MD-SOAR Maryland Shared Open Access Repository
Not a member yet
34521 research outputs found
Sort by
Modeling 2D van der Waals Materials with Homonuclear Bonds of Main Group Cations
No full textWe survey a range of 2D van der Waals (vdW) layered materials that contain homonuclear bonds of main group elements, specifically Ga–Ga, In–In, Si–Si, Ge–Ge, and P–P, using first-principles density functional theory (DFT) methods. The covalent bonding and geometries present in these materials can stabilize oxidation states that differ from those found in conventional semiconductors composed of main group elements, such as Si, InP, and GaAs. Since 2D vdW materials did not gain widespread use until recently, many have been excluded from the first-principles test sets developed over a decade prior, where the focus was on determining the accuracy of potentials used in different modeling methods. In this study, we benchmark the set by exploring a range of modeling methods that include GGA and meta-GGA exchange-correlation functionals commonly used in first-principles DFT methods, as well as exact exchange introduced using HSE06 and PBE0. We investigate their effects on the ground-state structure, electronic band structure, and computational cost and report on this benchmarking data. Our test set of 2D vdW materials contains multiple structural types that span binary, ternary, and quaternary compositions, including ferroic ground states. We found that the vdW-corrected GGA is capable of accurately capturing lattice constants (within ±2% relative to available experimental data) across various 2D vdW materials in our test set, with relatively low computational cost and turn-key compatibility with the open-source code Quantum Espresso. Additionally, vdW-corrected GGA can reliably identify stable ferroelectric and ferromagnetic ground states, can be used to determine trends in electronic band structure, and serve as a starting point for predicting more accurate band gaps. The predicted electronic band structure and corresponding projected density of states (PDOS) are crucial for establishing the connection between microscopic properties and atomic or molecular orbitals, which can, in turn, be used to predict novel functional 2D vdW materials.The authors wish to acknowledge DTRA through grant number HDTRA12410015. Calculations were performed, in part, using the UMBC High Performance Computing Facility (HPCF), supported by the National Science Foundation under the MRI grants CNS-0821258, CNS-1228778, and OAC1726023 and the SCREMS grant DMS-0821311. This research used the Theory and Computation facility of the Center for Functional Nanomaterials (CFN), which is a U.S. Department of Energy Office of Science User Facility, at Brookhaven National Laboratory under Contract No. DE-SC0012704.https://pubs.acs.org/doi/full/10.1021/acsorginorgau.5c0009
Narrating Belonging: The Effects of Digital Storytelling on Retaining STEM Students in Higher Education
No full textThis thesis explores how creative narrative practices can support students with limited visibility in STEM fields. Across three studies, the research examines the roles of relatable mentors and role models, and introduces digital storytelling as a method to foster academic engagement and personal growth. The first study identifies key challenges these students face, including limited faculty visibility and the absence of supportive mentors. The second study investigates students’ perspectives on role models, emphasizing the importance of shared experiences, attainable success, and transparent narratives. The third study evaluates the impact of a digital storytelling workshop where participants created multimedia stories about STEM professionals, leading to improvements in academic motivation, career confidence, and emotional connection. Together, these studies demonstrate that participatory storytelling serves as a valuable educational tool. By amplifying student voices and fostering deeper connections with role models, this research offers strategies to strengthen retention and engagement in STEM learning environments
LLM-Assisted Emergency Triage Benchmark: Bridging Hospital-Rich and MCI-Like Field Simulation
No full textGenAI4Health@NeurIPS 2025, The Second Workshop on GenAI for HealthPotential, Trust, and Policy Compliance, San Diego, California, December 6, 2025Research on emergency and mass casualty incident (MCI) triage has been limited by the absence of openly usable, reproducible benchmarks. Yet these scenarios demand rapid identification of the patients most in need, where accurate deterioration prediction can guide timely interventions. While the MIMIC-IV-ED database is openly available to credentialed researchers, transforming it into a triage-focused benchmark requires extensive preprocessing, feature harmonization, and schema alignment -- barriers that restrict accessibility to only highly technical users. We address these gaps by first introducing an open, LLM-assisted emergency triage benchmark for deterioration prediction (ICU transfer, in-hospital mortality). The benchmark then defines two regimes: (i) a hospital-rich setting with vitals, labs, notes, chief complaints, and structured observations, and (ii) an MCI-like field simulation limited to vitals, observations, and notes. Large language models (LLMs) contributed directly to dataset construction by (i) harmonizing noisy fields such as AVPU and breathing devices, (ii) prioritizing clinically relevant vitals and labs, and (iii) guiding schema alignment and efficient merging of disparate tables. We further provide baseline models and SHAP-based interpretability analyses, illustrating predictive gaps between regimes and the features most critical for triage. Together, these contributions make triage prediction research more reproducible and accessible -- a step toward dataset democratization in clinical AI.http://arxiv.org/abs/2509.2635
