Search

Large Language Models (LLMs) offer remarkable capabilities in code generation, natural language processing, and domain-specific reasoning. Their potential in aiding quantum software development remains underexplored, particularly for the PennyLane framework-a leading platform for hybrid quantum-classical computing. To address this gap, we introduce a novel, high-quality dataset comprising 3,347 PennyLane-specific code samples of quantum circuits and their contextual descriptions, specifically curated to train/fine-tune LLM-based quantum code assistance. Our key contributions are threefold: (1) the automatic creation and open-source release of a comprehensive PennyLane dataset leveraging quantum computing textbooks, official documentation, and open-source repositories; (2) the development of a systematic methodology for data refinement, annotation, and formatting to optimize LLM training efficiency; and (3) a thorough evaluation, based on a Retrieval-Augmented Generation (RAG) framework, demonstrating the effectiveness of our dataset in streamlining PennyLane code generation and improving quantum development workflows. Compared to existing efforts that predominantly focus on Qiskit, our dataset significantly broadens the spectrum of quantum frameworks covered in AI-driven code assistance. By bridging this gap and providing reproducible dataset-creation methodologies, we aim to advance the field of AI-assisted quantum programming, making quantum computing more accessible to both newcomers and experienced developers., Comment: 10 pages, 8 figures, 6 tables, submitted for review under IJCNN 2025

Enhancing the transport of oxygen anions in the cathode while maintaining surface stability is essential for improving the performance of intermediate-temperature solid oxide fuel cells (IT-SOFCs). This study investigates a novel cathode material candidate, Sr2CoNbO6-delta (SCNO), using density functional theory, molecular dynamics, and experimental characterization. The redox active Co cation at B-site and less reducible Nb cation at the B'-site together enhance both surface stability and electrocatalytic performance. SCNO is observed to have a higher concentration of oxygen vacancies and increased oxygen diffusivity on the surface. The surface stability of SCNO is further improved when simulated under compressive strain due to the GDC electrolyte substrate. These findings offer new insights into controlling Sr segregation in SCNO, contributing to a better understanding of its enhanced oxygen reduction reaction (ORR) activity and high surface stability. Subsequently, SCNO was synthesized to evaluate its potential as a cathode material in SOFCs. To assess its performance, symmetric cells with uniform dense thin films of varying thicknesses (40 and 80 nm) were fabricated using the pulsed laser deposition technique. Electrochemical impedance spectroscopy and distributed relaxation time analysis indicate that bulk oxygen ion diffusion is a limiting factor for the ORR in SCNO. The polarization resistance for the 40 and 80 nm dense thin film symmetric cells ranged between 0.329 - 0.241 ohm cm2 and 1.095 - 0.438 ohm cm2, respectively, within the temperature range of 773 - 973 K in an air atmosphere. The full cell configuration of NiO-GDC|GDC|SCNO demonstrated a significantly high peak power density of 0.633 W/cm2 at 973 K. This theory-guided design and experimental study suggest that SCNO is a promising candidate for IT-SOFC cathode materials., Comment: 31 pages, 9 figures, 4 tables

Automating cloud configuration and deployment remains a critical challenge due to evolving infrastructures, heterogeneous hardware, and fluctuating workloads. Existing solutions lack adaptability and require extensive manual tuning, leading to inefficiencies and misconfigurations. We introduce LADs, the first LLM-driven framework designed to tackle these challenges by ensuring robustness, adaptability, and efficiency in automated cloud management. Instead of merely applying existing techniques, LADs provides a principled approach to configuration optimization through in-depth analysis of what optimization works under which conditions. By leveraging Retrieval-Augmented Generation, Few-Shot Learning, Chain-of-Thought, and Feedback-Based Prompt Chaining, LADs generates accurate configurations and learns from deployment failures to iteratively refine system settings. Our findings reveal key insights into the trade-offs between performance, cost, and scalability, helping practitioners determine the right strategies for different deployment scenarios. For instance, we demonstrate how prompt chaining-based adaptive feedback loops enhance fault tolerance in multi-tenant environments and how structured log analysis with example shots improves configuration accuracy. Through extensive evaluations, LADs reduces manual effort, optimizes resource utilization, and improves system reliability. By open-sourcing LADs, we aim to drive further innovation in AI-powered DevOps automation., Comment: 17 pages with Appendix, 8 figures, and 7 tables. This paper is currently Under Review

