Search

There is growing interest in artificial intelligence (AI) tools, especially high-profile tools like ChatGPT, and these tools now appear to be part of the education experience for many high school students. To investigate students' use of AI tools for school assignments, their impressions of how using the tools might affect them cognitively and academically, and their thoughts on using AI tools to write their college admissions essays, ACT developed a survey in June 2023 which was administered to a large nationwide sample of students in Grades 10 through 12. In this study, almost half of the participating students reported that they had used AI tools, and the most common tool they used was ChatGPT. Among students who did not use AI tools, the top reason for not using them was having no interest in them. About two thirds of students also reported that they did not trust the information provided by AI tools, and a little over half indicated that they did not know enough about AI tools to use them. Students with higher academic performance were significantly more likely to use AI tools than were students with lower academic performance. Findings show that these tools appear to have much potential to enhance student learning, but there are concerns about appropriate use and potential negative outcomes.

Very metal-poor stars ([Fe/H]<-2) are important laboratories for testing stellar models and reconstructing the formation history of our galaxy. Asteroseismology is a powerful tool to probe stellar interiors and measure ages, but few asteroseismic detections are known in very metal-poor stars and none have allowed detailed modeling of oscillation frequencies. We report the discovery of a low-luminosity Kepler red giant (KIC 8144907) with high S/N oscillations, [Fe/H]=-2.66+/-0.08 and [alpha/Fe]=0.38+/-0.06, making it by far the most metal-poor star to date for which detailed asteroseismic modeling is possible. By combining the oscillation spectrum from Kepler with high-resolution spectroscopy we measure an asteroseismic mass and age of 0.79+/-0.02(ran)+/-0.01(sys) Msun and 12.0+/-0.6(ran)+/-0.4(sys) Gyr, with remarkable agreement across different codes and input physics, demonstrating that stellar models and asteroseismology are reliable for very metal-poor stars when individual frequencies are used. The results also provide a direct age anchor for the early formation of the Milky Way, implying that substantial star formation did not commence until redshift z~3 (if the star formed in-situ) or that the Milky Way has undergone merger events for at least ~12 Gyr (if the star was accreted by a dwarf satellite merger such as Gaia Enceladus)., Comment: 10 pages, 4 figures, accepted for publication in ApJ

The rapid advancement of quantum computing poses a significant threat to many current security algorithms used for secure communication, digital authentication, and information encryption. A sufficiently powerful quantum computer could potentially exploit vulnerabilities in these algorithms, rendering data in transit insecure. This threat is expected to materialize within the next 20 years. Immediate transition to quantum-resilient cryptographic schemes is crucial, primarily to mitigate store-now-decrypt-later attacks and to ensure the security of products with decade-long operational lives. This transition requires a systematic approach to identifying and upgrading vulnerable cryptographic implementations. This work developed a quantum assessment tool for organizations, providing tailored recommendations for transitioning their security protocols into a post-quantum world. The work included a systematic evaluation of the proposed solution using qualitative feedback from network administrators and cybersecurity experts. This feedback was used to refine the accuracy and usability of the assessment process. The results demonstrate its effectiveness and usefulness in helping organizations prepare for quantum computing threats. The assessment tool is publicly available at (https://quantum-watch.soton.ac.uk).

Asteroseismic modelling is a powerful way to derive stellar properties. However, the derived quantities are limited by built-in assumptions used in stellar models. This work presents a detailed characterisation of stellar model uncertainties in asteroseismic red giants, focusing on the mixing-length parameter $\alpha_{\rm MLT}$, the initial helium fraction $Y_{\rm init}$, the solar abundance scale, and the overshoot parameters. First, we estimate error floors due to model uncertainties to be $\approx$0.4\% in mass, $\approx$0.2\% in radius, and $\approx$17\% in age, primarily due to the uncertain state of $\alpha_{\rm MLT}$ and $Y_{\rm init}$. The systematic uncertainties in age exceed typical statistical uncertainties, suggesting the importance of their evaluation in asteroseismic applications. Second, we demonstrate that the uncertainties from $\alpha_{\rm MLT}$ can be entirely mitigated by direct radius measurements or partially through $\nu_{\rm max}$. Utilizing radii from Kepler eclipsing binaries, we determined the $\alpha_{\rm MLT}$ values and calibrated the $\alpha_{\rm MLT}$--[M/H] relation. The correlation observed between the two variables is positive, consistent with previous studies using 1-D stellar models, but in contrast with outcomes from 3-D simulations. Third, we explore the implications of using asteroseismic modelling to test the $\nu_{\rm max}$ scaling relation. We found that a perceived dependency of $\nu_{\rm max}$ on [M/H] from individual frequency modelling can be largely removed by incorporating the calibrated $\alpha_{\rm MLT}$--[M/H] relation. Variations in $Y_{\rm init}$ can also affect $\nu_{\rm max}$ predictions. These findings suggest that $\nu_{\rm max}$ conveys information not fully captured by individual frequencies, and that it should be carefully considered as an important observable for asteroseismic modelling., Comment: 18 pages, 7 figures; submitted to ApJ

Social determinants of health (SDOH) $-$ the myriad of circumstances in which people live, grow, and age $-$ play an important role in health outcomes. However, existing outcome prediction models often only use proxies of SDOH as features. Recent open data initiatives present an opportunity to construct a more comprehensive view of SDOH, but manually integrating the most relevant data for individual patients becomes increasingly challenging as the volume and diversity of public SDOH data grows. Large language models (LLMs) have shown promise at automatically annotating structured data. Here, we conduct an end-to-end case study evaluating the feasibility of using LLMs to integrate SDOH data, and the utility of these SDOH features for clinical prediction. We first manually label 700+ variables from two publicly-accessible SDOH data sources to one of five semantic SDOH categories. Then, we benchmark performance of 9 open-source LLMs on this classification task. Finally, we train ML models to predict 30-day hospital readmission among 39k heart failure (HF) patients, and we compare the prediction performance of the categorized SDOH variables with standard clinical variables. Additionally, we investigate the impact of few-shot LLM prompting on LLM annotation performance, and perform a metadata ablation study on prompts to evaluate which information helps LLMs accurately annotate these variables. We find that some open-source LLMs can effectively, accurately annotate SDOH variables with zero-shot prompting without the need for fine-tuning. Crucially, when combined with standard clinical features, the LLM-annotated Neighborhood and Built Environment subset of the SDOH variables shows the best performance predicting 30-day readmission of HF patients., Comment: 36 pages including references and appendix. This is a work in progress

