AI Researcher
Christian Rogers
Undergraduate researcher at the University of Utah studying mechanistic interpretability and medical applications of interpretable AI.
Undergraduate researcher at the University of Utah studying mechanistic interpretability and medical applications of interpretable AI.
My work spans two labs at the University of Utah, bridging AI interpretability with real-world medical applications.
Investigating the internal mechanisms of reward models used in reinforcement learning from human feedback (RLHF). This research aims to understand how reward models represent and process human preferences, contributing to safer and more transparent AI alignment.
Developing interpretable machine learning approaches for medical applications in telehealth, with a focus on stroke rehabilitation. This work leverages real-world sensor data to assess hand function recovery, enabling clinicians to make better-informed treatment decisions.
A sample of research and engineering projects. See the full list on the projects page.
Developing theory and experiments for tracing causal circuits across multiple ML model types in a Vision-Language-Action pipeline. January 2026 – Present.
Using Sparse Autoencoders and Transcoders to trace provenance, chain of thought, and reasoning of RAG LLM assistants in financial regulatory contexts. October 2025 – Present.
Open source digital circuit design teaching program and simulator. Guides students from logic gates through ALU design with mini-lessons and a cat mascot. Nov – Dec 2025.
Interested in collaboration or have a research inquiry? I'd love to hear from you.