Machine Learning (ML) techniques have been successfully applied to design various learned database index structures for both the one- and multi-dimensional spaces. Particularly, a class of traditional multi-dimensional indexes has been augmented with ML models to design ML-enhanced variants of their traditional counterparts. This paper focuses on the R-tree multi-dimensional index structure as it is widely used for indexing multi-dimensional data. The R-tree has been augmented with machine learning models to enhance the R-tree performance. The AI+R-tree is an ML-enhanced R-tree index structure that augments a traditional disk-based R-tree with an ML model to enhance the R-tree's query processing performance, mainly, to avoid navigating the overlapping branches of the R-tree that do not yield query results, e.g., in the presence of high-overlap among the rectangles of the R-tree nodes. We investigate the empirical tradeoffs in processing dynamic query workloads and in supporting updates over the AI+R-tree. Particularly, we investigate the impact of the choice of ML models over the AI+R-tree query processing performance. Moreover, we present a case study of designing a custom loss function for a neural network model tailored to the query processing requirements of the AI+R-tree. Furthermore, we present the design tradeoffs for adopting various strategies for supporting dynamic inserts, updates, and deletes with the vision of realizing a mutable AI+R-tree. Experiments on real datasets demonstrate that the AI+R-tree can enhance the query processing performance of a traditional R-tree for high-overlap range queries by up to 5.4X while achieving up to 99% average query recall., Comment: arXiv admin note: text overlap with arXiv:2207.00550

The injectivity of ReLU layers in neural networks, the recovery of vectors from clipped or saturated measurements, and (real) phase retrieval in $\mathbb{R}^n$ allow for a similar problem formulation and characterization using frame theory. In this paper, we revisit all three problems with a unified perspective and derive lower Lipschitz bounds for ReLU layers and clipping which are analogous to the previously known result for phase retrieval and are optimal up to a constant factor., Comment: 22 pages

Mainstream media, through their decisions on what to cover and how to frame the stories they cover, can mislead readers without using outright falsehoods. Therefore, it is crucial to have tools that expose these editorial choices underlying media bias. In this paper, we introduce the Media Bias Detector, a tool for researchers, journalists, and news consumers. By integrating large language models, we provide near real-time granular insights into the topics, tone, political lean, and facts of news articles aggregated to the publisher level. We assessed the tool's impact by interviewing 13 experts from journalism, communications, and political science, revealing key insights into usability and functionality, practical applications, and AI's role in powering media bias tools. We explored this in more depth with a follow-up survey of 150 news consumers. This work highlights opportunities for AI-driven tools that empower users to critically engage with media content, particularly in politically charged environments.

In this study, we evaluate methods to determine the frequency of species via quantity estimation from historical survey text. To that end, we formulate classification tasks and finally show that this problem can be adequately framed as a regression task using Best-Worst Scaling (BWS) with Large Language Models (LLMs). We test Ministral-8B, DeepSeek-V3, and GPT-4, finding that the latter two have reasonable agreement with humans and each other. We conclude that this approach is more cost-effective and similarly robust compared to a fine-grained multi-class approach, allowing automated quantity estimation across species., Comment: NoDaLiDa 2025, EcoNLP Workshop

Interpreting and controlling the internal mechanisms of large language models (LLMs) is crucial for improving their trustworthiness and utility. Recent efforts have primarily focused on identifying and manipulating neurons by establishing discrete mappings between neurons and semantic concepts. However, such mappings struggle to handle the inherent polysemanticity in LLMs, where individual neurons encode multiple, distinct concepts. This makes precise control challenging and complicates downstream interventions. Through an in-depth analysis of both encoder and decoder-based LLMs across multiple text classification datasets, we uncover that while individual neurons encode multiple concepts, their activation magnitudes vary across concepts in distinct, Gaussian-like patterns. Building on this insight, we introduce NeuronLens, a novel range-based interpretation and manipulation framework that provides a finer view of neuron activation distributions to localize concept attribution within a neuron. Extensive empirical evaluations demonstrate that NeuronLens significantly reduces unintended interference, while maintaining precise control for manipulation of targeted concepts, outperforming existing methods.

This paper investigates the significance of designing a reliable, intelligent, and true physical environment-aware precoding scheme by leveraging an accurately designed channel twin model to obtain realistic channel state information (CSI) for cellular communication systems. Specifically, we propose a fine-tuned multi-step channel twin design process that can render CSI very close to the CSI of the actual environment. After generating a precise CSI, we execute precoding using the obtained CSI at the transmitter end. We demonstrate a two-step parameters' tuning approach to design channel twin by ray tracing (RT) emulation, then further fine-tuning of CSI by employing an artificial intelligence (AI) based algorithm can significantly reduce the gap between actual CSI and the fine-tuned digital twin (DT) rendered CSI. The simulation results show the effectiveness of the proposed novel approach in designing a true physical environment-aware channel twin model.

This paper investigates deteriorations in knee and ankle dynamics during running. Changes in lower limb accelerations are analyzed by a wearable musculo-skeletal monitoring system. The system employs a machine learning technique to classify joint stiffness. A maximum-entropyrate method is developed to select the most relevant features. Experimental results demonstrate that distance travelled and energy expended can be estimated from observed changes in knee and ankle motions during 5 km runs.

