Digital Commons Network^™

Benchmarking Gaslighting Negation Attacks Against Multimodal Large Language Models, Bin Zhu, Yinxuan Gui, Huiyan Qi, Jingjing Chen, Chong-Wah Ngo, Ee-Peng Lim

Research Collection School Of Computing and Information Systems

Multimodal Large Language Models (MLLMs) have exhibited remarkable advancements in integrating different modalities, excelling in complex understanding and generation tasks. Despite their success, MLLMs remain vulnerable to conversational adversarial inputs. In this paper, we systematically study gaslighting negation attacks—a phenomenon where models, despite initially providing correct answers, are persuaded by user-provided negations to reverse their outputs, often fabricating justifications. We conduct extensive evaluations of state-of-the-art MLLMs across diverse benchmarks and observe substantial performance drops when negation is introduced. Notably, we introduce the first benchmark GaslightingBench, specifically designed to evaluate the vulnerability of MLLMs to negation arguments. GaslightingBench consists of multiple-choice …

The Stars Align: Modeling User Rating Calibration With Sparse Semantic Review Features, Rodrigo Alves, Antoine Ledent

Research Collection School Of Computing and Information Systems

User ratings are often treated as comparable across users, although identical scores may reflect different experiences. We study whether ratings can be viewed as user-specific discretizations of a shared semantic continuum derived from review text. Our method maps reviews into sparse semantic features with a sparse autoencoder and learns user-specific filters for each rating level. On Amazon Electronics, the learned embeddings align along a shared low-dimensional rating axis. Users differ mainly in how they anchor and partition this continuum, while preserving its overall ordinal structure. These findings support a semantic view of calibration beyond scalar bias correction.

Not Too Early, Not All At Once: Design Tensions In Ai-Mediated Self-Disclosure In Online Dating, Pei-Hua Tsai, Tianyi Zhang, Emran Bin Elias Poh, Anthony Tang, Yung-Ju Chang

Research Collection School Of Computing and Information Systems

Online dating relies on self-disclosure, yet initial conversations are fragile: users must navigate uncertainty around timing, boundaries, and reciprocity with little shared context. While advances in AI raise the possibility of mediating disclosure, how such support might reshape the experience of early-stage relational disclosure remains underexplored. We conducted 29 semi-structured interviews to examine how daters envision AI-mediated self-disclosure in online dating. Our findings surface recurring design tensions rather than simple opportunities or risks. Participants welcomed guidance that could pace disclosure, support reflection, and reduce social awkwardness, but stressed preserving agency and authorship. They valued interpretive assistance for sense-making of ambiguous …

Hide-And-Sweep: Detecting Concealed Cameras Via Led Illumination Sweeps, Jonghyuk Yun, Jaeyoung Moon, Yunseo Park, Sean Rui Xiang Tan, Byunghyun Kim, Rajesh Krishna Balan, Jun Han

Research Collection School Of Computing and Information Systems

Hidden cameras have increasingly infiltrated hotel and Airbnb rooms, posing serious privacy risks. Detecting such cameras is challenging because they are visually inconspicuous and often embedded inside everyday objects. Even worse, existing handheld detectors are manual and also rely on single-angle illumination and hence suffer from high false-positive rates. We present SweepLED (pronounced "sweepled")1, a practical hidden camera detection system that operates on a commodity smartphone augmented with an unobtrusive LED-embedded case. SweepLED performs LED sweeping - a controlled sequence of multi-angle illumination - while the user simply holds the phone still by hand, enabling the camera to capture how …

Rc-Nf: Robot-Conditioned Normalizing Flow For Real-Time Anomaly Detection In Robotic Manipulation, Shijie Zhou, Bin Zhu, Jiarui Yang, Xiangyu Zhao, Jingjing Chen, Yu-Gang Jiang

Research Collection School Of Computing and Information Systems

Recent advances in Vision-Language-Action (VLA) models have enabled robots to execute increasingly complex tasks. However, VLA models trained through imitation learning struggle to operate reliably in dynamic environments and often fail under Out-of-Distribution (OOD) conditions. To address this issue, we propose Robot-Conditioned Normalizing Flow(RC-NF), a real-time monitoring model for robotic anomaly detection and intervention that ensures the robot's state and the object's motion trajectory align with the task. RC-NF decouples the processing of task-aware robot and object states within the normalizing flow. It requires only positive samples for unsupervised training and calculates accurate robotic anomaly scores during inference through the …

Anatomical Domain Shifts: Test-Time Heterogeneous Adaptation For 3d Human Pose Prediction, Qiongjie Cui, Pan Zhou, Jingjing Chen, Na Zhao

