106261 research outputs found
Sort by
REVISITING FIRM-SPECIFIC EXPERIENCE: HOW MANAGERS SHAPE INTERNAL JOB TRANSITIONS
No full text53 pagesOrganizations often fill jobs through a combination of internal moves (e.g., promotions and transfers) and external hires. While internal hires are generally expected to benefit from their firm-specific experience upon entering new roles, this study examines how that experience may also create challenges—particularly in how managers handle internal transitions compared to external ones. While organizational policies often request that managers handle job transitions consistently across all new hires, I draw on Mintzberg’s (1971) theory and argue that managers—under time pressure—are more likely to take shortcuts with internal hires. Specifically, internal hires’ firm-specific knowledge tends to make these shortcuts appear more readily available and effective. I identify three aspects of the transition and predict that internal hires receive less formal evaluation, less structured onboarding, and are more often placed into roles with unfamiliar responsibilities. Using survey data from approximately 600 new hires and 200 hiring managers at a large healthcare organization, the findings support differences in onboarding and job assignments, while no significant difference is observed in the use of formal evaluation practices
Generative AI for Digital Pathology: Advancing Diffusion, Vision-Language, and Safety-Aware Models
No full text229 pagesThis dissertation investigates the transformative intersection of Artificial Intelligence (AI) and Digital Pathology, presenting both a comprehensive review and a series of methodological contributions that highlight the synergistic potential of Generative AI (GenAI) in this domain. It spans a broad range of tasks—from synthesizing histopathology images and modeling pathologists’ decision-making processes to generating synthetic diagnostic text for enhancing vision-language model (VLM) representation and alignment. The work further explores the integration of safety and context awareness into generative frameworks to significantly improve model performance and trustworthiness. Chapter 1 provides an in-depth review of recent advancements in AI for histopathology, encompassing both foundational techniques and state-of-the-art developments. It also surveys publicly available datasets that have been instrumental in driving progress in the field. Chapter 2 addresses the challenge of uncertainty in mitotic figure detection—a clinically significant yet inherently subjective task. To model this uncertainty, a probabilistic diffusion model is introduced to synthesize cell nuclei undergoing mitosis. This generative framework identifies key visual features that inform diagnostic decisions and offers a novel interpretability tool to support pathologists. Chapter 3 explores the application of VLMs in digital pathology, focusing on limitations related to the scarcity of large-scale image-caption datasets and the sensitivity of zero-shot classification to prompt design. To address these challenges, language rewrites generated by a large language model (LLM) are used to enrich an existing dataset, resulting in substantial performance gains. Additionally, a novel context modulation layer is proposed to enhance image-text alignment. Building on this foundation, Chapter 4 introduces an extended training paradigm for digital pathology VLMs that leverages synthetic captions. Alongside the standard CLIP loss, a new text-only loss is formulated using stochastic cosine similarity between original captions and various LLM-derived concepts (cell, organ, disease). The training also incorporates “safety-aware” incorrect captions that are LLM-generated texts that simulate clinically relevant misinterpretations to improve model robustness. The similarity between original and incorrect captions is explicitly penalized, guiding the model to better distinguish between meaningful and misleading descriptions. This strategy leads to significant improvements in both zero-shot classification and image-text retrieval tasks. Collectively, this dissertation advances the role of GenAI in digital pathology by tackling key challenges related to data scarcity, diagnostic uncertainty, and model reliability. The proposed methods contribute to the development of more robust, interpretable, and clinically trustworthy AI-assisted diagnostic systems
