Search

Esports offers a platform for players to engage in competitive and cooperative gaming with others remotely via the Internet. Despite these opportunities for social interaction, many players may still experience loneliness while playing online games. This study aims to enhance the social presence of partner players during online gameplay. The demonstration system, designed for 1-on-1 online competitive games, mutually transmits the partner's biosignals, through heartbeat-like vibrotactile stimuli. The system generates vibrotactile signals that represent two-dimensional emotions, arousal and valence, based on biosignals such as heart rate and electrodermal activity., Comment: Part of proceedings of 6th International Conference AsiaHaptics 2024

Dynamical quantum systems both driven by unitary evolutions and monitored through measurements have proved to be fertile ground for exploring new dynamical quantum matters. While the entanglement structure and symmetry properties of monitored systems have been intensively studied, the role of topology in monitored dynamics is much less explored. In this work, we investigate novel topological phenomena in the monitored dynamics through the lens of free-fermion systems. Free-fermion monitored dynamics were previously shown to be unified with the Anderson localization problem under the Altland-Zirnbauer symmetry classification. Guided by this unification, we identify the topological area-law-entangled phases in the former setting through the topological classification of disordered insulators and superconductors in the latter. As examples, we focus on 1+1D free-fermion monitored dynamics in two symmetry classes, DIII and A. We construct quantum circuit models to study different topological area-law phases and their domain walls in the respective symmetry classes. We find that the domain wall between topologically distinct area-law phases hosts dynamical topological modes whose entanglement is protected from being quenched by the measurements in the monitored dynamics. We demonstrate how to manipulate these topological modes by programming the domain-wall dynamics. In particular, for topological modes in class DIII, which behave as unmeasured Majorana modes, we devise a protocol to braid them and study the entanglement generated in the braiding process., Comment: 7 pages, 4 figures in the main text

In the complex realm of cognitive disorders, Alzheimer's disease (AD) and vascular dementia (VaD) are the two most prevalent dementia types, presenting entangled symptoms yet requiring distinct treatment approaches. The crux of effective treatment in slowing neurodegeneration lies in early, accurate diagnosis, as this significantly assists doctors in determining the appropriate course of action. However, current diagnostic practices often delay VaD diagnosis, impeding timely intervention and adversely affecting patient prognosis. This paper presents an innovative multi-omics approach to accurately differentiate AD from VaD, achieving a diagnostic accuracy of 89.25%. The proposed method segments the longitudinal MRI scans and extracts advanced radiomics features. Subsequently, it synergistically integrates the radiomics features with an ensemble of clinical, cognitive, and genetic data to provide state-of-the-art diagnostic accuracy, setting a new benchmark in classification accuracy on a large public dataset. The paper's primary contribution is proposing a comprehensive methodology utilizing multi-omics data to provide a nuanced understanding of dementia subtypes. Additionally, the paper introduces an interpretable model to enhance clinical decision-making coupled with a novel model architecture for evaluating treatment efficacy. These advancements lay the groundwork for future work not only aimed at improving differential diagnosis but also mitigating and preventing the progression of dementia.

In Petanque, each player aims to throw the boule closer to the jack. The closest boule to the jack among players will score the point. Currently, the distance of the boule to the jack is still measured using manual measurement tools such as measuring tape, string, and calipers. The manual measurement method is considered time-consuming and prone to inconsistent reading, which the ordinary referees and players conduct. A steady hand is required to hold the tape at two ends while squatting or kneeling. The technique of reading the measurement is also important to determine the accuracy of the length. This project aims to design and develop a prototype device that can measure the distance between jack and boule using a microcontroller and ultrasonic sensor technology. The device is expected to provide an instant measurement of the distance between the jack and the boule. The measurement data can be displayed on the mobile device to ease the user to view the result. This prototype device also counts the score points and determines the winner., Comment: 10 pages

As foundation models (FMs) play an increasingly prominent role in complex software systems, such as FM-powered agentic software (i.e., Agentware), they introduce significant challenges for developers regarding observability. Unlike traditional software, agents operate autonomously, using extensive data and opaque implicit reasoning, making it difficult to observe and understand their behavior during runtime, especially when they take unexpected actions or encounter errors. In this paper, we highlight the limitations of traditional operational observability in the context of FM-powered software, and introduce cognitive observability as a new type of required observability that has emerged for such innovative systems. We then propose a novel framework that provides cognitive observability into the implicit reasoning processes of agents (a.k.a. reasoning observability), and demonstrate the effectiveness of our framework in boosting the debuggability of Agentware and, in turn, the abilities of an Agentware through a case study on AutoCodeRover, a cuttingedge Agentware for autonomous program improvement.

Large language models (LLMs) have demonstrated remarkable capabilities in various scientific domains, from natural language processing to complex problem-solving tasks. Their ability to understand and generate human-like text has opened up new possibilities for advancing scientific research, enabling tasks such as data analysis, literature review, and even experimental design. One of the most promising applications of LLMs in this context is hypothesis generation, where they can identify novel research directions by analyzing existing knowledge. However, despite their potential, LLMs are prone to generating ``hallucinations'', outputs that are plausible-sounding but factually incorrect. Such a problem presents significant challenges in scientific fields that demand rigorous accuracy and verifiability, potentially leading to erroneous or misleading conclusions. To overcome these challenges, we propose KG-CoI (Knowledge Grounded Chain of Ideas), a novel system that enhances LLM hypothesis generation by integrating external, structured knowledge from knowledge graphs (KGs). KG-CoI guides LLMs through a structured reasoning process, organizing their output as a chain of ideas (CoI), and includes a KG-supported module for the detection of hallucinations. With experiments on our newly constructed hypothesis generation dataset, we demonstrate that KG-CoI not only improves the accuracy of LLM-generated hypotheses but also reduces the hallucination in their reasoning chains, highlighting its effectiveness in advancing real-world scientific research.