We discuss the interrelationship between the treatment of the elderly, the nature of production, and the transmission of culture. Respect for the elderly is endogenous. Parents cultivate an interest in consuming culture in their children; when they are older, children compensate their elders proportional to the degree to which their interests were previously cultivated. We show that this model is functionally equivalent to one in which cultural goods are transferred across generations. We focus on the relative well-being of the elderly and use the model to explain patterns in their relative well-being across societies. An important theme is that the cultivation of culture and norms for the respect and support of the elderly bear a nonlinear relationship with many economic variables, such as capital and or land intensity in production. We also discuss the interaction of property rights with production, assets such as productive resources, and relative treatment of the elderly. Insecurity of some types of property rights, such as rights over output, may benefit the elderly, while secure rights over productive resources may also benefit the elderly. We discuss how the elderly could be affected by demographic, technological and policy changes in both developing and developed economies.

Vision-and-Language Navigation (VLN) has gained increasing attention over recent years and many approaches have emerged to advance their development. The remarkable achievements of foundation models have shaped the challenges and proposed methods for VLN research. In this survey, we provide a top-down review that adopts a principled framework for embodied planning and reasoning, and emphasizes the current methods and future opportunities leveraging foundation models to address VLN challenges. We hope our in-depth discussions could provide valuable resources and insights: on one hand, to milestone the progress and explore opportunities and potential roles for foundation models in this field, and on the other, to organize different challenges and solutions in VLN to foundation model researchers., Comment: Authors contributed equally to this work, and supervisors contributed equal advising to this work

Large vision language models (LVLMs) often suffer from object hallucination, producing objects not present in the given images. While current benchmarks for object hallucination primarily concentrate on the presence of a single object class rather than individual entities, this work systematically investigates multi-object hallucination, examining how models misperceive (e.g., invent nonexistent objects or become distracted) when tasked with focusing on multiple objects simultaneously. We introduce Recognition-based Object Probing Evaluation (ROPE), an automated evaluation protocol that considers the distribution of object classes within a single image during testing and uses visual referring prompts to eliminate ambiguity. With comprehensive empirical studies and analysis of potential factors leading to multi-object hallucination, we found that (1) LVLMs suffer more hallucinations when focusing on multiple objects compared to a single object. (2) The tested object class distribution affects hallucination behaviors, indicating that LVLMs may follow shortcuts and spurious correlations.(3) Hallucinatory behaviors are influenced by data-specific factors, salience and frequency, and model intrinsic behaviors. We hope to enable LVLMs to recognize and reason about multiple objects that often occur in realistic visual scenes, provide insights, and quantify our progress towards mitigating the issues., Comment: Accepted to ALVR @ ACL 2024 | Project page: https://multi-object-hallucination.github.io/

During software development, poor design and implementation choices can detrimentally impact software maintainability. Design smells, recurring patterns of poorly designed fragments, signify these issues. Role-stereotypes denote the generic responsibilities that classes assume in system design. Although the concepts of role-stereotypes and design smells differ, both significantly contribute to the design and maintenance of software systems. Understanding the relationship between these aspects is crucial for enhancing software maintainability, code quality, efficient code review, guided refactoring, and the design of role-specific metrics. This paper employs an exploratory approach, combining statistical analysis and unsupervised learning methods, to understand how design smells relate to role-stereotypes across desktop and mobile applications. Analyzing 11,350 classes from 30 GitHub repositories, we identified several design smells that frequently co-occur within certain role-stereotypes. Specifically, three (3) out of six (6) role-stereotypes we studied are more prone to design smells. We also examined the variation of design smells across the two ecosystems, driven by notable differences in their underlying architecture. Findings revealed that design smells are more prevalent in desktop than in mobile applications, especially within the Service Provider and Information Holder role-stereotypes. Additionally, the unsupervised learning method showed that certain pairs or groups of role-stereotypes are prone to similar types of design smells. We believe these relationships are associated with the characteristic and collaborative properties between role-stereotypes. The insights from this research provide valuable guidance for software teams on implementing design smell prevention and correction mechanisms, ensuring conceptual integrity during design and maintenance phases., Comment: 32 pages, 8 figures

The integration of artificial intelligence (AI) in educational measurement has revolutionized assessment methods, enabling automated scoring, rapid content analysis, and personalized feedback through machine learning and natural language processing. These advancements provide timely, consistent feedback and valuable insights into student performance, thereby enhancing the assessment experience. However, the deployment of AI in education also raises significant ethical concerns regarding validity, reliability, transparency, fairness, and equity. Issues such as algorithmic bias and the opacity of AI decision-making processes pose risks of perpetuating inequalities and affecting assessment outcomes. Responding to these concerns, various stakeholders, including educators, policymakers, and organizations, have developed guidelines to ensure ethical AI use in education. The National Council of Measurement in Education's Special Interest Group on AI in Measurement and Education (AIME) also focuses on establishing ethical standards and advancing research in this area. In this paper, a diverse group of AIME members examines the ethical implications of AI-powered tools in educational measurement, explores significant challenges such as automation bias and environmental impact, and proposes solutions to ensure AI's responsible and effective use in education., Comment: 59 pages, 3 figures, a joint work of the Special Interest Group on Artificial Intelligence in Measurement and Education (AIME) from the National Council of Measurement in Education (NCME)

