varying AoI and transmission power combination. To tackle this variation, this paper proposes a meta-deep reinforcement learning (RL) approach that integrates deep Q-networks (DQNs) with model-agnostic meta-learning (MAML). DQNs determine optimal UAV decisions, while MAML enables scalability across varying objective functions. Numerical results indicate that the proposed algorithm converges faster and adapts to new objectives more effectively than traditional deep RL methods, achieving minimal AoI and transmission power overall., Comment: 10 pages, 8 figures

Modern power grids are transitioning towards power electronics-dominated grids (PEDG) due to the increasing integration of renewable energy sources and energy storage systems. This shift introduces complexities in grid operation and increases vulnerability to cyberattacks. This research explores the application of digital twin (DT) technology and machine learning (ML) techniques for anomaly detection in PEDGs. A DT can accurately track and simulate the behavior of the physical grid in real-time, providing a platform for monitoring and analyzing grid operations, with extended amount of data about dynamic power flow along the whole power system. By integrating ML algorithms, the DT can learn normal grid behavior and effectively identify anomalies that deviate from established patterns, enabling early detection of potential cyberattacks or system faults. This approach offers a comprehensive and proactive strategy for enhancing cybersecurity and ensuring the stability and reliability of PEDGs., Comment: preprint accepted to 2025 Conference of Young Researchers in Electrical and Electronic Engineering (2025 ElCon)

Advanced sensors and IoT devices have improved the monitoring and control of complex industrial enterprises. They have also created an interdependent fabric of geographically distributed process operations (clients) across these enterprises. Granger causality is an effective approach to detect and quantify interdependencies by examining how one client's state affects others over time. Understanding these interdependencies captures how localized events, such as faults and disruptions, can propagate throughout the system, possibly causing widespread operational impacts. However, the large volume and complexity of industrial data pose challenges in modeling these interdependencies. This paper develops a federated approach to learning Granger causality. We utilize a linear state space system framework that leverages low-dimensional state estimates to analyze interdependencies. This addresses bandwidth limitations and the computational burden commonly associated with centralized data processing. We propose augmenting the client models with the Granger causality information learned by the server through a Machine Learning (ML) function. We examine the co-dependence between the augmented client and server models and reformulate the framework as a standalone ML algorithm providing conditions for its sublinear and linear convergence rates. We also study the convergence of the framework to a centralized oracle model. Moreover, we include a differential privacy analysis to ensure data security while preserving causal insights. Using synthetic data, we conduct comprehensive experiments to demonstrate the robustness of our approach to perturbations in causality, the scalability to the size of communication, number of clients, and the dimensions of raw data. We also evaluate the performance on two real-world industrial control system datasets by reporting the volume of data saved by decentralization., Comment: Accepted in International Conference on Learning Representations (ICLR) 2025

We report the results of $^{139}$La NMR measurements in the non-centrosymmetric superconductor LaRhGe$_3$. This material crystallizes in a tetragonal structure without inversion symmetry and exhibits type-I superconductivity below 385 mK. We observed remarkably sharp NMR signals, indicating that the magnetic and electronic properties of the sample are extremely uniform in LaRhGe$_3$ despite the complex crystal structure. Our NMR results indicate that LaRhGe$_3$ is a weakly correlated semimetal in the normal state., Comment: 2 pages, 2 figures

Educational recommender systems (ERSs) play a crucial role in personalizing learning experiences and enhancing educational outcomes by providing recommendations of personalized resources and activities to learners, tailored to their individual learning needs. However, their effectiveness is often diminished by insufficient user control and limited transparency. To address these challenges, in this paper, we present the systematic design and evaluation of an interactive ERS, in which we introduce different levels of user control. Concretely, we introduce user control around the input (i.e., user profile), process (i.e., recommendation algorithm), and output (i.e., recommendations) of the ERS. To evaluate our system, we conducted an online user study (N=30) to explore the impact of user control on users' perceptions of the ERS in terms of several important user-centric aspects. Moreover, we investigated the effects of user control on multiple recommendation goals, namely transparency, trust, and satisfaction, as well as the interactions between these goals. Our results demonstrate the positive impact of user control on user perceived benefits of the ERS. Moreover, our study shows that user control strongly correlates with transparency and moderately correlates with trust and satisfaction. In terms of interaction between these goals, our results reveal that transparency moderately correlates and trust strongly correlates with satisfaction. Whereas, transparency and trust stand out as less correlated with each other., Comment: Published in IntRS'24: Joint Workshop on Interfaces and Human Decision Making for Recommender Systems, October 18, 2024, Bari (Italy)

