Search

Social determinants of health (SDoH) play a critical role in shaping health outcomes, particularly in pediatric populations where interventions can have long-term implications. SDoH are frequently studied in the Electronic Health Record (EHR), which provides a rich repository for diverse patient data. In this work, we present a novel annotated corpus, the Pediatric Social History Annotation Corpus (PedSHAC), and evaluate the automatic extraction of detailed SDoH representations using fine-tuned and in-context learning methods with Large Language Models (LLMs). PedSHAC comprises annotated social history sections from 1,260 clinical notes obtained from pediatric patients within the University of Washington (UW) hospital system. Employing an event-based annotation scheme, PedSHAC captures ten distinct health determinants to encompass living and economic stability, prior trauma, education access, substance use history, and mental health with an overall annotator agreement of 81.9 F1. Our proposed fine-tuning LLM-based extractors achieve high performance at 78.4 F1 for event arguments. In-context learning approaches with GPT-4 demonstrate promise for reliable SDoH extraction with limited annotated examples, with extraction performance at 82.3 F1 for event triggers., Comment: 12 pages, 2 figures and 3 tables. Accepted by LREC-COLING 2024

The LaAlO$_3$/SrTiO$_3$ interface hosts a plethora of gate-tunable electronic phases. Gating of LaAlO$_3$/SrTiO$_3$ interfaces are usually assumed to occur electrostatically. However, increasing evidence suggests that non-local interactions can influence and, in some cases, dominate the coupling between applied gate voltages and electronic properties. Here, we sketch quasi-1D ballistic electron waveguides at the LaAlO$_3$/SrTiO$_3$ interface as a probe to understand how gate tunability varies as a function of spatial separation. Gate tunability measurements reveal the scaling law to be at odds with the pure electrostatic coupling observed in traditional semiconductor systems. The non-Coulombic gating at the interface is attributed to the existence of a long-range nanoelectromechanical coupling between the gate and electron waveguide, mediated by the ferroelastic domains in SrTiO$_3$. The long-range interactions at the LaAlO$_3$/SrTiO$_3$ interface add unexpected richness and complexity to this correlated electron system.

Increasingly, more software services have been published onto the Internet, making it a big challenge to recommend services in the process of a scientific workflow composition. In this paper, a novel context-aware approach is proposed to recommending next services in a workflow development process, through learning service representation and service selection decision making behaviors from workflow provenance. Inspired by natural language sentence generation, the composition process of a scientific workflow is formalized as a step-wise procedure within the context of the goal of workflow, and the problem of next service recommendation is mapped to next word prediction. Historical service dependencies are first extracted from scientific workflow provenance to build a knowledge graph. Service sequences are then generated based on diverse composition path generation strategies. Afterwards, the generated corpus of composition paths are leveraged to study previous decision making strategies. Such a trained goal-oriented next service prediction model will be used to recommend top K candidate services during workflow composition process. Extensive experiments on a real-word repository have demonstrated the effectiveness of this approach., Comment: 14 pages, 8 figures. arXiv admin note: text overlap with arXiv:2205.11771

Medical imaging is critical to the diagnosis, surveillance, and treatment of many health conditions, including oncological, neurological, cardiovascular, and musculoskeletal disorders, among others. Radiologists interpret these complex, unstructured images and articulate their assessments through narrative reports that remain largely unstructured. This unstructured narrative must be converted into a structured semantic representation to facilitate secondary applications such as retrospective analyses or clinical decision support. Here, we introduce the Corpus of Annotated Medical Imaging Reports (CAMIR), which includes 609 annotated radiology reports from three imaging modality types: Computed Tomography, Magnetic Resonance Imaging, and Positron Emission Tomography-Computed Tomography. Reports were annotated using an event-based schema that captures clinical indications, lesions, and medical problems. Each event consists of a trigger and multiple arguments, and a majority of the argument types, including anatomy, normalize the spans to pre-defined concepts to facilitate secondary use. CAMIR uniquely combines a granular event structure and concept normalization. To extract CAMIR events, we explored two BERT (Bi-directional Encoder Representation from Transformers)-based architectures, including an existing architecture (mSpERT) that jointly extracts all event information and a multi-step approach (PL-Marker++) that we augmented for the CAMIR schema., Comment: Accepted at LREC-COLING 2024

