Digital Commons Network^™

Exploring Autonomous Agents: A Closer Look At Why They Fail When Completing Tasks, Ruofan Lu, Yichen Li, Yintong Huo

Research Collection School Of Computing and Information Systems

Autonomous agent systems powered by Large Language Models (LLMs) have demonstrated promising capabilities in automating complex tasks. However, current evaluations largely rely on success rates without systematically analyzing the interactions, communication mechanisms, and failure causes within these systems. To bridge this gap, we present a benchmark of 34 representative programmable tasks designed to rigorously assess autonomous agents. Using this benchmark, we evaluate three popular open-source agent frameworks combined with two LLM backbones, observing a task completion rate of approximately 50%. Through in-depth failure analysis, we develop a three-tier taxonomy of failure causes aligned with task phases, highlighting planning errors, task …

Envisioning Future Interactive Web Development: Editing Webpage With Natural Language, Truong Hai Dang, Jingyu Xiao, Yintong Huo

Research Collection School Of Computing and Information Systems

The evolution of web applications relies on iterative code modifications, a process that is traditionally manual and time-consuming. While Large Language Models (LLMs) can generate UI code, their ability to edit existing code from new design requirements (e.g., ”center the logo”) remains a challenge. This is largely due to the absence of large-scale, high-quality tuning data to align model performance with human expectations. In this paper, we introduce a novel, automated data generation pipeline that uses LLMs to synthesize a high-quality fine-tuning dataset for web editing, named Instruct4Edit. Our approach generates diverse instructions, applies the corresponding code modifications, and performs …

Swiping Rights: Hidden Costs Of Hidden Laws, Rachel Landy

Articles

While many legal regimes have undergone fundamental shifts to account for the online economy, one lags behind: industry-specific mandatory contract rules. These rules, frequently buried in decades-old, nonintuitive statutes, are often overly prescriptive and come with draconian penalties, hampering the goals of entrepreneurship and impeding startup growth.

Much has been written about the intersection of regulation and entrepreneurship. But sparse attention has been paid to the unintended consequences on early-stage companies of more subtle regulatory interventions like contract rules — especially those with inconsistent or conflicting requirements among the states. This Article begins to fill that gap and uses as …

The Irrelevance Of “Peak Fame” On Celebrity Candidate Success, Richard T. Longoria, Dana Reyes

Political Science Faculty Publications

Objective

Does it matter “when” a celebrity runs for office? The political scientists who study celebrity candidates need to consider an important insight from communications and media scholars. Fame is a time-dependent variable. Someone who was famous 40 years ago may no longer be widely known today.

Methods

A list of celebrity political candidates who ran for office in the United States between 1865 and 2022 was created to measure campaign success.

Results

In terms of winning an election, it does not matter whether a celebrity candidate runs at the peak of their fame or waits until decades later. What …

The Functional Connectome Mediates Associations Between Fitness And Cognition Across The Adult Lifespan, Juniad Merchant, Jeremy J. Purcell, Daniel D. Callow, Junyeon Won, Yash Kommula, Samuel Rosenberg, Yi Wei, John L. Woodard, Kristy A. Nielson, Stephen M. Rao, Sally Durgerian, Shuo Chen, J. Carson Smith

Psychology Faculty Research and Publications

Cardiorespiratory fitness (CRF) is increasingly recognized as essential for improving neural and cognitive function, especially in older age. Here, we examined a cross-sectional sample of the Human Connectome Project (HCP) Aging dataset (ages 36–100; N = 378) to test the hypothesis that functional connectivity profiles that are related to CRF are also associated with episodic memory and executive function across middle age and older adults. To test these hypotheses, we first used connectome-based predictive modeling (CPM) to identify functional connections (FCs) that were separately predictive of CRF (2 Minute Walk Test), episodic memory (Rey Auditory Verbal Learning Test), and executive …

Student Engagement In Early Education Classrooms: A Mixed Methods Study Examining Student And Teacher Perspectives, Jill Ann Derosa