Background: Multiparametric breast MRI data might improve tumor diagnostics, characterization, and treatment planning. Accurate alignment and delineation of images acquired at different field strengths such as 3T and 7T, remain challenging research tasks. Purpose: To address alignment challenges and enable consistent tumor segmentation across different MRI field strengths. Study type: Retrospective. Subjects: Nine female subjects with breast tumors were involved: six histologically proven invasive ductal carcinomas (IDC) and three fibroadenomas. Field strength/sequence: Imaging was performed at 3T and 7T scanners using post-contrast T1-weighted three-dimensional time-resolved angiography with stochastic trajectories (TWIST) sequence. Assessments: The method's performance for joint image registration and tumor segmentation was evaluated using several quantitative metrics, including signal-to-noise ratio (PSNR), structural similarity index (SSIM), normalized cross-correlation (NCC), Dice coefficient, F1 score, and relative sum of squared differences (rel SSD). Statistical tests: The Pearson correlation coefficient was used to test the relationship between the registration and segmentation metrics. Results: When calculated for each subject individually, the PSNR was in a range from 27.5 to 34.5 dB, and the SSIM was from 82.6 to 92.8%. The model achieved an NCC from 96.4 to 99.3% and a Dice coefficient of 62.9 to 95.3%. The F1 score was between 55.4 and 93.2% and the rel SSD was in the range of 2.0 and 7.5%. The segmentation metrics Dice and F1 Score are highly correlated (0.995), while a moderate correlation between NCC and SSIM (0.681) was found for registration. Data conclusion: Initial results demonstrate that the proposed method may be feasible in providing joint tumor segmentation and registration of MRI data acquired at different field strengths.

The Ontology Lookup Service (OLS) is an open source search engine for ontologies which is used extensively in the bioinformatics and chemistry communities to annotate biological and biomedical data with ontology terms. Recently there has been a significant increase in the size and complexity of ontologies due to new scales of biological knowledge, such as spatial transcriptomics, new ontology development methodologies, and curation on an increased scale. Existing Web-based tools for ontology browsing such as BioPortal and OntoBee do not support the full range of definitions used by today's ontologies. In order to support the community going forward, we have developed OLS4, implementing the complete OWL2 specification, internationalization support for multiple languages, and a new user interface with UX enhancements such as links out to external databases. OLS4 has replaced OLS3 in production at EMBL-EBI and has a backwards compatible API supporting users of OLS3 to transition., Comment: 4 pages plus references

This paper presents an efficient decoding approach for end-to-end automatic speech recognition (E2E-ASR) with large language models (LLMs). Although shallow fusion is the most common approach to incorporate language models into E2E-ASR decoding, we face two practical problems with LLMs. (1) LLM inference is computationally costly. (2) There may be a vocabulary mismatch between the ASR model and the LLM. To resolve this mismatch, we need to retrain the ASR model and/or the LLM, which is at best time-consuming and in many cases not feasible. We propose "delayed fusion," which applies LLM scores to ASR hypotheses with a delay during decoding and enables easier use of pre-trained LLMs in ASR tasks. This method can reduce not only the number of hypotheses scored by the LLM but also the number of LLM inference calls. It also allows re-tokenizion of ASR hypotheses during decoding if ASR and LLM employ different tokenizations. We demonstrate that delayed fusion provides improved decoding speed and accuracy compared to shallow fusion and N-best rescoring using the LibriHeavy ASR corpus and three public LLMs, OpenLLaMA 3B & 7B and Mistral 7B., Comment: Accepted to ICASSP2025

This paper presents the principal challenges and opportunities associated with computational biomechanics research. The underlying cognitive control involved in the process of human motion is inherently complex, dynamic, multidimensional, and highly non-linear. The dynamics produced by the internal and external forces and the body's ability to react to them is biomechanics. Complex and non-rigid bodies, needs a lot of computing power and systems to execute however, in the absence of adequate resources, one may rely on new technology, machine learning tools and model order reduction approaches. It is also believed that machine learning approaches can enable us to embrace this complexity, if we could use three arms of ML i.e. predictive modeling, classification, and dimensionality reduction. Biomechanics, since it deals with motion and mobility come with a huge set of data over time. Using computational (Computer Solvers), Numerical approaches (MOR) and technological advances (Wearable sensors), can let us develop computationally inexpensive frameworks for biomechanics focused studies dealing with a huge amount of data. A lot of misunderstanding arises because of extensive data, standardization of the tools to process this, database for the material property definitions, validation and verification of biomechanical models and analytical tools to model various phenomena using computational and modelling techniques. Study of biomechanics through computational simulations can improve the prevention and treatment of diseases, predict the injury to reduce the risk and hence can strengthen pivotal sectors like sports and lifestyle. This is why we choose to present all those challenges and problems associated with biomechanical simulation with complex geometries fail so as to help improve, analysis, performance and design for better lifestyle., Comment: 9 pages, 5 figures, 1 table, and conference paper