Dialog Structure Induction (DSI) is the task of inferring the latent dialog structure (i.e., a set of dialog states and their temporal transitions) of a given goal-oriented dialog. It is a critical component for modern dialog system design and discourse analysis. Existing DSI approaches are often purely data-driven, deploy models that infer latent states without access to domain knowledge, underperform when the training corpus is limited/noisy, or have difficulty when test dialogs exhibit distributional shifts from the training domain. This work explores a neural-symbolic approach as a potential solution to these problems. We introduce Neural Probabilistic Soft Logic Dialogue Structure Induction (NEUPSL DSI), a principled approach that injects symbolic knowledge into the latent space of a generative neural model. We conduct a thorough empirical investigation on the effect of NEUPSL DSI learning on hidden representation quality, few-shot learning, and out-of-domain generalization performance. Over three dialog structure induction datasets and across unsupervised and semi-supervised settings for standard and cross-domain generalization, the injection of symbolic knowledge using NEUPSL DSI provides a consistent boost in performance over the canonical baselines.

Traditional Deep Neural Network (DNN) quantization methods using integer, fixed-point, or floating-point data types struggle to capture diverse DNN parameter distributions at low precision, and often require large silicon overhead and intensive quantization-aware training. In this study, we introduce Logarithmic Posits (LP), an adaptive, hardware-friendly data type inspired by posits that dynamically adapts to DNN weight/activation distributions by parameterizing LP bit fields. We also develop a novel genetic-algorithm based framework, LP Quantization (LPQ), to find optimal layer-wise LP parameters while reducing representational divergence between quantized and full-precision models through a novel global-local contrastive objective. Additionally, we design a unified mixed-precision LP accelerator (LPA) architecture comprising of processing elements (PEs) incorporating LP in the computational datapath. Our algorithm-hardware co-design demonstrates on average <1% drop in top-1 accuracy across various CNN and ViT models. It also achieves ~ 2x improvements in performance per unit area and 2.2x gains in energy efficiency compared to state-of-the-art quantization accelerators using different data types., Comment: 2024 61st IEEE/ACM Design Automation Conference (DAC)

Particle-based fluid simulations have emerged as a powerful tool for solving the Navier-Stokes equations, especially in cases that include intricate physics and free surfaces. The recent addition of machine learning methods to the toolbox for solving such problems is pushing the boundary of the quality vs. speed tradeoff of such numerical simulations. In this work, we lead the way to Lagrangian fluid simulators compatible with deep learning frameworks, and propose JAX-SPH - a Smoothed Particle Hydrodynamics (SPH) framework implemented in JAX. JAX-SPH builds on the code for dataset generation from the LagrangeBench project (Toshev et al., 2023) and extends this code in multiple ways: (a) integration of further key SPH algorithms, (b) restructuring the code toward a Python package, (c) verification of the gradients through the solver, and (d) demonstration of the utility of the gradients for solving inverse problems as well as a Solver-in-the-Loop application. Our code is available at https://github.com/tumaer/jax-sph., Comment: Accepted at the ICLR 2024 Workshop on AI4Differential Equations In Science

We define and study a variant of QMA (Quantum Merlin Arthur) in which Arthur can make multiple non-collapsing measurements to Merlin's witness state, in addition to ordinary collapsing measurements. By analogy to the class PDQP defined by Aaronson, Bouland, Fitzsimons, and Lee (2014), we call this class PDQMA. Our main result is that PDQMA = NEXP; this result builds on the PCP theorem and complements the result of Aaronson (2018) that PDQP/qpoly = ALL. While the result has little to do with quantum mechanics, we also show a more "quantum" result: namely, that QMA with the ability to inspect the entire history of a hidden variable is equal to NEXP, under mild assumptions on the hidden-variable theory. We also observe that a quantum computer, augmented with quantum advice and the ability to inspect the history of a hidden variable, can solve any decision problem in polynomial time., Comment: 19 pages; v2: added detail to the proof of Theorem 5 and added a Main Ideas section