Theses & Dissertations

Student disengagement is a persistent issue in U.S. public schools and has been linked to negative outcomes for students. When students are engaged in learning, they are more likely to achieve academic success and socioemotional well-being in school. This study evaluated multiple perspectives on student engagement to gain a deeper understanding of how students and teachers in one school are making sense of engagement in their classrooms. Using a mixed-methods approach, I employed student surveys and conducted focus groups with both students and teachers to gather data on participants’ perceptions of student engagement. The findings from this study highlight that …

Context-Aware Hierarchical Taxonomy Generation For Scientific Papers Via Llm-Guided Multi-Aspect Clustering, Kun Zhu, Lizi Liao, Yuxuan Gu, Lei Huang, Xiaocheng Feng, Bing Qin

Research Collection School Of Computing and Information Systems

The rapid growth of scientific literature demands efficient methods to organize and synthesize research findings. Existing taxonomy construction methods, leveraging unsupervised clustering or direct prompting of large language models (LLMs), often lack coherence and granularity. We propose a novel context-aware hierarchical taxonomy generation framework that integrates LLM-guided multi-aspect encoding with dynamic clustering. Our method leverages LLMs to identify key aspects of each paper (e.g., methodology, dataset, evaluation) and generates aspect-specific paper summaries, which are then encoded and clustered along each aspect to form a coherent hierarchy. In addition, we introduce a new evaluation benchmark of 156 expert-crafted taxonomies encompassing 11.6k …

Why Stop At One Error? Benchmarking Llms As Data Science Code Debuggers For Multi-Hop And Multi-Bug Errors, Zhiyu Yang, Shuo Wang, Yukun Yan, Yang Deng

Research Collection School Of Computing and Information Systems

LLMs are transforming software development, yet current code generation and code repair benchmarks mainly assess syntactic and functional correctness in simple, single-error cases. LLMs’ capabilities to autonomously find and fix runtime logical errors in complex data science code remain largely unexplored. To address this gap, we introduce DSDBench: the Data Science Debugging Benchmark, the first benchmark for systematic evaluation of LLMs on multi-hop error tracing and multi-bug detection in data science code debugging. DSDBench adapts datasets from existing data science task benchmarks, such as DABench and MatPlotBench, featuring realistic data science debugging tasks with automatically synthesized multi-hop, multi-bug code snippets. …

A Comparison Of Comparisons: Evidence From An International Comparative Study Of 'Smart Cities', Kevin Ward, Teresa Abbruzzese, Tim Bunnell, Paolo Cardullo, I-Chun Catherine Chang, Byron Miller, Ramon Ribera-Fumaz, Haeran Shin, Zachary Spicer, Orlando Woods

Research Collection College of Integrative Studies

Every year the list lengthens of cities with some sort of ‘smart city’ public policy. In some, it emerges as the latest in a long line of urban digital and information communication policies. In others, the introduction of the notion of the ‘smart city’ marks a departure from past approaches to public policy. Additionally, the more studies emerge of actual smart city policies, then the less definitional agreement there seems to be. Nevertheless, that we have witnessed in the last two decades the ‘repeated instance’ of smart cities emerging in cities around the world seems incontrovertible. Like so much urban …

From Great “Liberator” To “Landlord Seeking Rent”: The Implications Of U.S. Reciprocal Tariffs Policy In Asia And Beyond, Henry S. Gao

Research Collection Yong Pung How School Of Law

The post-war international economic order was, to a large extent, underwritten by U.S. leadership. Nowhere was this more visible than in Asia, where the United States not only financed post-colonial development but also provided open access to its huge market, laying the foundation for export-led growth across the region. It underpinned regional stability through a blend of military, diplomatic, and economic engagement, including costly interventions in the Korean and Vietnam wars. That legacy, however, was fundamentally disrupted on April 2, 2025, when the Trump administration unveiled sweeping tariffs targeting key Asian economies. As Singapore’s defense minister wryly observed, the United …

Nationalism And Anglo-American Neo-Colonialism In Southeast Asia, 1945-1965, Wen-Qing (Wei Wenqing) Ngoei

Research Collection College of Integrative Studies

It is tempting to treat the United States’ ill-fated military intervention in Vietnam in the 1960s as the desperate final gasp of a moribund Western imperial system. Certainly, scholars have described Washington’s military debacle in Vietnam as the end of America’s “short-lived empire” in Southeast Asia, emblematic of the broader, decisive triumph of Southeast Asian nationalism over western colonialism. This reading of Southeast Asian countries attaining formal independence after 1945, with tacit assumptions that colonialism and nationalism exist in binary opposition, tends to ossify when the region’s history is viewed through the lens of U.S. defeat in Vietnam.

