http://www.rssboard.org/rss-specification 720 https://escholarship.org/uc/uci_etd/rss Recent eScholarship items from UC Irvine Electronic Theses and Dissertations Fri, 15 May 2026 04:53:11 +0000 https://escholarship.org/uc/item/16w5f4kk This dissertation consists of three empirical essays on macroeconomics and monetary economics. In the first chapter, I study how and why financial market reactions to macroeconomic news (e.g., payrolls, CPI) change over time. Focusing on interest rate responses, I document that market sensitivity increases during periods of high inflation, rising unemployment, and heightened Federal Reserve emphasis on macroeconomic conditions. To uncover the fundamental drivers of this variation, I distinguish between two hypotheses—changes in investor attention and changes in the perceived data dependence of the monetary policy rule—using an empirical strategy based on their differing implications for the transmission of monetary policy. Consistent with the attention hypothesis, I find that the impact of FOMC announcements is amplified in periods when market sensitivity to macroeconomic announcements is high. A standard New Keynesian model with cognitive discounting is used to illustrate the... https://escholarship.org/uc/item/16w5f4kk Thu, 14 May 2026 00:00:00 +0000 Wang, Xueliang https://escholarship.org/uc/item/35j567q6 Coastal ecosystems are increasingly exposed to interacting climate stressors, but predicting responses to those stressors remains challenging because ecological processes are complex and operate across multiple biological and spatial scales. While many studies focus on individual stressors, single levels of organization, or populations at a single location, fewer integrate organismal responses, community interactions, and ecosystem functioning. I conducted this research to provide insights into how the impacts of climate stressors extend across scales to affect nutrient cycling and thermal tolerance in intertidal coastal systems.Here I combined field manipulations, laboratory assays and trait-based analytical approaches to examine the effects of CO2, thermal, and desiccation stress on tidepool communities and intertidal macroalgae. First, I experimentally manipulated temperature and CO2 levels in natural tidepools to quantify changes in nitrate (NO₃⁻) and ammonium (NH₄⁺) fluxes... https://escholarship.org/uc/item/35j567q6 Wed, 13 May 2026 00:00:00 +0000 Rios Rojas, Beatriz Alejandra Alejandra https://escholarship.org/uc/item/55h3j20g This dissertation contains three chapters which study questions in public and labor economics relating to public assistance. Each of the three chapters make a connection to the Supplemental Nutrition Assistance Program (SNAP) from a distinct perspective and offer possible insights into public assistance more broadly.The first chapter examines the employment and program participation effects of SNAP work requirements. Work requirements to receive public assistance are widely implemented and debated policies. Despite this fact, there is not yet a consensus on how work requirements affect employment outcomes. I use administrative data from a state’s SNAP and Unemployment Insurance (UI) records to study this question. Exploiting the reintroduction of work requirements and the fact that only certain adults are subject to them, I find that work requirements reduce participation in SNAP and increase employment. I also find evidence that the work requirements have stronger effects on... https://escholarship.org/uc/item/55h3j20g Fri, 8 May 2026 00:00:00 +0000 De La Torre, Gamaliel https://escholarship.org/uc/item/3f60f3nf This thesis studies the asymptotic distribution of sandpile groups of random bipartite graphs in both the directed and undirected settings. Fix a prime p and let the two parts of the bipartition have sizes n and ⌈αn⌉, where 1 p < α ≤ 1. For directed Erdős-Rényi bipartite graphs, this thesis proves that the p-Sylow subgroup of the sandpile group converges to the conjectured Cohen-Lenstra distribution. For undirected Erdős-Rényi bipartite graphs, it proves the corresponding conjecture for odd primes p, where the conjectured limit is the symmetric Cohen-Lenstra distribution. In the undirected case with p = 2, the limiting distribution has been conjectured to be a variant of the symmetric Cohen-Lenstra distribution. Our results provide strong evidence for this conjecture.For a fixed finite abelian p-group G, the surjective moment of a random finite abelian pgroup X is the expected number of surjective homomorphisms from X to G. The surjective moment method seeks to identify a limiting... https://escholarship.org/uc/item/3f60f3nf Fri, 8 May 2026 00:00:00 +0000 Singhal, Deepesh https://escholarship.org/uc/item/4661t5js Plants and herbivores have been imposing reciprocal selection on one another for the past 350 million years, resulting in an incredible diversity of physical and chemical defensive traits that vary in type and magnitude, and within and across scales. Predicting patterns of plant defense is, however, challenging, due to the involvement of biotic and abiotic factors that interact in complex ways to shape plant-herbivore interactions. From the bottom-up, resource availability constrains investment in plant resistance and tolerance. From the top-down, predators of herbivores can shape plant defense through trophic cascades and by imposing selection on plant traits that mediate their effect. Unsurprisingly, there are still mechanisms and interactions within plant defenses that are not yet fully understood or characterized, which further complicates explaining variation and patterns. What remains as the forefront of this field is how interacting biotic and abiotic drivers independently... https://escholarship.org/uc/item/4661t5js Sat, 25 Apr 2026 00:00:00 +0000 Dean, Lydia S. https://escholarship.org/uc/item/09v1h7hk This dissertation argues that a key ingredient in the wave of ethnonationalism appearing across the world is the rhetorical weaponization of historical trauma. It analyzes two states, Croatia and India, extremely different from one another other than that they share a parallel traumatic history of having emerged from violent partition events. This work argues that the painful memory of these partitions lingered for generations as a political sore spot that remained unresolved. Contemporary ethnonationalism has emerged in both states via an eerily similar playbook, invoking that traumatic legacy and promising to provide hope, agency and redemptive retribution via a commitment to an exclusionary nationalism. Most shockingly, this wave of nationalism has been extremely effective at undermining established democratic orders. This work finds that it is this manipulation of trauma that has been foundational to the success and consolidation of these paticularist challenges to established... https://escholarship.org/uc/item/09v1h7hk Sat, 25 Apr 2026 00:00:00 +0000 Parikh, Kiyaan Samsamy https://escholarship.org/uc/item/6ns2r61x This dissertation consists of three essays that examine how information and institutions shape economic outcomes across urban development, public policy, and financial markets, using data-driven and quantitative methods. The first chapter studies the evolution of residential skyscraper development in Hong Kong. Using machine learning methods to address a high-dimensional, small-sample setting, the analysis identifies growth momentum as the primary driver of skyscraper construction. In contrast to traditional theories emphasizing economic fundamentals or business cycles, the findings highlight the role of strategic interaction among major developers, suggesting that oligopolistic competition generates a self-reinforcing dynamic in urban development. The second chapter examines the role of fiscal decentralization (FD) in managing large-scale health crises, focusing on the COVID-19 pandemic across OECD countries. The results reveal a nuanced relationship. During the early... https://escholarship.org/uc/item/6ns2r61x Thu, 23 Apr 2026 00:00:00 +0000 Hu, Jiangdong https://escholarship.org/uc/item/9xg0v169 Recovering the three-dimensional structure of objects, scenes, and people from images requires reasoning beyond what is directly visible---behind occluding surfaces, outside the field of view, and across time. This thesis develops methods for reconstructing shapes and motion across these domains, from static objects and scenes to human pose in dynamic settings. A recurring theme is fusing complementary representations: coordinate frame choices, perceptual and generative approaches, egocentric and exocentric perspectives. We begin by studying how coordinate systems and viewpoint choices shape learned representations for object shape prediction and pose estimation. We then develop multi-layer depth representations as a viewer-centered alternative to detection-based scene reconstruction. Combining viewpoint prediction with pose estimation produces representations that transfer across datasets. Building on these foundations, we integrate visual tracking with generative motion modeling... https://escholarship.org/uc/item/9xg0v169 Tue, 21 Apr 2026 00:00:00 +0000 Shin, Daeyun https://escholarship.org/uc/item/9s85f7v5 The project was motivated by the insurmountable gap between the way that sociology, and the social sciences more broadly, conceptualize slavery, and how a tradition of black thinkers conceptualize slavery. The discipline of sociology, particularly its United States institutional variant, cannot categorize slavery as a form of war - yet there is a rich archive of black thought that does conceptualize slavery as a form of war. Recently, black critical theorists have argued that racial slavery structures the entirety of what we call modernity. Analyzing the work of W.E.B. Du Bois, a belatedly canonized thinker in the discipline, can open up this gap between sociology and black thought. My first research question is: How does Du Bois conceptualize the relationship between war, blackness, and slavery? My second research question: Does Du Bois understand slavery as a form of permanent war that structures modernity? The methods in this project included content analysis, archival work,... https://escholarship.org/uc/item/9s85f7v5 Tue, 21 Apr 2026 00:00:00 +0000 Boko, Semassa https://escholarship.org/uc/item/1316m5p8 Generative modeling provides a principled way to learn underlying data distributions and has become a key tool for modern biological discovery. In genomics and single-cell omics, where measurements are heterogeneous, sparse, and shaped by stochastic regulatory processes, generative approaches enable representation learning from unsupervised or weakly supervised data, probabilistic integration across modalities, and counterfactual reasoning about how perturbations may change cellular programs. This dissertation develops and studies generative models for single-cell and genomic data, with an emphasis on multimodality and actionable biological interpretation. Chapter 2 introduces a weakly supervised cross-modality generative framework that aligns and translates between single-cell RNA-seq and ATAC-seq, enabling joint modeling when paired multiome measurements are limited. Chapter 3 evaluates whether contemporary instruction-tuned large language models (LLMs) can support researchers... https://escholarship.org/uc/item/1316m5p8 Mon, 20 Apr 2026 00:00:00 +0000 Liu, Junhao https://escholarship.org/uc/item/3fn9s8z0 Hypertension rates continue to rise across the United States, with Hispanic/Latino communities often having lower rates of control. In Hispanic/Latino communities, the literature has increasingly investigated the role of sociocultural influences, such as health-related social support (HRSS) and familism, in health behaviors and outcomes. This study thus aims to expand the literature by examining the association of medication non-adherence and blood pressure (BP) outcomes with health-related social control (HRSC), HRSS, and familism. This was a cross-sectional mixed-methods study with data collected from Hispanic/Latino adults with hypertension as part of the Mi Propio Camino study. The first phase of this study examined the association of HRSC (represented as persuasion and pressure), HRSS, and familism with medication non-adherence via separate logistic regressions and individual BP outcomes (i.e., systolic and diastolic BP) via linear regressions. Phase Two of this study... https://escholarship.org/uc/item/3fn9s8z0 Thu, 2 Apr 2026 00:00:00 +0000 Arizmendi De La Torre, Jessica https://escholarship.org/uc/item/1344b3ps Introduction: Over recent decades, wildfires have increased in size, frequency, and proximity to populated areas, accounting for up to 50% of PM₂.₅ in the Western United States (Burke et al., 2021; Congressional Research Service, 2023; Ghetu et al., 2022; National Interagency Fire Center, 2021; Orr et al., 2025). Wildfire exposure has been linked to adverse health outcomes such as cardiovascular mortality, asthma exacerbations, and negative birth outcomes (Feng et al., 2016; Orr et al., 2025; Sparks & Wagner, 2021). This study examines the relationship between wildfire-specific PM₂.₅ exposure during pregnancy and large-for-gestational-age (LGA) and small-for-gestational-age (SGA) birth outcomes. We hypothesized that maternal exposure to wildfire-specific PM₂.