Each year, selective American colleges sort through tens of thousands of applications to identify a first-year class that displays both academic merit and diversity. In the 2023-2024 admissions cycle, these colleges faced unprecedented challenges. First, the number of applications has been steadily growing. Second, test-optional policies that have remained in place since the COVID-19 pandemic limit access to key information historically predictive of academic success. Most recently, longstanding debates over affirmative action culminated in the Supreme Court banning race-conscious admissions. Colleges have explored machine learning (ML) models to address the issues of scale and missing test scores, often via ranking algorithms intended to focus on 'top' applicants. However, the Court's ruling will force changes to these models, which were able to consider race as a factor in ranking. There is currently a poor understanding of how these mandated changes will shape applicant ranking algorithms, and, by extension, admitted classes. We seek to address this by quantifying the impact of different admission policies on the applications prioritized for review. We show that removing race data from a developed applicant ranking algorithm reduces the diversity of the top-ranked pool without meaningfully increasing the academic merit of that pool. We contextualize this impact by showing that excluding data on applicant race has a greater impact than excluding other potentially informative variables like intended majors. Finally, we measure the impact of policy change on individuals by comparing the arbitrariness in applicant rank attributable to policy change to the arbitrariness attributable to randomness. We find that any given policy has a high degree of arbitrariness and that removing race data from the ranking algorithm increases arbitrariness in outcomes for most applicants., Comment: 25 pages, 8 figures

Spatio-temporal graph (STG) forecasting is a critical task with extensive applications in the real world, including traffic and weather forecasting. Although several recent methods have been proposed to model complex dynamics in STGs, addressing long-range spatio-temporal dependencies remains a significant challenge, leading to limited performance gains. Inspired by a recently proposed state space model named Mamba, which has shown remarkable capability of capturing long-range dependency, we propose a new STG forecasting framework named SpoT-Mamba. SpoT-Mamba generates node embeddings by scanning various node-specific walk sequences. Based on the node embeddings, it conducts temporal scans to capture long-range spatio-temporal dependencies. Experimental results on the real-world traffic forecasting dataset demonstrate the effectiveness of SpoT-Mamba., Comment: 6 pages, 2 figures, 3 tables. Spatio-Temporal Reasoning and Learning (STRL) Workshop at the 33rd International Joint Conference on Artificial Intelligence (IJCAI 2024)

This study presents a novel framework for 3D gaze tracking tailored for mixed-reality settings, aimed at enhancing joint attention and collaborative efforts in team-based scenarios. Conventional gaze tracking, often limited by monocular cameras and traditional eye-tracking apparatus, struggles with simultaneous data synchronization and analysis from multiple participants in group contexts. Our proposed framework leverages state-of-the-art computer vision and machine learning techniques to overcome these obstacles, enabling precise 3D gaze estimation without dependence on specialized hardware or complex data fusion. Utilizing facial recognition and deep learning, the framework achieves real-time, tracking of gaze patterns across several individuals, addressing common depth estimation errors, and ensuring spatial and identity consistency within the dataset. Empirical results demonstrate the accuracy and reliability of our method in group environments. This provides mechanisms for significant advances in behavior and interaction analysis in educational and professional training applications in dynamic and unstructured environments., Comment: 9 pages, 8 figures, conference, submitted to ICMI 2024

The growing availability of well-organized Electronic Health Records (EHR) data has enabled the development of various machine learning models towards disease risk prediction. However, existing risk prediction methods overlook the heterogeneity of complex diseases, failing to model the potential disease subtypes regarding their corresponding patient visits and clinical concept subgroups. In this work, we introduce TACCO, a novel framework that jointly discovers clusters of clinical concepts and patient visits based on a hypergraph modeling of EHR data. Specifically, we develop a novel self-supervised co-clustering framework that can be guided by the risk prediction task of specific diseases. Furthermore, we enhance the hypergraph model of EHR data with textual embeddings and enforce the alignment between the clusters of clinical concepts and patient visits through a contrastive objective. Comprehensive experiments conducted on the public MIMIC-III dataset and Emory internal CRADLE dataset over the downstream clinical tasks of phenotype classification and cardiovascular risk prediction demonstrate an average 31.25% performance improvement compared to traditional ML baselines and a 5.26% improvement on top of the vanilla hypergraph model without our co-clustering mechanism. In-depth model analysis, clustering results analysis, and clinical case studies further validate the improved utilities and insightful interpretations delivered by TACCO. Code is available at https://github.com/PericlesHat/TACCO., Comment: 11 pages, 5 figures, to be published in Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining

Recent advancements in general-purpose AI have highlighted the importance of guiding AI systems towards the intended goals, ethical principles, and values of individuals and groups, a concept broadly recognized as alignment. However, the lack of clarified definitions and scopes of human-AI alignment poses a significant obstacle, hampering collaborative efforts across research domains to achieve this alignment. In particular, ML- and philosophy-oriented alignment research often views AI alignment as a static, unidirectional process (i.e., aiming to ensure that AI systems' objectives match humans) rather than an ongoing, mutual alignment problem [429]. This perspective largely neglects the long-term interaction and dynamic changes of alignment. To understand these gaps, we introduce a systematic review of over 400 papers published between 2019 and January 2024, spanning multiple domains such as Human-Computer Interaction (HCI), Natural Language Processing (NLP), Machine Learning (ML), and others. We characterize, define and scope human-AI alignment. From this, we present a conceptual framework of "Bidirectional Human-AI Alignment" to organize the literature from a human-centered perspective. This framework encompasses both 1) conventional studies of aligning AI to humans that ensures AI produces the intended outcomes determined by humans, and 2) a proposed concept of aligning humans to AI, which aims to help individuals and society adjust to AI advancements both cognitively and behaviorally. Additionally, we articulate the key findings derived from literature analysis, including discussions about human values, interaction techniques, and evaluations. To pave the way for future studies, we envision three key challenges for future directions and propose examples of potential future solutions., Comment: 56 pages