In contrast, …

Progresssive Maculopathy In A Child Resulting From Vitreopapillary Traction On A Congenitally Anomalous Nerve, Ross M. Kennamer-Chapman, Lee M. Jampol, Rithwick Rajagopal

2020-Current year OA Pubs

PURPOSE: To present a case of worsening maculopathy and vision loss because of vitreopapillary traction associated with an anomalous optic nerve head in a pediatric patient successfully managed with pars plana vitrectomy.

METHODS: Retrospective case report.

RESULTS: A 13-year-old boy presented with unilateral maculopathy that slowly progressed over a 4-year period. As his visual acuity declined, he became increasingly symptomatic. The macular pathology occurred in the setting of vitreopapillary traction overlying multiple congenital optic nerve abnormalities, including a myelinated nerve fiber layer and a Bergmeister papilla. After pars plana vitrectomy to alleviate the traction, the patient's macular appearance improved, and …

Spanish Heritage Speakers And Latinidad: How Spanish Fluency Influences Self-Perceived Latinidad Among Heritage Speakers, David Alejandro Rivera Sandoval, Gladys Valle

Undergraduate Theses and Capstone Projects

Abstract of Research: This study examines 1) how heritage speakers' self-perceived proficiency and fluency in Spanish influence their identity as Latinos, and 2) the generational similarities and differences in their lived experiences. Through surveys and interviews, the relationship between Spanish proficiency and fluency and perceived Latinidad among nine Spanish heritage speakers (with Latin American ancestry) has been assessed. The research suggests that no statistically significant relationship was found between language proficiency or fluency and perceived Latinismo. However, qualitative findings revealed that participants construct their Latino identity based on cultural values, lived experiences, and a sense of belonging rather than language …

Correction: A Comprehensive Narrative Review Of The Challenges Surrounding Cutaneous Scc., Sarah C Beach, Austin S Cusick, Aaron S Farberg, Shannon C Trotter

Ambulatory and Primary Care Articles

No abstract provided.

Peptide Receptor Radionuclide Therapy-Induced Hypercortisolemic Crisis In Ectopic Cushing Syndrome, Raghavendra Pandit, Chethan Yamichannaiah, Anurag Ranjan Lila, Manjiri Karlekar, Saba Samad Memon, Tushar Bandgar

Faculty, Staff and Student Publications

A woman in her fourth decade of life with metastatic ectopic Cushing syndrome due to a thymic neuroendocrine tumor developed an acute adrenocorticotropin (ACTH)-mediated hypercortisolemic crisis 5 days after initiating lutetium-177-labeled DOTA-[Tyr3]-octreotate (177Lu-DOTATATE) peptide receptor radionuclide therapy (PRRT). She presented with hypertension, rapid weight gain (8 kg in a week), psychosis, and biochemical evidence of severe hypercortisolism: serum cortisol: 141 µg/dL (SI: 3889.7 nmol/L) (reference range, 5-25 µg/dL [SI: 137.9-689.7 nmol/L]), plasma ACTH greater than 2000 pg/mL (SI >440.4 pmol/L) (reference range, 10-60 pg/mL [SI: 2.2-13.3 pmol/L]), potassium: 3.4 mEq/L (reference range, 3.5-5.0 mEq/L). Low-dose intravenous etomidate (0.02 mg/kg/h) was …

Evidence-Informed Propositions For A Whole-Of-Program Assessment Approach, Janica Jamieson, Rachel Bacon, Andrea Bramley, Janeane Dart, Simone Gibson, Gemma Jenkins, Meredith Kennedy, Sue Kleve, Claire Palermo