Each LLM serving request goes through two phases. The first is prefill which processes the entire input prompt and produces the first output token and the second is decode which generates the rest of output tokens, one-at-a-time. Prefill iterations have high latency but saturate GPU compute due to parallel processing of the input prompt. In contrast, decode iterations have low latency but also low compute utilization because a decode iteration processes only a single token per request. This makes batching highly effective for decodes and consequently for overall throughput. However, batching multiple requests leads to an interleaving of prefill and decode iterations which makes it challenging to achieve both high throughput and low latency. We introduce an efficient LLM inference scheduler, Sarathi-Serve, to address this throughput-latency tradeoff. Sarathi-Serve introduces chunked-prefills which splits a prefill request into near equal sized chunks and creates stall-free schedules that adds new requests in a batch without pausing ongoing decodes. Stall-free scheduling unlocks the opportunity to improve throughput with large batch sizes while minimizing the effect of batching on latency. Furthermore, uniform batches in Sarathi-Serve ameliorate the imbalance between iterations resulting in minimal pipeline bubbles. Our techniques yield significant improvements in inference performance across models and hardware under tail latency constraints. For Mistral-7B on single A100 GPUs, we achieve 2.6x higher serving capacity and up to 3.7x higher serving capacity for the Yi-34B model on two A100 GPUs as compared to vLLM. When used with pipeline parallelism on Falcon-180B, Sarathi-Serve provides up to 5.6x gain in the end-to-end serving capacity. The source code for Sarathi-Serve is available at https://github.com/microsoft/sarathi-serve.

Image generation models are poised to become ubiquitous in a range of applications. These models are often fine-tuned and evaluated using human quality judgments that assume a universal standard, failing to consider the subjectivity of such tasks. To investigate how to quantify subjectivity, and the scale of its impact, we measure how assessments differ among human annotators across different use cases. Simulating the effects of ordinarily latent elements of annotators subjectivity, we contrive a set of motivations (t-shirt graphics, presentation visuals, and phone background images) to contextualize a set of crowdsourcing tasks. Our results show that human evaluations of images vary within individual contexts and across combinations of contexts. Three key factors affecting this subjectivity are image appearance, image alignment with text, and representation of objects mentioned in the text. Our study highlights the importance of taking individual users and contexts into account, both when building and evaluating generative models

Using the medium resolution spectrograph X-shooter, spectra of 235 OB and Wolf-Rayet (WR) stars in sub-solar metallicity environments have been secured. [...]This second paper focuses on the optical observations of 232 Magellanic Clouds targets. It describes the uniform reduction of the UVB (300 - 560 nm) and VIS (550 - 1020 nm) XShootU data as well as the preparation of advanced data products [...] . The data reduction of the raw data is based on the ESO CPL X-shooter pipeline. We paid particular attention to the determination of the response curves [...] We implemented slit-loss correction, absolute flux calibration, (semi-)automatic rectification to the continuum, and a correction for telluric lines. The spectra of individual epochs were corrected for the barycentric motion, re-sampled and co-added, and the spectra from the two arms were merged into a single flux calibrated spectrum covering the entire optical range with maximum signal-to-noise ratio. [...] We provide three types of data products: (i) two-dimensional spectra for each UVB and VIS exposure; (ii) one-dimensional UVB and VIS spectra before and after response-correction, as well as after applying various processing, including absolute flux calibration, telluric removal, normalisation and barycentric correction; and (iii) co-added flux-calibrated and rectified spectra over the full optical range, for which all available XShootU exposures were combined. For many of the targets, the final signal-to-noise ratio per resolution element is above 200 in both the UVB and the VIS co-added spectra. The reduced data and advanced scientific data products will be made available to the community upon publication of this paper. [...], Comment: 22 pages, 19 figures; accepted for publication in Astronomy & Astrophysics. Links to online tables and databases will be included upon publication

