Major William Fleshman

I'm a PhD student at Johns Hopkins University working with Benjamin Van Durme. My research focuses on efficient and composable adaptation of large language models — spanning LoRA routing and selection, reverse-engineered adaptation, and retrieval-augmented generation.

I care about both the practical and principled sides of deploying LLMs: building systems that can acquire knowledge continuously, forget data on demand, and serve diverse tasks without full retraining.

I am also a Senior Cyber Capability Developer in the Army, with experience in developing ML systems for cybersecurity and large-scale analysis.