₅ is associated with risks of LGA and SGA, and that the timing of exposure during pregnancy influences these outcomes. Methods: This retrospective cohort study analyzed 408,671 singleton live births recorded in Kaiser... https://escholarship.org/uc/item/1344b3ps Thu, 2 Apr 2026 00:00:00 +0000 Thomas, Ashley https://escholarship.org/uc/item/6154t66g While the United States is often known as a pro-gun country, most Americans do not own guns. More importantly, gun ownership status is not immutable, as some gun owners even decide to voluntarily relinquish their guns today. However, research on why and how people may relinquish their guns remains scarce. In response, this pilot study inductively investigated the pathways to voluntary gun relinquishment through in-depth interviews with three former gun owners. Preliminary findings showed that former gun owners relinquished their guns after witnessing firsthand the destructiveness of gun-related harms and developing the personalized fear of being responsible for the harms against themselves and their loved ones. These findings not only remain consistent with the existing literature on the opposition to gun ownership but also demonstrate the significance of recognizing voluntary gun relinquishment as a salient topic in gun research. Therefore, future research should speak with more... https://escholarship.org/uc/item/6154t66g Wed, 1 Apr 2026 00:00:00 +0000 Lee, Mary https://escholarship.org/uc/item/5hc6z9rp Autonomous Driving Systems (ADSes) are rapidly advancing and being deployed in safety-critical settings, making it essential to systematically identify safety violations before deployment. DoppelTest is a state-of-the-art approach for testing ADSes that utilizes efficient, cloned instances of the ADS itself to control vehicles in simulation, thereby generating highly interactive, bug-revealing scenarios without relying on simplified behavioral models. However, as DoppelTest was originally implemented specifically for Baidu Apollo, which we refer to as DoppelTest4Apollo, its applicability to other ADS architectures remains untested. In this thesis, I design, implement, and evaluate DoppelTest for Autoware (DoppelTest4AW), an open-source Level-4 ADS stack based on ROS 2, to assess DoppelTest’s transferability and effectiveness across different ADS platforms. I first identify and address the challenges of porting DoppelTest to Autoware, specifically the semantic differences in map... https://escholarship.org/uc/item/5hc6z9rp Wed, 1 Apr 2026 00:00:00 +0000 Liao, Xiang https://escholarship.org/uc/item/08f9j8sr DNA polymerases are central to biology and biotechnology due to their ability to faithfully propagate genetic information. Expanding the substrate scope of these enzymes beyond natural nucleic acids represents a major challenge in enzyme engineering and is essential to the practical use of new classes of genetic polymers. This thesis describes the directed evolution, technological development, and structural analysis of polymerases capable of synthesizing α-L-threofuranosyl nucleic acid (TNA), a synthetic genetic polymer with a noncanonical sugar backbone that is resistant to nucleases and acid-mediated degradation.To overcome the inherent incompatibility between natural polymerases and TNA substrates, an integrated evolutionary strategy was employed. Starting from a homologous recombination library, iterative rounds of selection and random mutagenesis were used to traverse the polymerase fitness landscape and identify mutation combinations that collectively enable efficient TNA... https://escholarship.org/uc/item/08f9j8sr Wed, 1 Apr 2026 00:00:00 +0000 Gross, Victoria https://escholarship.org/uc/item/1jj2v1ct Ice radio neutrino detectors are designed to record radio signals produced by neutrino interactions in large ice sheets, which produce Askaryan radio that may travel kilometers before detection. Previous and current detectors like ARIANNA, ARA, and RNO-G are have piloted and tested this experimental technique and shown the feasibility of detecting neutrino interactions. But due to the low neutrino flux at very high energies above  1018 eV, detection of neutrinos will require a much larger array than has currently been built. The planned IceCube-Gen2 radio array is such a planned experiment. It will utilize over 360 autonomous stations, 10x more than the next largest array, to probe the predicted cosmogenic neutrino flux. But as sensitivities increase to detect the extremely rare neutrino events, very rare background sources must also be contended with.Ice transmits radio, but as any medium may it can also refract and reflect said radio. This is already well understood at... https://escholarship.org/uc/item/1jj2v1ct Mon, 30 Mar 2026 00:00:00 +0000 Rice-Smith, Ryan https://escholarship.org/uc/item/5bg0n84t Turbulent combustion modeling within Reynolds-Averaged Navier-Stokes (RANS) and Large Eddy Simulations (LES) is a critical aspect of aerospace design, particularly for jet and rocket combustors and ground-based gas turbines. One common approach for modeling turbulent combustion is the flamelet model. While there has been substantial improvement to flamelet theory over its near 50-year history, there are still areas where models fail to include and predict critical physics. This work aims to address two of these areas. First, among other assumptions, existing flamelet models do not incorporate the three-dimensional effects of vortex stretching on the sub-grid fluid-chemical structure through a solution to the momentum equations. Including three-dimensional vortex stretching in the flamelet model produces a centrifugal effect which substantially increases the flammability limit predicted for the reacting mixture. This effect is explored in detail and incorporated into a new flamelet... https://escholarship.org/uc/item/5bg0n84t Sat, 28 Mar 2026 00:00:00 +0000 Hellwig, Wes https://escholarship.org/uc/item/7zc5r10d Long-distance commuting (LDC), defined as one-way work trips ≥50 miles, imposes social, health, economic, and climate burdens that are oftentimes unevenly distributed. Previous studies have focused on individual-level determinants of LDC behavior with less consideration given to population-level characteristics and neighboring influences, which need to be understood to inform regional-level solutions. This study explores associations at a population-level, both from a generation and an attraction perspective, between socio-demographic, household compositional, employment, and urban form factors and LDC behavior across four California regions (Los Angeles, San Francisco–Sacramento, San Diego, and Fresno). The shift in data aggregation, from individual to population-level, necessitated an examination of transformation techniques, including logit and additive log-ratio transformations, to ensure that explanatory variables met the assumptions of linear regression. Further, best practices... https://escholarship.org/uc/item/7zc5r10d Fri, 27 Mar 2026 00:00:00 +0000 Reinoehl, Montana Kyle Rodriguez https://escholarship.org/uc/item/0b53557g Introduction: Occupational stressors are workplace conditions that provoke harmful emotional and physical responses and, when chronic, contribute to anxiety and depression. These stressors are embedded within work organization through policies, procedures, and environmental arrangements. One increasingly prevalent yet understudied organizational arrangement is working from home (WFH). While WFH may reduce certain stressors, it may also introduce new risks for burnout and poor mental health. Evidence on WFH and mental health is mixed, particularly when comparing the COVID-19 and post-COVID-19 periods. As mental health morbidity continues to rise in the United States, it is critical to understand how WFH influences mental health and how its effects vary across periods of societal disruption. This study integrates established occupational stress frameworks to examine the association between WFH and severely poor mental health among U.S. workers during and after the COVID-19 pandemic.... https://escholarship.org/uc/item/0b53557g Fri, 27 Mar 2026 00:00:00 +0000 Borom, Marielle P https://escholarship.org/uc/item/9wb1q5zv Women with germline pathogenic variants in BRCA1 are predisposed to developing aggressive, early onset breast cancers, most commonly of the triple negative subtype. While defective DNA repair has long been considered the primary driver of BRCA1-associated tumorigenesis, accumulating evidence suggests that additional non-cell autonomous mechanisms contribute to disease initiation. In particular, BRCA1-mutant mammary glands exhibit pronounced epithelial hyperplasia and aberrant branching prior to tumor formation, raising the possibility that interactions between epithelial cells and the immune microenvironment play a causal role in early disease progression. In this dissertation, I investigate how immune cells specifically macrophages regulate premalignant remodeling in BRCA1-mutant mammary tissue. Using a genetically engineered mouse model of BRCA1 and p53 deficiency, I combine single-cell RNA-sequencing, spatial proteomic imaging, flow cytometry, and functional organoid... https://escholarship.org/uc/item/9wb1q5zv Thu, 26 Mar 2026 00:00:00 +0000 Prem Antony Samy, Angela Lincy https://escholarship.org/uc/item/9gs7w4xr Despite the general consensus that teachers’ learning of justice-centered instruction necessitates well-guided professional learning experiences across an extensive period of time, extending beyond one-shot professional development (PD), there are currently very few empirical studies that provide insight into teachers’ learning processes. This longitudinal qualitative case study explored the collective ideological sense-making processes of high school science teachers as they engaged in a cycle of co-designing, enacting, revising, and re-enacting an equity-centered unit over 2 years. The context was a long-term research-practice partnership project between a university and a progressive local school district that was committed to promoting equity, justice, and civic engagement. The participants were six high school biology teachers, all members of a co-design team, on one of the five teams in the larger project. The data consisted of 95.3 hours of video-recorded small-group team... https://escholarship.org/uc/item/9gs7w4xr Thu, 26 Mar 2026 00:00:00 +0000 Skoropad, Stephen R https://escholarship.org/uc/item/19x9h837 Modern deep neural networks are highly overparameterized and possess substantial structural redundancy, with many parameter configurations representing the same input–output function. While this redundancy enables expressive capacity and practical trainability, it also distorts optimization geometry, contributing to unstable training dynamics, and unintended behaviors such as memorization of training data. This work develops a structural perspective based on symmetry, scale invariance, and function-preserving reparameterizations to analyze how these properties shape optimization and information retention. Neural balancing is introduced as an explicit form of structural control that aligns interacting components through simple, function-preserving rescalings, with synaptic, neuronal, and channel-level methods demonstrating faster convergence, improved stability, and consistent gains across architectures and limited-data regimes. The same viewpoint is extended to memorization in... https://escholarship.org/uc/item/19x9h837 Thu, 26 Mar 2026 00:00:00 +0000 domingo, ian https://escholarship.org/uc/item/4dm2t083 Mobile Edge Computing (MEC) systems must dynamically balance inference accuracy, latency,and energy consumption under fluctuating wireless conditions and heterogeneous device constraints. In semantic Radio Access Network (RAN) slicing, this challenge is exacerbated by combinatorial configuration spaces, partial observability, and non-stationary channel behavior. Traditional optimization methods and purely correlation-driven contextual bandit approaches struggle to scale efficiently while maintaining Quality-of-Service (QoS) guarantees.This thesis proposes a causal-enhanced OMNIS framework that integrates Structural Causal Modeling (SCM) with contextual multi-armed bandit learning for adaptive semantic RAN slicing. By encoding explicit cause–effect relationships between wireless channel state (SNR), bit error rate, neural branch selection, latency, energy consumption, and inference accuracy, the proposed system performs counterfactual reasoning prior to action execution. This enables... https://escholarship.org/uc/item/4dm2t083 Wed, 25 Mar 2026 00:00:00 +0000 Samiullah, Omar https://escholarship.org/uc/item/0mt228d0 This thesis investigates the impact of power amplifier (PA) output power level on the bit error rate (BER) performance of a wireless communication system employing 4M-quadrature amplitude modulation (QAM) scheme. In practical wireless systems, PA output power plays a critical role in determining system performance due to the combined effects of noise and non-linear distortion. Insufficient transmitter output power reduces the received signal-to-noise ratio (SNR), leading to increased BER, while excessive transmitter output power requirement drives the PA into its nonlinear operating region, introducing distortion that also degrades BER performance. To capture this trade-off, a quantitative analytical framework is developed to characterize the combined influence of system noise and PA nonlinearity on BER. Based on this analysis, a closed-form expression is derived to determine the globally optimal PA output power that minimizes BER for a given wireless link. The analytical predictions... https://escholarship.org/uc/item/0mt228d0 Wed, 25 Mar 2026 00:00:00 +0000 Chae, Seungwoo https://escholarship.org/uc/item/9x0955k0 The eye lens is predominantly made up of proteins known as crystallins. βγ-Crystallins are the structural proteins that give the eye lens its refractivity, allowing the lens to focus light and the stability meant to last a lifetime. Modifications to the residues in crystallins caused by ageing or hereditary mutations can cause opacification of the lens, known as cataract. In this thesis, refractive index increment (dn/dc) and fluorescence emission are investigated with changes in the amino acid residues γS crystallin in variants compared to wild type. The dn/dc changes are seen with respect to changes in tryptophan residues. Tryptophan is highly polarizable, by modifying the tryptophan by either replacement or modification, changes are made to the local environment changing potential anisotropic polarizability. Tryptophan variants also change the fluorescence emission and quantum yield.This