We study the problem of learning mixtures of Gaussians with approximate differential privacy. We prove that roughly $kd^2 + k^{1.5} d^{1.75} + k^2 d$ samples suffice to learn a mixture of $k$ arbitrary $d$-dimensional Gaussians up to low total variation distance, with differential privacy. Our work improves over the previous best result [AAL24b] (which required roughly $k^2 d^4$ samples) and is provably optimal when $d$ is much larger than $k^2$. Moreover, we give the first optimal bound for privately learning mixtures of $k$ univariate (i.e., $1$-dimensional) Gaussians. Importantly, we show that the sample complexity for privately learning mixtures of univariate Gaussians is linear in the number of components $k$, whereas the previous best sample complexity [AAL21] was quadratic in $k$. Our algorithms utilize various techniques, including the inverse sensitivity mechanism [AD20b, AD20a, HKMN23], sample compression for distributions [ABDH+20], and methods for bounding volumes of sumsets., Comment: 52 pages. To appear in Neural Information Processing Systems (NeurIPS), 2024

This paper introduces a real-time Vehicle Collision Avoidance System (V-CAS) designed to enhance vehicle safety through adaptive braking based on environmental perception. V-CAS leverages the advanced vision-based transformer model RT-DETR, DeepSORT tracking, speed estimation, brake light detection, and an adaptive braking mechanism. It computes a composite collision risk score based on vehicles' relative accelerations, distances, and detected braking actions, using brake light signals and trajectory data from multiple camera streams to improve scene perception. Implemented on the Jetson Orin Nano, V-CAS enables real-time collision risk assessment and proactive mitigation through adaptive braking. A comprehensive training process was conducted on various datasets for comparative analysis, followed by fine-tuning the selected object detection model using transfer learning. The system's effectiveness was rigorously evaluated on the Car Crash Dataset (CCD) from YouTube and through real-time experiments, achieving over 98% accuracy with an average proactive alert time of 1.13 seconds. Results indicate significant improvements in object detection and tracking, enhancing collision avoidance compared to traditional single-camera methods. This research demonstrates the potential of low-cost, multi-camera embedded vision transformer systems to advance automotive safety through enhanced environmental perception and proactive collision avoidance mechanisms., Comment: Accepted at ICMLA 2024

Attosecond science has leveraged the highly nonlinear interactions between intense few-cycle laser pulses and matter, allowing for unprecedented observation and control of electron motion with remarkable temporal resolution. However, most existing experiments focusing on laser-controlled attosecond dynamics have dealt with quasi-bound electrons released in the ionization continua of atoms, molecules, or conduction bands in solid-state systems. Here, we employed the recently developed attomicroscopy imaging tool to investigate, visualize, and manipulate the motion of bound electrons in graphene. By adjusting the carrier-envelope phase and the field strength of the driving electric field, we were able to control both the amplitude and direction of the field-induced electron current between carbon atoms in graphene. This research opens new avenues for understanding and controlling dynamic, on-demand electron motion processes, including chemical reactions, molecular bonding, and the electronic properties of materials.

Traumatic Brain Injuries (TBIs) resulting from Road Traffic Crashes (RTCs) can have fatal and disabling effects on patients. In this study, we evaluated the TBIs outcomes of patients involved in RTCs and identify key contributing factors affecting these outcomes. Data on 207 patients recorded by physicians at a tertiary hospital in Bangladesh was collected. A random parameters multinomial logit model with heterogeneity in the means was utilized to assess patients outcomes in three categories: Non-surgical, Surgical, and Fatal. From the random parameters, the study found that male patients (55.48%) are more likely to experience surgical and fatal outcomes. Male motorcycle users have a higher probability of experiencing fatal consequences. Additionally, 60.94% of incidents on rural roads result in surgeries and fatalities, with nighttime crashes on these roads significantly increasing the likelihood of fatal outcomes. Key factors impacting the likelihood of TBIs outcomes include older age, pedestrian involvement, bus and truck crashes, speeding, wet pavements, overtaking, low visibility, and weekday crashes. The study identified two significant interaction variables that increase the probability of fatal outcomes from TBIs: the interactions between low visibility and bus involvement, and between overtaking and wet pavements. While these factors individually had a higher probability of leading to both surgical and fatal outcomes, together these factors increase the risks to fatalities. Overall, our findings provide more detailed insights about the impact of TBIs outcomes resulting from RTCs and emphasize the need to develop more effective measures to improve road safety and patient outcomes., Comment: 18 pages

This paper explores the potential of aerial reconfigurable intelligent surfaces (ARIS) to enhance coordinated multi-point non-orthogonal multiple access (CoMP-NOMA) networks. We consider a system model where a UAV-mounted RIS assists in serving multiple users through NOMA while coordinating with multiple base stations. The optimization of UAV trajectory, RIS phase shifts, and NOMA power control constitutes a complex problem due to the hybrid nature of the parameters, involving both continuous and discrete values. To tackle this challenge, we propose a novel framework utilizing the multi-output proximal policy optimization (MO-PPO) algorithm. MO-PPO effectively handles the diverse nature of these optimization parameters, and through extensive simulations, we demonstrate its effectiveness in achieving near-optimal performance and adapting to dynamic environments. Our findings highlight the benefits of integrating ARIS in CoMP-NOMA networks for improved spectral efficiency and coverage in future wireless networks., Comment: IEEE Globecom 2024

In this paper, we explore the existence of spherically symmetric strange quark configurations coupled with anisotropic fluid setup in the framework of modified Gauss-Bonnet theory. In this regard, we adopt two models such as \emph{(i)} $f(\mathcal{G})=\beta\mathcal{G}^2$, and \emph{(ii)} $f(\mathcal{G})=\delta_{1}\mathcal{G}^{x}(\delta_{2}\mathcal{G}^{y}+1)$, and derive the field equations representing a static sphere. We then introduce bag constant in the gravitational equations through the use of MIT bag model, so that the quarks' interior can be discussed. Further, we work out the modified equations under the use of Tolman IV ansatz to make their solution possible. Junction conditions are also employed to find the constants involved in the considered metric potentials. Afterwards, different values of model parameters and bag constant are taken into account to graphically exploring the resulting solutions. This analysis is done by considering five strange quark objects like Her X-I, LMC X-4, 4U 1820-30, PSR J 1614-2230, and Vela X-I. Certain tests are also applied on the developed models to check their physical feasibility. It is much interesting that this modified gravity under its both considered functional forms yield physically viable and stable results for certain parametric values., Comment: 34 pages, 8 figures

Deep Neural Networks (DNNs) suffer from significant retraining costs when adapting to evolving requirements. Modularizing DNNs offers the promise of improving their reusability. Previous work has proposed techniques to decompose DNN models into modules both during and after training. However, these strategies yield several shortcomings, including significant weight overlaps and accuracy losses across modules, restricted focus on convolutional layers only, and added complexity and training time by introducing auxiliary masks to control modularity. In this work, we propose MODA, an activation-driven modular training approach. MODA promotes inherent modularity within a DNN model by directly regulating the activation outputs of its layers based on three modular objectives: intra-class affinity, inter-class dispersion, and compactness. MODA is evaluated using three well-known DNN models and three datasets with varying sizes. This evaluation indicates that, compared to the existing state-of-the-art, using MODA yields several advantages: (1) MODA accomplishes modularization with 29% less training time; (2) the resultant modules generated by MODA comprise 2.4x fewer weights and 3.5x less weight overlap while (3) preserving the original model's accuracy without additional fine-tuning; in module replacement scenarios, (4) MODA improves the accuracy of a target class by 12% on average while ensuring minimal impact on the accuracy of other classes.