Multimodal Large Language Models for Enhanced Traffic Safety: A Comprehensive Review and Future Trends
No full textTraffic safety remains a critical global challenge, with traditional Advanced Driver-Assistance Systems (ADAS) often struggling in dynamic real-world scenarios due to fragmented sensor processing and susceptibility to adversarial conditions. This paper reviews the transformative potential of Multimodal Large Language Models (MLLMs) in addressing these limitations by integrating cross-modal data such as visual, spatial, and environmental inputs to enable holistic scene understanding. Through a comprehensive analysis of MLLM-based approaches, we highlight their capabilities in enhancing perception, decision-making, and adversarial robustness, while also examining the role of key datasets (e.g., KITTI, DRAMA, ML4RoadSafety) in advancing research. Furthermore, we outline future directions, including real-time edge deployment, causality-driven reasoning, and human-AI collaboration. By positioning MLLMs as a cornerstone for next-generation traffic safety systems, this review underscores their potential to revolutionize the field, offering scalable, context-aware solutions that proactively mitigate risks and improve overall road safety.http://arxiv.org/abs/2504.1613
TAG-EQA: Text-And-Graph for Event Question Answering via Structured Prompting Strategies
No full textLarge language models (LLMs) excel at general language tasks but often struggle with event-based questions-especially those requiring causal or temporal reasoning. We introduce TAG-EQA (Text-And-Graph for Event Question Answering), a prompting framework that injects causal event graphs into LLM inputs by converting structured relations into natural-language statements. TAG-EQA spans nine prompting configurations, combining three strategies (zero-shot, few-shot, chain-of-thought) with three input modalities (text-only, graph-only, text+graph), enabling a systematic analysis of when and how structured knowledge aids inference. On the TORQUESTRA benchmark, TAG-EQA improves accuracy by 5% on average over text-only baselines, with gains up to 12% in zero-shot settings and 18% when graph-augmented CoT prompting is effective. While performance varies by model and configuration, our findings show that causal graphs can enhance event reasoning in LLMs without fine-tuning, offering a flexible way to encode structure in prompt-based QA.We thank the reviewers for their detailed comments and suggestions. Some experiments were conducted on the UMBC HPCF, supported by the National Science Foundation under Grant No. CNS1920079. This material is also based on research that is in part supported by the Army Research Laboratory, Grant No. W911NF2120076, and by DARPA for the SciFy program under agreement number HR00112520301. The U.S. Government is authorized to reproduce and distribute reprints for Governmental purposes notwithstanding any copyright notation thereon. The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either express or implied, of DARPA or the U.S. Government.http://arxiv.org/abs/2510.0139
The Appalachian Mountains: A Geological and Archaeological History
No full textThe Appalachian Mountains are among the oldest and most geologically complex ranges on Earth, formed over a billion years through successive tectonic collisions, including the Grenville, Taconic, Acadian, and Alleghanian orogenies (Hatcher 2010; Tollo, Corriveau, McLelland, and Bartholomew 2004). These events produced a landscape of extraordinary structural diversity, featuring plateaus, ridges, valleys, and karst systems, profoundly shaping ecological systems and human history. This thesis examines the Appalachian Mountains as an interconnected system where geology, ecology, and culture are inseparably linked.
The project integrates geological, archaeological, paleontological, and cultural perspectives, focusing on the unglaciated Appalachian Plateau and Ridge and Valley regions. Archaeological case studies, arranged chronologically from the late Pleistocene to the late prehistoric period, reveal how landforms such as rockshelters, salt licks, and river valleys structured patterns of human mobility, subsistence, and settlement (Cobb & Nassaney 2002; Cremeens, Whisonant, and Davis 2003). Paleoenvironmental reconstructions and site analyses demonstrate that caves and karst zones served as ecological refugia and cultural landscapes, preserving evidence of everyday lifeways and ritual activity.