We present details on a new measurement of the muon magnetic anomaly, $a_\mu = (g_\mu -2)/2$. The result is based on positive muon data taken at Fermilab's Muon Campus during the 2019 and 2020 accelerator runs. The measurement uses $3.1$ GeV$/c$ polarized muons stored in a $7.1$-m-radius storage ring with a $1.45$ T uniform magnetic field. The value of $ a_{\mu}$ is determined from the measured difference between the muon spin precession frequency and its cyclotron frequency. This difference is normalized to the strength of the magnetic field, measured using Nuclear Magnetic Resonance (NMR). The ratio is then corrected for small contributions from beam motion, beam dispersion, and transient magnetic fields. We measure $a_\mu = 116 592 057 (25) \times 10^{-11}$ (0.21 ppm). This is the world's most precise measurement of this quantity and represents a factor of $2.2$ improvement over our previous result based on the 2018 dataset. In combination, the two datasets yield $a_\mu(\text{FNAL}) = 116 592 055 (24) \times 10^{-11}$ (0.20 ppm). Combining this with the measurements from Brookhaven National Laboratory for both positive and negative muons, the new world average is $a_\mu$(exp) $ = 116 592 059 (22) \times 10^{-11}$ (0.19 ppm)., Comment: 48 pages, 29 figures; 4 pages of Supplement Material; version accepted for publication in Physical Review D

Business research is facing the challenge of scattered knowledge, particularly in the realm of brand loyalty (BL). Although literature reviews on BL exist, they predominantly concentrate on the pre-sent state, neglecting future trends. Therefore, a comprehensive review is imperative to ascertain emerging trends in BL This study employs a bibliometric approach, analyzing 1,468 papers from the Scopus database. Various tools including R software, VOS viewer software, and Publish or Perish are utilized. The aim is to portray the knowledge map, explore the publication years, identify the top authors and their co-occurrence, reliable documents, institutions, subjects, research hotspots, and pioneering countries and universities in the study of BL. The qualitative section of this research identifies gaps and emerging trends in BL through Word Cloud charts, word growth analysis, and a review of highly cited articles from the past four years. Results showed that highly cited articles mention topics such as brand love, consumer-brand identification, and social networks and the U.S. had the most productions in this field. Besides, most citations were related to Keller with 1,173 citations. Furthermore, in the qualitative section, social networks and brand experiences were found to be of interest to researchers in the field. Finally, by introducing the antecedents and consequences of BL, the gaps and emerging trends in BL were identified, so as to present the di-rection of future research in this area., Comment: 29 pages, 7 tables, 9 figures

Quantum transport of charge or energy in networks with discrete sites is central to diverse quantum technologies, from molecular electronics to light harvesting and quantum opto-mechanical metamaterials. A one dimensional network can be viewed as waveguide. We show that if such waveguide is hybridised with a control unit that contains a few sites, then transmission through the waveguide depends sensitively on the motion of the sites in the control unit. Together, the hybrid waveguide and its control-unit form a Fano-Anderson chain whose Born-Oppenheimer surfaces inherit characteristics from both components: A bandstructure from the waveguide and potential energy steps as a function of site coordinates from the control-unit. Using time-dependent quantum wave packets, we reveal conditions under which the hybrid structure becomes transmissive only if the control unit contains mobile sites that induce non-adiabatic transitions between the surfaces. Hence, our approach provides functional synthetic Born-Oppenheimer surfaces for hybrid quantum technologies combining mechanic and excitonic elements, and has possible applications such as switching and temperature sensing.