Research Collection School Of Computing and Information Systems

The research frontier in human pose prediction (HPP) is advancing toward continual test-time adaptation (TTA), where models must self-adapt to dynamic test distributions. To date, the homeostatic continual TTA remains the sole viable solution, which isolates the model parameters and update domain-sensitive ones. Despite mitigating full-body domain gaps, human anatomical heterogeneity (domain shifts often localize to specific regions) is ignored. This anatomical-agnostic approach forces uniform parameter adaptation across kinematically distinct segments, causing: over-adaptation of stable regions and under-adaptation of shift-prone articulations. To address it, we introduce TT-HA, a novel Test-Time Heterogeneous Adaptation that implicitly estimates domain changes for anatomical segments, …

History To Future: Evolving Agent With Experience And Thought For Zero-Shot Vision-And-Language Navigation, Guangzhao Dai, Shuo Wang, Zihan Wang, Guo-Sen Xie, Yang Yang, Jinshan Pan, Qianru Sun, Xiangbo Shu

Research Collection School Of Computing and Information Systems

Vision-and-Language Navigation in Continuous Environment (VLN-CE) requires an agent to follow language instructions to navigate the target destination. With the advancement of large language models (LLMs), recent efforts have explored adapting them for zero-shot VLN-CE, offering a promising solution in addressing the drawbacks of poor generalization in the training-based paradigm. However, existing LLM-based works primarily perform naive reasoning for decision-making and lack feedback, e.g., reviewing historical errors and predicting future potentials. Consequently, it may suffer from continuous failure for those initial error tasks. In this paper, we rethink LLM-based zero-shot VLN-CE and propose a new paradigm, named EvoNav, to improve …

Task Complexity Matters: An Empirical Study Of Reasoning In Llms For Sentiment Analysis, Donghao Huang, Zhaoxia Wang

Research Collection School Of Computing and Information Systems

Large language models (LLMs) with reasoning capabilities have fueled a compelling narrative that reasoning universally improves performance across language tasks. We test this claim through a comprehensive evaluation of 504 configurations across seven model families—including adaptive, conditional, and reinforcement learning-based reasoning architectures—on sentiment analysis datasets of varying granularity (binary, five-class, and 27-class emotion). Our findings reveal that reasoning effectiveness is strongly task-dependent, challenging prevailing assumptions: (1) Reasoning shows task-complexity dependence—binary classification degrades up to -19.9 F1% points (pp), while 27-class emotion recognition gains up to  +16.0 pp; (2) Distilled reasoning variants underperform base models by 3–18 pp on simpler tasks, …

How Do Machine Learning Models Change?, Joel Castaño, Rafael Cabañas, Antonio Salmerón, David Lo, Silverio Martínez-Fernández

Research Collection School Of Computing and Information Systems

The proliferation of Machine Learning (ML) models and their open source implementations has transformed AI research and applications. Platforms like Hugging Face (HF) enable this evolving ecosystem, yet a large-scale longitudinal study of how these models change is lacking. This study addresses this gap by analyzing over 680,000 commits from 100,000 models and 2,251 releases from 202 of these models on HF using repository mining and longitudinal methods. We apply an extended ML change taxonomy to classify commits and use Bayesian networks to model temporal patterns in commit and release activities. Our findings show that commit activities align with established …

Heat Transfer And Micropolar Characteristic Analysis For Fluid Flow Over A Contracting Permeable Infinite-Length Cylinder: A Fractional Derivative Study, Khalid Al Homran

Theses

This study presents a comprehensive analysis of heat transfer and micropolar fluid characteristics for flow over a contracting permeable infinite-length cylinder using fractional derivative modeling. The micropolar fluid model accounts for microstructure effects that are prevalent in many practical applications, such as polymeric fluids, colloidal suspensions, and biological fluids. The inclusion of fractional derivatives allows the analysis to incorporate memory effects, which provide a more accurate depiction of the physical phenomena compared to classical integer-order models. The governing partial differential equations, representing the momentum, micropolar momentum and energy transport processes, are transformed into a set of nonlinear ordinary differential equations …

Developing A High-Resolution Off-Axis Common-Mode Digital Holographic Microscope, Lucy Cook

University Honors Theses

Off-axis digital holographic microscopy (DHM) is a powerful tool for 3D, non-invasive live-cell tracking without moving parts. However, traditional setups face an inherent dilemma: split-path interferometers offer high spatial resolution but poor temporal stability, while more stable common-mode configurations are historically limited to lower numerical aperture (NA) regimes. This thesis bridges that gap by scaling a common-mode DHM architecture into a high-resolution benchtop instrument featuring NA = 0.65 objectives, paired with a high-power 520 nm laser source to combat transmission losses and sustain imaging frame rates across an expanded optical footprint. We map the multi-variable design space required to satisfy …