To help MLOps engineers decide which operator to use in which deployment scenario, this study aims to empirically assess the accuracy vs latency trade-off of white-box (training-based) and black-box operators (non-training-based) and their combinations in an Edge AI setup. We perform inference experiments including 3 white-box (i.e., QAT, Pruning, Knowledge Distillation), 2 black-box (i.e., Partition, SPTQ), and their combined operators (i.e., Distilled SPTQ, SPTQ Partition) across 3 tiers (i.e., Mobile, Edge, Cloud) on 4 commonly-used Computer Vision and Natural Language Processing models to identify the effective strategies, considering the perspective of MLOps Engineers. Our Results indicate that the combination of Distillation and SPTQ operators (i.e., DSPTQ) should be preferred over non-hybrid operators when lower latency is required in the edge at small to medium accuracy drop. Among the non-hybrid operators, the Distilled operator is a better alternative in both mobile and edge tiers for lower latency performance at the cost of small to medium accuracy loss. Moreover, the operators involving distillation show lower latency in resource-constrained tiers (Mobile, Edge) compared to the operators involving Partitioning across Mobile and Edge tiers. For textual subject models, which have low input data size requirements, the Cloud tier is a better alternative for the deployment of operators than the Mobile, Edge, or Mobile-Edge tier (the latter being used for operators involving partitioning). In contrast, for image-based subject models, which have high input data size requirements, the Edge tier is a better alternative for operators than Mobile, Edge, or their combination.

Large Language Models (LLMs) have transformed how people interact with artificial intelligence (AI) systems, achieving state-of-the-art results in various tasks, including scientific discovery and hypothesis generation. However, the lack of a comprehensive and systematic evaluation framework for generating research ideas using LLMs poses a significant obstacle to understanding and assessing their generative capabilities in scientific discovery. To address this gap, we propose IdeaBench, a benchmark system that includes a comprehensive dataset and an evaluation framework for standardizing the assessment of research idea generation using LLMs. Our dataset comprises titles and abstracts from a diverse range of influential papers, along with their referenced works. To emulate the human process of generating research ideas, we profile LLMs as domain-specific researchers and ground them in the same context considered by human researchers. This maximizes the utilization of the LLMs' parametric knowledge to dynamically generate new research ideas. We also introduce an evaluation framework for assessing the quality of generated research ideas. Our evaluation framework is a two-stage process: first, using GPT-4o to rank ideas based on user-specified quality indicators such as novelty and feasibility, enabling scalable personalization; and second, calculating relative ranking based "Insight Score" to quantify the chosen quality indicator. The proposed benchmark system will be a valuable asset for the community to measure and compare different LLMs, ultimately advancing the automation of the scientific discovery process.

Countries in South Asia experience many catastrophic flooding events regularly. Through image classification, it is possible to expedite search and rescue initiatives by classifying flood zones, including houses and humans. We create a new dataset collecting aerial imagery of flooding events across South Asian countries. For the classification, we propose a fine-tuned Compact Convolutional Transformer (CCT) based approach and some other cutting-edge transformer-based and Convolutional Neural Network-based architectures (CNN). We also implement the YOLOv8 object detection model and detect houses and humans within the imagery of our proposed dataset, and then compare the performance with our classification-based approach. Since the countries in South Asia have similar topography, housing structure, the color of flood water, and vegetation, this work can be more applicable to such a region as opposed to the rest of the world. The images are divided evenly into four classes: 'flood', 'flood with domicile', 'flood with humans', and 'no flood'. After experimenting with our proposed dataset on our fine-tuned CCT model, which has a comparatively lower number of weight parameters than many other transformer-based architectures designed for computer vision, it exhibits an accuracy and macro average precision of 98.62% and 98.50%. The other transformer-based architectures that we implement are the Vision Transformer (ViT), Swin Transformer, and External Attention Transformer (EANet), which give an accuracy of 88.66%, 84.74%, and 66.56% respectively. We also implement DCECNN (Deep Custom Ensembled Convolutional Neural Network), which is a custom ensemble model that we create by combining MobileNet, InceptionV3, and EfficientNetB0, and we obtain an accuracy of 98.78%. The architectures we implement are fine-tuned to achieve optimal performance on our dataset.

Recently, Nagib et al (2024} utilized indicator functions of profiles of the streamwise normal stress to reveal the ranges of validity, in wall distance and Reynolds number, for each of two proposed models in Direct Numerical Simulations (DNS) of channel and pipe flows. A method more suited to experimental data is proposed here, as establishing accurate indicator functions is a challenge. The new method is outlined and used with the two leading models that propose either a logarithmic or power trend, for the normal stresses of turbulence in a fitting region of wall-bounded flows. The method, which is simple and robust, is used to evaluate each model over a wide range of Reynolds numbers by applying it to two of the prominent experimental data sets in zero-pressure-gradient boundary layers (ZPG) and pipe flows. A somewhat larger exponent for the power trend equal to 0.28, instead of 0.25, is found to slightly extend its range of validity. Correcting for outer intermittency in ZPG data also extends validity of the power trend to around half the boundary layer thickness. Projecting near-wall peak normal-stress values using both models based on data from locations in the fitting region, yields nearly identical results up to the highest Re available from these two experiments. The two relations provide projected values at Re=10M with a relative difference of approximately 4.5%; distinguishing between these values would require measurement accuracy well beyond our capabilities. Current results and those of Monkewitz and Nagib (2023) and Baxerras et al. (2024) on the mean flow, suggest that considerations of nonlinear growth of eddies from the wall and accounting for some viscous effects, are important for modeling wall-bounded flows. Such effects also apply to potential refinements of the logarithmic trend along ideas advanced by Deshpande et al. (2021)., Comment: 15 pages, 15 figures