Protecting software supply chains from malicious packages is paramount in the evolving landscape of software development. Attacks on the software supply chain involve attackers injecting harmful software into commonly used packages or libraries in a software repository. For instance, JavaScript uses Node Package Manager (NPM), and Python uses Python Package Index (PyPi) as their respective package repositories. In the past, NPM has had vulnerabilities such as the event-stream incident, where a malicious package was introduced into a popular NPM package, potentially impacting a wide range of projects. As the integration of third-party packages becomes increasingly ubiquitous in modern software development, accelerating the creation and deployment of applications, the need for a robust detection mechanism has become critical. On the other hand, due to the sheer volume of new packages being released daily, the task of identifying malicious packages presents a significant challenge. To address this issue, in this paper, we introduce a metadata-based malicious package detection model, MeMPtec. This model extracts a set of features from package metadata information. These extracted features are classified as either easy-to-manipulate (ETM) or difficult-to-manipulate (DTM) features based on monotonicity and restricted control properties. By utilising these metadata features, not only do we improve the effectiveness of detecting malicious packages, but also we demonstrate its resistance to adversarial attacks in comparison with existing state-of-the-art. Our experiments indicate a significant reduction in both false positives (up to 97.56%) and false negatives (up to 91.86%).

The unavailability of accurate boundary treatment methods for compressible Smoothed Particle Hydrodynamics (SPH) severely limits its ability to simulate flows in and around bodies. To this end, challenges specific to compressible flows with SPH are carefully considered. Based on these, robust and widely applicable boundary treatment methods for compressible SPH are proposed. These are accompanied by a novel technique to prevent particle penetration at boundaries. The proposed methods are shown to be significantly better than other recent approaches. A wide variety of test problems, many of which are not shown to be simulated with SPH thus far, are employed to highlight the strengths and weaknesses of the proposed methods. The implementation is open source and the results are automated in the interest of reproducibility. Overall, this research contributes to the advancement of SPH as a viable alternative to mesh-based methods for compressible flow simulations., Comment: 47 pages, 22 figures, 2 tables

Current approaches to prosthetic control are limited by their reliance on traditional methods, which lack real-time adaptability and intuitive responsiveness. These limitations are particularly pronounced in assistive technologies designed for individuals with diverse cognitive states and motor intentions. In this paper, we introduce a framework that leverages Reinforcement Learning (RL) with Deep Q-Networks (DQN) for classification tasks. Additionally, we present a preprocessing technique using the Common Spatial Pattern (CSP) for multiclass motor imagery (MI) classification in a One-Versus-The-Rest (OVR) manner. The subsequent 'csp space' transformation retains the temporal dimension of EEG signals, crucial for extracting discriminative features. The integration of DQN with a 1D-CNN-LSTM architecture optimizes the decision-making process in real-time, thereby enhancing the system's adaptability to the user's evolving needs and intentions. We elaborate on the data processing methods for two EEG motor imagery datasets. Our innovative model, RLEEGNet, incorporates a 1D-CNN-LSTM architecture as the Online Q-Network within the DQN, facilitating continuous adaptation and optimization of control strategies through feedback. This mechanism allows the system to learn optimal actions through trial and error, progressively improving its performance. RLEEGNet demonstrates high accuracy in classifying MI-EEG signals, achieving as high as 100% accuracy in MI tasks across both the GigaScience (3-class) and BCI-IV-2a (4-class) datasets. These results highlight the potential of combining DQN with a 1D-CNN-LSTM architecture to significantly enhance the adaptability and responsiveness of BCI systems., Comment: 23 pages, 1 figure, 6 tables