MULTI-MODAL SYNCHROTRON-BASED INVESTIGATIONS INTO ORGANOMINERAL INTERACTIONS IN REDOX-DYNAMIC WATER SYSTEMS
No full text218 pagesBiogeochemical dynamics of manganese (Mn) and iron (Fe) in redox-dynamic environments play important roles in regulating water quality in natural and nature-based systems. Manganese is a contaminant of concern in drinking water systems, and its mobilization and speciation in redox-stratified water columns is regulated by hydrological patterns that are changing in a warming climate. Interactions between Mn or Fe oxide minerals and organic matter can play an important role in either stabilizing or destabilizing organic carbon, depending on the redox state and transformations of Mn and Fe, with implications for the bioavailability of carbon to fuel microbial activity in denitrifying bioreactor systems. The work described in this thesis integrates field-scale observations, geochemical modeling, and multi-modal synchrotron techniques to elucidate how redox-dependent dynamics of Mn and Fe influence water quality in systems including drinking water reservoirs and denitrifying biofilter systems. In a weakly stratified drinking water reservoir, prolonged stratification during drought conditions was found to enhance manganese release, complicating treatment process. We monitored Mn in the Ithaca Reservoir during two contrasting years: a wet year in 2021 and a dry-to-moderate drought year in 2022. The results showed that decreased streamflow during the dry year increased stratification and raised Mn concentrations in the hypolimnion. The contrasting conditions allow us to study how stratification impacted the mobilization and precipitation pathways for Mn and we found that mixed calcium-manganese carbonates, rather than MnCO3, may contribute to Mn cycling. These findings emphasize the role of hydrological and geochemical interactions in controlling manganese solubility and water quality under changing climatic conditions. In nature-based systems like denitrifying woodchip bioreactors (WBRs), redox-active iron and manganese minerals were shown to enhance the oxidative decomposition of woody lignocellulosic biomass, releasing bioavailable carbon that supports microbial denitrification. The interplay between dissolved oxygen exposure and mineral-driven oxidative processes significantly improved wood degradation and nitrate removal efficiency, providing a new perspective on how organo-mineral interactions impact coupled carbon and nutrient cycling in the environment. Furthermore, experiments with mineral-coated woodchips revealed contrasting effects of iron and manganese under varying redox conditions: oxic-anoxic cycling accelerated lignocellulose breakdown, while under strictly anoxic conditions minerals stabilized organic matter and slowed decomposition. In conclusion, these findings advance the understanding of Mn cycling in reservoirs and the dual roles of Mn and Fe in enhancing the efficiency of wood degradation and nitrate removal in bioreactors. They provide actionable insights for optimizing reservoir management and denitrifying bioreactor performance, addressing key environmental challenges in water quality improvement, nutrient cycling, and carbon cycling.2026-06-1
Data from: Chiroptical Strain Sensors from Electrospun Cadmium Sulfide Quantum-Dot Fibers
No full textPlease cite as: Chandi Hansadi Jayamaha Mudalige, Thomas Ugras, Kirt Page, Richard Robinson, Tobias Hanrath, Larissa Shepherd. (2025) Data from: Chiroptical Strain Sensors from Electrospun Cadmium Sulfide Quantum-Dot Fibers [dataset] Cornell University Library eCommons Repository. https://doi.org/10.7298/d8f3-q379These files contain data supporting the results found in figures reported in Jayamaha et. al., "Chiroptical Strain Sensors from Electrospun Cadmium Sulfide Quantum-Dot Fibers". In this work, we report, highly sensitive and reversible strain sensors from chiroptically active fibers embedded in elastomeric films. The chiroptical response on stretching is indicated quantitatively as reversible changes in magnitude, spectral position (wavelength), and sign in circular dichroism (CD) and linear dichroism (LD) signals and qualitatively as a prominent change in the birefringence features under cross-polarizers.Cornell Center for Materials Research (CCMR), supported by the NSF MRSEC program (DMR-1719875).
Functional Materials Beamline (FMB) at the Cornell High Energy Synchrotron Source (CHESS), supported by the Air Force Research Laboratory (AFRL) under agreement FA8650-22-2-5200.
