Other Electrical and Computer Engineering Commons^™

Biosecure-Llm Framework: Protecting Llms From Cyberbiosecurity Threats And The Case For Independent Ai Safety Governance, Xavier-Lewis Palmer, Lucas Potter, Srdjan Lesaja, Sotirios Karathanasis, Mohammad Ghasemigol

Computer Science Faculty Publications

Large Language Models (LLMs) are becoming critical infrastructure in scientific, healthcare, and governmental contexts. As frontier AI laboratories increasingly partner with government agencies, a fundamental question arises: Who should control the safety and policy-enforcement layers that constrain model behavior? Current safety mechanisms (LLM guardrails) are typically designed for generic "harmlessness" and operate by detecting semantic patterns and refusing requests. However, they are inadequate governance instruments because they cannot implement auditable, domain-specific controls tied to external regulatory policy objects (e.g., control lists or rules governing personally identifying information). Even a perfectly aligned model is not able to express institution-specific policy without …

Development Of Periodic Plasmonic Nano-Structures For Enhanced Labeled Bio-Sensing Systems, Kyle Zackary Smith

Graduate Theses, Dissertations, and Problem Reports (ETD)

The biomedical industry has seen sustained growth over the past half century, with a continually increasing demand for flexible, easy-to-use, and cost-effective tools. One large area of commercial interest has been point-of-use or point-of-care diagnostics, using optical based Lab-On-Chip (LOC) style systems. Label and label-free fluorescence detection systems are common benchtop modalities that have seen recent integration into these portable, cost-effective LOC applications. However, despite their maturity, there are still opportunities to improve device characteristics, specifically in reference to throughput, limit-of-detection (LOD), and hybrid integration (along with associated costs).

Optical research avenues at WVU have focused on improving these systems …

Fairness Without Demographic Attributes In Medical Vision–Language Models, Ahsan Habib Akash

Graduate Theses, Dissertations, and Problem Reports (ETD)

Vision–language models (VLMs) and the embedding models that underlie them have become the default representation layer for multimodal artificial intelligence. Trained by contrastive alignment over enormous, loosely curated image–text corpora, they inherit the demographic skew of that data, and they encode it in the geometry of the shared embedding space itself. The consequence is a model that performs unevenly across demographic groups even when no protected attribute is ever supplied as an input. In consumer applications this is an equity problem; in medicine it is a safety problem, because an embedding that is systematically less discriminative for one subpopulation translates …

Data Augmentation And The Reliability Of Conformal Prediction For Uncertainty Quantification In Medical Imaging, Rizwan Ahamed

Graduate Theses, Dissertations, and Problem Reports (ETD)

The safe clinical deployment of deep learning models for high-stakes medical imaging tasks requires more than high average accuracy; it requires demonstrable, per-case reliability. Uncertainty quantification (UQ) provides the missing signal that tells a clinician when a model prediction can be trusted and when a case should be escalated for expert review. Among UQ approaches, conformal prediction (CP) is especially attractive because it produces prediction sets that are guaranteed, under the assumption of exchangeability, to contain the true label with a user chosen probability, and it does so without assumptions about the model or the data distribution. This report first …

State-Dependent Queueing For Adaptive Signal Control: A Simulation-Based Performance Evaluation, Shaimaa Alseddiek, Usama Elrawy Shahdah, Hala B. Nafea, Hossam El-Din Moustafa, El-Said Ahmed Marzouk, Mohamed M. Ashour

Mansoura Engineering Journal

Urban traffic congestion persists as a critical challenge to transportation system efficiency, sustainability, and safety. Traditional queuing models utilizing fixed service rates inadequately represent the dynamic feedback between congestion and capacity in real vehicular flow. State-Dependent Queuing Models (SDQMs) address this limitation by modelling service rate as a function of queue length or density. This research advances SDQM application for adaptive traffic signal control through development of a calibrated state-dependent departure rate implemented within a microscopic simulation environment using SUMO and TraCI. Six control strategies including fixed-time, actuated, and two SDQM variants were evaluated across traffic demands ranging from undersaturated …