Electronic Health Records (EHRs) contain rich patient information and are crucial for clinical research and practice. In recent years, deep learning models have been applied to EHRs, but they often rely on massive features, which may not be readily available for all patients. We propose HTP-Star, which leverages hypergraph structures with a pretrain-then-finetune framework for modeling EHR data, enabling seamless integration of additional features. Additionally, we design two techniques, namely (1) Smoothness-inducing Regularization and (2) Group-balanced Reweighting, to enhance the model's robustness during fine-tuning. Through experiments conducted on two real EHR datasets, we demonstrate that HTP-Star consistently outperforms various baselines while striking a balance between patients with basic and extra features., Comment: CHIL 2024

The integration of language and 3D perception is crucial for developing embodied agents and robots that comprehend and interact with the physical world. While large language models (LLMs) have demonstrated impressive language understanding and generation capabilities, their adaptation to 3D environments (3D-LLMs) remains in its early stages. A primary challenge is the absence of large-scale datasets that provide dense grounding between language and 3D scenes. In this paper, we introduce 3D-GRAND, a pioneering large-scale dataset comprising 40,087 household scenes paired with 6.2 million densely-grounded scene-language instructions. Our results show that instruction tuning with 3D-GRAND significantly enhances grounding capabilities and reduces hallucinations in 3D-LLMs. As part of our contributions, we propose a comprehensive benchmark 3D-POPE to systematically evaluate hallucination in 3D-LLMs, enabling fair comparisons among future models. Our experiments highlight a scaling effect between dataset size and 3D-LLM performance, emphasizing the critical role of large-scale 3D-text datasets in advancing embodied AI research. Notably, our results demonstrate early signals for effective sim-to-real transfer, indicating that models trained on large synthetic data can perform well on real-world 3D scans. Through 3D-GRAND and 3D-POPE, we aim to equip the embodied AI community with essential resources and insights, setting the stage for more reliable and better-grounded 3D-LLMs. Project website: https://3d-grand.github.io, Comment: Project website: https://3d-grand.github.io

Large Language Models (LLMs) have shown promising results on various language and vision tasks. Recently, there has been growing interest in applying LLMs to graph-based tasks, particularly on Text-Attributed Graphs (TAGs). However, most studies have focused on node classification, while the use of LLMs for link prediction (LP) remains understudied. In this work, we propose a new task on LLMs, where the objective is to leverage LLMs to predict missing links between nodes in a graph. This task evaluates an LLM's ability to reason over structured data and infer new facts based on learned patterns. This new task poses two key challenges: (1) How to effectively integrate pairwise structural information into the LLMs, which is known to be crucial for LP performance, and (2) how to solve the computational bottleneck when teaching LLMs to perform LP. To address these challenges, we propose LinkGPT, the first end-to-end trained LLM for LP tasks. To effectively enhance the LLM's ability to understand the underlying structure, we design a two-stage instruction tuning approach where the first stage fine-tunes the pairwise encoder, projector, and node projector, and the second stage further fine-tunes the LLMs to predict links. To address the efficiency challenges at inference time, we introduce a retrieval-reranking scheme. Experiments show that LinkGPT can achieve state-of-the-art performance on real-world graphs as well as superior generalization in zero-shot and few-shot learning, surpassing existing benchmarks. At inference time, it can achieve $10\times$ speedup while maintaining high LP accuracy.

Recent advancements in foundation models (FMs) have unlocked new prospects in autonomous driving, yet the experimental settings of these studies are preliminary, over-simplified, and fail to capture the complexity of real-world driving scenarios in human environments. It remains under-explored whether FM agents can handle long-horizon navigation tasks with free-from dialogue and deal with unexpected situations caused by environmental dynamics or task changes. To explore the capabilities and boundaries of FMs faced with the challenges above, we introduce DriVLMe, a video-language-model-based agent to facilitate natural and effective communication between humans and autonomous vehicles that perceive the environment and navigate. We develop DriVLMe from both embodied experiences in a simulated environment and social experiences from real human dialogue. While DriVLMe demonstrates competitive performance in both open-loop benchmarks and closed-loop human studies, we reveal several limitations and challenges, including unacceptable inference time, imbalanced training data, limited visual understanding, challenges with multi-turn interactions, simplified language generation from robotic experiences, and difficulties in handling on-the-fly unexpected situations like environmental dynamics and task changes., Comment: First Vision and Language for Autonomous Driving and Robotics Workshop (VLADR @ CVPR 2024)

We present a set of 1194 Type-1 Rosseland-mean opacity tables for four different metallicity mixtures. These new Los Alamos OPLIB atomic radiative opacity tables are an order of magnitude larger in number than any previous opacity table release, and span regimes where previous opacity tables have not existed. For example, the new set of opacity tables expands the metallicity range to $Z$\,=\,10$^{-6}$ to $Z$\,=\,0.2 which allows improved accuracy of opacities at low and high metallicity, increases the table density in the metallicity range $Z$\,=\,10$^{-4}$ to $Z$\,=\,0.1 to enhance the accuracy of opacities drawn from interpolations across neighboring metallicities, and adds entries for hydrogen mass fractions between $X$\,=\,0 and $X$\,=\,0.1 including $X$\,=\,$10^{-2}, 10^{-3}, 10^{-4}, 10^{-5}, 10^{-6}$ that can improve stellar models of hydrogen deficient stars. We implement these new OPLIB radiative opacity tables in \MESA, and find that calibrated solar models agree broadly with previously published helioseismic and solar neutrino results. We find differences between using the new 1194 OPLIB opacity tables and the 126 OPAL opacity tables range from $\approx$\,20--80\% across individual chemical mixtures, up to $\approx$\,8\% and $\approx$\,15\% at the bottom and top of the solar convection zone respectively, and $\approx$\,7\% in the solar core. We also find differences between standard solar models using different opacity table sources that are on par with altering the initial abundance mixture. We conclude that this new, open-access set of OPLIB opacity tables does not solve the solar modeling problem, and suggest the investigation of physical mechanisms other than the atomic radiative opacity., Comment: 28 pages, 12 (13) figures. Accepted for Publication in ApJ