Additional support was provided by NSF grant CHE-200358
Overexpression of Sorghum sHSP and heat tolerance in Arabidopsis thaliana
No full textIntroduction: Climate change is driving shifts in weather patterns, leading to an increased frequency of sudden heat wave events that pose major risks to staple food crops. Understanding the mechanisms plants use in thermal stress mitigation becomes crucial for agricultural sustainability. Plants physiologically adapt to novel stressors primarily through a large family of heat shock proteins (HSPs), which are classified by their approximate molecular weight into Hsp100, Hsp90, Hsp70, Hsp60, and small heat shock proteins (sHSPs, <60 kDa). Small heat shock proteins typically exist as multimeric complexes within plant cells, remaining inactive until external stress triggers their ATP-independent dissociation. Upon activation, the residues of sHSPs bind to exposed hydrophobic regions on misfolded proteins, preventing protein aggregation, while larger chaperones facilitate refolding or degradation. Small heat shock proteins were selected for this study due to their consistent expression in maturing seeds and their protective function during heat stress and desiccation events. This study investigates the transgenic overexpression of the 16.9 kDa sHSP from the highly heat-tolerant staple grain Sorghum bicolor (SsHSP) in the model organism Arabidopsis thaliana. The objective of this study is to understand the impact of SsHSP overexpression on thermotolerance, providing insight into cross-species enhancement of heat stress resilience in plants
Cultural Embeddedness and Economic Motivations in Smallholder Cattle Farming: A Mixed-Methods Study from KwaZulu-Natal, South Africa
No full textThis study examines the cultural and economic motivations of cattle farming among smallholder farmers in KwaZulu-Natal, South Africa. Drawing on Polanyi's (1944) embeddedness theory and Granovetter's work (1985) on socio-cultural values' influence on economic behavior, it investigates how cattle serve dual roles as both cultural assets and economic resources. Survey data from 59 smallholder farmers in rural and peri-urban locations was analyzed using an ordinal embeddedness score measuring cultural versus economic motivations. Ordinal logistic regression revealed that geographic location significantly predicts cultural embeddedness (OR = 0.219, p = 0.042), with 83% of rural farmers displaying high cultural embeddedness compared to 17% of peri-urban farmers. Age showed no significant influence on motivation. Notably, 41% of farmers demonstrated mixed embeddedness, expressing both cultural and economic motivations simultaneously. Cultural and economic factors alone could not explain herd size, breed selection, or farmer association participation, suggesting additional influences on these decisions. Cattle fulfill ceremonial purposes, provide social status through practices like lobola, and serve as savings and insurance. Rural farmers maintained stronger cultural motivations while peri-urban farmers expressed primarily economic motivations. This geographic pattern aligns with differences in market access and infrastructure between locations. The study demonstrates how embeddedness theory helps explain the interaction between location and farmer motivations, revealing that cultural and economic functions of cattle are balanced differently depending on geographic context
ASSESSING THE FEASIBILITY OF LOCAL TARPS AS ADDITIONAL POST-CALIBRATION TARGETS FOR FIELD NDVI TIMESERIES
No full textThe Normalized Difference Vegetation Index (NDVI) is a widely used indicator of vegetation vigor, yet its reliability in agricultural time series is limited by atmospheric and radiometric inconsistencies. This study evaluates two local correction methods using pseudo-invariant reference tarps to reduce noise in Planet Scope NDVI measurements for maize fields in Ithaca, New York during the 2023 growing season. The first approach, the Satellite Tarp Residual Method, applied residuals from linear regressions of satellite-derived tarp NDVI to adjust field NDVI. The second approach, the Spectral–Satellite Tarp Correction, predicted spectral tarp NDVI for satellite acquisition dates via linear models, computed correction terms, and applied them to field NDVI. Validation against spectral ground-truth NDVI from Elementary Measurement Areas (EMAs) employed Dynamic Time Warping (DTW) to assess temporal profile similarity. Results indicate that the Satellite Tarp Residual Method provided the smallest DTW distance (0.6064) compared to the uncorrected NDVI (0.6091), with the best performance achieved by subtracting red tarp residuals. The Spectral–Satellite Tarp Correction yielded minimal improvement and, in some cases, degraded temporal agreement due to interpolation errors. Findings suggest that temporally coincident satellite-based reference measurements offer more effective short-term corrections than predicted spectral values when spectral data are temporally sparse. Future work should integrate satellite temporal coverage with spectral spectral precision through hybrid or non-linear correction frameworks to improve NDVI time series reliability for precision agriculture
Identifying genetic determinants of Mycobacterium tuberculosis acid growth arrest by transposon mutagenesis coupled with next-generation sequencing (Tn-seq).
No full textMycobacterium tuberculosis, the causative agent of tuberculosis, remains the leading global infectious disease killer. Adaptation of Mycobacterium tuberculosis to acidic niches within the host during infection is vital to establish the disease. Here, we present a high-density transposon mutant sequencing library data set identifying genetic determinants of acid growth arrest to serve as a resource