The frame scaling problem is: given vectors $U := \{u_{1}, ..., u_{n} \} \subseteq \mathbb{R}^{d}$, marginals $c \in \mathbb{R}^{n}_{++}$, and precision $\varepsilon > 0$, find left and right scalings $L \in \mathbb{R}^{d \times d}, r \in \mathbb{R}^n$ such that $(v_1,\dots,v_n) := (Lu_1 r_1,\dots,Lu_nr_n)$ simultaneously satisfies $\sum_{i=1}^n v_i v_i^{\mathsf{T}} = I_d$ and $\|v_{j}\|_{2}^{2} = c_{j}, \forall j \in [n]$, up to error $\varepsilon$. This problem has appeared in a variety of fields throughout linear algebra and computer science. In this work, we give a strongly polynomial algorithm for frame scaling with $\log(1/\varepsilon)$ convergence. This answers a question of Diakonikolas, Tzamos and Kane (STOC 2023), who gave the first strongly polynomial randomized algorithm with poly$(1/\varepsilon)$ convergence for the special case $c = \frac{d}{n} 1_{n}$. Our algorithm is deterministic, applies for general $c \in \mathbb{R}^{n}_{++}$, and requires $O(n^{3} \log(n/\varepsilon))$ iterations as compared to $O(n^{5} d^{11}/\varepsilon^{5})$ iterations of DTK. By lifting the framework of Linial, Samorodnitsky and Wigderson (Combinatorica 2000) for matrix scaling to frames, we are able to simplify both the algorithm and analysis. Our main technical contribution is to generalize the potential analysis of LSW to the frame setting and compute an update step in strongly polynomial time that achieves geometric progress in each iteration. In fact, we can adapt our results to give an improved analysis of strongly polynomial matrix scaling, reducing the $O(n^{5} \log(n/\varepsilon))$ iteration bound of LSW to $O(n^{3} \log(n/\varepsilon))$. Additionally, we prove a novel bound on the size of approximate frame scaling solutions, involving the condition measure $\bar{\chi}$ studied in the linear programming literature, which may be of independent interest., Comment: Comments welcome

Hot electrons and holes generated from the decay of localized surface plasmons in metallic nanoparticles can be harnessed for applications in solar energy conversion and sensing. In this paper, we study the generation of hot carriers in large spherical gold-silver alloy nanoparticles using a recently developed atomistic modelling approach that combines a solution of Maxwell's equations with large-scale tight-binding simulations. We find that hot-carrier properties depend sensitively on the alloy composition. Specifically, nanoparticles with a large gold fraction produce hot carriers under visible light illumination while nanoparticles with a large silver fraction require higher photon energies to produce hot carriers. Moreover, most hot carriers in nanoparticles with a large gold fraction originate from interband transitions which give rise to energetic holes and "cold" electrons near the Fermi level. Increasing the silver fraction enhances the generation rate of hot carriers from intraband transitions which produce energetic electrons and "cold" holes. These findings demonstrate that alloy composition is a powerful tuning parameter for the design of nanoparticles for applications in solar energy conversion and sensing that require precise control of hot-carrier properties., Comment: 19 pages, 5 figures

Due to recent development in quantum computing, the invention of a large quantum computer is no longer a distant future. Quantum computing severely threatens modern cryptography, as the hard mathematical problems beneath classic public-key cryptosystems can be solved easily by a sufficiently large quantum computer. As such, researchers have proposed PQC based on problems that even quantum computers cannot efficiently solve. Generally, post-quantum encryption and signatures can be hard to compute. This could potentially be a problem for IoT, which usually consist lightweight devices with limited computational power. In this paper, we survey existing literature on the performance for PQC in resource-constrained devices to understand the severeness of this problem. We also review recent proposals to optimize PQC algorithms for resource-constrained devices. Overall, we find that whilst PQC may be feasible for reasonably lightweight IoT, proposals for their optimization seem to lack standardization. As such, we suggest future research to seek coordination, in order to ensure an efficient and safe migration toward IoT for the post-quantum era., Comment: 13 pages, 3 figures and 7 tables. Formatted version submitted to ACM Computer Surveys

A species' genetic code or genome encodes valuable evolutionary, biological, and phylogenetic information that aids in species recognition, taxonomic classification, and understanding genetic predispositions like drug resistance and virulence. However, the vast number of potential species poses significant challenges in developing a general-purpose whole genome classification tool. Traditional bioinformatics tools have made notable progress but lack scalability and are computationally expensive. Machine learning-based frameworks show promise but must address the issue of large classification vocabularies with long-tail distributions. In this study, we propose addressing this problem through zero-shot learning using TEPI, Taxonomy-aware Embedding and Pseudo-Imaging. We represent each genome as pseudo-images and map them to a taxonomy-aware embedding space for reasoning and classification. This embedding space captures compositional and phylogenetic relationships of species, enabling predictions in extensive search spaces. We evaluate TEPI using two rigorous zero-shot settings and demonstrate its generalization capabilities qualitatively on curated, large-scale, publicly sourced data., Comment: Accepted to IEEE JBHI