We study a flaring activity of the HSP Mrk421 that was characterized from radio to very-high-energy (VHE; E $>0.1$TeV) gamma rays with MAGIC, Fermi-LAT, Swift, XMM-Newton and several optical and radio telescopes. These observations included, for the first time for a gamma-ray flare of a blazar, simultaneous X-ray polarization measurements with IXPE. We find substantial variability in both X-rays and VHE gamma rays throughout the campaign, with the highest VHE flux above 0.2 TeV occurring during the IXPE observing window, and exceeding twice the flux of the Crab Nebula. However, the VHE and X-ray spectra are on average softer, and the correlation between these two bands weaker that those reported in previous flares of Mrk421. IXPE reveals an X-ray polarization degree significantly higher than that at radio and optical frequencies. The X-ray polarization angle varies by $\sim$100$^\circ$ on timescales of days, and the polarization degree changes by more than a factor 4. The highest X-ray polarization degree reaches 26%, around which a X-ray counter-clockwise hysteresis loop is measured with XMM-Newton. It suggests that the X-ray emission comes from particles close to the high-energy cutoff, hence possibly probing an extreme case of the Turbulent Extreme Multi-Zone model. We model the broadband emission with a simplified stratified jet model throughout the flare. The polarization measurements imply an electron distribution in the X-ray emitting region with a very high minimum Lorentz factor, which is expected in electron-ion plasma, as well as a variation of the emitting region size up to a factor of three during the flaring activity. We find no correlation between the fluxes and the evolution of the model parameters, which indicates a stochastic nature of the underlying physical mechanism. Such behaviour would be expected in a highly turbulent electron-ion plasma crossing a shock front., Comment: Submitted to Astronomy and Astrophysics. Corresponding authors: Axel Arbet-Engels, Lea Heckmann, David Paneque

Finding test cases to kill the alive mutants in Mutation testing needs to calculate the Reachability, Infection and Propagation(RIP) conditions and full test specification. In this paper, a formal approach to calculate RIP conditions is proposed. The Dijkestra's weakest precondition predicate transformer (wp(_,_)) is used to calculate infection and propagation conditions. The rc(_) function is defined to calculate the reachability conditions generated by each statement. Four programs and their mutants are examined as running examples and as case studies to show the applicability of the method.

This work presents steps toward the design of Majorana zero modes (MZM) in InAsP quantum dots (QD) embedded in an InP semiconducting nanowire in contact with a p-type superconductor described by the Kitaev Hamiltonian. The single particle spectrum is obtained from million atom atomistic calculations with QNANO and many-electron spectra using exact diagonalization (ED) and the hybrid Variational Quantum Eigensolver (VQE) method. A variational ansatz is constructed to capture the ground state of the system by utilizing a generalized form of the analytical solution for a particular set of parameters. By systematically deviating from the analytically solvable regime while maintaining the system in the topological phase (TP), the effectiveness of the variational function in reproducing the correct ground state and topological properties of the system is evaluated. This is done through a quantum algorithm for a many-body state containing MZM. The results are compared with exact solution in topological phase and demonstrate the capability of VQE, along with classical simulations, to accurately model the many-body spectra in topologically nontrivial state.

Predicting low-energy molecular conformations given a molecular graph is an important but challenging task in computational drug discovery. Existing state-of-the-art approaches either resort to large scale transformer-based models that diffuse over conformer fields, or use computationally expensive methods to generate initial structures and diffuse over torsion angles. In this work, we introduce Equivariant Transformer Flow (ET-Flow). We showcase that a well-designed flow matching approach with equivariance and harmonic prior alleviates the need for complex internal geometry calculations and large architectures, contrary to the prevailing methods in the field. Our approach results in a straightforward and scalable method that directly operates on all-atom coordinates with minimal assumptions. With the advantages of equivariance and flow matching, ET-Flow significantly increases the precision and physical validity of the generated conformers, while being a lighter model and faster at inference. Code is available https://github.com/shenoynikhil/ETFlow., Comment: NeurIPS 2024

This article, is devoted to $n$-EP and $n$-hypo-EP operators. We give some characteristic properties of these two classes and various links with other known classes in the literature, especially the classes of EP, SD, hypo-EP and $n$-normal operators.

The rapid expansion of foundation models (FMs), such as large language models (LLMs), has given rise to FMware--software systems that integrate FMs as core components. While building demonstration-level FMware is relatively straightforward, transitioning to production-ready systems presents numerous challenges, including reliability, high implementation costs, scalability, and compliance with privacy regulations. This paper provides a thematic analysis of the key obstacles in productionizing FMware, synthesized from industry experience and diverse data sources, including hands-on involvement in the Open Platform for Enterprise AI (OPEA) and FMware lifecycle engineering. We identify critical issues in FM selection, data and model alignment, prompt engineering, agent orchestration, system testing, and deployment, alongside cross-cutting concerns such as memory management, observability, and feedback integration. We discuss needed technologies and strategies to address these challenges and offer guidance on how to enable the transition from demonstration systems to scalable, production-ready FMware solutions. Our findings underscore the importance of continued research and multi-industry collaboration to advance the development of production-ready FMware.

In this paper, we introduce a novel modification to the BB84 Quantum Key Distribution (QKD) protocol, aimed at enhancing its efficiency through the use of quantum memory and delayed measurement. In the standard BB84 protocol, the receiver immediately measures the qubits sent by the sender using randomly chosen bases. Due to mismatches between the sender and receiver's bases, a significant portion of the qubits are discarded, reducing the overall key generation rate. Our proposed protocol allows the receiver to store the received qubits in quantum memory and defer measurement until after the sender reveals her basis choices, effectively eliminating the need to discard mismatched qubits. This modification improves the key generation efficiency while maintaining the core security features of the standard BB84 protocol. By avoiding the unnecessary loss of qubits, our protocol achieves a higher secret key rate without introducing additional vulnerabilities. We present a detailed step-by-step explanation of the delayed measurement process. Although this approach does not alter the security guarantees of BB84, it represents a significant improvement in efficiency, making the protocol more viable for large-scale quantum communication networks.