Dementia is associated with various cognitive impairments and typically manifests only after significant progression, making intervention at this stage often ineffective. To address this issue, the Prediction and Recognition of Cognitive Decline through Spontaneous Speech (PROCESS) Signal Processing Grand Challenge invites participants to focus on early-stage dementia detection. We provide a new spontaneous speech corpus for this challenge. This corpus includes answers from three prompts designed by neurologists to better capture the cognition of speakers. Our baseline models achieved an F1-score of 55.0% on the classification task and an RMSE of 2.98 on the regression task., Comment: 2 pages, no figure, conference

Transmission Expansion Planning (TEP) is the process of optimizing the development and upgrade of the power grid to ensure reliable, efficient, and cost-effective electricity delivery while addressing grid constraints. To support growing demand and renewable energy integration, energy storage is emerging as a pivotal asset that provides temporal flexibility and alleviates congestion. This paper presents a TEP model that incorporates the sizing and siting of short-duration storage. With a focus on high spatial resolution, the model is applied to a 2,000-bus synthetic Texas power system, offering detailed insights into geographic investment and operational patterns. To maintain computational feasibility, a simple yet effective storage candidates (SC) method is introduced, significantly reducing the search space. Results highlight that transmission investments are primarily driven by renewable energy expansion, while storage investments are shaped by renewable curtailment and load-shedding events, with their primary function being peak load shaving. The findings underscore the importance of co-optimizing transmission and storage to minimize costs and enhance grid reliability. However, limitations in the ability of the SC method to identify optimal storage locations to meet long-term needs suggest opportunities for future research, including dynamic candidate selection and hybrid optimization techniques.

Link equivalence up to isotopy in a 3-space is the problem that lies at the root of knot theory, and is important in 3-dimensional topology and geometry. We consider its restriction to alternating links, given by two alternating diagrams with $n_1$ and $n_2$ crossings, and show that this problem has polynomial algorithm in terms of $max\{n_1, n_2\}$. For the proof, we use Tait flyping conjectures, observations from the work of Lackenby, Menasco, Sundberg and Thistlethwaite on alternating links, and algorithmic complexity of some problems from graph theory and topological graph theory., Comment: 12 pages, 3 figures

The standardization of an interface for dense linear algebra operations in the BLAS standard has enabled interoperability between different linear algebra libraries, thereby boosting the success of scientific computing, in particular in scientific HPC. Despite numerous efforts in the past, the community has not yet agreed on a standardization for sparse linear algebra operations due to numerous reasons. One is the fact that sparse linear algebra objects allow for many different storage formats, and different hardware may favor different storage formats. This makes the definition of a FORTRAN-style all-circumventing interface extremely challenging. Another reason is that opposed to dense linear algebra functionality, in sparse linear algebra, the size of the sparse data structure for the operation result is not always known prior to the information. Furthermore, as opposed to the standardization effort for dense linear algebra, we are late in the technology readiness cycle, and many production-ready software libraries using sparse linear algebra routines have implemented and committed to their own sparse BLAS interface. At the same time, there exists a demand for standardization that would improve interoperability, and sustainability, and allow for easier integration of building blocks. In an inclusive, cross-institutional effort involving numerous academic institutions, US National Labs, and industry, we spent two years designing a hardware-portable interface for basic sparse linear algebra functionality that serves the user needs and is compatible with the different interfaces currently used by different vendors. In this paper, we present a C++ API for sparse linear algebra functionality, discuss the design choices, and detail how software developers preserve a lot of freedom in terms of how to implement functionality behind this API., Comment: 43 pages

Matter-RADiation interaction SIMulation (MRADSIM) is an innovative modular software toolkit developed to simulate the effects of radiation on electronic components, human beings and various materials. It incorporates innovative features aimed at enhancing parametric precision, reducing computational time, and introducing supplementary functions for tailored calculations across diverse applications including the applications required for space missions. Notably, MRADSIM is distinguished as the pioneering simulation toolkit to integrate cutting-edge Artificial Intelligence and Machine Learning (AI/ML) algorithms, with the primary objective of effectively recognizing potential radiation-induced issues and facilitating the implementation of mitigation strategies to avert catastrophic failures in mission-critical systems, whether terrestrial or space-based. The distinctive attributes of MRADSIM, coupled with its early adoption by researchers from the National Institute for Nuclear Physics of Italy (INFN), significantly contribute to the toolkits added value., Comment: 11 pages, 12 figures, 3 tables, 75th International Astronautical Congress (IAC), Milan, Italy, 14-18 October

