ProxySPEX: Inference-Efficient Interpretability via Sparse Feature Interactions in LLMs (NeurIPS, 2025)
SPEX: Scaling Feature Interaction Explanations for LLMs (ICML, 2025)
Learning to Understand: Identifying Interactions via the Möbius Transform (NeurIPS, 2024)
Convolutional Learning on Multigraphs (IEEE Transactions on Signal Processing, 2023; ICASSP, 2023)
Convolutional Filtering and Neural Networks with Non-Commutative Algebras (IEEE Transactions on Signal Processing, 2023; ICASSP, 2024)
Equitable Optimization of U.S. Airline Route Networks (Computers, Environment and Urban Systems, 2023; Andrew P. Sage Memorial Conference, 2022) - Best Paper Award in Climate and Transportation (Sage)
Democratizing Aviation Emissions Estimation: Development of an Open-Source, Data-Driven Methodology (ICRAT, 2022) - Best Paper Award in Economics, Policy, and Equity
Learning Connectivity for Data Distribution in Robot Teams (IROS, 2021)
Research Experience
Interned with Apple and Uber AI.
Education
Ph.D. student in EECS at University of California, Berkeley, advised by Kannan Ramchandran.
Background
Research focuses on developing trustworthy machine learning, emphasizing methods that interpret and explain the complex decision-making processes of foundation models, using techniques from signal processing and game theory.