Generative adversarial networks (GANs) learn a latent space whose samples can be mapped to real-world images. Such latent spaces are difficult to interpret. Some earlier supervised methods aim to create an interpretable latent space or discover interpretable directions that require exploiting data labels or annotated synthesized samples for training. However, we propose using a modification of vector quantization called space-filling vector quantization (SFVQ), which quantizes the data on a piece-wise linear curve. SFVQ can capture the underlying morphological structure of the latent space and thus make it interpretable. We apply this technique to model the latent space of pretrained StyleGAN2 and BigGAN networks on various datasets. Our experiments show that the SFVQ curve yields a general interpretable model of the latent space that determines which part of the latent space corresponds to what specific generative factors. Furthermore, we demonstrate that each line of SFVQ's curve can potentially refer to an interpretable direction for applying intelligible image transformations. We also showed that the points located on an SFVQ line can be used for controllable data augmentation.

Parkinson's disease (PD) is a progressive neurological disorder that impacts the quality of life significantly, making in-home monitoring of motor symptoms such as Freezing of Gait (FoG) critical. However, existing symptom monitoring technologies are power-hungry, rely on extensive amounts of labeled data, and operate in controlled settings. These shortcomings limit real-world deployment of the technology. This work presents LIFT-PD, a computationally-efficient self-supervised learning framework for real-time FoG detection. Our method combines self-supervised pre-training on unlabeled data with a novel differential hopping windowing technique to learn from limited labeled instances. An opportunistic model activation module further minimizes power consumption by selectively activating the deep learning module only during active periods. Extensive experimental results show that LIFT-PD achieves a 7.25% increase in precision and 4.4% improvement in accuracy compared to supervised models while using as low as 40% of the labeled training data used for supervised learning. Additionally, the model activation module reduces inference time by up to 67% compared to continuous inference. LIFT-PD paves the way for practical, energy-efficient, and unobtrusive in-home monitoring of PD patients with minimal labeling requirements., Comment: 11 pages

We propose a game-theoretic framework to model and optimize user engagement in cooperative activities over social networks. While traditional diffusion models suggest that individuals are only influenced by their neighbors, empirical evidence shows that diffusion alone does not fully explain network evolution, and non-diffusion factors play a significant role in network growth. We model network participation and resource-sharing as strategic games involving boundedly rational players to address this gap between the analytical models and empirical evidence. Specifically, we employ Log-Linear Learning (LLL), a version of noisy best response, to capture players' decision-making strategies. By incorporating stochastic decision models like LLL, our framework integrates both diffusion and non-diffusion dynamics into network evolution dynamics. Through equilibrium analysis and simulations, we demonstrate that our model aligns with theoretical predictions from existing analytical frameworks and empirical observations across various initial network configurations. Our second contribution is a novel method for selecting anchor nodes to enhance user participation. This approach allows system designers to identify anchor nodes and compute their incentives in real time under a more realistic information requirement constraints as compared to the existing approaches. The proposed approach adapts to changing network conditions by reallocating resources from less impactful to more influential nodes. Furthermore, the method is resilient to anchor node failures, ensuring sustained and continuous network participation., Comment: 16 pages, 8 figures

We study the dynamics and interactions of the solitonic domain walls that occur in realistic axion electrodynamics models including the Chern-Simons interaction, $a\epsilon_{\mu\nu\lambda\sigma}F^{\mu\nu} F^{\lambda\sigma}$, between an axion $a(x)$ of mass $m_a$, and a massless U(1) gauge field, e.g. EM, interacting with strength $\alpha=e^2/4\pi$ with charged matter, e.g. electron-positron pairs. In particular, in the presence of a U(1) gauge-and-matter relativistic thermal plasma we study the friction experienced by the walls due to the Chern-Simons term. Utilizing the linear response method we include the collective effects of the plasma, as opposed to purely particle scattering across the wall (as is done in previous treatments) which is valid only in the thin wall regime that is rarely applicable in realistic cases. We show that the friction depends on the Lorentz-$\gamma$-factor-dependent inverse thickness of the wall in the plasma frame, $\ell^{-1} \sim \gamma m_a$, compared to the three different plasma scales, the temperature $T$, the Debye mass $m_D\sim\sqrt{\alpha} T$, and the damping rate $\Gamma \sim \alpha^2 T$, and elucidate the underlying physical intuition for this behavior. (For friction in the thin-wall-limit we correct previous expressions in the literature.) We further consider the effects of long-range coherent magnetic fields that are possibly present in the early universe and compare their effect with that of thermal magnetic fields. Finally, we briefly discuss the possible early universe consequences of our results for domain wall motion and network decay, stochastic gravitational wave production from domain wall networks, and possible primordial black hole production from domain wall collapse, though a more complete discussion of these topics is reserved for a companion paper., Comment: 56 Pages, 8 Figures

Planar microcavity polaritons have recently emerged as a promising technology for improving several performance characteristics of organic light-emitting diodes, photodiodes and photovoltaics. To form polaritons and achieve enhanced performance, traditional microcavities with high reflectivity mirrors are fabricated by energy-intensive physical vapor deposition methods, which restrict their use in applications requiring flexibility and low cost. Here, for the first time, we demonstrate a dielectric all-solution-processed polariton microcavity consisting of Rhodamine 6G films in a poly(vinyl alcohol) matrix, exhibiting more than 400 meV Rabi-splitting and photoluminescence with uniform dispersion along the lower polariton mode. Our fully automated deposition and annealing fabrication protocol played a key role in preventing interlayer mixing and producing high optical-quality polariton microcavities, enabling us to observe enhanced scattering of reservoir excitons to the lower polariton and to explore the effects of strong coupling on bimolecular interactions. Notably, we found that polariton microcavities exhibit a more than 10-fold increase in the critical excitation density for bimolecular annihilation compared to bare Rhodamine 6G films. This enhancement can only be partially attributed to the sub-3-fold measured enhancement in radiative lifetime, highlighting the critical role of strong coupling in the influence of molecular dynamics., Comment: 25 pages, 4 figures, supporting information

In this paper, we find the Hochschild cohomology groups of the universal associative conformal envelope $U(3)$ of the Virasoro Lie conformal algebra with respect to associative locality $N=3$ on the generator with coefficients in all finite modules. In order to obtain this result, we construct the Anick resolution via the algebraic discrete Morse theory and Gr\"obner--Shirshov basis.