Supramolecular Assembly In Short Peptide Systems For Selective Metabolite Recognition And Drug Nanoencapsulation, Maithreyi Ramakrishnan

Dissertations, Theses, and Capstone Projects

Short peptides can form adaptive supramolecular assemblies, and understanding how minimal sequences organize around neurometabolites or hydrophobic cancer drugs enables the rational design of functional materials. This thesis combines molecular dynamics with experimental validation to establish design rules linking peptide sequence to emergent structure and function. Chapter 1 outlines the molecular determinants governing peptide assembly. Chapter 2 reviews computational workflows that reveal sequence-dependent conformations and supramolecular organization. Chapter 3 applies these principles to Dynamic Peptide Libraries which identify tetrapeptides that selectively interact with neurometabolites. Chapter 4 extends the same interaction-driven framework to design tryptophan-rich pentapeptides that co-assemble with kinase inhibitors …

Radiotheranostics For Gynecological Pathologies, Joni Sebastiano

Dissertations, Theses, and Capstone Projects

Molecular imaging, specifically positron emission tomography (PET), is vital for detecting disease and understanding its biological makeup. The exploitation of radiolabeled antibodies for PET imaging has proven to be indispensable to the field of molecular imaging over the last several decades. The development of radioimmunoconjugates for use in immunoPET has not only allowed for the sensitive, specific, and high-resolution detection of disease, but has also enabled a deeper understanding of the disease biology, ultimately guiding more personalized therapeutic strategies. Most typically, this technology is harnessed for the diagnosis and treatment of cancer, however more recently, the field has explored the …

Algorithmic Problems In Automorphic Orbits Of Free Groups, Siobhan B. O'Connor

Dissertations, Theses, and Capstone Projects

One of the fundamental problems in the field of combinatorial group theory is telling when two group presentations represent isomorphic groups. Since applying a free group automorphism to the set of relators of a presentation gives an isomorphic group, we want to be able to quickly decide when looking at a relator whether a given word can be sent to it via an automorphism. We give a hands-on introduction to the automorphisms of free groups using patterns of colored beads. We show that you can make this decision correctly in constant time on average by looking for "orbit-blocking" words that …

Parkinson’S Disease Phenotype Stratification Using Multiple Correspondence Analysis, Kelly Astudillo

Dissertations, Theses, and Capstone Projects

Parkinson’s disease (PD) is the second most common neurodegenerative disorder, with over 12 million people projected to be affected by 2040 (Dorsey et al., 2018). Deep phenotyping and stratification can provide useful information regarding PD pathogenesis and can aid in the  development of disease modifying therapies that aim to delay the progression or prevent the onset of neurodegeneration (Blandini et al., 2019; Smith & Schapira, 2022). Utilizing multivariate methods such as multiple correspondence analysis (MCA) permits for the simultaneous analysis of distinct data modalities. To the best of our knowledge, MCA has not been previously used to explore phenotype patterns …

A Predictive Coding Account Of Spatial Working Memory Following Prophylactic Levetiracetam Administration Prior To Traumatic Brain Injury, Omeima Mutwali

Dissertations, Theses, and Capstone Projects

Traumatic brain injury (TBI) symptom prevention and remediation is an important area of research that would benefit vulnerable groups, including active-duty and veteran soldiers. These patients can sustain penetrative forces in fields of combat or in training, which result in focal lesions that trigger inflammatory and degenerative processes in the brain. Both primary and secondary injuries are associated with changes to cognition, behavior and affective state. This disease poses increased risk of epileptogenesis, as well. Given these outcomes, prior research has evaluated levetiracetam (LEV) as a prophylactic treatment for seizures, cognitive deficits and negative emotionality. LEV acts as a presynaptic …

Considering Numerical Relativity For Upgrades In Mcfacts Merger Dynamics, Shawn K. Ray

Dissertations, Theses, and Capstone Projects

Binary black hole mergers (BBH), some likely originating from active galactic nuclei (AGN), are observed by ground-based detectors such as LIGO-Virgo-KAGRA (LVK) and will be detected by future space-based detectors such as the Laser Interferometer Space Antenna (LISA) and the Laser Interferometer Lunar Antenna (LILA). The dynamics of AGN may be responsible for a significant fraction of BBH mergers soon to be observed by LISA and LILA, yet the dynamics of BBH mergers within AGN disks remains ambiguous despite their appearance in gravitational wave observations. Monte carlo For AGN Channel Testing and Simulation (McFACTS) is an open-source software used to …