thesis also looks at implementation of AI-assisted grading and the impact of the feedback... https://escholarship.org/uc/item/9x0955k0 Tue, 24 Mar 2026 00:00:00 +0000 Abiad, Amanda https://escholarship.org/uc/item/2bq1c3np Many biomedical, environmental, and sociological studies have one or more covariates that are subject to limit of detection (LOD) which are often left-censored measurements. Handling LOD covariates becomes more difficult in longitudinal settings, such as repeated measures, where a subject is measured repeatedly over time. Traditional approaches to deal with censored covariates include complete case analysis and substitution methods. Although complete case analysis is a valid approach, loss of observations leads to inefficiency. Substitution methods are simple, but may introduce biases.This dissertation proposes a semiparametric framework for addressing estimation methods in the presence of LOD covariates and in a longitudinal setting. We utilize cubic B-splines for flexible estimation of the log of the hazard, Beran-type estimator for the bivariate case, and adapting these approaches to a repeated measures setting. In the presence of a single LOD covariate, we fit the log of the... https://escholarship.org/uc/item/2bq1c3np Sat, 21 Mar 2026 00:00:00 +0000 Kwon, Jimmy https://escholarship.org/uc/item/8t21d5ps This dissertation is centered around exploring the reactivity of redox-active metals bearing redox-active ligands. Also discussed is the use of a tetradentate ligand scaffold to investigate the influences of coordination geometry and ancillary ligands on the spectral and electrochemical properties of metal-complexes bearing multiple redox-active ligands.Chapter 2 describes the synthesis of a new tetradentate, N2P2, redox-active ligand featuring a phenylenediamine backbone and flexible arms bearing diphenylphosphino moieties. The ligand was successfully bound to a ruthenium (II) metal-center and subsequently demonstrated the oxidative addition of iodine. Reduction of this diiodide species in the present of select ancillary ligands led to the isolation of a series of fully reduced complexes. By comparing solid-state bond metrics and the chemical shifts of select NMR resonances, clear trends emerge that serve to benchmark and implicate ligand-localized electrons in these redox-reactions.... https://escholarship.org/uc/item/8t21d5ps Fri, 20 Mar 2026 00:00:00 +0000 Stofan, Jake Andrew https://escholarship.org/uc/item/84b6s6p3 The extracellular matrix (ECM) is a dynamic network essential for structural support and cellular regulation, with alterations in its properties linked to diseases such as cancer, fibrosis, and vascular disorders. Fibrin, a naturally derived ECM component, exhibits strain stiffening and viscoelasticity, making it a valuable scaffold for studying cell-ECM interactions and mechanical force propagation. This dissertation presents three complementary studies that together establish a non-invasive toolkit for longitudinal imaging and mechanical characterization of fibrillar ECM scaffolds.In collaboration with Dr. Peter Chang, a deep-learning model was trained to predict fluorescently labeled fibers from label-free reflection confocal and transmission images, eliminating the need for exogenous dyes and enabling repeated imaging without compromising scaffold integrity. This virtual labeling framework was then extended to reconstruct nuclei, actin cytoskeleton, and vascular channels from... https://escholarship.org/uc/item/84b6s6p3 Fri, 20 Mar 2026 00:00:00 +0000 Albassri, Sarah https://escholarship.org/uc/item/7b17w7kv Liquid-phase transmission electron microscopy (LP-TEM) enables real-time imaging of nanoscale processes in liquids by sealing a thin liquid layer between electron-transparent membranes; in this setting, the imaging beam also deposits energy into the liquid and drives radiolysis, tightly coupling image quality to chemistry. This dissertation develops complementary experimental, computational, and instrumentation workflows to translate LP-TEM observations into mechanistic understanding and scalable electron-beam polymerization.Electrochemical liquid electron microscopy is applied to an electrically fueled active molecular system in which a cysteine-based precursor is oxidized to a gelator that self-assembles into ~20–22 nm fibers. A low-dose protocol (≤1 e⁻ Å⁻² s⁻¹) separates beam-induced products from electrochemically driven assembly, while segmentation and DSSIM analytics reveal two growth fronts—one advancing at approximately constant velocity and a second that accelerates with... https://escholarship.org/uc/item/7b17w7kv Fri, 20 Mar 2026 00:00:00 +0000 Gibson, Wyeth https://escholarship.org/uc/item/5zv791bs Alpha herpesviruses (αHVs), including human pathogens, herpes simplex virus types 1 and 2 (HSV-1, HSV-2), varicella-zoster virus (VZV), as well as the animal pathogen, pseudorabies virus (PRV), initially infect mucosal epithelial cells. Subsequently, they invade peripheral neurons, establishing lifelong latent infections in the peripheral nervous system (PNS). While mucosal epithelia antiviral defenses are well characterized, the mechanisms governing neuronal antiviral responses remain poorly understood. Due to their postmitotic status and cellular polarity, neurons depend on antiviral responses that do not provoke inflammatory damage. Despite this constraint, neurons are exposed to a complex cytokine milieu produced by infected mucosal epithelia. Type III interferons (e.g IFN lambda; IFN-λ) are among the most abundant cytokines secreted by infected mucosa. This dissertation research elucidates mechanisms of how neurons respond to IFN-λ signaling, identifies key neuronal antiviral... https://escholarship.org/uc/item/5zv791bs Fri, 20 Mar 2026 00:00:00 +0000 Salazar, Stephanie https://escholarship.org/uc/item/4sf9d4hq Real-world robotic systems must make reliable decisions under structured, multimodal, and dynamically evolving uncertainty. To operate effectively, these systems require a capacity to both quantify their own latent doubt and act strategically to reduce it—a dual requirement that current methods struggle to meet. This dissertation develops a scalable and active probabilistic inference framework grounded in Bayesian principles to enable trustworthy autonomy in complex, unstructured environments. The proposed framework integrates parametric recursive inference, nonparametric trajectory-level estimation, active belief-space decision-making, and human–robot collaborative reasoning within a unified probabilistic foundation. The work first establishes a parametric recursive inference framework that unifies probabilistic filtering with optimization-based posterior approximation, enabling structured and computationally efficient uncertainty propagation under nonlinearities. It then... https://escholarship.org/uc/item/4sf9d4hq Fri, 20 Mar 2026 00:00:00 +0000 Seo, Minwon https://escholarship.org/uc/item/4bq9w2xd This dissertation examines the role of the Holy Spirit in shaping knowledge, labor, gender formation, and survival within a Black Baptist congregation in Los Angeles. Based on ethnographic fieldwork, it argues that the Holy Spirit operates not only as theological belief but also as a social actor and an epistemological framework. I develop the concept of the epistemology of spirit to describe how congregants understand divine presence as a source of knowledge that challenges dominant Western assumptions about rationality, agency, and the boundaries of the human as a racialized and gendered subject.Situating Family of Faith within Los Angeles as a carceral city structured by antiblack policing and neoliberal redevelopment, this dissertation explores how this congregation moves beyond respectability politics to articulate spiritual frameworks for navigating structural vulnerability. It examines how churchwomen shape digital worship and archive through gendered labor in the era of... https://escholarship.org/uc/item/4bq9w2xd Fri, 20 Mar 2026 00:00:00 +0000 Azzara, Monique https://escholarship.org/uc/item/92w8f1hq This dissertation develops self-contained localization and navigation for dismounted agents operating in indoor, GPS-denied settings, with a particular focus on firefighting and emergency response. The work addresses four tightly coupled challenges: accurate personnel localization using foot-mounted inertial sensors, alignment of locally estimated trajectories to globally interpretable reference frames, navigation and guidance in both known and unknown indoor environments, and localization of distant objects of interest. First, this dissertation advances foot-mounted inertial navigation by addressing limitations of conventional Zero-Velocity Update (ZUPT) methods. An activity-dependent adaptive ZUPT framework was introduced, using a time-series machine-learning model to detect stance phases across diverse pedestrian motions and subjects. In addition, the idealized assumption of zero foot velocity during stance was revisited, and a velocity estimation model was proposed to... https://escholarship.org/uc/item/92w8f1hq Thu, 19 Mar 2026 00:00:00 +0000 Sangenis Rafart, Eudald https://escholarship.org/uc/item/6wq9q7j0 Proteins rely on precise structural features to maintain stability and function, particularly in long‑lived systems such as the vertebrate eye lens. This dissertation brings together methodological development, biophysical characterization, and research‑based pedagogy to investigate how protein structure can be probed, interpreted, and taught.The first component of this work focuses on developing and formalizing experimental and analytical workflows for protein biophysics, culminating in a methods‑oriented study modeled in the style of a Current Protocols article. This work integrates circular dichroism spectroscopy, differential scanning fluorimetry, dynamic light scattering, and complementary structural tools into a coherent framework for assessing protein stability, folding, and aggregation. By standardizing these approaches, I established a practical toolkit that supports both my subsequent crystallin studies and future investigations of protein variants in the laboratory.The... https://escholarship.org/uc/item/6wq9q7j0 Thu, 19 Mar 2026 00:00:00 +0000 Crosby, Marquise GeVon https://escholarship.org/uc/item/62v042hs Extensive use of per- and polyfluoroalkyl substances (PFAS) in industrial and consumer products has led to widespread environmental contamination and growing risks to water resources. Mitigating human health impacts from PFAS exposure requires reliable predictions of their environmental mass loading, transport, and fate, which are critical in populated areas but challenging due to their numerous sources and complex interactions with soil, water, and air-water interfaces. Wastewater treatment plants are major conduits between numerous waste generators and the environment. Once released, floodplains act as critical interfaces between surface water and groundwater. This dissertation addresses key knowledge gaps in understanding these important channels for PFAS migration in urban settings by (1) characterizing patterns of historical release of PFAS from global wastewater treatment plants (WWTPs), (2) uncovering contamination and hysteretic adsorption mechanisms of PFAS in sediments... https://escholarship.org/uc/item/62v042hs Thu, 19 Mar 2026 00:00:00 +0000 Cookson, Esther https://escholarship.org/uc/item/61w0f45s Cell-cell communication (CCC) is essential for coordinating cellular behaviors during development, tissue remodeling, and homeostasis in multicellular organisms. Morphogens and growth factors can be delivered over short distances through paracrine signaling, or systemically to act over long distances, to regulate critical biological processes including growth, immune responses and survival. Despite extensive studies of these signaling mechanisms, a key question remains: how do cells deliver signals with spatial precision and ensure communication with their correct target cells within complex tissue environments? Emerging evidence suggests that CCC mediated by long, thin cellular protrusions provide an effective mechanism for spatially precise signal transmission via direct contact between cells. Two classes of these cellular protrusions are present in the zebrafish skin, termed airinemes and cytonemes, both of which mediate Delta-Notch signaling and play vital roles during postembryonic... https://escholarship.org/uc/item/61w0f45s Thu, 19 Mar 2026 00:00:00 +0000 Wang, Yi https://escholarship.org/uc/item/50c7107r Reinforcement Learning (RL) is a field of dynamic control that leverages Machine Learning (ML) and has been employed in diverse areas such as robotics, games, navigation, and industrial automation. Currently, many state-of-the-art RL algorithms rely on deep neural networks. However, deep learning faces significant challenges: high computational demands due to costly backpropagation and gradient-based methods, extreme sensitivity to noise in data, networks, and underlying hardware, and a lack of human-like cognitive support for long-term memorization and transparency. To address these limitations, this dissertation turns to brain-inspired learning paradigms, specifically HyperDimensional Computing (HDC), which is gaining traction for its high efficiency, robust operation, and ability to be implemented on lightweight hardware. We develop a suite of HDC-based RL algorithms that demonstrate the viability of this approach. Our contributions include an HDC-based Q-learning algorithm... https://escholarship.org/uc/item/50c7107r Thu, 19 Mar 2026 00:00:00 +0000 Issa, Mariam Ali https://escholarship.org/uc/item/4pd5p4ph First, this dissertation discusses an improvement to the self-consistent phonons algorithm (SCP). The SCP method allows one to include anharmonic effects when treating a many body quantum system at thermal equilibrium. The system is then described by an effective temperature-dependent harmonic Hamiltonian, which can be used to estimate its various