Despite advancements in medical care, hip fractures impose a significant burden on individuals and healthcare systems. This paper focuses on the prediction of hip fracture risk in older and middle-aged adults, where falls and compromised bone quality are predominant factors. We propose a novel staged model that combines advanced imaging and clinical data to improve predictive performance. By using CNNs to extract features from hip DXA images, along with clinical variables, shape measurements, and texture features, our method provides a comprehensive framework for assessing fracture risk. A staged machine learning-based model was developed using two ensemble models: Ensemble 1 (clinical variables only) and Ensemble 2 (clinical variables and DXA imaging features). This staged approach used uncertainty quantification from Ensemble 1 to decide if DXA features are necessary for further prediction. Ensemble 2 exhibited the highest performance, achieving an AUC of 0.9541, an accuracy of 0.9195, a sensitivity of 0.8078, and a specificity of 0.9427. The staged model also performed well, with an AUC of 0.8486, an accuracy of 0.8611, a sensitivity of 0.5578, and a specificity of 0.9249, outperforming Ensemble 1, which had an AUC of 0.5549, an accuracy of 0.7239, a sensitivity of 0.1956, and a specificity of 0.8343. Furthermore, the staged model suggested that 54.49% of patients did not require DXA scanning. It effectively balanced accuracy and specificity, offering a robust solution when DXA data acquisition is not always feasible. Statistical tests confirmed significant differences between the models, highlighting the advantages of the advanced modeling strategies. Our staged approach could identify individuals at risk with a high accuracy but reduce the unnecessary DXA scanning. It has great promise to guide interventions to prevent hip fractures with reduced cost and radiation., Comment: 29 pages, 5 figures, 6 tables

Barium (Ba) stars help to verify asymptotic giant branch (AGB) star nucleosynthesis models since they experienced pollution from an AGB binary companion and thus their spectra carry the signatures of the slow neutron capture process (s process). For 180 Ba stars, we searched for AGB stellar models that match the observed abundance patterns. We employed three machine learning algorithms as classifiers: a Random Forest method, developed for this work, and the two classifiers used in our previous study. We studied the statistical behaviour of the s-process elements in the observational sample to investigate if the AGB models systematically under- or overpredict the abundances observed in the Ba stars and show the results in the form of violin plots of the residuals between spectroscopic abundances and model predictions. We find a significant trend in the residuals that implies an underproduction of the elements Nb, Mo, and Ru in the models relative to the observations. This may originate from a process (e.g. the intermediate neutron-capture process, i process) at the metallicity of the Ba stars not yet included in the AGB models. Correlations are found between the residuals of these elements, suggesting a common origin for the deviations. In addition, there is a weak metallicity dependence of their residuals. The s-process temperatures derived with the [Zr/Fe] - [Nb/Fe] thermometer have an unrealistic value for the majority of our stars. The most likely explanation is that at least a fraction of these elements are not produced in a steady-state s process, and instead may be due to processes not included in the AGB models. The mass distribution of the identified models confirms that our sample of Ba stars was polluted by low-mass AGB stars. Most of the matching AGB models require low accreted mass, but a few systems with high accreted mass are needed to explain the observations. (abridged), Comment: 20 pages, 20 figures

Planet formation takes place in protoplanetary discs around young T-Tauri stars. PDS 70 is one of the first confirmed examples of a system where the planets are currently forming in gaps in the disc, and can be directly imaged. One of the main early influences on planet formation is the lifetime of the protoplanetary disk, which is limited by the intense stellar X-ray and UV radiation. Stellar coronal activity and accretion of material onto the star are both potential sources of XUV radiation. Previous \textit{Swift} observations detected UV emission, which were consistent with a low rate of accretion. We present follow up observations with the XMM-Newton observatory, which observed PDS 70 simultaneously in X-ray and UV in order to determine intensity of XUV radiation in the system, and identify if the source is coronal, accretion, or both. We detect a strong source in both X-ray and UV, with an average X-ray 0.2-12 keV luminosity of $1.37\times10^{30}\ \mathrm{erg\ s}^{-1}$, and a possible flare which increased the luminosity to $2.8\times10^{30}\ \mathrm{erg\ s}^{-1}$. The UV flux density is in excess of what would be expected from chromospheric emission, and supports the interpretation that PDS 70 has continuing weak accretion less than $\sim10^{-10}\ \mathrm{M_{\odot}\ yr^{-1}}$. The implications of the detected X-ray and UV radiation are that the disc is likely to be in the final stages of dispersal, and will be completely evaporated in the next million years, bringing an end to the primary planet formation process., Comment: 17 pages, Published in MNRAS

Humans are efficient language learners and inherently social creatures. Our language development is largely shaped by our social interactions, for example, the demonstration and feedback from caregivers. Contrary to human language learning, recent advancements in large language models have primarily adopted a non-interactive training paradigm, and refined pre-trained models through feedback afterward. In this work, we aim to examine how corrective feedback from interactions influences neural language acquisition from the ground up through systematically controlled experiments, assessing whether it contributes to learning efficiency in language models. We introduce a trial-and-demonstration (TnD) learning framework that incorporates three components: student trials, teacher demonstrations, and a reward conditioned on language competence at various developmental stages. Our experiments reveal that the TnD approach accelerates word acquisition for student models of equal and smaller numbers of parameters, and we highlight the significance of both trials and demonstrations. We further show that the teacher's choices of words influence students' word-specific learning efficiency, and a practice-makes-perfect effect is evident by a strong correlation between the frequency of words in trials and their respective learning curves. Our findings suggest that interactive language learning, with teacher demonstrations and student trials, can facilitate efficient word learning in language models.

