yogesh

Hey! I'm a M.S. grad student researcher working on representation learning for health and trustworthy AI.

I work with Prof. Jingjing Zou on learning from wearable sensor data, including W-JEPA, a self-supervised approach to wearable representation learning. I’m interested in how these models generalize across people, sensor placements, and datasets. I also work on AI safety and interpretability, including whether natural language autoencoders can provide reliable oversight of language models.

Previously, I completed my BS in Computer Science Honors and Data Science at UW-Madison, where I completed my senior honors thesis with Prof. Fred Sala. I also worked with Prof. Vikas Singh on multi-modal emotion understanding and conformal prediction, and with Prof. Junjie Hu on VLM benchmarking for safety-critical scenarios.

I was also the Technical Lead at the Wisconsin AI Safety Initiative (WAISI) and received the Student Employee of the Year for Innovation and Technology award (and MASEA Runner Up 2025).

Research Interests: Representation Learning · Multimodal Models · Machine Learning for Health · Interpretability

Thanks for stopping by. Feel free to reach out if you want to chat about research, or anything else.