The central argument of this project is that the Appalachian landscape was a dynamic force, not a static backdrop. It actively shaped human migration and cultural development over nearly 20,000 years. By arranging sites chronologically, this study shows a direct cause-and-effect trajectory from early, mobile hunter-gatherers to increasingly sedentary and socially complex societies. The Appalachians were not a monolithic barrier. Instead, they formed a resource-rich mosaic of environments whose unique geology repeatedly catalyzed cultural transformation. In doing so, this thesis not only situates Appalachia within deep geological time but also foregrounds its role in shaping the long arc of human history in eastern North America
Cultivating preservice teachers’ critical love via exploration of names
No full textNames are deeply intertwined with identity, carrying cultural, linguistic, and familial significance. Grounded in Sealey-Ruiz’s concept of critical love, this study explores a naming project designed for predominantly white preservice teachers at a rural U.S. institution in the Appalachian Highlands. Drawing on Hammond’s culture tree framework and Bishop’s metaphor of books as mirrors, windows, and sliding glass doors, the project moved beyond surface-level cultural awareness to interrogate how names reflect belonging and exclusion. Findings revealed that many preservice teachers engaged with their own naming histories. Reading and discussing multicultural children’s literature catalysed their deeper reflection. They grappled with the emotional weight of name mispronunciations, anglicization, and erasure, recognising how these experiences shape their prospective young students’ sense of self. This study highlights the need for teacher preparation programmes to integrate culturally sustaining pedagogies that cultivate linguistic responsiveness and critical love, ensuring that future educators affirm students’ full identities.https://www.tandfonline.com/doi/full/10.1080/13540602.2025.257709
Tiffany Kelly for Gaithersburg Mayor
No full textThe City of Gaithersburg in Montgomery County, MD, is holding elections November 4. Two candidates are running for mayor: incumbent Jud Ashman and challenger Tiffany Kelly. In accompanying double-header episodes, Sunil Dasgupta talks with community activist Kelly about need gaps in the city. tiffanyisrunning.com Music By Adam Bobrow.https://open.spotify.com/episode/284texEW2ub4lncRV8C05
Determining functionals and data assimilation and a novel regularity criterion for the three-dimensional navier–stokes equations
No full textIn this paper we present two results: (1) a data assimilation algorithm for the 3D Navier–Stokes equation (3D NSE) using nodal data and, as a consequence, (2) a novel regularity criterion for the 3D NSE based on finitely many observations of the velocity. The data assimilation algorithm we employ utilizes nudging, a method based on a Newtonian relaxation scheme motivated by feedback control. The observations, which may be either modal, nodal or volume elements, are drawn from a weak solution of the 3D NSE and are collected almost everywhere in time over a finite grid, and our results, including the regularity criterion, hold for data of any of the aforementioned forms. The regularity criterion we propose follows from our data assimilation algorithm and is hence intimately connected to the notion of determining functionals (modes, nodes and volume elements). To the best of our knowledge, all existing regularity criteria require knowing the solution of the 3D NSE almost everywhere in space. Our regularity criterion is fundamentally different from any preexisting regularity criterion as it is based on finitely many observations (modes, nodes and volume elements). We further prove that the regularity criterion we propose is both a necessary and sufficient condition for regularity. Thus, our result can be viewed as a natural generalization of the notion of determining modes, nodes and volume elements as well as the asymptotic tracking property of the nudging algorithm for the 2D NSE to the 3D setting.https://link.springer.com/article/10.1007/s40687-025-00530-
Quantum On Track: UMBC Researchers Demonstrate Feasibility Of Using Quantum Devices To Manage Urban Train Scheduling, Using A Baltimore Transit Line As A Model
No full textphotographers: Marlayn Demond and Elijah DavisA new study led by UMBC researchers—and focused on Baltimore’s Light RailLink, a hybrid tram-rail network sharing roads with cars inside Baltimore.https://umbc.edu/stories/quantum-on-track-for-train-scheduling