dynamic and static properties. We combine SCP with ab initio (AI) potential energy evaluation, in which case the numerical bottleneck of AI-SCP is the evaluation of Gaussian averages of the AI potential energy and its derivatives. These averages are computed efficiently by the quasi-Monte Carlo method utilizing low-discrepancy sequences leading to a fast convergence with respect to the number, S, of the AI energy evaluations. Moreover, a further substantial (an-order-of-magnitude) improvement in efficiency is achieved once a numerically cheap approximation of the AI potential is available. This is based on using a perturbation theory-like... https://escholarship.org/uc/item/4pd5p4ph Thu, 19 Mar 2026 00:00:00 +0000 Schiltz, Colin https://escholarship.org/uc/item/3sx819mk Programming scalable system-on-chip architectures composed of processor arrays remains challenging due to the complexity of partitioning applications, assigning tasks and data, and realizing communication under strict locality constraints. Regular grid-based architectures offer scalability and predictable communication, yet mapping real applications onto such platforms is often manual, error-prone, and difficult to validate. This dissertation presents the Grid of Processing Cells Compiler (GPCC), an automated framework that maps application task graphs extracted from C++ specifications onto a truly scalable two-dimensional grid of processing cells with local memories and neighbor-only communication. GPCC automatically analyzes application graphs, performs task and data partitioning, places tasks onto grid locations, and generates consistent on-chip and off-chip communication, producing executable System TLM-2.0 architecture models for system-level simulation and evaluation.GPCC... https://escholarship.org/uc/item/3sx819mk Thu, 19 Mar 2026 00:00:00 +0000 Wang, Yutong https://escholarship.org/uc/item/3pv3t1rg This dissertation investigates the design, implementation, and pedagogical implications of generative artificial intelligence (AI) in writing education through a multiyear, multi-context research program. Drawing on the sustained development and classroom deployment of a generative AI writing tutor (PapyrusAI), the study examines how AI systems are shaped by instructional values, how students engage with them in practice, and how their use relates to learning-relevant outcomes. Across three empirical studies spanning secondary and postsecondary settings, the dissertation integrates qualitative design analysis, discourse-level interaction analytics, and quantitative modeling of student engagement and self-efficacy. The first study documents a long-term participatory design process, analyzing artifacts, observations, and educator feedback to trace how pedagogical priorities, institutional constraints, and perceptions of AI influenced system features and implementation decisions... https://escholarship.org/uc/item/3pv3t1rg Thu, 19 Mar 2026 00:00:00 +0000 Ritchie, Daniel https://escholarship.org/uc/item/3n86b2n9 This study of reading recessed narrative in the modernist novel proceeds in the spirit of Roland Barthes’ “image of a triumphant plural, unimpoverished by any constraint of representation” (S/Z 5). . . his invitation to “renewal of the entrances to the text” by the reader to create a “plural text: the same and new” (S/Z 13, 16). The texts examined include the works of Joyce, Woolf, Stein, and Faulkner. Recessed character narration, a term this project aims to define, assign, and promote, is the catalyst to the final transformative alchemy of each of these novels to arrive at the fulfillment of their impossible. This project aims to examine the operations and literary implications of reading recessed narrative and to apply it to some of the most distinguished and studied modernist novels to see how and when it occurs and to what effect; and, finally, to begin to grasp how, as readers and critics, we attend to these new readings. Among the cases I argue: it is through the language... https://escholarship.org/uc/item/3n86b2n9 Thu, 19 Mar 2026 00:00:00 +0000 O'Reilly, Bridget Winifred https://escholarship.org/uc/item/2bh1h7hs The work described in this dissertation focuses on the overall theme of bio-inspired coordination complexes, taking inspiration from the intricate cofactors found within the active sites of Fe-containing metalloenzymes. Part of this work discusses the preparation and properties of Fe complexes supported by the redox-active [SNS] pincer-type ligand, while the other part discusses impacts of the secondary coordination sphere on the reactivity of an Fe(IV)–oxido complex supported by a tripodal ligand framework.Chapter 2 presents a series of Fe–thiolate complexes supported by the [SNS] framework of the general formula, {[SNS]Fe(dppe)}–/0/+, over three redox states. Previous work with this ligand platform demonstrated its capability to bridge multiple metal centers to afford heterobi- and heterotrimetallic complexes. In addition, when coordinated to Group 10 metals, this ligand also proved capable of supplying an oxidizing equivalent and participating in hydrogen atom transfer. The... https://escholarship.org/uc/item/2bh1h7hs Thu, 19 Mar 2026 00:00:00 +0000 Jaimes, Jennifer Lynn https://escholarship.org/uc/item/1w8579ck This dissertation consists of three chapters in urban and public economics that examine how housing policies affect residential development, property values, and neighborhood change. Chapter 1 studies the effects of upzoning in the context of Los Angeles’s Transit Oriented Communities (TOC) program. Housing affordability is a pressing challenge in the United States, and local governments are increasingly exploring policies to increase housing supply. Using property characteristics and consumer trace data, I show that easing zoning requirements increases housing supply, as measured by the number of units. I also find that upzoning leads to higher house prices and demographic shifts, including a greater share of in-migrant households that are non-Hispanic White. Additionally, I find delayed increases in housing units near TOC areas, suggesting spillover effects beyond the treated zones. These findings are robust to a triple-difference-in-differences specification. Chapter... https://escholarship.org/uc/item/1w8579ck Thu, 19 Mar 2026 00:00:00 +0000 Tanrisever, Zeynep Idil https://escholarship.org/uc/item/1mq7d0wq Esta tesis examina el problema de la ficción en el Siglo de Oro español a partir de la asociación, formulada en el siglo XVI, entre la lectura de “malos libros” y la melancolía. El estudio reconstruye las ideas sobre la lectura y la literatura que permiten comprender esa vinculación y sitúa el problema en el contexto humanista en el que se pensaban los efectos de la ficción. A partir de la historia de las ideas sobre la lectura desde San Agustín hasta Cervantes, la tesis reconstruye el marco conceptual en el que esa asociación se formula. En San Agustín, la lectura se entiende como ejercicio de cuidado y consolación del alma; esta concepción se desarrolla en la Edad Media, donde la palabra escrita se figura como alimento que se deglute y rumia para modificar el yo. A partir de la lectura de ficción como acto de compasión en Agustín y de su referencia a la fábula de Píramo y Tisbe, se analiza la configuración medieval de la literatura como discurso sentimental, desde el paradigma... https://escholarship.org/uc/item/1mq7d0wq Thu, 19 Mar 2026 00:00:00 +0000 Marcos Villahoz, Juan Nelson https://escholarship.org/uc/item/91t2w28j Practicing Kingship: Scenes of Counsel as Heuristic Drama in Gower and Shakespeare,” examines the hybrid model of medieval English kingship, integrating divine right with feudal reciprocity, where counsel serves as the essential mechanism for restraining monarchical authority, ensuring virtuous decision-making aligned with the common profit, and fostering collaborative governance between king and aristocrats. Originating from the author's reflections on misconceptions about Henry de Bracton's legal principles—initially misinterpreted as endorsing absolute rule—the project reevaluates Shakespeare's King Lear through the lens of medieval kingship, arguing that Lear's tragic downfall arises from his misrecognition of the hybrid structure of his own kingship, manifested in his illegal division of the kingdom and rejection of counsel from Kent and Cordelia. Drawing on Walter Ullmann's dual kingship framework, which posits an uneasy tension between theocratic and feudal models of kingship,... https://escholarship.org/uc/item/91t2w28j Wed, 18 Mar 2026 00:00:00 +0000 Cassidy, Christopher Mariano https://escholarship.org/uc/item/7mp355pr Twisted tensor products are a unifying construction in noncommutative algebra, encompassing smash products, group rings, Ore extensions, tensor products, and more. Generalized AS-regular (twisted Calabi--Yau) algebras are viewed as noncommutative analogues of polynomial rings, which play an important role in the active research area of noncommutative geometry. In this project, we aim to construct a new class of generalized AS-regular (twisted Calabi--Yau) algebras through twisted tensor products.This dissertation is mostly devoted to computing the Ext-algebra of twisted tensor products, and we restrict our study to Koszul rings. For a well-behaved Koszul algebra (or more generally a Koszul ring), being generalized Artin--Schelter regular or twisted Calabi--Yau is equivalent to its Ext-algebra being Frobenius (respectively quasi-Frobenius).Moreover, in computing the Ext-algebra, Koszulness allows us to leverage Koszul duality instead of using the traditional resolution approach.There... https://escholarship.org/uc/item/7mp355pr Wed, 18 Mar 2026 00:00:00 +0000 Zhang, Maureen https://escholarship.org/uc/item/38h5d6sb Climate-driven change and human pressures are rapidly altering coastal upwelling ecosystems through shifts in carbonate chemistry, nutrient supply, oxygen availability, and food-web structure, yet biological monitoring often struggles to scale from organism condition to community-level impacts. A key unresolved question is whether DNA-based approaches can provide sensitive, transferable indicators of these stressors across diverse taxa and highly dynamic oceanographic gradients. Here, I integrate two omics approaches, metabarcoding of mesozooplankton communities and metagenomics of water-column microbes, to evaluate indicator potential, identify candidate targets, and clarify how spatiotemporal gradients structure coastal biological responses.In the Southern California Bight, mitochondrial cytochrome c oxidase subunit 1 metabarcoding of mesozooplankton (>200 µm) across 20 nearshore locations and four seasons revealed communities dominated by copepods and krill and structured... https://escholarship.org/uc/item/38h5d6sb Wed, 18 Mar 2026 00:00:00 +0000 Bandy, Ashton Margaret https://escholarship.org/uc/item/0gq7400m High-stakes decision-making domains demand Machine Learning (ML) models to not only be accurate but also fair, transparent, and robust. This dissertation presents a comprehensive study on learning fairness-aware decision models for critical applications, with a focus on undergraduate admissions. We begin by highlighting the limitations of traditional admissions review process and classical ML model usage. These processes struggle with issues pertaining to performance and scalability, can also inadvertently reinforce societal biases. To address the aforementioned issues, we first develop deep learning classifiers with regularization techniques to predict admissions outcomes, demonstrating improved predictive accuracy over baseline models. Furthermore, we leverage interpretability techniques to validate model decisions. Building on this foundation, we introduce an adversarial debiasing framework that substantially mitigates discriminatory patterns in model predictions. It ensures... https://escholarship.org/uc/item/0gq7400m Wed, 18 Mar 2026 00:00:00 +0000 Priyadarshini, Amisha https://escholarship.org/uc/item/89b68351 Smartphones enable users to execute both security-critical and potentially harmful applications, creating an imperative for a secure environment to safeguard sensitive operations. While manufacturers have addressed this through Trusted Execution Environments (TEEs) such as TrustZone, these systems typically have a large Trusted Computing Base (TCB) including complex hardware and software. Consequently, they remain vulnerable to attacks via hardware vulnerabilities (such as side channels), architectural flaws, and software exploits. To mitigate these risks, we present a novel, physically-isolated TEE architecture for smartphones designed to minimize the TCB size. We introduce a split-trust hardware design featuring statically-partitioned domains connected through a formally verified hardware mailbox. By enforcing strict physical isolation, this architecture ensures provably exclusive access to computation and I/O resources on a temporary basis. This fundamental shift eliminates... https://escholarship.org/uc/item/89b68351 Mon, 16 Mar 2026 00:00:00 +0000 Chen, Mingyi https://escholarship.org/uc/item/7615g2t8 This dissertation investigates how Reconfigurable Intelligent Surfaces (RISs) can be designed, modeled, and optimized to operate effectively in realistic wireless environments. The work begins by introducing a wave-controlled metasurface-based RIS architecture in which standing-wave voltages on a transmission line bias multiple varactor-loaded passive reflecting elements simultaneously, reducing wiring complexity while preserving the ability to form high-quality beams and nulls. An electromagnetic modeling framework is developed to characterize this architecture, capturing mutual coupling, material losses, and nonlinear device behavior. Two sampling circuits are proposed to convert the standing-wave voltages into DC voltages to reverse bias the varactor diodes -- a passive rectifier-based envelope detector, and a sample-and-hold (S/H) circuit. Using these realistic