Inductive Biases In Field-Level Cosmological Inference From Galaxy Catalogs, James O'Connor Baldwin

Dissertations, Theses, and Capstone Projects

We perform field-level likelihood-free inference of the matter density parameter Ωm from simulated galaxy catalogs using machine learning models with differing inductive biases. Using features extracted from hydrodynamic simulations in the CAMELS suite, we investigate how both observable choice and model architecture govern the extraction of cosmological information. We consider galaxy positions and line-of-sight peculiar velocities, both separately and in combination, and compare permutation-invariant Deep Sets, implemented with either standard multilayer perceptrons (MLPs) or Kolmogorov–Arnold Networks (KANs), to graph neural networks (GNNs) implemented with MLPs, which explicitly encode spatial relations. We evaluate inference performance under both in-distribution and out-of-distribution (OOD) …

Legal Ethics Of Ai Snake Oil: Navigating The Hype, Harm, And Hope Of Legal Ai, Drew Simshaw

Michigan Law Review

A review of AI Snake Oil.By Arvind Narayanan and Sayash Kapoor.

From 5g To 6g: A Survey On Security, Privacy, And Standardization Pathways, Mengmeng Yang, Youyang Qu, Thilina Ranbaduge, Chandra Thapa, Nazatul Haque Sultan, Ming Ding, Hajime Suzuki, Wei Ni, Sharif Abuadbba, David Smith, Paul Tyler, Josef Pieprzyk, Thierry Rakotoarivelo, Xinlong Guan, Sirine Mrabet

Research outputs 2022 to 2026

The vision for 6G aims to enhance network capabilities, supporting an intelligent digital ecosystem where artificial intelligence (AI) is a key. However, the expansion of 6G raises critical security and privacy concerns due to the increased integration of IoT devices, edge computing, and AI. This survey provides a comprehensive overview of 6G protocols with a focus on security and privacy, identifying risks that have not been experienced in preceding 5G systems, and presenting mitigation strategies. While many vulnerabilities from earlier generations persist, the introduction of AI/ML introduces novel risks like model inversion and malicious manipulation of AI. Vulnerabilities in emerging …

Geometric Characterization Of Ideals In Bipolar Semigroups, Kittipong Laipaporn, Rasimate Maungchang, David M. Cook, Prathomjit Khachorncharoenkul

Research outputs 2022 to 2026

This paper develops a geometric framework for analyzing the ideal structure of the bipolar semigroup ��={(−��,��)∣��,��∈ℝ+0} under coordinate-wise addition. Subsets of B are interpreted as planar regions, allowing ideals to be described in terms of boundary behavior. In particular, we prove that the complement of a simply connected region is an ideal of the commutative additive semigroup (��,+) if and only if its boundary contains no strictly decreasing segment. This provides a direct and visually verifiable criterion for ideality, linking algebraic structure to geometric shape. Each ideal can be written as a union of translates of the form ��+��, with …

Polarimetric Terahertz Imaging For The Measurement Of Birefringence In Plastic, Rachel Cohen

Theses

Birefringence offers a promising way to observe stress concentration in materials such as glass and plastic, and thereby to identify weaknesses. Polarimetric imaging can be used to measure the birefringence of material, so long as the material is transparent to the light being used for the imaging. In this research, 2D Terahertz imaging was investigated as a means of measuring the birefringence of plastics that are opaque to visible light but transparent to THz radiation, for the eventual purpose of analyzing the residual stress present. In order to do so, two separate terahertz cameras were characterized for potential use in …

Assessing The Effectiveness Of Tilt-Informed Assessments In Calculus I, Samuel Horelick

Inquiry: The Journal of the Virginia Community Colleges

Transparent Design in Learning and Teaching (TILT) is widely promoted as an evidence-based framework intended to clarify expectations, promote equity, and improve student learning. While prior research reports positive outcomes across many disciplines, less is known about how transparency functions in quantitative, problem-solving courses such as calculus, where students often value efficiency and autonomy. This study examines the effects of a TILT-informed assignment redesign in two sections of Calculus I at a Virginia Community College System institution. One section completed a traditional assignment, while the other completed an equivalent task redesigned to make the purpose, task, and evaluation criteria explicit. …

From Process Retrieval To Understanding: A Look Into Curiosity And Metacognition In My Physics Classrooms, Nahuel Acosta Burroso