Software development is a dynamic activity that requires engineers to work effectively with tools, processes, and collaborative teams. As a result, the presence of friction can significantly hinder productivity, increase frustration, and contribute to low morale among developers. By contrast, higher satisfaction levels are positively correlated with higher levels of perceived productivity. Hence, understanding the factors that cause bad experiences for developers is critical for fostering a positive and productive engineering environment. In this research, we employed a mixed-method approach, including interviews, surveys, diary studies, and analysis of developer telemetry data to uncover and triangulate common factors that cause "bad days" for developers. The interviews involved 22 developers across different levels and roles. The survey captured the perception of 214 developers about factors that cause them to have "bad days," their frequency, and their impact on job satisfaction. The daily diary study engaged 79 developers for 30 days to document factors that caused "bad days" in the moment. We examined the telemetry signals of 131 consenting participants to validate the impact of bad developer experience using system data. Findings from our research revealed factors that cause "bad days" for developers and significantly impact their work and well-being. We discuss the implications of these findings and suggest future work.

We consider multi-draft speculative sampling, where the proposal sequences are sampled independently from different draft models. At each step, a token-level draft selection scheme takes a list of valid tokens as input and produces an output token whose distribution matches that of the target model. Previous works have demonstrated that the optimal scheme (which maximizes the probability of accepting one of the input tokens) can be cast as a solution to a linear program. In this work we show that the optimal scheme can be decomposed into a two-step solution: in the first step an importance sampling (IS) type scheme is used to select one intermediate token; in the second step (single-draft) speculative sampling is applied to generate the output token. For the case of two identical draft models we further 1) establish a necessary and sufficient condition on the distributions of the target and draft models for the acceptance probability to equal one and 2) provide an explicit expression for the optimal acceptance probability. Our theoretical analysis also motives a new class of token-level selection scheme based on weighted importance sampling. Our experimental results demonstrate consistent improvements in the achievable block efficiency and token rates over baseline schemes in a number of scenarios.

Extreme-Mass-Ratio Inspirals (EMRIs) are one of the key targets for future space-based gravitational wave detectors, such as LISA. The scientific potential of these sources can only be fully realized with fast and accurate waveform models. In this work, we extend the \textsc{FastEMRIWaveform} (\texttt{FEW}) framework by providing fully relativistic waveforms at adiabatic order for circular, equatorial orbits in Kerr spacetime, for mass ratios up to $10^{-3}$. We study the importance of including relativistic corrections in the waveform for both vacuum and non-vacuum environments. For EMRIs in vacuum, we find that non-relativistic waveforms can result in $\sim 35\%$ error in the predicted source's horizon redshift. By developing relativistic non-vacuum EMRI waveforms, we demonstrate significant improvements in detecting environmental effects. Our analysis shows that incorporating relativistic corrections enhances constraints on accretion disks, modeled through power-law torques, and improves the constraints on disk parameters (error $\sim6\%$), representing a significant improvement from previous estimates. We also estimated the evidence for models in a scenario where ignoring the accretion disk causes bias in parameter estimation (PE) and report a $\log_{10}$ Bayes factor of $1.1$ in favor of the accretion disk model. Additionally, in a fully relativistic setup, we estimate the parameters of superradiant scalar clouds with high accuracy, achieving errors below $5\%$ for the scalar cloud's mass and below $0.5\%$ for the ultralight field's mass. These results demonstrate that incorporating relativistic effects greatly enhances the accuracy and reliability of waveform predictions, essential for PE and model selection., Comment: 22 pages, 13 figures

FPGAs are now ubiquitous in cloud computing infrastructures and reconfigurable system-on-chip, particularly for AI acceleration. Major cloud service providers such as Amazon and Microsoft are increasingly incorporating FPGAs for specialized compute-intensive tasks within their data centers. The availability of FPGAs in cloud data centers has opened up new opportunities for users to improve application performance by implementing customizable hardware accelerators directly on the FPGA fabric. However, the virtualization and sharing of FPGA resources among multiple users open up new security risks and threats. We present a novel fault attack methodology capable of causing persistent fault injections in partial bitstreams during the process of FPGA reconfiguration. This attack leverages power-wasters and is timed to inject faults into bitstreams as they are being loaded onto the FPGA through the reconfiguration manager, without needing to remain active throughout the entire reconfiguration process. Our experiments, conducted on a Pynq FPGA setup, demonstrate the feasibility of this attack on various partial application bitstreams, such as a neural network accelerator unit and a signal processing accelerator unit., Comment: Accepted for presentation in the 2024 IEEE Asian Test Symposium (ATS)

Domain adaptive semantic segmentation is the task of generating precise and dense predictions for an unlabeled target domain using a model trained on a labeled source domain. While significant efforts have been devoted to improving unsupervised domain adaptation for this task, it is crucial to note that many models rely on a strong assumption that the source data is entirely and accurately labeled, while the target data is unlabeled. In real-world scenarios, however, we often encounter partially or noisy labeled data in source and target domains, referred to as Generalized Domain Adaptation (GDA). In such cases, we suggest leveraging weak or unlabeled data from both domains to narrow the gap between them, resulting in effective adaptation. We introduce the Generalized Gaussian-mixture-based (GenGMM) domain adaptation model, which harnesses the underlying data distribution in both domains to refine noisy weak and pseudo labels. The experiments demonstrate the effectiveness of our approach.

Imaging atmospheric Cherenkov telescopes (IACTs) are used to observe very high-energy photons from the ground. Gamma rays are indirectly detected through the Cherenkov light emitted by the air showers they induce. The new generation of experiments, in particular the Cherenkov Telescope Array Observatory (CTAO), sets ambitious goals for discoveries of new gamma-ray sources and precise measurements of the already discovered ones. To achieve these goals, both hardware and data analysis must employ cutting-edge techniques. This also applies to the LST-1, the first IACT built for the CTAO, which is currently taking data on the Canary island of La Palma. This paper introduces a new event reconstruction technique for IACT data, aiming to improve the image reconstruction quality and the discrimination between the signal and the background from misidentified hadrons and electrons. The technique models the development of the extensive air shower signal, recorded as a waveform per pixel, seen by CTAO telescopes' cameras. Model parameters are subsequently passed to random forest regressors and classifiers to extract information on the primary particle. The new reconstruction was applied to simulated data and to data from observations of the Crab Nebula performed by the LST-1. The event reconstruction method presented here shows promising performance improvements. The angular and energy resolution, and the sensitivity, are improved by 10 to 20% over most of the energy range. At low energy, improvements reach up to 22%, 47%, and 50%, respectively. A future extension of the method to stereoscopic analysis for telescope arrays will be the next important step., Comment: Accepted in A&A