models, we develop several algorithms to operate the wave-controlled RIS to enhance the received signal power at... https://escholarship.org/uc/item/7615g2t8 Mon, 16 Mar 2026 00:00:00 +0000 Ben Itzhak, Gal https://escholarship.org/uc/item/6r96v3f3 Northern Mexico has experienced increased temperatures because of anthropogenic climate change. While it is anticipated to become drier and more water-stressed by the end of the 21st century, the magnitude and spatial extent of precipitation changes associated with warming remain largely uncertain. While new paleoclimate records can help fill these knowledge gaps, few records exist, and far fewer records look at temperature and precipitation change together. Further, the interpretation of cave-based paleoclimate records benefits from detailed fieldwork and precise characterization of cave geometry and the overlying bedrock but has historically been constrained by conventional mapping approaches and limited participation of local community partners. In this dissertation, I use two new approaches to address these challenges: (1) a multi-proxy speleothem framework that combines high-precision fluid inclusion microthermometry with conventional stable isotope (δ¹⁸O, δ¹³C) and trace... https://escholarship.org/uc/item/6r96v3f3 Mon, 16 Mar 2026 00:00:00 +0000 Pahl, Bryant Labajo https://escholarship.org/uc/item/0nn81170 Catecholate-type ligands on transition metals have demonstrated the ability to act as proton and electron reservoirs, enabling diverse chemical transformations. This dissertation investigates how redox-active amidophenolate ligands control the electronic structure, reactivity and physical properties of transition metal complexes, spanning five-coordinate ruthenium complexes as well as group IV charge-transfer dyes.Chapter 2 examines the synthesis of a new series of five-coordinate ruthenium (II) amidophenolate complexes, LnRu(Phap)(PR3)m (L = N2, MeCN, CO, C2H4, Phap = 3,5-di-tert-butyl-(6-phenylamidophenolate), PR3 = PPh3, PMe3, diphenylphosphinoethane). The geometry of these complexes is controlled by the rich electron density donated by the amidophenolate ligand as well as the ancillary phosphines’ steric bulk. X-ray diffraction analysis supported by DFT calculations reveal a high level of covalency between the metal and amidophenolate which leads to intermediate ligand oxidation... https://escholarship.org/uc/item/0nn81170 Mon, 16 Mar 2026 00:00:00 +0000 Lee, Younju https://escholarship.org/uc/item/9928m25p Simulations are a powerful tool for training reinforcement learning (RL) agents, yet policies learned in simulation often fail when deployed under conditions that differ from training, whether due to surprising behavior from other agents, unseen situations, or incorrectly modeled dynamics. This dissertation develops methods that make RL agents more robust and adaptable across such mismatches. For learning strong strategies against unknown opponents in competitive games, we introduce three population-based methods: Pipeline PSRO for scalable parallel equilibrium computation with convergence guarantees, Extensive-Form Double Oracle (XDO) for convergence linear in the number of information states, and Self-Play PSRO for efficient population growth. Connecting equilibrium computation to robustness, we propose Feasible Adversarial Robust RL (FARR), which casts robust training as a two-player game that automatically filters out irrelevantly difficult environment conditions. For sim-to-real... https://escholarship.org/uc/item/9928m25p Sat, 14 Mar 2026 00:00:00 +0000 Lanier, John Banister https://escholarship.org/uc/item/7cj8p2xk Single-cell and spatial multi-omic profiling boosts biological discoveries by measuring gene expression, chromatin state, and tissue context across substantial cellular and tissue heterogeneity; however, most studies remain partially observed: modalities are often missing or unpaired, measurements are sparse, and distributions shift across protocols and cohorts. These properties induce central computer-science challenges in robust representation learning, cross-modality translation under partial observability, and interpretable reasoning over high-dimensional regulatory mechanisms. This work develops learning and inference frameworks capable of modeling the scale and heterogeneity of consortium datasets, where close collaboration between computational and experimental scientists is essential to convert large cohort resources into reproducible analysis pipelines and robust, interpretable model outputs. We first develop robust epigenetic representations and leverage them to... https://escholarship.org/uc/item/7cj8p2xk Sat, 14 Mar 2026 00:00:00 +0000 Xu, Siwei https://escholarship.org/uc/item/7vf130rf Viral infections of the central nervous system (CNS) resulting in pathological damage and autoimmune neuroinflammatory diseases suspected to be triggered by viral infections, as in the case of multiple sclerosis (MS), share a critical reliance on chemokine-driven immune cell infiltration and dysregulated peripheral immune responses. The role of neutrophils in potentiating these dysregulated responses remains to be further elucidated. Neutrophils, historically understudied in MS due to low detection in postmortem tissue, are increasingly implicated in disease onset, lesion expansion, and relapse severity. Elevated neutrophil chemokines (CXCL1, CXCL2, CXCL5, CXCL8) in patients correlate with relapse activity, clinical disability, and blood-brain barrier (BBB) disruption. Pre-clinical models of MS, including EAE, cuprizone, and viral-induced infection, demonstrate that neutrophils facilitate BBB breakdown, can potentially mediate demyelination, and can act redundantly with monocyte-derived... https://escholarship.org/uc/item/7vf130rf Fri, 13 Mar 2026 00:00:00 +0000 Olivarria, Gema Melissa https://escholarship.org/uc/item/6xn7x1qd Power-to-gas-to-power (P2P) offers a pathway to convert surplus renewable electricity into hydrogen, store it across time, and reconvert it to electricity when grid value and reliability needs are highest. This thesis focuses on the generation stage of an integrated P2P system at the University of California, Irvine, experimentally evaluating a 50 kW proton exchange membrane fuel cell (PEMFC) module coupled with a Thermal Care NQA25 air-cooled chiller and a Chroma 17040 battery emulator, and developing a steady-state stack model to interpret performance. Experiments were organized into three categories: steady-state and start-up characterization, value-based load following using Time-Dependent Valuation (TDV) signals from the California Energy Commission, and renewable-profile load complementing using NREL System Advisor Model data. Characterization confirmed stable operation of the system across the intended setpoints, with cold- and hot-starts exhibiting sub-second electrical... https://escholarship.org/uc/item/6xn7x1qd Fri, 13 Mar 2026 00:00:00 +0000 Moreno Ruiz, Esteban https://escholarship.org/uc/item/5h80x4dk This dissertation examines how institutional policies and information shocks influence racial disparities and decision-making in criminal justice and education. The first chapter evaluates a 2021 policy in Berkeley, California, that banned low-level traffic stops. Using synthetic difference-in-differences, I find that while the policy reduced non-moving violation stops and narrowed racial disparities for Black drivers, it also led to a decline in contraband discovery. This suggests a concrete trade-off between achieving racial equity and maintaining traditional investigative outputs. The second chapter provides the first empirical evaluation of the DEA’s "Operation Pipeline" training. Analyzing data from 2008 to 2015 via interacted fixed effects, I find no statistically significant average impact on drug seizures or racial disparities, suggesting that anecdotal enforcement "spikes" may represent spurious heterogeneity rather than scalable program effects. The third... https://escholarship.org/uc/item/5h80x4dk Fri, 13 Mar 2026 00:00:00 +0000 Anderson, Emily Hope https://escholarship.org/uc/item/5gh7z496 The phenomenon of neutrino flavor conversion whereby the flavor of a neutrino particle can change between its time of production and later detection was the first definitive evidence of physics beyond the Standard Model. Some of the oscillation parameters used to describe this conversion are not yet well measured, leaving important questions still open regarding flavor conversion both in vacuum and as neutrinos travel through matter. NOvA is a long-baseline neutrino oscillation experiment that uses Fermilab’s predominantly νµ NuMI beam. A 14 kton oil-based liquid scintillator far detector 810 km away is used to measure neutrino oscillation through the νµ disappearance and νe appearance channels. Super-K is a 50 kton water Cherenkov detector, which measures the disappearance of νe produced during solar fusion. The high density environment of the sun decreases the νe survival probability at higher energies observable in Super-K compared to the vacuum-dominated oscillations at lower... https://escholarship.org/uc/item/5gh7z496 Fri, 13 Mar 2026 00:00:00 +0000 Yankelevich, Alejandro Jaime https://escholarship.org/uc/item/3f8363sn Exoplanets—planets orbiting stars other than the Sun—are now known by the thousands, primarily through the transit and radial-velocity methods. A transit is a small, periodic dimming that occurs when a planet passes in front of its star; radial velocities detect the star’s Doppler shift due to the planet’s gravity. The central challenge in modern exoplanet studies is reducing measurement uncertainty well enough to translate these indirect detections into reliable physical properties. Because the planet signal is encoded in the starlight, errors in the light curve or stellar parameters propagate directly into planet radii, masses, and densities. This dissertation builds a coherent path from observations to those properties by combining forward modeling of crowded-field photometry, uniform stellar parameters, and consistent inference across both individual systems and populations. The overarching aim is to improve the accuracy and internal consistency of exoplanet measurements as... https://escholarship.org/uc/item/3f8363sn Fri, 13 Mar 2026 00:00:00 +0000 Han, Te https://escholarship.org/uc/item/3c03v54g This dissertation brings together three essays in applied microeconomics that examine how policy, institutions, and information frictions shape economic behavior in markets and the public sector. Spanning environmental taxation, political economy, and pricing under asymmetric information, the chapters share a unifying focus on how agents respond to incentives when operating under institutional and informational constraints.The first chapter, “The Effect of Aviation Fuel Taxes on Airline Operations: A Brazilian Case Study,” studies how airlines adjust operational decisions in response to fuel taxation. Exploiting variation in state-level aviation fuel taxes in Brazil, the paper analyzes impacts on route structure, fuel consumption, pricing, and emissions-related outcomes. The results show that airlines respond not only through cost pass-through but also through network reallocation and strategic operational adjustments, highlighting the incidence and environmental effectiveness... https://escholarship.org/uc/item/3c03v54g Fri, 13 Mar 2026 00:00:00 +0000 Son, Jiwon https://escholarship.org/uc/item/2t5264c5 Air-sea gas exchange is the process by which gases are transported between the atmosphere and ocean boundary layer. The rate of gas exchange is important in calculating the global budgets of climate-active trace gases. The rate of gas exchange is influenced by multiple factors in the atmosphere and ocean boundary layers, such as wind speed, sea surface temperature, and ocean/atmospheric concentration. In most models, the transfer velocity (kw) is parameterized as having a quadratic relationship to wind speed with a scaling parameter obtained from inversion of radiocarbon in a global ocean model (Wanninkhof, 2014). However, bubbles produced by breaking waves have been shown to be an important component of air-sea gas transfer of CO2 (Reichl & Deike, 2020). In this dissertation, the two parameterizations are compared using a global inverse circulation ocean model (OCIM2) to evaluate if the inclusion of bubble-mediated gas exchange will improve model performance. This study... https://escholarship.org/uc/item/2t5264c5 Fri, 13 Mar 2026 00:00:00 +0000 Al Ali, Shayma Ali https://escholarship.org/uc/item/4p20m0tj Over-the-counter (OTC) markets are the main trading venues for many fixed-income and derivative instruments. Their decentralized and often non-anonymous structure creates search frictions and renders bargaining important for the determination of transaction costs and other measures of liquidity. This dissertation develops a unified set of search-theoretic models to study recent (and not so recent) evolutions of trading practices in OTC markets.Chapter 1 formalizes a decentralized asset market where investors form long-term trading relationships with dealers. Relationships partially substitute for search by granting investors sustained access to intermediation services, and they also mitigate a holdup problem that arises when investors cannot contract over future trades. As relationships become more stable, equilibrium asset holdings become more dispersed, investors trade in larger quantities, and aggregate volume rises through both extensive and intensive margin effects. Intermediation... https://escholarship.org/uc/item/4p20m0tj Wed, 11 Mar 2026 00:00:00 +0000 Maciocco, Alex William https://escholarship.org/uc/item/6498n2k0 Software developers commit code changes throughout software development and maintenance using version control systems. These code changes are usually accompanied by commit messages written in natural language, which are free-form textual descriptions of the corresponding change. These commit messages play a vital role in