Active wavelength-scale optoelectronic components are widely used in photonic integrated circuitry, however coherent sources of light -- namely optical lasers -- remain the most challenging component to integrate. Semiconductor nanowire lasers represent a flexible class of light source where each nanowire is both gain material and cavity; however, strong coupling between these properties and the performance leads to inhomogeneity across the population. While this has been studied and optimized for individual material systems, no architecture-wide insight is available. Here, nine nanowire laser material systems are studied and compared using 55,516 nanowire lasers to provide statistically robust insight into performance. These results demonstrate that, while it may be important to optimise internal quantum efficiency for certain materials, cavity effects are always critical. Our study provides a roadmap to optimize the performance of nanowire lasers made from any material: this can be achieved by ensuring a narrow spread of lengths and end-facet reflectivities.

Stellar spin down is a critical yet poorly understood component of stellar evolution. In particular, results from the Kepler Mission imply that mature age, solar-type stars have inefficient magnetic braking, resulting in a stalled spin down rate. However, a large number of precise asteroseismic ages are needed for mature ($\geq$ 3Gyr) stars in order to probe the regime where traditional and stalled spin-down models differ. In this paper, we present a new asteroseismic benchmark star for gyrochronology discovered using reprocessed Kepler short cadence data. KIC 11029516 (Papayu) is a bright ($K_{p}$ = 9.6 mag) solar-type star with well-measured rotation period (21.1$\pm$0.8 days) from spot modulation using 4 years of Kepler long cadence data. We combine asteroseismology and spectroscopy to obtain $T_{eff}=5888\pm100$ K, $\rm{[Fe/H]} = 0.30 \pm 0.06\,$ dex, $M = 1.24 \pm 0.05 M_{\odot}$, $R = 1.34 \pm 0.02 R_{\odot}$ and age of 4.0 $\pm$ 0.4 Gyr, making Papayu one of the most similar stars to the Sun in terms of temperature and radius with an asteroseismic age and a rotation period measured from spot modulation. We find that Papayu sits at the transition of where traditional and weakened spin-down models diverge. A comparison with stars of similar zero-age main-sequence temperatures supports previous findings that weakened spin-down models are required to explain the ages and rotation periods of old solar-type stars.

Computational marketing has become increasingly important in today's digital world, facing challenges such as massive heterogeneous data, multi-channel customer journeys, and limited marketing budgets. In this paper, we propose a general framework for marketing AI systems, the Neural Optimization with Adaptive Heuristics (NOAH) framework. NOAH is the first general framework for marketing optimization that considers both to-business (2B) and to-consumer (2C) products, as well as both owned and paid channels. We describe key modules of the NOAH framework, including prediction, optimization, and adaptive heuristics, providing examples for bidding and content optimization. We then detail the successful application of NOAH to LinkedIn's email marketing system, showcasing significant wins over the legacy ranking system. Additionally, we share details and insights that are broadly useful, particularly on: (i) addressing delayed feedback with lifetime value, (ii) performing large-scale linear programming with randomization, (iii) improving retrieval with audience expansion, (iv) reducing signal dilution in targeting tests, and (v) handling zero-inflated heavy-tail metrics in statistical testing., Comment: KDD 2024

Text-to-speech (TTS) development for African languages such as Luganda is still limited, primarily due to the scarcity of high-quality, single-speaker recordings essential for training TTS models. Prior work has focused on utilizing the Luganda Common Voice recordings of multiple speakers aged between 20-49. Although the generated speech is intelligible, it is still of lower quality than the model trained on studio-grade recordings. This is due to the insufficient data preprocessing methods applied to improve the quality of the Common Voice recordings. Furthermore, speech convergence is more difficult to achieve due to varying intonations, as well as background noise. In this paper, we show that the quality of Luganda TTS from Common Voice can improve by training on multiple speakers of close intonation in addition to further preprocessing of the training data. Specifically, we selected six female speakers with close intonation determined by subjectively listening and comparing their voice recordings. In addition to trimming out silent portions from the beginning and end of the recordings, we applied a pre-trained speech enhancement model to reduce background noise and enhance audio quality. We also utilized a pre-trained, non-intrusive, self-supervised Mean Opinion Score (MOS) estimation model to filter recordings with an estimated MOS over 3.5, indicating high perceived quality. Subjective MOS evaluations from nine native Luganda speakers demonstrate that our TTS model achieves a significantly better MOS of 3.55 compared to the reported 2.5 MOS of the existing model. Moreover, for a fair comparison, our model trained on six speakers outperforms models trained on a single-speaker (3.13 MOS) or two speakers (3.22 MOS). This showcases the effectiveness of compensating for the lack of data from one speaker with data from multiple speakers of close intonation to improve TTS quality., Comment: Presented at the AfricaNLP workshop at ICLR 2024

Linking (aligning) biomedical concepts across diverse data sources enables various integrative analyses, but it is challenging due to the discrepancies in concept naming conventions. Various strategies have been developed to overcome this challenge, such as those based on string-matching rules, manually crafted thesauri, and machine learning models. However, these methods are constrained by limited prior biomedical knowledge and can hardly generalize beyond the limited amounts of rules, thesauri, or training samples. Recently, large language models (LLMs) have exhibited impressive results in diverse biomedical NLP tasks due to their unprecedentedly rich prior knowledge and strong zero-shot prediction abilities. However, LLMs suffer from issues including high costs, limited context length, and unreliable predictions. In this research, we propose PromptLink, a novel biomedical concept linking framework that leverages LLMs. It first employs a biomedical-specialized pre-trained language model to generate candidate concepts that can fit in the LLM context windows. Then it utilizes an LLM to link concepts through two-stage prompts, where the first-stage prompt aims to elicit the biomedical prior knowledge from the LLM for the concept linking task and the second-stage prompt enforces the LLM to reflect on its own predictions to further enhance their reliability. Empirical results on the concept linking task between two EHR datasets and an external biomedical KG demonstrate the effectiveness of PromptLink. Furthermore, PromptLink is a generic framework without reliance on additional prior knowledge, context, or training data, making it well-suited for concept linking across various types of data sources. The source code is available at https://github.com/constantjxyz/PromptLink.