Generating accurate code review comments remains a significant challenge due to the inherently diverse and non-unique nature of the task output. Large language models pretrained on both programming and natural language data tend to perform well in code-oriented tasks. However, large-scale pretraining is not always feasible due to its environmental impact and project-specific generalizability issues. In this work, first we fine-tune open-source Large language models (LLM) in parameter-efficient, quantized low-rank (QLoRA) fashion on consumer-grade hardware to improve review comment generation. Recent studies demonstrate the efficacy of augmenting semantic metadata information into prompts to boost performance in other code-related tasks. To explore this in code review activities, we also prompt proprietary, closed-source LLMs augmenting the input code patch with function call graphs and code summaries. Both of our strategies improve the review comment generation performance, with function call graph augmented few-shot prompting on the GPT-3.5 model surpassing the pretrained baseline by around 90% BLEU-4 score on the CodeReviewer dataset. Moreover, few-shot prompted Gemini-1.0 Pro, QLoRA fine-tuned Code Llama and Llama 3.1 models achieve competitive results (ranging from 25% to 83% performance improvement) on this task. An additional human evaluation study further validates our experimental findings, reflecting real-world developers' perceptions of LLM-generated code review comments based on relevant qualitative metrics.

Objectives: Age and gender estimation is crucial for various applications, including forensic investigations and anthropological studies. This research aims to develop a predictive system for age and gender estimation in living individuals, leveraging dental measurements such as Coronal Height (CH), Coronal Pulp Cavity Height (CPCH), and Tooth Coronal Index (TCI). Methods: Machine learning models were employed in our study, including Cat Boost Classifier (Catboost), Gradient Boosting Machine (GBM), Ada Boost Classifier (AdaBoost), Random Forest (RF), eXtreme Gradient Boosting (XGB), Light Gradient Boosting Machine (LGB), and Extra Trees Classifier (ETC), to analyze dental data from 862 living individuals (459 males and 403 females). Specifically, periapical radiographs from six teeth per individual were utilized, including premolars and molars from both maxillary and mandibular. A novel ensemble learning technique was developed, which uses multiple models each tailored to distinct dental metrics, to estimate age and gender accurately. Furthermore, an explainable AI model has been created utilizing SHAP, enabling dental experts to make judicious decisions based on comprehensible insight. Results: The RF and XGB models were particularly effective, yielding the highest F1 score for age and gender estimation. Notably, the XGB model showed a slightly better performance in age estimation, achieving an F1 score of 73.26%. A similar trend for the RF model was also observed in gender estimation, achieving a F1 score of 77.53%. Conclusions: This study marks a significant advancement in dental forensic methods, showcasing the potential of machine learning to automate age and gender estimation processes with improved accuracy.

Severe collisions can result from aggressive driving and poor road conditions, emphasizing the need for effective monitoring to ensure safety. Smartphones, with their array of built-in sensors, offer a practical and affordable solution for road-sensing. However, the lack of reliable, standardized datasets has hindered progress in assessing road conditions and driving patterns. This study addresses this gap by introducing a comprehensive dataset derived from smartphone sensors, which surpasses existing datasets by incorporating a diverse range of sensors including accelerometer, gyroscope, magnetometer, GPS, gravity, orientation, and uncalibrated sensors. These sensors capture extensive parameters such as acceleration force, gravitation, rotation rate, magnetic field strength, and vehicle speed, providing a detailed understanding of road conditions and driving behaviors. The dataset is designed to enhance road safety, infrastructure maintenance, traffic management, and urban planning. By making this dataset available to the community, the study aims to foster collaboration, inspire further research, and facilitate the development of innovative solutions in intelligent transportation systems., Comment: 29 pages, 14 Figures, journal paper, submitted into Scientific Data Journal

Incorporating advanced information and communication technologies into smart grids (SGs) offers substantial operational benefits while increasing vulnerability to cyber threats like false data injection (FDI) attacks. Current SG attack analysis tools predominantly employ formal methods or adversarial machine learning (ML) techniques with rule-based bad data detectors to analyze the attack space. However, these attack analytics either generate simplistic attack vectors detectable by the ML-based anomaly detection models (ADMs) or fail to identify critical attack vectors from complex controller dynamics in a feasible time. This paper introduces MISGUIDE, a novel defense-aware attack analytics designed to extract verifiable multi-time slot-based FDI attack vectors from complex SG load frequency control dynamics and ADMs, utilizing the Gurobi optimizer. MISGUIDE can identify optimal (maliciously triggering under/over frequency relays in minimal time) and stealthy attack vectors. Using real-world load data, we validate the MISGUIDE-identified attack vectors through real-time hardware-in-the-loop (OPALRT) simulations of the IEEE 39-bus system., Comment: 12 page journal

