about/
Hi, I'm Usha (she/they)! :)
I'm a second year computer science Ph.D. student at Harvard, coadvised by Prof. Hima Lakkaraju and Prof. Hanspeter Pfister. I am grateful to be supported by the Kempner Institute Graduate Fellowship.
Before this, I completed my undergraduate education at Brown University, where I was advised by the wonderful Chen Sun. I also interned at Microsoft Research during that time.
My research interests generally lie in the realm of machine learning interpretability, and I am largely motivated by downstream applications in healthcare and fairness. If you want to chat about research or academia, please reach out to me at usha[_]bhalla[@]g[.]harvard[.]edu, especially if you are a woman/minority student considering grad school!
research/
- U. Bhalla*, A. Oesterling*, S. Srinivas, F. Calmon, H. Lakkaraju. "Interpreting CLIP with Sparse Linear Concept Embeddings (SpLiCE)" Under review.- U. Bhalla*, S. Srinivas*, H. Lakkaraju. "Discriminative Feature Attributions: A Bridge between Post Hoc Explainability and Inherent Interpretability" NeurIPS 2023.
- T. Yun, U. Bhalla, E. Pavlick, C. Sun. “Do Vision-Language Pretrained Models Learn Primitive Concepts?” Transactions on Machine Learning Research.
- N. Semmineh*, U. Bhalla*, L. Bell, A. Stokes, M. Lee, L. Hu, J.L. Boxerman, C. Quarles. “Analysis of Accuracy and Precision of Recommended Protocols for Dynamic Susceptibility Contrast MRI for Brain Metastases.” ISMRM 2020.
tidbits/
Let's play five truths and a lie!1. I have a twin sister.
2. I grew up in Arizona.
3. I've been dancing Bharatanatyam since I was six.
4. I'm a golden dragon and a leo (the best possible pairing).
5. The last book I read was Stay True by Hua Hsu.
6. Simon helped make this website!
