Adithya Giri

Background

About Adithya Giri

I am a standing sophomore EECS (Electrical Engineering and Computer Science) major at UC Berkeley with a Regents and Chancellor’s Scholarship who has worked extensively in ML. I’ve worked as a paid research intern at NASA Ames for more than a year, researching how to improve the performance of their models which use a star light intensity signals in identifying a very rare type of celestial body, and researching the use of adversarial learning to improve the robustness of NASA’s planet classification models which use those same star light signals. I’m also conducting research on using inductive biases from convolutional neural networks to improve the performance of transformer based vision models. My first paper from my research was in the Top 300 nationwide in the Science Talent Search, and a continuation of that work is in the process of being published at ICLR as a single author paper. Relevant coursework I’m taking this semester includes Math 54, CS 61A, and CS 61B.