modern software development by serving as one of the communication channels through which developers communicate the context of a change to collaborators.Since it is evident that commit messages play an important role in developer communication, accurately and comprehensively measuring their quality is essential to ensure they bolster communication among developers. To do so, literature has defined what semantically constitutes a Good commit message where they proposed that a Good commit message should include both a summary of the changes in the commit (noted as What) and a description of the reasons or justifications for the changes (noted as Why). In this definition,... https://escholarship.org/uc/item/6498n2k0 Mon, 9 Mar 2026 00:00:00 +0000 Li, Jiawei https://escholarship.org/uc/item/05g4z3ff Digital health systems increasingly sense physiology, behavior, and context in everyday life, creating the opportunity to deliver timely, individualized support. Yet in high-stakes settings, average-case models can be unsafe: individuals have sparse and drifting data, cohorts are heterogeneous, multimodal streams are missing, and fluent LLM explanations can be ungrounded in personal evidence. This thesis defines trustworthy personalization as a coupled objective of (i) accurate individual-level prediction under these constraints and (ii) interpretability that is traceable to the same personal signals used by the system. To meet this objective, the thesis advances a two-track agenda across five papers. Track 1 (predictive personalization) develops data-efficient personal models for longitudinal sensing via (i) week-ahead loneliness forecasting from multimodal wearable/mobile data with individual evidence attribution, (ii) week-ahead affect forecasting that fuses wearable... https://escholarship.org/uc/item/05g4z3ff Mon, 9 Mar 2026 00:00:00 +0000 Yang, Zhongqi https://escholarship.org/uc/item/2c76s9nf In clinical research, practitioners often specify simple models where most covariate effects are modeled linearly in the covariates due to the desire for inference on estimands with simple interpretations. However, these models are nearly guaranteed to be misspecified for the true model. Under misspecification, we target parameters other than standard parameters most practitioners believe they are estimating. One response to this is to continue to fit these simple but ultimately misspecified models with rigorous understanding of what they target and when these targets might still have utility in applied analysis. While there is literature discussing the consequences of misspecification, there are gaps in (i) our understanding of the behavior of standard inferential frameworks for these parameters under certain study designs and (ii) our understanding of situations in which these estimands may yield misleading scientific conclusions. An alternative response to this is to argue... https://escholarship.org/uc/item/2c76s9nf Fri, 6 Mar 2026 00:00:00 +0000 Birnbaum, Adam https://escholarship.org/uc/item/88r011tj Aging is a leading factor in degenerative pathologies throughout the body. The cellular factors underlying aging manifest as several phenotypes that must be understood to characterize and prevent aging-related diseases. This dissertation presents two studies that seek to address age-related diseases in different ways.The first study attempts to validate a new model of age-related macular degeneration to improve studies of the disease's metabolic aspects. Age-related macular degeneration is the leading cause of visual impairment among older adults and involves a breakdown of the supportive retinal pigment epithelial cells found at the rear of the retina. Previous studies have indicated that mitochondrial dysfunction and metabolic imbalances may play a key role in the pathology of age-related macular degeneration. Studying this is challenging due to the confounding effects of human nuclear genetics and mitochondrial activity. Therefore, the Kenney lab generated cytoplasmic hybrids... https://escholarship.org/uc/item/88r011tj Fri, 20 Feb 2026 00:00:00 +0000 Dohl, Jacob https://escholarship.org/uc/item/8sd06596 3D geometric understanding and generation are fundamental problems in computer vision, with broad applications in augmented and virtual reality, human computer interaction, and autonomous systems, and they form a core component of learning world models. Recent advances in deep learning have led to substantial progress across many 3D tasks, yet achieving accuracy, geometric consistency, and computational efficiency simultaneously remains a key challenge. This dissertation addresses this challenge by developing efficient and principled learning frameworks that explicitly leverage geometric structure across multiple 3D representations.The thesis first studies explicit surface geometry and introduces an optimal transport based framework for learning diffeomorphic mesh deformations. By modeling surfaces as probability measures and employing sliced Wasserstein distance, this approach defines efficient and geometrically meaningful discrepancies between meshes, enabling accurate cortical... https://escholarship.org/uc/item/8sd06596 Sat, 14 Feb 2026 00:00:00 +0000 Le, Thanh Tung https://escholarship.org/uc/item/8sf0075t The early stage of mammalian development is a critical phase that is vulnerable to environmental factors. It is increasingly recognized that early life adversity (ELA) predisposes people to neuropsychiatric disorders later in life (Merrick et al. 2019; Waters and Gould 2022). Understanding how ELA impacts the brain can lead to advances in treatments and therapeutics for a variety of neuropsychiatric diseases. Here, we explore how developmental perturbation impacts synapses and start to examine the molecular machinery that is involved in the development and function of excitatory and inhibitory synapses in the striatum of the adult mouse. The striatum receives input from virtually all major cortical regions and has outputs to several mid-brain regions that regulate behavior, making it a central hub for sensory processing and behavioral response (Peters et al. 2021; Reig and Silberberg 2014; Hunnicutt et al. 2016). The mouse dorsal striatum can be subdivided into dorsolateral (DLS)... https://escholarship.org/uc/item/8sf0075t Wed, 11 Feb 2026 00:00:00 +0000 de Carvalho, Gregory Belo https://escholarship.org/uc/item/7354w61s This thesis examines the implicit messaging of Nathan Dunn’s Chinese Museum, which exhibited Dunn’s collection of Chinese objects and mannequins in Philadelphia from 1838 to 1841. The visuals of these dioramas—which featured fifty life-sized figures of real people in China—drew inspiration from conventions of representations of empire, Orient, and commodity, concretizing nineteenth-century Americans’ existing conceptions of China as a material phenomenon. Inspired by the rising popularity of didactic natural history and ethnographic museums, Dunn determined that the best way to educate Americans about Chinese culture would be through an exhibit displaying his collection of Chinese objects in context, reinscribing them with their original meanings that had been deformed by the process of trans-continental trade. Unfortunately, Dunn’s intentions were thwarted by his own actions. By presenting the “authentic” vision of China as an assemblage of material goods that could be imported... https://escholarship.org/uc/item/7354w61s Fri, 6 Feb 2026 00:00:00 +0000 Wilson, Chloe Jayne https://escholarship.org/uc/item/8gv8j62w Neuroinflammation is a hallmark of multiple neurodegenerative conditions, including Alzheimer’s, Parkinson’s, Huntington’s disease, spinal cord/brain injury, and stroke. Microglia, the immune cells of the central nervous system (CNS), are the first to initiate inflammation by proliferating, migrating to regions of disease, and secreting pro-inflammatory cytokines. While this initial wave of inflammation is critical for proper disease response, microglia often continue inflammatory cytokine production into the chronic phase of disease/injury where it can be detrimental to neuroregeneration and neurorepair. Although rodent models have been a mainstay for basic and translational research in the CNS, recent work highlights profound differences between human and rodent microglia. New protocols for the differentiation of human induced pluripotent stem cells (hiPSC) to microglia (iMG) present one solution for this disconnect.My thesis work utilized iMG to investigate the basic mechanisms... https://escholarship.org/uc/item/8gv8j62w Thu, 5 Feb 2026 00:00:00 +0000 Sakthivel, Pooja S https://escholarship.org/uc/item/8cg6q4hv In this dissertation, we study stability, robustness, and fairness in two foundational frameworks of algorithmic game theory: matchings under preferences and cooperative games. Classical results in both settings focus on identifying stable outcomes and establishing their structural and algorithmic properties. However, stability alone does not address robustness to changes in the input or fairness at the level of individual agents. Motivated by these gaps, we investigate how stability-based solution concepts behave under perturbations and how fairness can be incorporated while retaining stability. We first study robustness in two-sided matching markets. We introduce robust stable matchings, which are matchings that remain stable across multiple instances defined on the same set of agents, capturing stability under changes to preferences. We characterize when the structural, algorithmic, and geometric properties of stable matchings persist, parametrized by the number of agents... https://escholarship.org/uc/item/8cg6q4hv Tue, 3 Feb 2026 00:00:00 +0000 Gangam, Rohith Reddy https://escholarship.org/uc/item/8vq1k30p For decades, the world has delayed the development and deployment of large-scale efforts to switch from fossil fuels to renewable energy sources. Industries such as cement, steel, and glass production, which are continuously expanding, contribute to gigatons of CO2 annual emissions and are nearly impossible to decarbonize. Consequently, before the end of this century, the involvement of technology to capture CO2 directly from the atmosphere (DAC) at an annual ~37 gigaton scale is now a requirement to avoid the direst consequences of global warming. In this context, Ca-based oxides have demonstrated outstanding potential due to their selective and high CO2 uptake in the reversible reaction: CaO(s) + CO2(g) ↔ CaCO3(s) + 178 kJ/mol at 298 K. Numerous efforts have been made to lower operating temperatures and achieve faster reaction kinetics, including the use of smaller particle sizes at the microscale and nanoscale. Unfortunately, much of the solid-gas reactions remain unclear,... https://escholarship.org/uc/item/8vq1k30p Sat, 31 Jan 2026 00:00:00 +0000 Martinez, Jenny https://escholarship.org/uc/item/8ft4099r Deep learning provides a powerful framework for addressing inverse problems in particle physics, where unobserved theoretical quantities must be inferred from noisy, imperfect detector observations. Across multiple projects, we introduce attention-based transformer models that explicitly encode the physical symmetries underlying particle interactions. Symmetry Preserving Attention Networks (SPANets) leverage these properties to perform efficient jet–parton assignment, achieving state-of-the-art accuracy in some of the most combinatorially challenging reconstruction tasks. Extensions of SPANet incorporate additional event level information, enabling joint signal–background discrimination and kinematic regression, which improve sensitivity to rare and hypothetical events at the LHC. We further develop generative transformer architectures to create Variational Latent Diffusion (VLD) models for conditional generative modeling of parton- and particle-level kinematics, enabling fully... https://escholarship.org/uc/item/8ft4099r Sat, 31 Jan 2026 00:00:00 +0000 Shmakov, Alexander Konstantinovich https://escholarship.org/uc/item/1kw0h6kj This thesis investigates the vortex fluidic device (VFD) as a thin-film microfluidic platform for engineering nanoscale carriers for topical and ocular drug delivery. The first project focuses on encapsulating the hydrophobic bioactive α-bisabolol (ABS) in Tween 20 emulsions for dermal application. The second project explores the self-assembly of 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) liposomes as potential vehicles for protein delivery in age-related macular degeneration (AMD) therapy.For ABS, a coarse oil-in-water emulsion premix (ABS/Tween 20/water, 1:1:8 w/w/v) was prepared by brief sonication and then processed under VFD continuous flow at 45° tilt angle and 25 °C. Rotational speed (6000 to 9000 rpm), flow rate (0.1 to 0.5 mL/min), and rotational direction (clockwise, CW; counterclockwise, CCW) were systematically varied. Droplet size and polydispersity were quantified by dynamic light scattering (DLS), and long-term emulsion stability was benchmarked against... https://escholarship.org/uc/item/1kw0h6kj Sat, 31 Jan 2026 00:00:00 +0000 Zhou, Yve https://escholarship.org/uc/item/0pz7s6tq Proton exchange membrane water electrolyzers (PEMWEs) are a leading technology for green hydrogen production, offering high efficiency and compatibility with renewable energy sources. However, the reliance on iridium-based catalysts for the oxygen evolution reaction (OER) presents cost and durability challenges that must be addressed to enable large-scale deployment. To reduce iridium loading a design of adjacent components to the catalyst layer is needed, such as introduction of the microporous layer (MPL), a fine titanium later to provide good electric contact to the catalyst layer. Furthermore, to ensure adequate contact between iridium-based catalyst layer and MPL a compression study is needed. Lastly, morphology of the catalyst layer can be tailored by introduction of pore-formers that increase iridium oxide surface area and enable better oxygen removal. This is achieved by