Online convex optimization (OCO) is a powerful tool for learning sequential data, making it ideal for high precision control applications where the disturbances are arbitrary and unknown in advance. However, the ability of OCO-based controllers to accurately learn the disturbance while maintaining closed-loop stability relies on having an accurate model of the plant. This paper studies the performance of OCO-based controllers for linear time-invariant (LTI) systems subject to disturbance and model uncertainty. The model uncertainty can cause the closed-loop to become unstable. We provide a sufficient condition for robust stability based on the small gain theorem. This condition is easily incorporated as an on-line constraint in the OCO controller. Finally, we verify via numerical simulations that imposing the robust stability condition on the OCO controller ensures closed-loop stability.

ExoplANETS-A is an EU Horizon-2020 project with the primary objective of establishing new knowledge on exoplanet atmospheres. Intimately related to this topic is the study of the host-stars radiative properties in order to understand the environment in which exoplanets lie. The aim of this work is to exploit archived data from space-based observatories and other public sources to produce uniform sets of stellar data that can establish new insight on the influence of the host star on the planetary atmosphere. We have compiled X-ray and UV luminosities, which affect the formation and the atmospheric properties of the planets, and stellar parameters, which impact the retrieval process of the planetary-atmosphere's properties and its errors. Our sample is formed of all transiting-exoplanet systems observed by HST or Spitzer. It includes 205 exoplanets and their 114 host-stars. We have built a catalogue with information extracted from public, online archives augmented by quantities derived by the Exoplanets-A work. With this catalogue we have implemented an online database which also includes X-ray and OHP spectra and TESS light curves. In addition, we have developed a tool, exoVOSA, which is able to fit the spectral energy distribution of exoplanets. We give an example of using the database to study the effects of the host-star high-energy emission on the exoplanet atmosphere. The sample has a planet radius valley which is located at 1.8 Earth radii, in agreement with previous studies. Multiplanet systems in our sample were used to test the photoevaporation model and we find that out of 14 systems, only one significant case poses a contradiction to it (K2-3). In summary, the exoplanet and stellar resources compiled and generated by ExoplANETS-A form a sound basis for current JWST observations and for future work in the era of Ariel., Comment: 14 pages

While a number of knowledge graph representation learning (KGRL) methods have been proposed over the past decade, very few theoretical analyses have been conducted on them. In this paper, we present the first PAC-Bayesian generalization bounds for KGRL methods. To analyze a broad class of KGRL models, we propose a generic framework named ReED (Relation-aware Encoder-Decoder), which consists of a relation-aware message passing encoder and a triplet classification decoder. Our ReED framework can express at least 15 different existing KGRL models, including not only graph neural network-based models such as R-GCN and CompGCN but also shallow-architecture models such as RotatE and ANALOGY. Our generalization bounds for the ReED framework provide theoretical grounds for the commonly used tricks in KGRL, e.g., parameter-sharing and weight normalization schemes, and guide desirable design choices for practical KGRL methods. We empirically show that the critical factors in our generalization bounds can explain actual generalization errors on three real-world knowledge graphs., Comment: 32 pages, 3 figures, 4 tables, The 41st International Conference on Machine Learning (ICML 2024)

Investigating children's embodied learning in mixed-reality environments, where they collaboratively simulate scientific processes, requires analyzing complex multimodal data to interpret their learning and coordination behaviors. Learning scientists have developed Interaction Analysis (IA) methodologies for analyzing such data, but this requires researchers to watch hours of videos to extract and interpret students' learning patterns. Our study aims to simplify researchers' tasks, using Machine Learning and Multimodal Learning Analytics to support the IA processes. Our study combines machine learning algorithms and multimodal analyses to support and streamline researcher efforts in developing a comprehensive understanding of students' scientific engagement through their movements, gaze, and affective responses in a simulated scenario. To facilitate an effective researcher-AI partnership, we present an initial case study to determine the feasibility of visually representing students' states, actions, gaze, affect, and movement on a timeline. Our case study focuses on a specific science scenario where students learn about photosynthesis. The timeline allows us to investigate the alignment of critical learning moments identified by multimodal and interaction analysis, and uncover insights into students' temporal learning progressions.

We introduce a Bayesian optimization approach to guide the sputter deposition of molybdenum thin films, aiming to achieve desired residual stress and sheet resistance while minimizing susceptibility to stochastic fluctuations during deposition. Thin films are pivotal in numerous technologies, including semiconductors and optical devices, where their properties are critical. Sputter deposition parameters, such as deposition power, vacuum chamber pressure, and working distance, influence physical properties like residual stress and resistance. Excessive stress and high resistance can impair device performance, necessitating the selection of optimal process parameters. Furthermore, these parameters should ensure the consistency and reliability of thin film properties, assisting in the reproducibility of the devices. However, exploring the multidimensional design space for process optimization is expensive. Bayesian optimization is ideal for optimizing inputs/parameters of general black-box functions without reliance on gradient information. We utilize Bayesian optimization to optimize deposition power and pressure using a custom-built objective function incorporating observed stress and resistance data. Additionally, we integrate prior knowledge of stress variation with pressure into the objective function to prioritize films least affected by stochastic variations. Our findings demonstrate that Bayesian optimization effectively explores the design space and identifies optimal parameter combinations meeting desired stress and resistance specifications.