Integrated Sensing and Communication (ISAC) is a technology paradigm that combines sensing capabilities with communication functionalities in a single device or system. In vehicle-to-everything (V2X) sidelink, ISAC can provide enhanced safety by allowing vehicles to not only communicate with one another but also sense the surrounding environment by using sidelink signals. In ISAC-capable V2X sidelink, the random resource allocation results in an unstructured and sparse distribution of time and frequency resources in the received orthogonal frequency division multiplexing (OFDM) grid, leading to degraded radar detection performance when processed using the conventional 2D-FFT method. To address this challenge, this paper proposes a high-resolution off-grid radar target detection algorithm irrespective of the OFDM grid structure. The proposed method utilizes the Newtonized orthogonal matching pursuit (NOMP) algorithm to effectively detect weak targets masked by the sidelobes of stronger ones and accurately estimates off-grid range and velocity parameters with minimal resources through Newton refinements. Simulation results demonstrate the superior performance of the proposed NOMP-based target detection algorithm compared to existing compressed sensing (CS) methods in terms of detection probability, resolution, and accuracy. Additionally, experimental validation is performed using a bi-static radar setup in a semi-anechoic chamber. The measurement results validate the simulation findings, showing that the proposed algorithm significantly enhances target detection and parameter estimation accuracy in realistic scenarios.

We consider the problem of stragglers in distributed computing systems. Stragglers, which are compute nodes that unpredictably slow down, often increase the completion times of tasks. One common approach to mitigating stragglers is work replication, where only the first completion among replicated tasks is accepted, discarding the others. However, discarding work leads to resource wastage. In this paper, we propose a method for exploiting the work completed by stragglers rather than discarding it. The idea is to increase the granularity of the assigned work, and to increase the frequency of worker updates. We show that the proposed method reduces the completion time of tasks via experiments performed on a simulated cluster as well as on Amazon EC2 with Apache Hadoop., Comment: This paper has been accepted for publication in IEEE Transactions on Services Computing. The initial results presented in this paper appeared in the proceedings of the Allerton Conference on Communication, Control, and Computing in 2023

Mobile devices such as smartphones, laptops, and tablets can often connect to multiple access networks (e.g., Wi-Fi, LTE, and 5G) simultaneously. Recent advancements facilitate seamless integration of these connections below the transport layer, enhancing the experience for apps that lack inherent multi-path support. This optimization hinges on dynamically determining the traffic distribution across networks for each device, a process referred to as \textit{multi-access traffic splitting}. This paper introduces \textit{NetworkGym}, a high-fidelity network environment simulator that facilitates generating multiple network traffic flows and multi-access traffic splitting. This simulator facilitates training and evaluating different RL-based solutions for the multi-access traffic splitting problem. Our initial explorations demonstrate that the majority of existing state-of-the-art offline RL algorithms (e.g. CQL) fail to outperform certain hand-crafted heuristic policies on average. This illustrates the urgent need to evaluate offline RL algorithms against a broader range of benchmarks, rather than relying solely on popular ones such as D4RL. We also propose an extension to the TD3+BC algorithm, named Pessimistic TD3 (PTD3), and demonstrate that it outperforms many state-of-the-art offline RL algorithms. PTD3's behavioral constraint mechanism, which relies on value-function pessimism, is theoretically motivated and relatively simple to implement., Comment: NeurIPS (Datasets and Benchmarks)

The safe and effective deployment of Large Language Models (LLMs) involves a critical step called alignment, which ensures that the model's responses are in accordance with human preferences. Prevalent alignment techniques, such as DPO, PPO and their variants, align LLMs by changing the pre-trained model weights during a phase called post-training. While predominant, these post-training methods add substantial complexity before LLMs can be deployed. Inference-time alignment methods avoid the complex post-training step and instead bias the generation towards responses that are aligned with human preferences. The best-known inference-time alignment method, called Best-of-N, is as effective as the state-of-the-art post-training procedures. Unfortunately, Best-of-N requires vastly more resources at inference time than standard decoding strategies, which makes it computationally not viable. In this work, we introduce Speculative Rejection, a computationally-viable inference-time alignment algorithm. It generates high-scoring responses according to a given reward model, like Best-of-N does, while being between 16 to 32 times more computationally efficient., Comment: NeurIPS 2024

Metallic materials such as brass, copper, and aluminum are used in numerous applications, including industrial manufacturing. The vibration characteristics of these objects are unique and can be used to identify these objects from a distance. This research presents a methodology for detecting and classifying these metallic objects using the vibration dynamics induced by their micro-Doppler signatures. The proposed approach utilizes image processing techniques to extract pivotal features from spectrograms. These spectrograms originate from micro-Doppler signatures of data collected during controlled laboratory experiments where signals were transmitted towards vibrating metal sheets, and the ensuing reflections were recorded using a software-defined radio (SDR). The spectrogram data was augmented using geometric transformation to train a convolutional neural network (CNN) based machine learning model for object classification. The results indicate that the proposed CNN model achieved an accuracy of more than 95% in classifying metals into brass, copper, and aluminum. This research could be used to understand the foundations of classifying spectrogram images using micro-Doppler signatures for its applications towards enhancing the sensing capabilities in industrial and defense applications., Comment: 11 pages, to be published in the May 2025 issue of the IEEE Instrumentation & Measurement Magazine

Effective communication within universities is crucial for addressing the diverse information needs of students, alumni, and external stakeholders. However, existing chatbot systems often fail to deliver accurate, context-specific responses, resulting in poor user experiences. In this paper, we present KatzBot, an innovative chatbot powered by KatzGPT, a custom Large Language Model (LLM) fine-tuned on domain-specific academic data. KatzGPT is trained on two university-specific datasets: 6,280 sentence-completion pairs and 7,330 question-answer pairs. KatzBot outperforms established existing open source LLMs, achieving higher accuracy and domain relevance. KatzBot offers a user-friendly interface, significantly enhancing user satisfaction in real-world applications. The source code is publicly available at \url{https://github.com/AiAI-99/katzbot}.