introduction of nanobubbles into catalyst layer inks to create nano-scale pores. This study examines... https://escholarship.org/uc/item/0pz7s6tq Wed, 28 Jan 2026 00:00:00 +0000 Ebrahimpour Tolouei, Nadia https://escholarship.org/uc/item/55r871qq Over the past decade, with the rapid growth of social media, it has become easier for some people to hide behind a screen and spread hate online. Hate speech can be in forms of posts and comments. Frequent exposure to hate speech can lead to mental health problems such as depression. Therefore, the problem of hate speech detection has attracted the attention of many researchers in the past decade.Hate speech has three dimensions. The first dimension is the expression which can be explicit or implicit. The second dimension is modality which can be unimodal or multimodal. The third dimension is related to the target of the hate speech. If the target is a specific group it is called cyber-bullying. It can also be more fine-grained such as misogyny and racism.I contribute to the problem of hate speech detection in various ways. I propose a method based on compact BERT models to detect cyber-bullying, tackling the challenge of informal language, hashtags and emojis and imbalanced dataset.... https://escholarship.org/uc/item/55r871qq Fri, 23 Jan 2026 00:00:00 +0000 Behzadi, Mitra https://escholarship.org/uc/item/2gr0d13f This thesis presents two perspectives on the mathematical theory of quantum error correction. In the first perspective, we ask how to relate the plethora of quantum error-correcting codes that have been discovered since the 1990's. We address this question by developing a unifying framework for quantum error-correcting codes in the languages of category theory and real algebraic geometry. Our framework builds upon the Knill-Laflamme quantum error-correcting condition and incorporates all codes meeting this condition, including standard CSS, stabilizer and surface codes, and the more recent development of quantum Tanner codes. We produce an overall diagram of categories and functors that describes both the practical and theoretical ways in which these codes are constructed, e.g. surface codes from topology. Furthering the connection to real algebraic geometry, we develop the concept of a quantum error form and use it to characterize geometrically all quantum codes correcting... https://escholarship.org/uc/item/2gr0d13f Fri, 23 Jan 2026 00:00:00 +0000 Marks, Adam https://escholarship.org/uc/item/8zq6n7sm Fault tolerance is essential for scalable quantum computation, and surface codes are the leading candidates due to high error thresholds and nearest-neighbor compatibility. Their chief limitation is the physical-qubit overhead per logical qubit. This thesis focuses on the defect-based approach, where logical information is encoded by introducing holes (defects) in a planar surface code. I formulate a distance-constrained design-optimization problem that trades off logical-qubit density against logical error rate, and employ stochastic search (e.g., simulated annealing and genetic algorithms) to explore large layout spaces. Through simulations, I compare closed versus partially open defects, quantify sensitivity to key hyperparameters, and identify regimes where increased density is achievable without violating target distance constraints. The results provide practical guidelines for distance-aware layout synthesis and clarify when defect geometry yields genuine density gains versus... https://escholarship.org/uc/item/8zq6n7sm Wed, 21 Jan 2026 00:00:00 +0000 Sayedsalehi, Samira https://escholarship.org/uc/item/87q5b2gn With finger proprioceptive deficits affecting over 50% of stroke survivors, yet often remaining unevaluated and untreated, targeting these deficits could be key to improving hand function recovery. Traditional clinical methods for assessing proprioception lack precision, while existing robotic devices are often complex and inaccessible. This study evaluated the use of PINKIE, a rapidly manufacturable, portable version of the FINGER exoskeleton robot, for measuring finger proprioception deficits. Three aims guided the project: (1) validate PINKIE against FINGER using the Crisscross proprioception test; (2) examine correlations between PINKIE proprioceptive errors and performance on a finger dexterity task (button fastening); and (3) test associations between PINKIE errors and a self-reported hand activity questionnaire. In 20 unimpaired participants, proprioceptive errors measured with PINKIE were significantly larger than, but also significantly correlated with, those measured... https://escholarship.org/uc/item/87q5b2gn Tue, 20 Jan 2026 00:00:00 +0000 Rojas, Raymond Diaz https://escholarship.org/uc/item/7580929w The increasing power density and integration complexity of modern semiconductor devices necessitate advanced thermal management solutions to ensure reliable performance, particularly in heterogeneous packaging configurations. Conventional copper heat spreaders, while effective for general heat dissipation, inadvertently promote thermal crosstalk between high-power heat sources and adjacent temperature-critical regions. This study presents an innovative topology optimization approach to design heat spreaders that actively control internal heat flow, significantly reducing thermal crosstalk and minimizing material usage. Two optimization methodologies are explored: a thermo-mechanical strategy utilizing porous copper fillers to balance thermal isolation and structural integrity, and a thermo-fluidic strategy leveraging fluid-filled channels optimized for convective heat removal. Results from the thermo-mechanical phase highlight a clear trade-off, with higher porosity fillers enhancing... https://escholarship.org/uc/item/7580929w Mon, 19 Jan 2026 00:00:00 +0000 Shaikh, Zubair https://escholarship.org/uc/item/0xn6908g The global energy transition demands technologies capable of decarbonizing industrial sectors that are poorly served by direct electrification. These sectors require not just electrons, but molecules—chemical feedstocks that can replace fossil-derived inputs in high-temperature and high-throughput processes. Solid oxide ceramics (SOCs) represent a versatile and promising platform for this challenge, uniquely offering reversible operation between electrolysis and fuel cell modes. In electrolysis mode, SOCs efficiently convert water and carbon dioxide into hydrogen or syngas, while in fuel cell mode, they generate electricity from a wide range of fuels, including hydrogen and hydrocarbons. Unlike polymer electrolyte membrane (PEM) and alkaline electrolyzers, which are typically unidirectional and limited to low-temperature operation, SOCs operate at elevated temperatures that enable higher electrical efficiency, broader fuel flexibility, and integration with industrial waste heat... https://escholarship.org/uc/item/0xn6908g Mon, 19 Jan 2026 00:00:00 +0000 Rose, Christian https://escholarship.org/uc/item/6sk0g26r Romania’s militaristic and fascist Iron Guard movement–also known as the Legion of the Archangel Michael–called for a violent, spiritual crusade to reclaim the nation’s blood and soul. This thesis considers how the Iron Guard conceived of a “new man” and “legionary woman” through topics like ultranationalism, physical education, and mysticism. The movement’s forceful rejection of urbanism and foreign influences culminated in a deeply anti-semitic philosophy that asked its male followers to embark on a journey of regeneration through discipline, labor, and violence. The Iron Guard asked “legionary sisters” to defend against invading cultural forces, yet largely confined women’s participation to domestic and spiritual duties. To understand the psychic appeal of Romanian fascism, this thesis will wander through rural Romanian towns, unpacking how the cult of death and spectacles of the “new man” manifested on the bodies of the legionaries. The study analyzes several publications... https://escholarship.org/uc/item/6sk0g26r Wed, 14 Jan 2026 00:00:00 +0000 Block, Josh https://escholarship.org/uc/item/8kq180mg Random Telegraph Noise (RTN) has become a major source of time-dependent variability in deeply scaled CMOS and emerging memory technologies. It arises from the random capture and emission of charge carriers by defects located in the oxide or at the channel interface, leading to discrete fluctuations in device current or threshold voltage. While these effects were negligible in older, larger-geometry technologies, modern transistors, containing only a small number of carriers in the channel experience noticeable parameter shifts from individual traps, affecting the performance of analog, mixed-signal, and memory circuits.This thesis studies the physical mechanisms of RTN, its statistical characteristics, and its impact on circuit-level operation. At the device level, energy-band models and trap kinetics are used to describe capture and emission processes, including their dependence on bias and temperature. At the circuit level, the conversion of RTN into timing jitter, phase noise,... https://escholarship.org/uc/item/8kq180mg Mon, 12 Jan 2026 00:00:00 +0000 Narayanan, Thyagarajan https://escholarship.org/uc/item/84n223d4 The distributed training of machine learning models via gradient descent is generally conducted by iteratively computing the local gradients of a loss function and aggregating them across all processors. Communicating these gradients during aggregation is often a major cost but sparsification techniques can greatly improve efficiency. One such technique, Top-k gradient compression, ensures that only the k largest components of each local gradient are sent. However, effectively scaling this method can be challenging. The standard ring AllReduce algorithm, which is frequently used to aggregate dense gradients, lacks a counterpart that is optimized for sparse data. Notably, ring algorithms are contention-free, which generally make them easier to scale than other collective communication algorithms. Thus, in practice, the ring AllGather algorithm, which can be trivially adapted for sparse data, may be used instead, even though its bandwidth costs are proportional to the number of... https://escholarship.org/uc/item/84n223d4 Thu, 25 Dec 2025 00:00:00 +0000 Arunachalam, Anshul https://escholarship.org/uc/item/5b38t262 Self-renewing hematopoietic stem cells (HSCs) produce billions of blood cells a day to maintain peripheral blood and immune cells in circulation. Inflammation can alter the balance of steady-state hematopoiesis, and disrupted hematopoiesis can lead to blood disorders or cancers. Over a lifetime of cell divisions, HSCs may acquire somatic mutations that provide a competitive advantage over their wild-type counterparts. The clonal expansion of a mutant HSC population is termed “clonal hematopoiesis” (CH). CH is linked to an increased overall risk of mortality due to incidences of cardiovascular disease and transformation into hematological malignancy. Tobacco and nicotine use remain the leading preventable drivers of cancer risk, and both direct and secondhand exposure to combustible cigarettes or electronic nicotine devices perturbs immune function and hematopoiesis. The World Health Organization estimates more than 100 million people across the world are using electronic cigarettes,... https://escholarship.org/uc/item/5b38t262 Thu, 25 Dec 2025 00:00:00 +0000 Sullivan, Jeanette https://escholarship.org/uc/item/1bf8j2r7 English academic writing is a critical yet challenging skill for learners in English as a Foreign Language (EFL) contexts. The rapid integration of digital tools, accelerated by the COVID-19 pandemic, has transformed writing instruction; however, evidence of its pedagogical effectiveness remains fragmented and often overlooks teacher perceptions and genre-specific impacts. This three-study dissertation addresses these gaps by investigating the role of digital tools in EFL writing instruction within the EFL higher education context, employing a multi-method approach to triangulate evidence from student outcomes, meta-analytic synthesis, and teacher experiences.Study 1 conducted a classroom experiment with 111 Chinese undergraduates, comparing infographic-based pre-writing to traditional outlining. Results showed that infographic creation significantly improved summary-writing quality, source use, and self-efficacy, but not opinion writing, highlighting the genre-sensitive nature... https://escholarship.org/uc/item/1bf8j2r7 Thu, 25 Dec 2025 00:00:00 +0000 Lin, Xiao https://escholarship.org/uc/item/7pw56096 This study examines how Human-Centered AI (HCAI) and Applied Artificial Intelligence (AI) programs at universities balance technical and human-centered goals and incorporate learning theories within required courses. Twenty programs—including bachelor’s, master’s, certificates, and minors—were analyzed, with courses coded for human-centered (H1–H5) and technical (T1–T4) goals and learning-theory orientation (L1–L4). Results show that 70% of programs prioritize human-centered competencies—particularly communication, collaboration, and ethical literacy—while technical goals are emphasized in most “Applied” programs, especially at the bachelor’s and master’s levels. Learning-theory analysis indicates that larger programs maintain balanced integration of Cognitivism, Constructivism, Situated Learning, and Sociocultural Theory, whereas smaller programs and minors emphasize cognitive and constructivist approaches with limited sociocultural representation. Program title, degree level,... https://escholarship.org/uc/item/7pw56096 Wed, 24 Dec 2025 00:00:00 +0000 Nassery, Sarah https://escholarship.org/uc/item/7b4564qb According to the master narrative framework (Hammack, 2010), individuals construct a sense of identity alongside societal perspectives of what it means to live an acceptable life course. Navigating a life course alongside societal ideals could be particularly complex for transgender individuals as they face pervasive cultural