UC Riverside

This thesis examines how queer Chicanx artists negotiate belonging under inner-cultural control through lowrider visual culture in The Q-Sides exhibition. The first chapter, “Up in Flames: Iconoclastic Por Vida,” focuses on Manuel Paul’s mural Por Vida in San Francisco’s Mission District. I begin with a formal analysis of the three panels of the mural to show how Paul transforms traditional Chicanx imagery to represent queer intimacy, transgender visibility, and lesbian autonomy within a historically masculine cultural context. The chapter then goes into the mural's presence on Instagram, revealing how comments and reactions demonstrate digital control of visibility, and uses Koritha Mitchell’s “know-your-place aggression” to interpret efforts to regulate queer visibility in the community. It also traces how these issues led to acts of vandalism and arson, framing the mural’s destruction within a broader history of policing the intersectional queer and Latinx bodies in public spaces. The chapter concludes by considering the afterlife of Por Vida through photography and The Q-Sides exhibition, especially Fred Rowe’s photographic print and House of MORE!, exploring how queer Chicanx artists utilize reproduction and performance to reclaim agency and preserve collective memory that confronts the erasure of queerness in Chicanx spaces. The second chapter, “Disidentifying the Lowrider: Portraits from The Q-Sides,” explores how The Q-Sides photographic series reimagines lowrider visual culture from a queer perspective. I take photography, portraiture, and performance into account in how the photographs aim to reclaim visibility for queer and trans Chicanx and Latinx subjects within a traditionally heteronormative and masculine cultural context. I analyze two photographic prints from the twelve photographs in The Q-Sides, Volume 10 and Volume 3, by artists Vero Majano, Brown Angel, and Kari Orvik, to demonstrate how José Esteban Muñoz’s concept of disidentification embeds itself within the series. As with working on a lowrider, the artists rework the familiar visual language of East Side Story mixtape covers and modify imagery associated with heteronormative Chicanx identity into symbols of queer belonging. Chapter Two also contextualizes the photographs within the broader histories of criminalizing lowriders and the policing of Chicanx fashion. By looking at local queer and trans community members as documentary-style portraits, The Q-Sides broadens the visual archive of queer Chicanx photography. This series of photographs uses the medium of photography as a tool for collective affirmation for individuals of these intersecting identities. Together, these chapters demonstrate that queer Chicanx visual culture does not merely respond to complexities of visibility but actively reshapes the conditions of visibility imposed upon it. Artists in The Q-Sides use various mediums such as muralism, photography, and drag to assert queer belonging in lowrider visuals. Their art demonstrates that visibility is never neutral and interlocks with issues within systems that continue to control bodies over race and gender. By exploring how queer Chicanx artists address these challenges, this thesis contends that visual culture serves not only as a form of representation but also as a means of survival by self-representation in modern California.

The ability to simulate biological phenomena across a wide range of scales has been steadily improving for decades. The atom-level approach known as Molecular Dynamics (MD) has proven remarkably useful for rationally engineering proteins, studying peptide binding, and detailing protein folding and unfolding pathways. Though the utility of MD is difficult to overstate, its simulations are expensive, relying on integrating equations of motion across exceedingly small timescales. This creates a large discrepancy between the timescales accessible to MD, on the order of microseconds, and those of biomolecular phenomena, on the order of milliseconds to seconds. The work presented here serves as a deep dive into studying patterns in MD data, formulating approaches that extrapolate phenomena from short simulations and construct a platform by which new samples may be efficiently generated. The first half of this dissertation focuses on analyzing MD data from complicated molecular machinery, beginning with a deep dive into a CRISPR Associated Transposon (CAST) system called Cascade-TniQ. A set of approaches is implemented to unveil subtle binding sites that influence conformational changes and to demonstrate synchronous motions, yielding a detailed mechanistic basis for the thermodynamics of the DNA-binding process. A second deep dive into CRISPR/Cas13a inspired the development of a method studying cross-talk efficiency known as the Signal-to-Noise Ratio, detailing the dynamic allostery responsible for activating the complex upon DNA binding. The second half introduces a neural network, SurfNet, informed by these deep dives into Cascade-TniQ and CRISPR/Cas13a. SurfNet learns a function of atomic coordinates known as a collective variable (CV) to guide MD-based enhanced sampling methods. These methods efficiently explore the space of CV outputs but require good CVs, an unsolved problem for most proteins. SurfNet was validated on the prototypical test systems of Alanine Dipeptide and Chignolin using On-the-fly Probability Enhanced Sampling (OPES). Ultimately, the work presented here is meant to assist the sample generation and analysis journeys that computational biophysicists embark on to learn as much as we can about the brilliant molecular machines known as proteins.

