62588 research outputs found
Sort by
Collaborative Spoken Word Poetry with AI
No full textStructure for leading an in-class creative activity, based in iterative theatrical devising and driven by group collaboration with generative AI.This Lehigh AI Project provides a structure for a creative activity in order to experiment with and reflect upon the potential of AI for artistic collaboration. Over the course of a 75-minute in-class exercise, students (or other participants) engage in an iterative artistic process with text- and image-generating AI applications. Over a sequence of steps, students develop a creative work that incorporates poetic text, digital imagery, and embodied performance. The process combines individual effort and group decision-making, entangling both in artistic collaboration with AI. The exercise concludes with a presentation of the work to their peers and a reflective discussion about creation, authorship, and originality. The design of this exercise is modeled after general devising techniques in theatrical practice as well as the \u27cut-up\u27 method popularized by William S. Burroughs. This project presents a structured walkthrough of the creative activity, documenting the instructions, reasoning, and supplemental notes for each step in order for an instructor to guide the class in AI artmaking and reflection.</p
Exploring the Integration of AI into Journalistic Pre-Professional Workflows and Its Impact on Co-Intelligence - Project Summary
No full textThis project explores how pre-professional journalists integrate AI into workflows, assessing its impact on ethics, creativity, and productivity.This research examines the integration of artificial intelligence (AI) into the workflows of pre-professional journalists and explores its implications for journalistic ethics. With a focus on human-AI employment, the study seeks to understand how AI tools are perceived, adopted, and used by student journalists. Participants from a collegiate student newspaper will engage in an survey to first better understand the newsroom attitudes and culture around AI knowledge and engagement. The study will analyze how varying levels of familiarity with AI and general technology savviness influence the integration of AI into journalistic workflows, and how these factors affect the perceived quality and ethics of the resulting work.
Following that, a small subsection of students will participate in a one-hour observational think-aloud session. During these sessions, participants will verbalize their cognitive processes and decision-making while using AI tools to assist with writing tasks. This approach will capture not only the practical application of AI but also the ethical considerations that arise as journalists interact with these technologies.
By comparing AI-assisted writing to traditional writing methods, the research will assess the cognitive impact of AI and explore its potential as an augmentation tool, rather than a replacement for human journalists.
Ultimately, the research aims to provide insights into the sustainable, ethical implementation of AI in newsrooms, offering guidelines that preserve journalistic integrity while enhancing productivity and creativity. These findings will contribute to the broader conversation on the evolving relationship between AI and journalism, helping to shape future best practices for ethical AI adoption in the field
High dimensional behavior of time constant in the first-passage percolation model
No full textThe percolation model is a foundational tool for understanding stochastic processes in graph structures, where vertices or edges in a lattice are randomly occupied to form clusters representing network connectivity. This model has broad applications, from material science to epidemiology, by capturing how processes evolve through random media. Building on percolation theory, the first-passage percolation (FPP) model incorporates a temporal element by assigning random passage times to edges, enabling the study of minimal traversal times across networks in high-dimensional spaces.This dissertation studies the high-dimensional behavior of the time constant in FPP, a measure representing the time required to traverse a unit distance across various directional vectors. We begin by analyzing the asymptotic behavior of the time constant along the sub-diagonal and the general directions when the passage time distribution is Exponential-like. Then, we explore the geometric properties of the limit shape and prove that its boundary does not contain any 2epsilon-flat edge in the axis direction if the dimension is sufficiently high. We employ a generating function approach to study the time constant for the nearly-diagonal directions.
We further derive the order of the time constant along the sub-diagonal direction when the passage times follow a Gamma-like distribution, thereby extending the work of Lu.