Research outputs 2022 to 2026

Aim: To develop evidence-informed propositions for implementing whole-of-program assessment across dietetic education settings, in response to emerging education theory, technological advances such as generative artificial intelligence, and the need for meaningful learning and trustworthy progression decisions. Methods: A four-stage multiple methods study, grounded in pragmatism, was conducted. Initial propositions were developed through a systematic scoping review of programmatic assessment literature. Thematic analysis was used to identify patterns, which were iteratively refined into themes and then preliminary propositions. These were further refined through workshops with dietetic educators, who provided feedback based on their expertise. A repeated systematic scoping review incorporated recent …

Clinician-Led Development And Feasibility Of A Neural Network For Assessing 3d Dental Cavity Preparations Assisted By Conversational Ai, Mohammed El-Hakim, Haitham Khaled, Amr Fawzy, Robert Anthonappa

Research outputs 2022 to 2026

Introduction: Artificial intelligence is emerging in dental education, but its use in preclinical assessment remains limited. Large language models like ChatGPT^® V4.5 enable non-programmers to build AI models through real-time guidance, addressing the coding barrier. Aim: This study aims to empower clinician-led, low-cost, AI-driven assessment models in preclinical restorative dentistry and to evaluate the technical feasibility of using a neural network to score 3D cavity preparations. Methods: Twenty mandibular molars (tooth 46), each with two carious lesions, were prepared and scored by two expert examiners using a 20-point rubric. The teeth were scanned with a Medit i700^® and …

Testing If Vitamin K1 Reduces The Progression Of Non-Severe Calcific Aortic Stenosis: Design And Rationale Of The Prevention Of Aortic Stenosis Progression Phylloquinone Ossification Reduction Trial (Passport), William A. Courtney, Jarryd Walker, Aindreas Dorai-Raj, Tom Gilbert, Adil Rajwani, Jamie W. Bellinge, Jonathan M. Hodgson, Graham S. Hillis, Carl J. Schultz

Research outputs 2022 to 2026

Aims Calcific aortic stenosis (CAS) is the most common heart valvulopathy in high-income countries. There is no known treatment for CAS other than replacement of the valve in severe, symptomatic disease. Observational studies and a small openlabel randomized trial have reported that vitamin K1 supplementation may reduce the progression rate of calcification and obstruction in CAS. Methods and results 3PASSPORT(ACTRN12622000447752) will be a prospective, randomized, double-blind, placebo-controlled clinical trial investigating if nutritional supplementation with 10 mg of vitamin K1 can reduce the rate of valvular calcification and haemodynamic progression in CAS. Patients identified to have mild or moderate CAS based …

The Role Of Atmospheric Turbulence On Cloud Droplet Distribution And Droplet Impaction On An Airfoil, Arash Shad, Hashnayne Ahmed, ‪Nadim Zgheib, S. Balachandar, S. A. Sherif

Mechanical Engineering Faculty Publications

We conduct turbulence-informed Euler–Lagrange simulations of droplet-laden flow to examine the effects of atmospheric turbulence on droplet impingement characteristics on a NACA 0012 airfoil. We implement statistical overloading, whereby we inject millions of droplets, ranging in diameter from 1 to 160 microns, upstream of the airfoil at an average streamwise velocity of 120 m/s. Within a fraction of a second, these droplets would have either impinged on the airfoil or continued downstream. While the incoming air flow is maintained uniform, droplet injection is informed by ambient turbulence, affecting both the droplet velocity distribution as well as the size-dependent preferential concentration …

More Alike Than Not: The Open Access Preferences Of Humanities Scholars, Rachel E. Scott, Ana Dubnjakovic

Faculty and Staff Publications – Milner Library

Several studies have noted that humanists have not been as quick or enthusiastic in their adoption of Open Access (OA) as their colleagues in other disciplines. This article leverages the Ithaka S+R US 2021 Faculty Survey to provide contextualized analysis of the OA and Open Education Resources (OER) preferences of humanities scholars, as well as some practices related to OA and OER, relative to their colleagues in other disciplines. Findings suggest that although humanists do stand apart in many OA preferences, the small effect sizes render these differences less important than previously suspected. The implications of these findings are considered …