Visual Question Answer (VQA) poses the problem of answering a natural language question about a visual context. Bangla, despite being a widely spoken language, is considered low-resource in the realm of VQA due to the lack of a proper benchmark dataset. The absence of such datasets challenges models that are known to be performant in other languages. Furthermore, existing Bangla VQA datasets offer little cultural relevance and are largely adapted from their foreign counterparts. To address these challenges, we introduce a large-scale Bangla VQA dataset titled ChitroJera, totaling over 15k samples where diverse and locally relevant data sources are used. We assess the performance of text encoders, image encoders, multimodal models, and our novel dual-encoder models. The experiments reveal that the pre-trained dual-encoders outperform other models of its scale. We also evaluate the performance of large language models (LLMs) using prompt-based techniques, with LLMs achieving the best performance. Given the underdeveloped state of existing datasets, we envision ChitroJera expanding the scope of Vision-Language tasks in Bangla.

The proliferation of transliterated texts in digital spaces has emphasized the need for detecting and classifying hate speech in languages beyond English, particularly in low-resource languages. As online discourse can perpetuate discrimination based on target groups, e.g. gender, religion, and origin, multi-label classification of hateful content can help in comprehending hate motivation and enhance content moderation. While previous efforts have focused on monolingual or binary hate classification tasks, no work has yet addressed the challenge of multi-label hate speech classification in transliterated Bangla. We introduce BanTH, the first multi-label transliterated Bangla hate speech dataset comprising 37.3k samples. The samples are sourced from YouTube comments, where each instance is labeled with one or more target groups, reflecting the regional demographic. We establish novel transformer encoder-based baselines by further pre-training on transliterated Bangla corpus. We also propose a novel translation-based LLM prompting strategy for transliterated text. Experiments reveal that our further pre-trained encoders are achieving state-of-the-art performance on the BanTH dataset, while our translation-based prompting outperforms other strategies in the zero-shot setting. The introduction of BanTH not only fills a critical gap in hate speech research for Bangla but also sets the stage for future exploration into code-mixed and multi-label classification challenges in underrepresented languages.

Text classification is a quintessential and practical problem in natural language processing with applications in diverse domains such as sentiment analysis, fake news detection, medical diagnosis, and document classification. A sizable body of recent works exists where researchers have studied and tackled text classification from different angles with varying degrees of success. Graph convolution network (GCN)-based approaches have gained a lot of traction in this domain over the last decade with many implementations achieving state-of-the-art performance in more recent literature and thus, warranting the need for an updated survey. This work aims to summarize and categorize various GCN-based Text Classification approaches with regard to the architecture and mode of supervision. It identifies their strengths and limitations and compares their performance on various benchmark datasets. We also discuss future research directions and the challenges that exist in this domain.

We study extreme events in a finite-size 2D Abelian sandpile model, specifically focusing on avalanche area and size. Employing the approach of Block Maxima, the study numerically reveals that the rescaled distributions for the largest avalanche size and area converge into the Gumbel and Weibull family of Generalized Extreme Value (GEV) distributions respectively. Numerically, we propose scaling functions for extreme avalanche activities that connect the activities on different length scales. With the help of data collapse, we estimate the precise values of these scaling exponents. The scaling function provides an understanding of the intricate dynamics within the sandpile model, shedding light on the relationship between system size and extreme event characteristics. The findings presented in this paper give valuable insights into the extreme behaviour of the Abelian sandpile model and offer a framework to understand the statistical properties of extreme events in complex systems., Comment: 7 pages, 4 figures

The paradigm of retrieval-augmented generated (RAG) helps mitigate hallucinations of large language models (LLMs). However, RAG also introduces biases contained within the retrieved documents. These biases can be amplified in scenarios which are multilingual and culturally-sensitive, such as territorial disputes. In this paper, we introduce BordIRLines, a benchmark consisting of 720 territorial dispute queries paired with 14k Wikipedia documents across 49 languages. To evaluate LLMs' cross-lingual robustness for this task, we formalize several modes for multilingual retrieval. Our experiments on several LLMs reveal that retrieving multilingual documents best improves response consistency and decreases geopolitical bias over using purely in-language documents, showing how incorporating diverse perspectives improves robustness. Also, querying in low-resource languages displays a much wider variance in the linguistic distribution of response citations. Our further experiments and case studies investigate how cross-lingual RAG is affected by aspects from IR to document contents. We release our benchmark and code to support further research towards ensuring equitable information access across languages at https://huggingface.co/datasets/borderlines/bordirlines.