Finally, applying Dhar\u27s cluster exploration method to the case of exponentially distributed passage times, we find that the non-backtracking first visiting time to the associated hyperplanes is concentrated, evidenced by the in-probability convergence and the moment estimates.</p
Incorporating Large Language Models in Public Health Education - Project Summary
No full textThe goal of this assignment is to produce student generated epidemic models that can address a healthcare issue using—entirely—ChatGPT.There is an immediate need for public health practitioners to incorporate mathematical modeling into infectious diseases decision making. This is demonstrated by a continued increase in the use of forecasting models by state and federal health agencies such as the CDC. However, the typical public health curriculum does not require mathematical modeling and programming of epidemic models. In my class, Outbreak Science and Public Health forecasting, one week will be devoted to the use of Large Language Models to construct epidemic models that can incorporate temporal public health decisions. The only restriction is that students ask the LLM to construct a compartmental model, a typical model used in infectious disease modeling. Students will be assessed on the feasibility of the model, the ability of the model to include practical public health decisions or interventions, and the student\u27s ability to apply what they have learned in class to describe the model to others. In addition to the instructor, the Chief of the Division of Infectious Diseases and Chief Infection Control and Prevention Officer from the Lehigh Valley Health Network have agreed to evaluate student proposals. This exercise represents a paradigm shift in public health education and how LLM tools can be used to construct epidemic models that are impactful and improve evidence-based public health decision making.</p
Type II Flux Compactifications and Orientifold Intersections
No full textIt is well known that the universe is expanding, seemingly driven by anear-constant scalar potential. Much work has been done attempting to
construct such potentials using flux compactifications in string theory se-
tups. This thesis will discuss a nongeometric model and its associated scalar
potential. It will then discuss and classify local minima of the scalar poten-
tial. Certain ingredients in the model, namely D-branes and O-planes, are
discussed in further detail. In addition to their application in flux compact-
ifications, intersections of these objects are used when building models of
particle physics. These intersections will be discussed. It will be shown that
supergravity solutions for two perpendicularly intersecting localized sources
in flat space do not exist for a generic diagonal metric Ansatz. Sections 2-6
are based on work from[1]. Sections 7-9 are based on work from[2].
1
</p
An Integrated Language and Memory Approach to Conceptual-Lexical Retrieval
No full textLanguage production and memory researchers both study the long term effects of semantic retrieval. Both fields have routinely observed the same semantic interference phenomenon, in which generation from memory impairs subsequent access to semantically related representations. However, to date there has been very little communication across these literatures and our understanding of conceptual-lexical retrieval is fragmented across fields. In this dissertation I propose that the language production and memory literatures are ostensibly studying different aspects of the same semantic interference phenomenon and that an integration of the separate but complementary accounts of conceptual-lexical retrieval is long overdue. Thus the goal of this dissertation was to lay the groundwork for further integrative efforts. I developed a verbal integrated incremental learning account of conceptual-lexical access to serve as a blueprint for future model development and tested a subset of the resulting hypotheses across three empirical components. In Component 1, Retrieval-Induced Semantic Interference, I developed a modified version of the retrieval practice paradigm, which would serve as the primary methodology for the dissertation. I shed light on how interference unfolds over time by implementing an internally structured continuous naming design in each phase, and the effect of past generative activity was assessed in the final phase by comparing related picture naming latencies to unrelated experimental controls. The results confirmed that cumulative interference occurs during a standard memory manipulation (category-stem cued retrieval) and that this interference carries forward to a subsequent phase of continuous picture naming. These findings thus use a fine-grained temporal analysis to highlight the common basis of adaptive learning processes and show that related retrieval affects subsequent lexical access, regardless of the specific task. Component 2, Adaptive Learning After Semantic Interference: Does Interference Promote Enhanced Relearning?, explored the memory claim of enhanced relearning (Storm et al., 2008; Hulbert & Anderson, 2015; Ritvo et al., 2019) in which relearning after semantic interference results in heightened accessibility compared to both initial levels and unpunished controls. I examined the unfolding of learning and relearning over generation, relearning, and assessment phases and found evidence for restoration but not enhancement of accessibility for past competitors. This finding calls into question the generality of enhanced relearning and suggests that it may chiefly apply to novel learning contexts rather than acting on preexisting conceptual-lexical networks. Component 3, Durability and Locus of Retrieval-Induced Semantic Interference, sought to investigate the durability of retrieval-induced semantic interference and the respective contributions of conceptual versus conceptual-lexical adaptive learning. Experiments 3A and 3B used a continuous picture naming procedure in two phases separated by a short or overnight retention interval to assess the durability of semantic interference as well as the possible role of