stereotypes of having “abnormal” or societally disruptive identities (Bradford & Syed, 2019; Gazzola & Morrison, 2014). The present study extends existing literature by expanding on minimally researched societally proliferated master narratives that have been identified in previous literature, as well as developing new master narrative themes discussed in the coming out narratives of 73 transgender participants. A flexible abductive approach to thematic analysis was used to identify two broader society-focused master narratives: (1) an individual should be their authentic self, and (2) family should provide unconditional support, along with one transgender-focused... https://escholarship.org/uc/item/7b4564qb Wed, 24 Dec 2025 00:00:00 +0000 Schlaupitz, Caleb Alexander https://escholarship.org/uc/item/2w10m7xb Smiling, a facial expression observed surprisingly often during the experience of acute pain, may play a pivotal role in shaping physiological and psychological responses. This study explores the influence of smiling on objective (i.e., heart rate) and subjective responses (i.e., pain reports and state affect) during a painful cold temperature task (N = 57, Mage = 20.80, SD = 1.50). Participants who spontaneously smiled during the painful task had lower heart rates throughout the study and reported higher levels of positive affect during recovery from pain. By examining the significance of this facial expression, this study offers valuable insights into ways in which smiling contributes to the promotion of positive physiological and psychological outcomes within the challenging context of pain. https://escholarship.org/uc/item/2w10m7xb Wed, 24 Dec 2025 00:00:00 +0000 Luu, Jazlyn H https://escholarship.org/uc/item/0k5971pc Purpose: The present study explores lifetime childhood trauma load, perceptions of neighborhood safety, and their interaction as predictors of subjective and objective sleep outcomes at age 11-13.Methods: Participants (N = 7,191) were from the Adolescent Brain Cognitive Development (ABCD) Study, the largest longitudinal study exploring child development. Childhood trauma load, perceived neighborhood safety, and subjective sleep measures were assessed via caregiver reports. Objective sleep measures captured in a subsample of participants (N = 3,672) were derived from child Fitbit data.Results: Greater childhood trauma load and lower perceived neighborhood safety were significantly associated with increased sleep disturbance on nearly all subjective measures, even after adjusting for sleep disturbance one year prior and sociodemographic characteristics. In contrast, greater trauma load was only significantly associated with one objective metric (increased bedtime variability), but... https://escholarship.org/uc/item/0k5971pc Wed, 24 Dec 2025 00:00:00 +0000 Canady, Margaret T https://escholarship.org/uc/item/86h51394 In William Shakespeare’s A Midsummer Night’s Dream, Nick Bottom is “translated” into an ass after he enters a hawthorn brake: a thorny bush historically used as a hedge to protect fields, cattle, and fairyland, according to British and Irish folklore. In the play, the hawthorn brake is used as a tiring-house, a theatrical space where actors enter to change costumes and wait for their cue. Given the perennial associations of hawthorns with fairy lore and other symbolism connected to May Day and Midsummer celebrations since ancient times, I investigate whether Shakespeare uses the hawthorn as a portal to fairyland, one that Nick Bottom crosses unknowingly, becoming part donkey and experiencing a night of fantasy, magic, and desire among the forest fairies. This inquiry also allows for examining of the blurred line between theater and fairyland, and how these boundaries can be manipulated to enhance theatrical and emotional effects. To support my argument, I draw on folklore about... https://escholarship.org/uc/item/86h51394 Fri, 19 Dec 2025 00:00:00 +0000 Soligo, Marcia https://escholarship.org/uc/item/82x776k4 Tilt is a gaming phenomenon where an emotional response, typically negative, leads to deterioration of gameplay; tilt is a form of emotional dysregulation, wherein players are selecting dysregulatory regulation strategies as part of a downward spiral in emotional self-regulation. The present study first surveyed young adult video game players (n=165) on their triggers and responses to tilt. A subset of participants (n=30) were also interviewed to provide qualitative context and explore how players make sense of tilt. Omnibus statistical tests, correlations, and partial correlations were conducted to examine how triggers and responses vary by demographics, gameplay habits, personality, and social context of play. Higher length of experience correlated positively with being triggered by opponents. Gamer identity correlated positively with planning responses and behavioral toxicity. Emotional regulation tendency indicates that tilt is indeed a form of emotional dysregulation. Genre... https://escholarship.org/uc/item/82x776k4 Fri, 19 Dec 2025 00:00:00 +0000 Wu, Minerva https://escholarship.org/uc/item/68r0z6n9 Understanding our ocean requires analyzing it as it naturally exists—in three dimensions. Empirical orthogonal functions (EOFs) are often used to analyze dominant patterns of climate variability. They have been used extensively to discover, quantify, and predict important climate phenomena such as El Niño-Southern Oscillation, the North Atlantic Oscillation, and the Pacific Decadal Oscillation. However, computing three-dimensional (3D) EOFs for large datasets has been computationally prohibitive due to computer memory limitations when implementing traditional methods (i.e., singular value decomposition) to obtain EOFs. These limitations have restricted EOF calculations to a single layer, a small spatial domain, or coarse resolution. This dissertation presents two approaches that enable 3D EOF computation for large climate datasets using a regular personal computer with limited memory (e.g., 16 GB RAM). The first is the temporal covariance method, and the second is the partition... https://escholarship.org/uc/item/68r0z6n9 Fri, 19 Dec 2025 00:00:00 +0000 Lafarga, Danielle https://escholarship.org/uc/item/12n1c86x In this thesis, I discuss Iris Murdoch’s philosophical concept of moral attention, drawing primarily from her essay, “The Idea of Perfection.” I suggest that her thinking is a distinctive example of philosophical synthesis that bridges multiple intellectual traditions. Writing against a mid-twentieth century movement that Murdoch characterized as an “existentialist-behaviorist” view that reduced moral life to isolated moments of choice, Murdoch offers an account of moral development that is centered around the patient and continuous work of loving attention. Drawing on Simone Weil's spiritual psychology, Wittgensteinian descriptive method, and Platonic metaphysics, Murdoch proposed that moral progress occurs primarily through inner transformation. This thesis argues that Murdoch has a synthetic intellectual method that integrates different writers commonly considered to belong to both continental and analytic philosophical traditions as well as from literature, religious thought,... https://escholarship.org/uc/item/12n1c86x Fri, 19 Dec 2025 00:00:00 +0000 Flanigan, Thomas https://escholarship.org/uc/item/9z16k10r Studies have estimated that the global economic burden of neurological disorders costs nearly $1.9 trillion per year and is projected to increase by 4.3% annually. While this fact highlights the macroeconomic burden of public health disorders, these disorders are just as debilitating on a personal level, affecting the patient and their caretakers.Neurons are the principal cell of the nervous system, they transmit and detect electrochemical signals through their axons and dendrites, respectively. Unlike many other cells, neurons are non-mitotic, presenting an indisputable challenge for renewal and repair when systems become damaged. Fortunately, neurons can regenerate their axons, dendrites, and even both simultaneously; highlighting a remarkable capacity for repair. However, for cells to initiate repair, a repair stimulus must first be conveyed and subsequently detected. Indeed, cell damage induces several changes in cell homeostasis including organelle fragmentation, ion imbalance,... https://escholarship.org/uc/item/9z16k10r Thu, 18 Dec 2025 00:00:00 +0000 Hwu, Patrick T https://escholarship.org/uc/item/330862n6 Seasonal excesses and deficits of renewable energy may provide an opportunity to meet peak load demand through the conversion to hydrogen and reconversion of hydrogen into electricity, otherwise known as power-to-gas-to-power (P2P). This work details the installation, commissioning, and evaluation of the operation of the production subsystem of a greater P2P system located at the University of California, Irvine. The production subsystem includes a commercially available alkaline electrolyzer (AEC) system for hydrogen production and associated air-cooled chiller for cooling demands, with a combined total power of approximately 90 kWAC. The operating production rate is shown to significantly impact the system energy intensity as it varies from 65-95 kWhAC/ kgH2. Completed characterization testing shows important gas purity metrics like the H2 in O2 percentage and deoxidation (DE-OX) reactor temperature to vary from 0.2-1.2 %V and 50-100 °C, respectively. Emulation of renewable... https://escholarship.org/uc/item/330862n6 Wed, 17 Dec 2025 00:00:00 +0000 Laviguer, Bobby https://escholarship.org/uc/item/80h8h6w3 The rapid proliferation of advanced medical imaging modalities, such as Magnetic Resonance Imaging (MRI) and Computed Tomography (CT), has generated vast volumes of three-dimensional data that are central to modern clinical practice. While these datasets offer unprecedented opportunities for diagnosis, treatment planning, and disease monitoring, they also pose major computational challenges: robust methods are required to capture anatomical variability, ensure topological plausibility, and generalize across patient populations. Traditional approaches, relying on hand-crafted features and iterative optimization, are often limited in efficiency and scalability. More recently, deep learning has transformed medical image analysis by enabling end-to-end frameworks that automatically extract hierarchical features and model complex anatomical structures.This dissertation advances the state of 3D medical image analysis across three key research areas. First, in image registration, it... https://escholarship.org/uc/item/80h8h6w3 Mon, 15 Dec 2025 00:00:00 +0000 Han, Kun https://escholarship.org/uc/item/3b9956fr This dissertation develops a cohesive framework for diagnosing chiral spin liquids (CSLs) in microscopic electronic models, employing as a primary tool the density matrix renormalization group algorithm (DMRG) on finite cylinders. The central results concern the half-filled triangular-lattice Hofstadter–Hubbard model at a flux per plaquette Φ△ = π/2. On long YC cylinders we find an integer quantum Hall (IQH) band insulator at weak U/t, a CSL at intermediate U/t, and evidence consistent with a continuous IQH→CSL transition near Uc/t ≃ 11.5. Identification is multi-pronged: the ES shows a single chiral branch in the CSL versus counter-propagating structure in the IQH; boundary-twist charge pumps are quantized in IQH and vanish in the CSL, while opposite twists for ↑, ↓ yield a finite neutral response; and entanglement entropy fits versus circumference give a topological contribution compatible with a bosonic ν = 1/2 CSL, accounting for finite-entanglement systematics. Critical behavior... https://escholarship.org/uc/item/3b9956fr Mon, 15 Dec 2025 00:00:00 +0000 Miksian Magaldi, Rafael https://escholarship.org/uc/item/1qm8r975 This dissertation argues that understanding play requires embodied participation rather than detached observation. To study play, one must play. Through thousands of hours of participant observation and autoethnographic engagement within karaoke venues across Orange County, California, and beyond, I examine karaoke as a play medium using play as a method of inquiry. By engaging directly, authentically with the karaoke, and within my community, I was able to pursue a search for the play itself, a phenomenological engagement with play as lived experience rather than abstract theory. As part of this inquiry, I contend that karaoke is a uniquely revealing activity for exploring the possibilities of play. Karaoke’s multilayered frames, game, show, activity, and party, operate not as ontological categories but as dynamic states of appropriated play media reframed within a single ecology of karaoke play, each organizing distinct rhetorics of procedure, performance, leisure, and affect.... https://escholarship.org/uc/item/1qm8r975 Sun, 14 Dec 2025 00:00:00 +0000 Bryan, Jeffrey https://escholarship.org/uc/item/84t9w1hc Among the various strategies for mitigating CO2 emissions, carbon capture and utilization (CCU) offers dual benefits of greenhouse gas reduction and productive use of captured carbon. Ammonia-based CO2 capture represents a promising CCU pathway for post-combustion applications, enabling both carbon sequestration and the formation of nitrogen-rich compounds suitable for fertilizer production.This dissertation investigates the gas-phase reaction between CO2 and NH3 at low temperatures and atmospheric pressure, under concentrations representative of flue gas. The study integrates kinetic simulations, laboratory-scale continuous-flow reactor experiments and product characterization; with the objective of measuring CO2 capture, capture efficiency, and product composition. CO2 capture is quantified using a nondispersive infrared (NDIR) analyzer, suspended particles are measured and quantified using a light scattering particle analyzer (ORION), and the collected solids are characterized... https://escholarship.org/uc/item/84t9w1hc Fri, 12 Dec 2025 00:00:00 +0000 Santamaria, Jose