Despite a two-fold increase in doctoral attainment among Latinas in the U.S. over the past two decades, disparities in completion persist compared to other racial-ethnic groups. A lack of social support in academia contributes to isolation and intentions to drop out among Latina doctoral students, highlighting the need to better understand and address their experiences. Guided by Latina/o Critical Theory, Community Cultural Wealth, and Ecological Systems Theory, this dissertation examined how Latina PhD students in U.S. social science programs navigate relationships with mentors, advisors, and peers, focusing on support and conflict. It also explored their use of social media for support and their interactions with institutional resources.This study addressed five research questions related to culturally responsive mentoring, advisor conflict navigation, peer support and conflict resolution, online social support, and institutional support. Using a multi-method design, Study A collected survey data from 214 Latina PhD students, and Study B conducted plática-style focus groups with 40 first-generation Latina PhD students to examine the mechanisms and cultural factors shaping relationship dynamics in academia. Notably, participants reported limited use of social media and minimal institutional programming to support relationship-building, and thus, these areas were examined descriptively. Findings highlight the complex, culturally embedded ways in which Latina PhD students navigate academic relationships. For mentoring, three themes emerged: adaptation and resources, relational mentoring, and critically conscious mentoring. Advisor conflict navigation was shaped by cultural dissonance, persistence, and resistance, and navigating whiteness and belonging. Peer support revealed five key forms, including academic mentoring, co-processing of racialized and inequality-based experiences, emotional support, companionship, and shared identity and understanding. Additionally, online emotional support was perceived as significantly higher than informational support. These findings offer insight into how Latina PhD students persist and navigate systems not designed with them in mind and contribute to the existing literature by demonstrating that relationship navigation is a dynamic, culturally grounded process shaped by social networks, cultural values, and hierarchical structures within academia.

I study the problem of partial identification in the presence of nuisance functions in terms of moment inequality models in three dimensions. First of all, I discuss the inference for partially identified parameters of interest in unconditional moment inequalities while there are point identified nuisance functions. The parameters of interest are partially identified by the confidence set of the true parameters of interest that satisfies the uniform asymptotic size control property. Secondly, I apply the method to develop a novel structure model to infer the partially idenfied interaction effects of radio stations in the U.S. while the conditional choice probabilities serve as the nuisance functions. Our empirical results indicate negative interaction effects among different radio commercial markets. In the third dimension, I discuss the inference for partially identified parameters of interest in conditional moment inequalities while there are point identified nuisance functions. The necessary parameters of interest are also partially identified by the confidence set of the true parameters of interest that satisfies the uniform asymptotic size control property.

