About Me

I am a fourth-year Ph.D. student in the Department of Statistics at University of Washington advised by Zaid Harchaoui and Alex Luedtke. My research interests involve generative modeling, optimal transport, gradient flows, and deep learning in general. I am interested in application of optimal transport and gradient flows in statistical inference and explainable artificial intelligence. At UW, I am fortunate to be working with Soumik Pal on optimal transport.

I did my undergraduate in Mathematics from Indian Institute of Technology Kanpur with a minor in English literature. At IITK, I worked on developing methods for output analysis and convergence diagnostics of Markov chain Monte Carlo and importance sampling with Dootika Vats.

Contact me at medhaaga [at] uw [dot] edu.

News

• Gave a talk titled Pre-trained Transformer Learning as Heat Flows at JMM 2025. [Slides]
• Presented a poster titled Pretrained Transformers are Heat Flows at ASA SLDS 2024. Won the best poster award! [Slides]
• Presented a poster titled Iterated Schrödinger Bridge Approximation to Wasserstein Gradient Flows at SIAM MSD 2024. [Slides]