sleep-based consolidation. In contrast to the prior two components, I only found retrieval-induced semantic interference after nocturnal sleep, and this interference was only manifest as increased production errors. Experiments 3C and 3D explored the origin and locus of retrieval-induced semantic interference. These experiments introduced a novel variation of a pre-lexical picture matching task to elicit conceptual-level as distinct from conceptual-lexical adaptive learning. To assess the impact of this learning, the final assessment phase involved either generative picture naming or a continuation of picture matching. In both experiments, the initial picture matching phase produced cumulative facilitation over each related generation. In Experiment 3C, the following picture naming phase produced cumulative interference, but there was no effect of the previous phase. Experiment 3D did not find significant cumulative processes nor cross-phase facilitation in the final matching assessment phase. The results of this component indicate that more work is needed to ascertain the durability of interference and that interference may only manifest when lexical retrieval is a required stage of processing. The dissertation concludes with an update of the blueprint for an integrated memory and language production account of conceptual-lexical retrieval. Together, the present work shows that assuming a combined memory-language approach to the study of retrieval-induced semantic interference is not only illuminating, but necessary to fully characterize the dynamics of the conceptual-lexical interface. </p
Navigating the Corporate Sustainability Reporting Landscape: Evaluating the Impact of ESG Disclosure on Corporate Decision-Making
No full textThis study examines the relationship between corporate greenhouse gas (GHG) emissions and financial performance, the prevalence of various climate reporting frameworks, and the influence of leadership involvement on ESG (Environmental, Social, Governance) reporting scope. Using data from 15 companies, this research investigates whether a correlation exists between GHG emissions and financial performance, identifies the most frequently used reporting frameworks among GRI, SASB, GHG Protocol, and TCFD, and assesses how leadership roles such as Chief Sustainability Officer (CSO) impact the scope of ESG reporting. The findings reveal a significant relationship between GHG emissions and financial performance, indicating that oftentimes reduced emissions are associated with increased revenue. The study also finds that SASB and GHG Protocol are the most widely adopted frameworks, reflecting a trend towards greater transparency and accountability in sustainability reporting. Additionally, the presence of a CSO is linked to reduced GHG emissions. This research underscores the importance of integrating robust reporting frameworks and dedicated sustainability leadership to drive environmental and financial success. This study contributes to the understanding of how sustainability reporting practices influence corporate performance and provides actionable insights for advancing corporate sustainability efforts. </p
NON-VIRAL CRISPR GENE EDITING IN MESENCHYMAL STEM CELLS FOR IMMUNOMODULATORY CELL THERAPY
No full textMusculoskeletal disorders are extremely prevalent and a leading cause of chronic pain and disability. Immunomodulatory cell therapy has potential to improve treatment of musculoskeletal disorders such as osteoarthritis and rheumatoid arthritis. Mesenchymal stem cells (MSCs) function in musculoskeletal tissue repair and have immunosuppressive properties. However, the in vivo survival and differentiation of transplanted cells are limited, in part due to inflammatory microenvironments associated with tissue injury. Here, we use CRISPR tools to activate the expression of anti-inflammatory cytokines in MSCs to modulate innate immune cell phenotype. Specifically, interleukin-4 (IL-4) and interleukin-10 (IL-10) are targeted due to their known role in pro-regenerative polarization of macrophages. Macrophage polarization markers are evaluated in vitro after treatment with gene edited MSC trophic factors. The results assess the therapeutic potential of CRIPSR activation methods to increase the immunomodulatory function of MSCs. </p
MICROFLUIDIC SYSTEM FOR IN VITRO ASSESSMENT OF VASCULAR CELL RESPONSE TO REGENERATIVE NANOTHERAPIES
No full textAbdominal aortic aneurysms (AAAs) are localized, rupture-prone expansions of the abdominal aorta wall. In this condition, structural extracellular matrix (ECM) proteins of the aorta wall—specifically elastic fibers and collagen fibers, which impart elasticity and stiffness, respectively—are gradually degraded by overexpressed matrix metalloproteinases (MMPs) following an injury stimulus. We aim to deliver therapeutics to the AAA wall using polymer nanoparticles (NPs) capable of stimulating on-site matrix regeneration and repair. This study aimed to determine how NP shape and size impact endocytosis and transmigration through the endothelial cell (EC) layer from circulation into the medial layer of the AAA wall.First, rod-shaped NPs were created by mechanically stretching poly(lactic-co- glycolic acid) (PLGA) NPs embedded in a polyvinyl alcohol (PVA) film, with longer rod- shaped NPs formed depending on the degree to which the PVA films were stretched. A live/dead assay revealed that our PLGA NPs are safe and do not cause cell death. Immunofluorescence staining showed that cytokine activation induces endothelial dysfunction in ECs by increasing the expression of the inflammatory marker Integrin αVβ3 and decreasing the expression of the adhesion protein vascular endothelial (VE)-cadherin. We demonstrated that this disruption enables greater EC uptake and translocation of NPs. Fluorescence studies indicate high endothelial transmigration and endocytosis with rod- shaped NPs in cytokine-activated ECs compared to healthy control cells, supporting the potential advantages of using higher aspect ratio (AR) NPs for accumulation at the aneurysm site. We also showed that the mechanisms of NP transmigration across an activated EC