Machine learning has been extensively applied for various classical software testing activities such as test generation, minimization, and prioritization. Along the same lines, recently, there has been interest in applying quantum machine learning to software testing. For example, Quantum Extreme Learning Machines (QELMs) were recently applied for testing classical software of industrial elevators. However, most studies on QELMs, whether in software testing or other areas, used ideal quantum simulators that fail to account for the noise in current quantum computers. While ideal simulations offer insight into QELM's theoretical capabilities, they do not enable studying their performance on current noisy quantum computers. To this end, we study how quantum noise affects QELM in three industrial and real-world classical software testing case studies, providing insights into QELMs' robustness to noise. Such insights assess QELMs potential as a viable solution for industrial software testing problems in today's noisy quantum computing. Our results show that QELMs are significantly affected by quantum noise, with a performance drop of 250% in regression tasks and 50% in classification tasks. Although introducing noise during both ML training and testing phases can improve results, the reduction is insufficient for practical applications. While error mitigation techniques can enhance noise resilience, achieving an average 3.0% performance drop in classification, but their effectiveness varies by context, highlighting the need for QELM-tailored error mitigation strategies.

Large Language Models (LLMs) have demonstrated remarkable success as general-purpose task solvers across various fields, including NLP, healthcare, finance, and law. However, their capabilities remain limited when addressing domain-specific problems, particularly in downstream NLP tasks. Research has shown that models fine-tuned on instruction-based downstream NLP datasets outperform those that are not fine-tuned. While most efforts in this area have primarily focused on resource-rich languages like English and broad domains, little attention has been given to multilingual settings and specific domains. To address this gap, this study focuses on developing a specialized LLM, LlamaLens, for analyzing news and social media content in a multilingual context. To the best of our knowledge, this is the first attempt to tackle both domain specificity and multilinguality, with a particular focus on news and social media. Our experimental setup includes 19 tasks, represented by 52 datasets covering Arabic, English, and Hindi. We demonstrate that LlamaLens outperforms the current state-of-the-art (SOTA) on 16 testing sets, and achieves comparable performance on 10 sets. We make the models and resources publicly available for the research community.(https://huggingface.co/QCRI), Comment: LLMs, Multilingual, Language Diversity, Large Language Models, Social Media, News Media, Specialized LLMs, Fact-checking, Media Analysis

Efficient sampling in biomolecular simulations is critical for accurately capturing the complex dynamical behaviors of biological systems. Adaptive sampling techniques aim to improve efficiency by focusing computational resources on the most relevant regions of phase space. In this work, we present a framework for identifying the optimal sampling policy through metric driven ranking. Our approach systematically evaluates the policy ensemble and ranks the policies based on their ability to explore the conformational space effectively. Through a series of biomolecular simulation case studies, we demonstrate that choice of a different adaptive sampling policy at each round significantly outperforms single policy sampling, leading to faster convergence and improved sampling performance. This approach takes an ensemble of adaptive sampling policies and identifies the optimal policy for the next round based on current data. Beyond presenting this ensemble view of adaptive sampling, we also propose two sampling algorithms that approximate this ranking framework on the fly. The modularity of this framework allows incorporation of any adaptive sampling policy making it versatile and suitable as a comprehensive adaptive sampling scheme., Comment: 33 pages, 6 figures

Topic modelling is a pivotal unsupervised machine learning technique for extracting valuable insights from large document collections. Existing neural topic modelling methods often encode contextual information of documents, while ignoring contextual details of candidate centroid words, leading to the inaccurate selection of topic words due to the contextualization gap. In parallel, it is found that functional words are frequently selected over topical words. To address these limitations, we introduce CAST: Corpus-Aware Self-similarity Enhanced Topic modelling, a novel topic modelling method that builds upon candidate centroid word embeddings contextualized on the dataset, and a novel self-similarity-based method to filter out less meaningful tokens. Inspired by findings in contrastive learning that self-similarities of functional token embeddings in different contexts are much lower than topical tokens, we find self-similarity to be an effective metric to prevent functional words from acting as candidate topic words. Our approach significantly enhances the coherence and diversity of generated topics, as well as the topic model's ability to handle noisy data. Experiments on news benchmark datasets and one Twitter dataset demonstrate the method's superiority in generating coherent, diverse topics, and handling noisy data, outperforming strong baselines.

This study investigates the impact of emotionally enriched AI feedback on student engagement and emotional responses in higher education. Leveraging the Control-Value Theory of Achievement Emotions, we conducted a randomized controlled experiment involving 425 participants where the experimental group received AI feedback enhanced with motivational elements, while the control group received neutral feedback. Our findings reveal that emotionally enriched feedback is perceived as more beneficial and helps reduce negative emotions, particularly anger, towards receiving feedback. However, it had no significant impact on the level of engagement with feedback or the quality of student work. These results suggest that incorporating emotional elements into AI-driven feedback can positively influence student perceptions and emotional well-being, without compromising work quality. Our study contributes to the growing body of research on AI in education by highlighting the importance of emotional considerations in educational technology design.

The aggressiveness of prostate cancer, the most common cancer in men worldwide, is primarily assessed based on histopathological data using the Gleason scoring system. While artificial intelligence (AI) has shown promise in accurately predicting Gleason scores, these predictions often lack inherent explainability, potentially leading to distrust in human-machine interactions. To address this issue, we introduce a novel dataset of 1,015 tissue microarray core images, annotated by an international group of 54 pathologists. The annotations provide detailed localized pattern descriptions for Gleason grading in line with international guidelines. Utilizing this dataset, we develop an inherently explainable AI system based on a U-Net architecture that provides predictions leveraging pathologists' terminology. This approach circumvents post-hoc explainability methods while maintaining or exceeding the performance of methods trained directly for Gleason pattern segmentation (Dice score: 0.713 $\pm$ 0.003 trained on explanations vs. 0.691 $\pm$ 0.010 trained on Gleason patterns). By employing soft labels during training, we capture the intrinsic uncertainty in the data, yielding strong results in Gleason pattern segmentation even in the context of high interobserver variability. With the release of this dataset, we aim to encourage further research into segmentation in medical tasks with high levels of subjectivity and to advance the understanding of pathologists' reasoning processes., Comment: 58 pages, 15 figures (incl. supplementary)