Maintaining numerical stability in machine learning models is crucial for their reliability and performance. One approach to maintain stability of a network layer is to integrate the condition number of the weight matrix as a regularizing term into the optimization algorithm. However, due to its discontinuous nature and lack of differentiability the condition number is not suitable for a gradient descent approach. This paper introduces a novel regularizer that is provably differentiable almost everywhere and promotes matrices with low condition numbers. In particular, we derive a formula for the gradient of this regularizer which can be easily implemented and integrated into existing optimization algorithms. We show the advantages of this approach for noisy classification and denoising of MNIST images., Comment: Accepted at ICML24 Workshop: Differentiable Almost Everything: Differentiable Relaxations, Algorithms, Operators, and Simulators

Personalized Federated Learning (pFL) holds immense promise for tailoring machine learning models to individual users while preserving data privacy. However, achieving optimal performance in pFL often requires a careful balancing act between memory overhead costs and model accuracy. This paper delves into the trade-offs inherent in pFL, offering valuable insights for selecting the right algorithms for diverse real-world scenarios. We empirically evaluate ten prominent pFL techniques across various datasets and data splits, uncovering significant differences in their performance. Our study reveals interesting insights into how pFL methods that utilize personalized (local) aggregation exhibit the fastest convergence due to their efficiency in communication and computation. Conversely, fine-tuning methods face limitations in handling data heterogeneity and potential adversarial attacks while multi-objective learning methods achieve higher accuracy at the cost of additional training and resource consumption. Our study emphasizes the critical role of communication efficiency in scaling pFL, demonstrating how it can significantly affect resource usage in real-world deployments.

Mid-infrared (MIR) spectroscopy is a powerful technique employed for a variety of applications, including gas sensing, industrial inspection, astronomy, surveillance, and imaging. Thin-film narrowband interference filters, targeted to specific absorption bands of target molecules, are commonly deployed for cost-effective MIR sensing systems. These devices require complex and time-consuming fabrication processes. Also, their customization on the micro-scale for emerging miniaturized applications is challenging. Plasmonic nanostructure arrays operating in reflection and transmission modes have been developed for MIR. However, they experience undesirable characteristics, such as broad spectra and low reflection/transmission efficiencies. All-dielectric metasurfaces have low intrinsic losses and have emerged as a substitute for plasmonic metasurfaces in MIR spectroscopy. Nevertheless, they typically operate only in reflection mode. In this work, we present a hybrid metal-dielectric metasurface for MIR spectroscopy operating in transmission mode. The metasurface is composed of germanium (Ge) atop aluminum (Al) cylinders, and we show that the transmission response arises because of the hybridization of modes arising from the Ge and the Al structures. The presented metasurface has a high transmission efficiency of 80 % at $\lambda = 2.6\ \mu\text{m}$, and a narrow full-width-at-half-maximum of $0.4\ \mu\text{m}$. We show numerical simulations, successful fabrication using a straightforward fabrication method, and deployment as the in-line optical filter in a CO$_2$ gas detection with a limit of detection of ~0.04% (a few hundred ppm). Our work demonstrates the potential for hybrid metasurfaces as in-line gas sensing optical filters in MIR spectroscopy.

We conduct a comprehensive \textit{ab initio} investigation of electron-electron interactions within the pyrochlore structures of R$_2$Ru$_2$O$_7$, R$_2$Ir$_2$O$_7$, Ca$_2$Ru$_2$O$_7$, and Cd$_2$Ru$_2$O$_7$, where R denotes a rare-earth element. Utilizing a multiorbital Hubbard model, we systematically explore the effects of various rare-earth elements and applied high pressure on the correlation strength in these compounds. Our calculations on the Coulomb interaction parameter $U$ and the bandwidth $W$ reveal that the chemical pressure for R$_2$Ru$_2$O$_7$ and R$_2$Ir$_2$O$_7$ leads to an unusual increase in $U/W$ ratio, hence, increase in correlation strength. Contrary to conventional understanding of bandwidth control, our study identifies that the Hubbard $U$ is more influential than the bandwidth $W$ behind the metal-insulator landscape of R$_2$Ru$_2$O$_7$ and R$_2$Ir$_2$O$_7$, leading to an interaction-controlled metal-insulator transition. We also find unexpected behavior in physical pressure. Whereas physical pressure leads to a decrease in the correlation strength $U/W$ as usual in R$_2$Ru$_2$O$_7$, the effect is notably small in Ca$_2$Ru$_2$O$_7$ and Cd$_2$Ru$_2$O$_7$, which provides an important clue to understanding unusual pressure-induced metal-insulator transition observed experimentally.