Cardiac glycosides (CGs) are toxic compounds produced or sequestered by certain plant and animal species for defense against enemies. CGs derive their toxicity to animals from inhibiting the Na+/K+-ATPase (NKA), a critical ion pump necessary for maintaining cell osmotic equilibrium and membrane potential. While some organisms that carry or otherwise encounter CGs evolved resistance through target site insensitivity (TSI) via substitutions in the NKA’s alpha subunit, others appear to have evolved alternative resistance mechanisms. For instance, milkweed herbivores such as the monarch butterfly and one of their natural enemies, the parasitic nematode Steinernema carpocapsae, exhibit TSI. However, the congener Steinernema feltiae shows partial CG resistance without some of the large-effect TSI substitutions, suggesting an alternative protective strategy. Likewise, some resistant insects, such as the milkweed tussock moth Euchaetes egle, lack TSI substitutions entirely. These observations point to the existence of possible additional CG resistance mechanisms, such as toxin efflux via ABC transporters or metabolic detoxification. Our study aims to uncover such alternative resistance mechanisms through studying wild isolates of Caenorhabditis elegans in the Caenorhabditis Natural Diversity Resource (CaeNDR). By conducting toxicity assays with ouabain, a model CG, we have identified resistant C. elegans strains. We are applying genome-wide association studies (GWAS) to determine genes and genetic variants linked to CG resistance. Our findings will shed light on the genetic architecture underlying CG resistance and emphasize the need to explore resistance mechanisms beyond TSI. Our research has broader implications for chemical ecology, evolutionary biology, and drug resistance studies.

Eukaryotic chromatin is packaged within the nucleus in the form of a highly organized structure, with nucleosomes as basic units. Each nucleosome is comprised of an octamer of histone proteins, wrapped by 147-base pair DNA. The chromatin environment, orchestrated by high-order nucleosome assembly, critically influences various genome activities, such as epigenetic regulation. Increasing evidence has indicated that the multivalent interaction between an epigenetic modifier or effector and the nucleosome array dictates the epigenetic states, and consequently, genome organization and gene activity. To date, due to the complexity of high-order nucleosome assembly, the structure and mechanism of epigenetic complexes have mostly been characterized at the mono- and/or dinucleosome level, which has limited our mechanistic understanding of their nucleosome interactions in a native chromatin environment. To extend this limit, we have reconstituted an epigenetically modified tetranucleosome for protein interaction. In preliminary studies, we measured the nucleosome-binding affinity of two heterochromatin factors using the electrophoretic mobility shift assay (EMSA). Our long-term goal is to deepen the understanding of how multivalent protein-nucleosome interactions contribute to epigenetic modification and genome organization in cells.

Conservation of natural species requires an understanding of organisms' geographic ranges and existential threats that occur within these ranges. A species’ rarity or commonness is ultimately determined by taxonomy which defines boundaries between closely related species. As rare species tend to be at inherently greater extinction risk than widespread ones, taxonomy interacts heavily with conservation. Similarly, hybridization strongly interacts with taxonomy because it points to limitations in species boundaries. Hybridization also interacts directly with conservation in that hybrid individuals are difficult to classify in terms of legal conservation status, and outbreeding may be beneficial or harmful to species’ persistence depending on context. Thus, the topics of conservation, taxonomy, and hybridization are heavily intertwined, but they have not been expressly studied together as a conceptual framework. To develop this conceptual framework, I use the study system of manzanitas to investigate how these topics interact at three scales of observation. At the population level, I show that detecting hybridization led to the discovery of a new species of manzanita and ongoing hybridization poses a threat to its persistence. At higher scales of observation, hybridization could only be inferred indirectly but remained a prominent consideration. At the subregion scale, I show that the genetic distinctiveness of federally endangered Del Mar manzanita could only be observed after broad sampling and considering all neighboring manzanita species because they appear to be linked by hybridization. The study defines the range of Del Mar manzanita which is limited to the Coast of San Diego County. At the bioregional scale, I evaluate the phylogenetic relationships of 88% of species in the genus of manzanitas and find at least partial support for the majority investigated species. I can identify which taxa seem well-understood and which taxa require further study such as several widespread tetraploid species. Overall, the study suggests that for diverse systems with frequent hybridization, higher scales of observation are useful for identifying general patterns and complications such as cryptic hybridization, but taxonomic revision is better reserved until cases can be evaluated at small or intermediate scales, especially at the population scale where it is easier to directly observe hybridization.

