George Stoica

I'm a Machine Learning PhD Candidate from the School of Interactive Computing at Georgia Institute of Technology, and am co-advised by Professors Judy Hoffman and Ranjay Krishna. Since September 2024, I've been visiting the RAIVN Lab at the University of Washington. I'm also actively involved with research at Ai2. I'm grateful to be supported by an NSF Graduate Research Fellowship and a Google PhD Fellowship.

I am broadly interested in efficient model development. My current research focuses on large-scale training efficiency, from both the data and objective perspectives. My past work includes improving training for generative models, gradient-free post-training adaptation to new tasks via model merging, designing reliable datasets for training and evaluation, and efficient architecture development for downstream tasks.

Previously, I was a Research Assistant in the Machine Learning Department at Carnegie Mellon University advised by Professor Barnabás Póczos. I hold a Bachelors degree in Statistics and Machine Learning from Carnegie Mellon University with School of Computer Science and University Honors.