Scaling the capacity of language models has consistently proven to be a reliable approach for improving performance and unlocking new capabilities. Capacity can be primarily defined by two dimensions: the number of model parameters and the compute per example. While scaling typically involves increasing both, the precise interplay between these factors and their combined contribution to overall capacity remains not fully understood. We explore this relationship in the context of sparse Mixture-of-Experts (MoEs), which allow scaling the number of parameters without proportionally increasing the FLOPs per example. We investigate how varying the sparsity level, i.e., the fraction of inactive parameters, impacts model's performance during pretraining and downstream few-shot evaluation. We find that under different constraints (e.g., parameter size and total training compute), there is an optimal level of sparsity that improves both training efficiency and model performance. These results provide a better understanding of the impact of sparsity in scaling laws for MoEs and complement existing works in this area, offering insights for designing more efficient architectures.

Prostate cancer (PCa) detection using deep learning (DL) models has shown potential for enhancing real-time guidance during biopsies. However, prostate ultrasound images lack pixel-level cancer annotations, introducing label noise. Current approaches often focus on limited regions of interest (ROIs), disregarding anatomical context necessary for accurate diagnosis. Foundation models can overcome this limitation by analyzing entire images to capture global spatial relationships; however, they still encounter challenges stemming from the weak labels associated with coarse pathology annotations in ultrasound data. We introduce Cinepro, a novel framework that strengthens foundation models' ability to localize PCa in ultrasound cineloops. Cinepro adapts robust training by integrating the proportion of cancer tissue reported by pathology in a biopsy core into its loss function to address label noise, providing a more nuanced supervision. Additionally, it leverages temporal data across multiple frames to apply robust augmentations, enhancing the model's ability to learn stable cancer-related features. Cinepro demonstrates superior performance on a multi-center prostate ultrasound dataset, achieving an AUROC of 77.1% and a balanced accuracy of 83.8%, surpassing current benchmarks. These findings underscore Cinepro's promise in advancing foundation models for weakly labeled ultrasound data., Comment: accepted to IEEE ISBI 2025

Femtosecond laser light can transfer spin angular momentum between magnetic subspecies that exhibit hybridized valence bands within an alloy or compound, and represents the fastest route for manipulating the magnetization of a material. To date, ultrafast spin transfer has predominantly been explained in terms of the initial and final states available for laser excitation. Here, by comparing the measured and calculated dynamics across the entire $M$-edges of two very similar Heusler compounds, $Co_2MnGa$ and $Co_2MnGe$ as well as a sample of elemental Co, we find that simply accounting for the initial and final electron states available for laser excitation cannot alone explain the experimental observations. The influence of spin lifetimes must also be included, due to the shifting of the Fermi level upon replacing Ga with Ge, or the presence of crystalline disorder. This explains why the ordered $L2_1$ phase of $Co_2MnGa$ demonstrates strong laser-induced magnetic signal enhancements across the entire Co-edge, while similar enhancements were not observed in partially disordered $Co_2MnGe$. Although intra-site spin-transfers were expected in the minority channel in pure Co due to the presence of many more available states in the minority channel above the Fermi level, no such signal was observed due to very short few-femtosecond spin lifetimes in a metal. Finally, we identify key regions in the magnetic asymmetry where a transiently enhanced signal could be misinterpreted as a light-induced spin-transfer signature.