Vision–Language–Action (VLA) models have become a dominant paradigm for generalist robot manipulation, yet most are effectively memoryless: they map the current observation and a language instruction to an action without an explicit mechanism for retaining and reusing information across long-horizon, partially observable episodes. This thesis makes two contributions toward memory-augmented VLA policies. First, we unify three heterogeneous VLA backbones—pi0.5, Cosmos, and DreamZero— under a single BasePolicy interface, so the same evaluation and rollout infrastructure applies uniformly to all three. Second, we introduce DynaMem, an input-level, single-point dynamic memory injection mechanism that augments any backbone behind this interface without modifying its internal architecture or requiring backbone-specific engineering. We evaluate the resulting system with closed-loop rollouts on the RoboMemArena (RMA) benchmark. Preliminary experiments indicate that input-level memory injection improves task completion on memory-dependent manipulation tasks while maintaining compatibility with standard evaluation pipelines.Our experiments yield a careful negative result: when DynaMem is fine-tuned on top of a converged behavior-cloned π0.5 policy, task performance is statistically indistinguishable from a no-memory control, and zeroing the memory at inference time does not degrade performance—the policy learns to ignore the injected tokens. We trace this to a structural property of behavior cloning from plateaued checkpoints: the training loss can be minimized without consulting memory, so the zero-initialized memory pathway receives no gradient pressure. The injection overhead itself is negligible (under 10 ms per memory write and 12 µs per control step). We distill these findings into design guidelines for future memory-augmented VLA training.

This research focuses on the San Gabriel Valley (SGV), a region located east of Los Angeles County where there was once a growing community of Japanese farmers starting in the late-19th century. In recording Japanese American history, less emphasis is placed on surrounding local areas that tend to be overlooked in favor of dense urban areas, like the historic district of Little Tokyo, where there are more written sources and records available. This paper argues that a multiple methodological approach through oral history and quantitative data analysis can be utilized to reveal complex local histories often absent from official archives. These methodologies are examined separately, but also collectively, to uncover these less-researched histories of Japanese Americans in the SGV. The goal of this work is to provide future pathways for multimethodological research to occur and to open further dialogue that recognizes the significance of local histories within a larger national framework.

This dissertation examines the maintenance of Asturian, a minoritized Romance language of northern Spain, among a digitally engaged and largely diasporic community of speakers. Drawing on a mixed-methods design that combines an online survey of sixty-four participants with twenty-fi ve semi-structured follow-up interviews, the study characterizes a speaker profile it terms the Asturian Heritage Speaker (AHS). Its central argument is that the diasporic Asturian speaker resembles the heritage speaker of Spanish in the United States far more closely than the neofalante of Galician or the euskaldunberri of Basque, the new-speaker fi gures produced by the other minoritized languages of the Iberian Peninsula.The analysis advances through three interlocking questions. The first concerns the new speaker and literacy: because Asturian lacks the institutional infrastructure (standardization, co-official status, and systematic schooling) that has generated new speakers elsewhere, it produces speakers who acquire the language orally at home but rarely consolidate it in writing. Nearly sixty percent of participants placed themselves at the two lowest proficiency levels, and fewer than half had ever attended a formal Asturian class, which most described as an elective offered, when at all, at the contingent margins of a Spanish-centered curriculum. The second and analytical core addresses diaspora and semantic maintenance, arguing that the language is sustained lexically and affectively rather than systematically: it concentrates in the domains of family and intimacy, is carried by a semantically cohesive core of words reproduced with notable consistency across participants, and is conditioned by physical proximity to the region. The third examines co-officiality, where overwhelming support coexists with an internal structure in which opposition is confined to the lowest-proficiency participants. Reading quantitative distributions alongside open-ended justifications, the study concludes that attitudes toward the language are negotiated through competing conceptions of what Asturian is and what it is for, returning to the structural absence from which the analysis began.