Thalamocortical Structural Covariation Networks Are Related To Familial Risk For Schizophrenia In The Context Of Lower Nuclei Volume Estimates In Patients: An Enigma Study, Annalisa Lella, Linda A. Antonucci, Roberta Passiatore, Loredana Bellantuono, Pierluigi Selvaggi, Teresa Popolizio, Guido Di Sciascio, Alessandro Saponaro, Patrizia Ricci, John Blangero

School of Medicine Publications

Background

Structural brain differences in the thalamus and the cortex have been widely reported in schizophrenia (SCZ) relative to neurotypical control individuals (NCs). Most previous studies examined the thalamusas a whole as a single region of interest. In addition, findings in individuals at familial high risk for SCZ (FHRs) remain inconclusive. Here, we investigated whether local and network-wide thalamic-related structural alterations vary as a function of familial risk for SCZ.

Methods

Structural magnetic resonance imaging scans were obtained from 5197 participants (NC, n = 3409; FHR, n = 257; SCZ, n = 1531) across 32 cross-sectional samples within the …

Recommendations From Multi-Disciplinary Professionals For Survivor-Informed And Comprehensive Care For Human Trafficking Survivors, Christina Tsoi, Timothy B. Smith, Elizabeth Cutrer-Párraga, Devan Clayton, Joshua M. Marshall, Jamila Mastny

Faculty Publications

Background: Human trafficking affects millions of people worldwide with multiple adverse outcomes including psychopathology. Although research on human trafficking has become abundant in other academic disciplines (e.g., public health, criminology, social work), healthcare research specific to the mental health treatment of survivors remains limited. Objective: The purpose of this study was to gather recommendations from professionals about mental health treatment of trafficking survivors. Method: Semi-structured interviews were conducted with 21 multidisciplinary professionals working with trafficking survivors. Data were analyzed using qualitative content analysis methodology. Results: An overarching theme derived from the data concerned the recommendation to implement trafficking survivor-informed care, …

Potentially Toxic Metals, Anthropogenic Microfibres And Other Microplastics In Sediments From A Small, Urban Estuary (Plym, Southwest England), Kwong Ricky Li, Andrew Turner

School of Geography, Earth and Environmental Sciences

Anthropogenic microfibres (AMFs) and microplastics (MPs) are pervasive and ubiquitous environmental contaminants of growing concern. However, their study in geosolids, like sediments, is often conducted in the absence of other important geochemical and anthropogenic markers. In the present study, surface and subsurface sediments from the Plym Estuary, southwest England, have been analysed for AMFs and MPs along with organic matter (as loss on ignition; LOI), calcium, iron, manganese and potentially toxic metals (PTMs). The latter were heterogeneously distributed throughout the estuary because of multiple inputs and variations in sediment granulometry. AMFs were mainly constructed of cellulosic and petroleum-based polymers and …

Image Captioning Through The Lens Of The Gricean Maxims: Generating Meaningful And Relevant Image Descriptions, Annika Lindh

Doctoral

Image captioning models enable us to automatically generate natural language image descriptions for previously unseen images. It combines the two fields of computer vision and natural language generation, allowing models to interpret the con tent of an image and communicate that knowledge through natural language text.

Research into image captioning has the potential benefit of reducing the gap in digital information availability between fully sighted individuals and those who are visually impaired. However, automatically generated captions often fail to provide the required level of detail and specificity to achieve this goal. Furthermore, current standard evaluation methods are insufficient at measuring …

Risk Factors To Predict Mortality In People With Seizures Reattending Emergency Departments: A Dual-Centre Study, Rohit Shankar, Madeline Dale, Sarah Lennard, William Henley, Brendan Mclean, Mark Jadav, Richard Laugharne, Gurjit Chohan, Sarah Trippier, Laura Sewell, Jamie Ford, Sarah Davis

Peninsula Medical School

Background: Recurrent seizure-related emergency department (ED) attendances are common and associated with poor outcomes. We aimed to identify factors associated with repeated seizure-related ED use and to develop a mortality risk score among individuals re-attending with seizures. Methods: We conducted a retrospective cohort study across two district general hospitals in South-West England, Royal Cornwall Hospitals NHS Trust (RCH; 2015–2018) and Royal United Hospitals Bath NHS Trust (RUH; 2018–2022). Adults (≥18 years) with ≥2 seizure-related ED attendances were included. Negative binomial regression was used to examine demographic and clinical predictors of repeat attendances. A risk score based on these predictors was …

