Presentations

Introduces large language models in healthcare, the motivation for using them in clinical settings, and the broader research context behind medical NLP and healthcare AI.

Covers transformer architecture and training foundations, giving students a practical understanding of how modern language models are built and why those design choices matter.

Focuses on multimodal healthcare models, validation metrics, vision-language models, and hands-on sessions with open and proprietary models for healthcare applications.

Walks through the medical LLM lifecycle, from development and validation to implementation planning, with emphasis on the real constraints of healthcare deployment.

An introductory AI Methods Lab meeting presentation covering the history of LLMs, how they work, what they can do, and how multimodal AI fits into the broader landscape.

A Methods Program Meetup talk focused on diffusion language models in general, and the foundations needed to reason about their behavior.

A presentation on our own implementation of diffusion language models, LoRA-adapted diffusion, emphasizing efficient and flexible discrete diffusion built on top of pretrained LLMs.

A presentation on how large language models can support research workflows in data science and biostatistics, which includes three practicals - academic writing, automated data analysis, and free to structured text.