Recent years witnessed remarkable progress in computational histopathology, largely fueled by deep learning. This brought the clinical adoption of deep learning-based tools within reach, promising significant benefits to healthcare, offering a valuable second opinion on diagnoses, streamlining complex tasks, and mitigating the risks of inconsistency and bias in clinical decisions. However, a well-known challenge is that deep learning models may contain up to billions of parameters; supervising their training effectively would require vast labeled datasets to achieve reliable generalization and noise resilience. In medical imaging, particularly histopathology, amassing such extensive labeled data collections places additional demands on clinicians and incurs higher costs, which hinders the art's progress. Addressing this challenge, researchers devised various strategies for leveraging deep learning with limited data and annotation availability. In this paper, we present a comprehensive review of deep learning applications in histopathology, with a focus on the challenges posed by data scarcity over the past decade. We systematically categorize and compare various approaches, evaluate their distinct contributions using benchmarking tables, and highlight their respective advantages and limitations. Additionally, we address gaps in existing reviews and identify underexplored research opportunities, underscoring the potential for future advancements in this field., Comment: 36 pages

Passive visible light communication (VLC) modulates light propagation or reflection to transmit data without directly modulating the light source. Thus, passive VLC provides an alternative to conventional VLC, enabling communication where the light source cannot be directly controlled. There have been ongoing efforts to explore new methods and devices for modulating light propagation or reflection. The state-of-the-art has broken the 100 kbps data rate barrier for passive VLC by using a digital micro-mirror device (DMD) as the light modulating platform, or transmitter, and a photo-diode as the receiver. We significantly extend this work by proposing a massive spatial data channel framework for DMDs, where individual channels can be decoded in parallel using an event camera at the receiver. For the event camera, we introduce event processing algorithms to detect numerous channels and decode bits from individual channels with high reliability. Our prototype, built with off-the-shelf event cameras and DMDs, can decode up to $\sim$2,000 parallel channels, achieving a data transmission rate of 1.6 Mbps, markedly surpassing current benchmarks by 16x., Comment: 14 pages, 21 figures, nonacm

When assisting people in daily tasks, robots need to accurately interpret visual cues and respond effectively in diverse safety-critical situations, such as sharp objects on the floor. In this context, we present M-CoDAL, a multimodal-dialogue system specifically designed for embodied agents to better understand and communicate in safety-critical situations. The system leverages discourse coherence relations to enhance its contextual understanding and communication abilities. To train this system, we introduce a novel clustering-based active learning mechanism that utilizes an external Large Language Model (LLM) to identify informative instances. Our approach is evaluated using a newly created multimodal dataset comprising 1K safety violations extracted from 2K Reddit images. These violations are annotated using a Large Multimodal Model (LMM) and verified by human annotators. Results with this dataset demonstrate that our approach improves resolution of safety situations, user sentiment, as well as safety of the conversation. Next, we deploy our dialogue system on a Hello Robot Stretch robot and conduct a within-subject user study with real-world participants. In the study, participants role-play two safety scenarios with different levels of severity with the robot and receive interventions from our model and a baseline system powered by OpenAI's ChatGPT. The study results corroborate and extend the findings from automated evaluation, showing that our proposed system is more persuasive and competent in a real-world embodied agent setting.

Monolingual data, being readily available in large quantities, has been used to upscale the scarcely available parallel data to train better models for automatic translation. Self-learning, where a model is made to learn from its output, is one approach to exploit such data. However, it has been shown that too much of this data can be detrimental to the performance of the model if the available parallel data is comparatively extremely low. In this study, we investigate whether the monolingual data can also be too little and if this reduction, based on quality, has any effect on the performance of the translation model. Experiments have shown that on English-German low-resource NMT, it is often better to select only the most useful additional data, based on quality or closeness to the domain of the test data, than utilizing all of the available data.

Ensuring that Large Language Models (LLMs) generate text representative of diverse sub-populations is essential, particularly when key concepts related to under-represented groups are scarce in the training data. We address this challenge with a novel clustering-based active learning framework, enhanced with knowledge distillation. The proposed framework transforms the intermediate outputs of the learner model, enabling effective active learning for generative tasks for the first time. Integration of clustering and knowledge distillation yields more representative models without prior knowledge of underlying data distribution and overbearing human efforts. We validate our approach in practice through case studies in counter-narration and style transfer. We construct two new datasets in tandem with model training, showing a performance improvement of 2%-10% over baseline models. Our results also show more consistent performance across various data subgroups and increased lexical diversity, underscoring our model's resilience to skewness in available data. Further, our results show that the data acquired via our approach improves the performance of secondary models not involved in the learning loop, showcasing practical utility of the framework.

Large Language Models excel in tasks like natural language understanding and text generation. Prompt engineering plays a critical role in leveraging LLM effectively. However, LLMs black-box nature hinders its interpretability and effective prompting engineering. A wide range of model explanation approaches have been developed for deep learning models, However, these local explanations are designed for single-output tasks like classification and regression,and cannot be directly applied to LLMs, which generate sequences of tokens. Recent efforts in LLM explanation focus on natural language explanations, but they are prone to hallucinations and inaccuracies. To address this, we introduce PromptExp , a framework for multi-granularity prompt explanations by aggregating token-level insights. PromptExp introduces two token-level explanation approaches: 1. an aggregation-based approach combining local explanation techniques, and 2. a perturbation-based approach with novel techniques to evaluate token masking impact. PromptExp supports both white-box and black-box explanations and extends explanations to higher granularity levels, enabling flexible analysis. We evaluate PromptExp in case studies such as sentiment analysis, showing the perturbation-based approach performs best using semantic similarity to assess perturbation impact. Furthermore, we conducted a user study to confirm PromptExp's accuracy and practical value, and demonstrate its potential to enhance LLM interpretability., Comment: 11 pages

Despite the significant influx of prompt-tuning techniques for generative vision-language models (VLMs), it remains unclear how sensitive these models are to lexical and semantic alterations in prompts. In this paper, we evaluate the ability of generative VLMs to understand lexical and semantic changes in text using the SugarCrepe++ dataset. We analyze the sensitivity of VLMs to lexical alterations in prompts without corresponding semantic changes. Our findings demonstrate that generative VLMs are highly sensitive to such alterations. Additionally, we show that this vulnerability affects the performance of techniques aimed at achieving consistency in their outputs.