Multicomponent One-Dimensional Quantum Droplets Across The Mean-Field Stability Regime, I. A. Englezos, P. Schmelcher, S. I. Mistakidis

Physics Faculty Research & Creative Works

The Lee-Huang-Yang (LHY) energy correction at the edge of the mean-field stability regime is known to give rise to beyond mean-field structures in a wide variety of systems. In this work, we analytically derive the LHY energy for two-, three-and four-component one-dimensional bosonic short-range interacting mixtures across the mean-field stability regime. For varying intercomponent attraction in the two-component setting, quantitative deviations from the original LHY treatment emerge being imprinted in the droplet saturation density and width. On the other hand, for repulsive interactions an unseen early onset of phase-separation occurs for both homonuclear and heteronuclear mixtures. Closed LHY expressions for …

Bending Fatigue In Additively Manufactured Metals: A Review Of Current Research And Future Directions, Md Bahar Uddin, Sriram Praneeth Isanaka, Frank Liou

Mechanical and Aerospace Engineering Faculty Research & Creative Works

Metal additive manufacturing (MAM), also referred to as 3D printing, has proven remarkable in the fabrication of complex metal components in multiple sectors. However, the assessment of this revolutionary process through bending fatigue is frequently impeded due to concerns about mechanical and physical conditions of the printed components. The unique layer-by-layer production process results in varied microstructures, anisotropy, and intrinsic defects that considerably differ from traditionally manufactured wrought metals. This review article aims to integrate and evaluate historical and contemporary research on the bending fatigue of additively manufactured materials. More specifically, the impact of process parameters, build orientation, surface conditions, …

Generative Ai And Empirical Software Engineering: A Paradigm Shift, Christoph Treude, Margaret-Anne Storey

Research Collection School Of Computing and Information Systems

The widespread adoption of generative AI in software engineering marks a paradigm shift, offering new opportunities to design and utilize software engineering tools while influencing both developers and the artifacts they create. Traditional empirical methods in software engineering, including quantitative, qualitative, and mixed-method approaches, are well established. However, this paradigm shift introduces novel data types and redefines many concepts in the software engineering process. The roles of developers, users, agents, and researchers increasingly overlap, blurring the distinctions between these social and technical actors within the field. This paper examines how integrating AI into software engineering challenges traditional research paradigms. It …

Defects4c: Benchmarking Large Language Model Repair Capability With C/C++ Bugs, Jian Wang, Xiaofei Xie, Qiang Hu, Shangqing Liu, Jiongchi Yu, Jiaolong Kong, Yi Li

Research Collection School Of Computing and Information Systems

Automated Program Repair (APR) plays a critical role in enhancing the quality and reliability of software systems. While substantial progress has been made in Java-based APR, largely facilitated by benchmarks like Defects4J, there remains a significant gap in research on C/C++ program repair, despite the widespread use of C/C++ and the prevalence of associated vulnerabilities. This gap is primarily due to the lack of high-quality, open-source benchmarks tailored for C/C++. To address this issue, we introduce Defects4C, a comprehensive and executable benchmark specifically designed for C/C++ program repair. Our dataset is constructed from real-world C/C++ repositories and includes a large …

Mmlu-Prox: A Multilingual Benchmark For Advanced Large Language Model Evaluation, Weihao Xuan, Et. Al.

Research Collection School Of Computing and Information Systems

Existing large language model (LLM) evaluation benchmarks primarily focus on English, while current multilingual tasks lack parallel questions that specifically assess cross-lingual reasoning abilities. This dual limitation makes it challenging to assess LLMs’ performance in the multilingual setting comprehensively. To fill this gap, we introduce MMLU-ProX, a comprehensive benchmark covering 29 languages, built on an English benchmark. Each language version consists of 11,829 identical questions, enabling direct cross-lingual comparisons. Additionally, to meet efficient evaluation needs, we provide a lite version containing 658 questions per language. To ensure the high quality of MMLU-ProX, we employ a rigorous development process that involves …