Critical and Creative Thinking Capstones Collection

In the modern high school physics classroom, the rapid proliferation of generative artificial intelligence (GenAI) has created a unique educational challenge. While these tools can solve complex problems quickly, they often lead to cognitive offloading, where students bypass the internal dialogue and deep thinking needed for true understanding. This project follows an action research journey that shifts the focus from restricting technology to cultivating curiosity and metacognition. I define curiosity as the gap identifier that appears when a student realizes they do not know something, while metacognition acts as the navigator that helps them decide what to do next. By …

Adaptive Multimodal Smart Home Control On A Raspberry Pi 5 Using Hand Gestures, Voice Cues, And User Feedback, Vaibhav Bora

Theses

A real time multimodal smart home control system deployed on a Raspberry Pi 5 is presented. The system combines hand gestures, short voice cues, and proximity aware interaction to execute household commands such as light brightness control, fan speed adjustment, and stop or kill switch actions. Lightweight gesture and keyword spotting voice classifiers were trained offline and exported to TensorFlow Lite for efficient on device inference. For more natural spoken phrases, the system additionally integrates a locally deployed pretrained offline ASR component rather than a speech recognizer trained from scratch. Using a USB camera and microphone, the system operates fully …

Sound Detection Of Memory Leaks In Llvm Ir Programs Using Accumulation Analysis, Robert Blacha

Theses

Resource leaks occur when a limited resource such as memory is allocated by a program and needlessly held past the point of use. Leaks can lead to a degradation of services which can be specifically triggered with malicious behavior, for example abusing a memory leak in a program to cause a server to slow down and crash for a denial-of-service attack.

Prior work has demonstrated that accumulation analysis provides a sound detection of resource leaks with a working implementation for programs written in Java. While useful, current implementations are limited to programs written in Java, which has a garbage collector, …

Perceptual And Geometric Advances In Crowd Simulation, Bilas Talukdar

Dissertations

Simulating realistic crowd motion remains a fundamental challenge in computer graphics and multi-agent systems, as it requires modeling both physically plausible interactions and perceptually natural behaviors. Existing crowd simulation methods typically employ simplified geometric abstractions, most commonly circular agent representations, and model navigation using either analytical interaction formulations (e.g., force, velocity, or constraint-based methods) or learned policies derived through reinforcement learning. Despite their effectiveness, these approaches often overlook detailed geometric structure and do not explicitly account for perceptual realism. This dissertation addresses these challenges by improving the realism of virtual crowd simulation through two key advancements: perceptual preference learning and …

Differential-Geometric Methods For Neural Signed Distance Fields: Parameterized Surface Extraction And Curvature Regularization For Cad Models, Haotian Yin

Dissertations

Neural signed distance fields have emerged as a powerful framework for representing three-dimensional geometry through continuous and differentiable neural functions. Their flexibility, resolution independence, and compatibility with gradient-based optimization make them especially attractive for surface reconstruction and geometric learning. However, despite these advantages, two fundamental challenges remain for engineering-grade applications. First, higher-order geometric properties such as curvature are difficult to model reliably during training and often require computationally expensive second-order differentiation. Second, while neural signed distance fields provide implicit surface representations, they do not directly yield a globally consistent forward map or parameterization for downstream geometric processing.

This dissertation addresses …

Holistic Dram Enhancements: From Intrinsic In-Memory Operations To Robust Security Mechanisms, Ranyang Zhou

Dissertations

Dynamic Random-Access Memory (DRAM) is both the performance bottleneck and a critical security boundary of modern computing systems. Its physical properties make it an attractive substrate for near-data computation—yet those same properties expose it to disturbance-based hardware attacks. This dissertation argues that these two dimensions are not independent: the architectural choices that make DRAM efficient also reshape its threat landscape. Addressing both requires a unified approach to memory architecture and security co-design.

The first part of this dissertation attacks the memory wall through four processing-in-DRAM (PIM) frameworks. ReD-LUT and LT-PIM unify lookup-table arithmetic with charge-sharing logic, achieving up to 37.8x …

Anonymity And Accountability In Secure Messaging, Erin Kenney

Dissertations

Encypted messaging has become more and more prevalent as time moves on, and its benefits in assuring privacy cannot be overstated, but it also brings along with it concerns on how to moderate platforms where all messages are hidden. Message Franking, followed by Traceback systems, addressed these concerns by allowing the sender of a message to be proven when reported, even for forwarded messages in the case of Traceback, however these systems damage the privacy guarantees that originally motivated encrypted messaging to begin with.

In practice, even without those concerns encrypted messaging alone is not enough to prevent the most …