The Internet of Things (IoT) establishes connectivity between billions of heterogeneous devices that provide a variety of essential everyday services. The IoT faces several challenges, including energy efficiency and scalability, that require consideration of enabling technologies such as network softwarization. This technology is an appropriate solution for IoT, leveraging Software Defined Networking (SDN) and Network Function Virtualization (NFV) as two main techniques, especially when combined with Machine Learning (ML). Although many efforts have been made to optimize routing in softwarized IoT, the existing solutions do not take advantage of distributed intelligence. In this paper, we propose to optimize routing in softwarized IoT networks using Federated Deep Reinforcement Learning (FDRL), where distributed network softwarization and intelligence (i.e., FDRL) join forces to improve routing in constrained IoT networks. Our proposal introduces the combination of two novelties (i.e., distributed controller design and intelligent routing) to meet the IoT requirements (mainly performance and energy efficiency). The simulation results confirm the effectiveness of our proposal compared to the conventional counterparts.

The generalization performance of deep neural networks (DNNs) is a critical factor in achieving robust model behavior on unseen data. Recent studies have highlighted the importance of sharpness-based measures in promoting generalization by encouraging convergence to flatter minima. Among these approaches, Sharpness-Aware Minimization (SAM) has emerged as an effective optimization technique for reducing the sharpness of the loss landscape, thereby improving generalization. However, SAM's computational overhead and sensitivity to noisy gradients limit its scalability and efficiency. To address these challenges, we propose Gradient-Centralized Sharpness-Aware Minimization (GCSAM), which incorporates Gradient Centralization (GC) to stabilize gradients and accelerate convergence. GCSAM normalizes gradients before the ascent step, reducing noise and variance, and improving stability during training. Our evaluations indicate that GCSAM consistently outperforms SAM and the Adam optimizer in terms of generalization and computational efficiency. These findings demonstrate GCSAM's effectiveness across diverse domains, including general and medical imaging tasks.

We present an Arabic crossword puzzle generator from a given text that utilizes advanced language models such as GPT-4-Turbo, GPT-3.5-Turbo and Llama3-8B-Instruct, specifically developed for educational purposes, this innovative generator leverages a meticulously compiled dataset named Arabic-Clue-Instruct with over 50,000 entries encompassing text, answers, clues, and categories. This dataset is intricately designed to aid in the generation of pertinent clues linked to specific texts and keywords within defined categories. This project addresses the scarcity of advanced educational tools tailored for the Arabic language, promoting enhanced language learning and cognitive development. By providing a culturally and linguistically relevant tool, our objective is to make learning more engaging and effective through gamification and interactivity. Integrating state-of-the-art artificial intelligence with contemporary learning methodologies, this tool can generate crossword puzzles from any given educational text, thereby facilitating an interactive and enjoyable learning experience. This tool not only advances educational paradigms but also sets a new standard in interactive and cognitive learning technologies. The model and dataset are publicly available., Comment: This paper has been accepted for presentation at LoResLM @ COLING 2025

Multimodal Large Language Models (MLLMs) have achieved significant advancements in tasks like Visual Question Answering (VQA) by leveraging foundational Large Language Models (LLMs). However, their abilities in specific areas such as temporal understanding, which is crucial for comprehending real-world dynamics, remain underexplored. To address this, we propose a challenging evaluation benchmark named TemporalVQA, consisting of two parts: (1) Temporal Order Understanding and (2) Time-lapse Estimation. The first part requires MLLMs to determine the sequence of events by analyzing temporally consecutive video frames. The second part presents image pairs with varying time differences, framed as multiple-choice questions, asking MLLMs to estimate the time-lapse between images with options ranging from seconds to years. Our evaluations of advanced MLLMs, including models like GPT-4o and Gemini-1.5-Pro, reveal significant challenges: GPT-4o achieved only 43.8% average consistent accuracy in temporal order tasks and 70% in time-lapse estimation, with open-source models performing even less effectively. These findings underscore the limitations of current MLLMs in visual temporal understanding and reasoning, highlighting the need for further improvements in their temporal capabilities. Our dataset can be found at https://huggingface.co/datasets/fazliimam/temporal-vqa., Comment: Our dataset can be found at \url{https://huggingface.co/datasets/fazliimam/temporal-vqa}