A single-lined spectroscopic orbit for the M giant in the X-ray binary IGR J16194-2810 is determined from a time-series of optical spectra. The spectroscopic orbital period of 192.5 days is twice that of the photometric period, confirming that the M giant in the system is an ellipsoidal variable. The giant is identified as a first ascent giant approaching the red giant tip. The primary is a neutron star (NS) with its M giant companion filling its Roche lobe verifying the system classification as a Low-Mass X-ray binary (LMXB). Stellar C, N, O and Fe abundances are derived for the M giant with the C, N, O values typical for a field giant with [Fe/H] = -0.14. The system does not have a large kick velocity. Models for the evolution of the system into a binary NS-white dwarf (WD) are presented. The X-ray properties are discussed in the context of this model. This binary is a rare example of a luminous, long orbital period, LMXB early in the transient ellipsoidal phase., Comment: accepted for publication in ApJ

Modern cloud has turned data services into easily accessible commodities. With just a few clicks, users are now able to access a catalog of data processing systems for a wide range of tasks. However, the cloud brings in both complexity and opportunity. While cloud users can quickly start an application by using various data services, it can be difficult to configure and optimize these services to gain the most value from them. For cloud providers, managing every aspect of an ever-increasing set of data services, while meeting customer SLAs and minimizing operational cost is becoming more challenging. Cloud technology enables the collection of significant amounts of workload traces and system telemetry. With the progress in data science (DS) and machine learning (ML), it is feasible and desirable to utilize a data-driven, ML-based approach to automate various aspects of data services, resulting in the creation of autonomous data services. This paper presents our perspectives and insights on creating autonomous data services on Azure. It also covers the future endeavors we plan to undertake and unresolved issues that still need attention., Comment: SIGMOD Companion of the 2023 International Conference on Management of Data. 2023

Developing effective biomedical retrieval models is important for excelling at knowledge-intensive biomedical tasks but still challenging due to the deficiency of sufficient publicly annotated biomedical data and computational resources. We present BMRetriever, a series of dense retrievers for enhancing biomedical retrieval via unsupervised pre-training on large biomedical corpora, followed by instruction fine-tuning on a combination of labeled datasets and synthetic pairs. Experiments on 5 biomedical tasks across 11 datasets verify BMRetriever's efficacy on various biomedical applications. BMRetriever also exhibits strong parameter efficiency, with the 410M variant outperforming baselines up to 11.7 times larger, and the 2B variant matching the performance of models with over 5B parameters. The training data and model checkpoints are released at \url{https://huggingface.co/BMRetriever} to ensure transparency, reproducibility, and application to new domains., Comment: Work in progress. The model and data will be uploaded to \url{https://github.com/ritaranx/BMRetriever}

Many children in foster care experience trauma that is rooted in unstable family relationships. Other members of the foster care system like foster parents and social workers face secondary trauma. Drawing on 10 years of Reddit data, we used a mixed methods approach to analyze how different members of the foster care system find support and similar experiences at the intersection of two Reddit communities - foster care, and abuse. Users who cross this boundary focus on trauma experiences specific to different roles in foster care. While representing a small number of users, boundary crossing users contribute heavily to both communities, and, compared to matching users, receive higher scores and more replies. We explore the roles boundary crossing users have both in the online community and in the context of foster care. Finally, we present design recommendations that would support trauma survivors find communities more suited to their personal experiences.

Mental health in children and adolescents has been steadily deteriorating over the past few years. The recent advent of Large Language Models (LLMs) offers much hope for cost and time efficient scaling of monitoring and intervention, yet despite specifically prevalent issues such as school bullying and eating disorders, previous studies on have not investigated performance in this domain or for open information extraction where the set of answers is not predetermined. We create a new dataset of Reddit posts from adolescents aged 12-19 annotated by expert psychiatrists for the following categories: TRAUMA, PRECARITY, CONDITION, SYMPTOMS, SUICIDALITY and TREATMENT and compare expert labels to annotations from two top performing LLMs (GPT3.5 and GPT4). In addition, we create two synthetic datasets to assess whether LLMs perform better when annotating data as they generate it. We find GPT4 to be on par with human inter-annotator agreement and performance on synthetic data to be substantially higher, however we find the model still occasionally errs on issues of negation and factuality and higher performance on synthetic data is driven by greater complexity of real data rather than inherent advantage.

The nature of the minute-to-hour long Fast X-ray Transients (FXTs) localised by telescopes such as Chandra, Swift, and XMM-Newton remains mysterious, with numerous models suggested for the events. Here, we report multi-wavelength observations of EP240315a, a 1600 s long transient detected by the Einstein Probe, showing it to have a redshift of z=4.859. We measure a low column density of neutral hydrogen, indicating that the event is embedded in a low-density environment, further supported by direct detection of leaking ionising Lyman-continuum. The observed properties are consistent with EP240315a being a long-duration gamma-ray burst, and these observations support an interpretation in which a significant fraction of the FXT population are lower-luminosity examples of similar events. Such transients are detectable at high redshifts by the Einstein Probe and, in the (near) future, out to even larger distances by SVOM, THESEUS, and Athena, providing samples of events into the epoch of reionisation., Comment: 41 pages, 7 figures, submitted

In this work is considered a spectral problem, involving a second order term on the domain boundary: the Laplace-Beltrami operator. A variational formulation is presented, leading to a finite element discretization. For the Laplace-Beltrami operator to make sense on the boundary, the domain is smooth: consequently the computational domain (classically a polygonal domain) will not match the physical one. Thus, the physical domain is discretized using high order curved meshes so as to reduce the \textit{geometric error}. The \textit{lift operator}, which is aimed to transform a function defined on the mesh domain into a function defined on the physical one, is recalled. This \textit{lift} is a key ingredient in estimating errors on eigenvalues and eigenfunctions. A bootstrap method is used to prove the error estimates, which are expressed both in terms of \textit{finite element approximation error} and of \textit{geometric error}, respectively associated to the finite element degree $k\ge 1$ and to the mesh order~$r\ge 1$. Numerical experiments are led on various smooth domains in 2D and 3D, which allow us to validate the presented theoretical results.

SPHEREx, the Spectro-Photometer for the History of the Universe, Epoch of Reionization, and ices Explorer, is a NASA MIDEX mission planned for launch in 2024. SPHEREx will carry out the first all-sky spectral survey at wavelengths between 0.75 micron and 5 micron with spectral resolving power ~40 between 0.75 and 3.8 micron and ~120 between 3.8 and 5 micron At the end of its two-year mission, SPHEREx will provide 0.75-to-5 micron spectra of each 6.2"x6.2" pixel on the sky - 14 billion spectra in all. This paper updates an earlier description of SPHEREx presenting changes made during the mission's Preliminary Design Phase, including a discussion of instrument integration and test and a summary of the data processing, analysis, and distribution plans.