work
teaching
Fall 2026: CSC 280, Introduction to AI; Curriculum design for Ethics in AI
research
Fallibility, Persuadability, and Correctability of Large Language Models Under Sustained Conversational Misinformation Pressure
August 2026
Nature Scientific Reports
We found that in a multi-turn conversation, even relatively recent LLMs could be convinced to repeat misinformation in agreement with a user. Our findings show that this tendency increases with the persistence of the user and the length of the conversation. In some cases, an oscillation between agreeing and disagreeing with misinformation was observed.
My Master's Thesis explored ways to explain the results of AI Patent search tools that are opaque, a problem for inventors and patent examiners. I created a system to generate boolean queries that would reproduce the results of an AI-based patent search tool, and iteratively improve the query using machine learning and search tools to explore improvements to the query.
Boolean explanations bring interpretability to artificial intelligence search in prior art assessment April 2026 PLOS One
As part of a Kaggle competition, we created a system that would reproduce the results of an AI patent search tool using a boolean query. The query was generated using linguistic heuristics to accurately reproduce the results of the AI search, and thus offer an explanation to patent examiners and inventors.
Around-Body Versus On-Body Motion Sensing; A Comparison of Efficacy Across a Range of Body Movements and Scales November 2024 bioengineering
We compared two motion evaluation systems: an on-body physical sensor system, and a visual detection system making use of colored markers. The systems were compared across a number of motions and scales, to determine whether a more accessible colored marker system using an ordinary smartphone camera could reproduce the accuracy of an on-body sensor system. We found that the around-body marker system generally performed on-par with the on-body sensor system.