layer depend on NP AR.Further studies explored NP transport dynamics between cytokine-activated endothelial cells (ECs) and underlying aneurysmal smooth muscle cells (aSMCs) in a simulated aneurysm wall environment, using a microfluidic device. The study examined the effects of NP shape (spherical vs. rod-shaped) and Doxycycline (Dox) release on aSMC and EC phenotypes, as well as markers of elastic matrix homeostasis, MMP activity, and cytokine expression. Key findings revealed that rod-shaped NPs transmigrate more effectively through activated ECs, accumulate earlier in aSMCs, and predominantly remain in the extracellular space. This distribution enhances Dox release, improving aSMC phenotype, promoting matrix homeostasis, and reducing MMP and cytokine activity, indicating potential therapeutic benefits.These results highlight the potential of using NP shape as a modality to enhance NP permeation into the aneurysm wall. This study also contributes to our understanding of the mechanisms likely engaged by NPs in penetrating the endothelial lining of aneurysmal wall segments for enhanced delivery of therapeutics to support better outcomes for patients with AAAs.</p
Phase Transitions of Rayleigh-Taylor Instability in Elastic-Plastic Soft Materials
No full textRayleigh-Taylor instability (RTI) is originally a hydrodynamic instability and the majority of the literature focuses on RTI between two Newtonian fluids. RTI in Newtonian fluids occurs when two different fluids sharing an interface are accelerated toward each other perpendicular to their shared interface. The small perturbations at the interface grow rapidly into spikes which eventually manifest themselves into mushroom-shaped structures as the instability develops. However, when the driving conditions (i.e. acceleration, pressure, temperature, etc.) are extreme, RTI can be achieved when at least one of the materials is an EP solid or a non-Newtonian material and there are various industrial geological and astrophysical applications of RTI in solids. In a more generalized definition, RTI can be observed between two materials with different densities when they are accelerated toward each other while their density and pressure gradients are in opposite directions and perpendicular to their shared interface. In the case of EP solids or non-Newtonian media, the mechanical strength of the material (i.e. the yield strength and elasticity) mitigates the RT growth. In the case of Hookean solids (i.e. metals), the driving conditions to initiate the instability alter the mechanical properties of the solids where they transition into a plasma state and achieve plastic flow. The very first experiments studying RTI in EP solids utilized explosives in order to achieve the required extreme conditions to initiate the instability for a metal, where under these conditions the mechanical properties of the solid became ambiguous, and the measurement process for the instability was extremely challenging. In order to study RTI in the presence of mechanical strength, and with more ease, soft materials were utilized by numerous experimentalists where the mechanical behavior of the non-Newtonian soft materials was analogous to the EP solids in extreme driving conditions. Utilization of soft materials to study RTI in EP materials provided more ease and accuracy in material characterization and measuring process of the RT growth. In this dissertation, RTI in EP solids is studied with a non-Newtonian soft solid (mayonnaise) in tandem with air, where the mechanical properties of mayonnaise are measured with various rheological measurements. Its time-dependent nature and shear-thinning characteristics require various rheological measurements which must be chosen according to the application (i.e. the type of the experiment) it is going to be used. In this dissertation, three different types of experiments are presented in order to study the three distinct phases of RTI in EP solids: (a) pure elastic regime, (b) stable plastic regime, and (c) instability regime as well as the transition criteria between each of them. The experiments in this dissertation are conducted with the rotating wheel experimental setup, where the centrifugal acceleration harnessed from the horizontally rotating disk provides the driving conditions to initiate RTI between the soft material and air. The centrifugal acceleration, which is the driving acceleration for RT growth, increases linearly with a constant acceleration rate that can be altered, which allows for investigation of the effects of the acceleration rise times or in other words the acceleration rate. As the driving acceleration increases, at first, the perturbation amplitude grows elastically until it reaches a limit where the soft material begins to exhibit plastic behavior which also signifies the first dramatic decrease in the mechanical strength while still behaving predominantly solid-like. Once the transition to the stable plastic regime is achieved the perturbation cannot fully recover the sustained RT growth anymore despite it can still recover a significant amount of growth once the disk is brought to rest. On the contrary, after transitioning to the stable plastic regime, if the sample is kept accelerated it reaches a point where the yield strength of the material is exceeded, where the material yields and begins to behave predominantly viscously, which makes the perturbation to transition to the instability regime where any growth sustained is not reversible anymore. The transition criteria and the growth patterns of each RTI regime are studied with time-dependent acceleration profiles, for different perturbation geometries. The effects of the initial perturbation dimensions, the acceleration rate, the yield strength, the shear modulus, and the material density are investigated in detail for each RTI phase and the transitions between them. Furthermore, non-dimensional parameters are developed in order to allow for a broader interpretation of the experimental data presented to aid computational and analytical modeling as well as forming a basis for future experimental studies utilizing soft materials