Dhruva Karkada

• PhD Student, Physics Department
  

dkarkada@berkeley.edu Personal Site

Current Research

Despite fundamental differences between deep learning systems and biological brains, deep learning remains a theoretically- and experimentally-accessible playground for understanding learning as a general phenomenon. My work probes deep learning systems empirically and theoretically, with the end goal of understanding general principles underlying how such systems learn and compute. My recent work centers on understanding the origin and structure of vector representations in language models. I ultimately hope to trace these learned representations back to the statistics of the training data, to help understand how deep learning achieves better sample complexity than lazy methods like kernel machines.

Bio

I grew up in Texas and studied physics, astronomy, and computer science at UT Austin. I joined Berkeley as a physics PhD student in 2021, excited to use ideas from physics to understand information processing in neural networks. In my free time, I enjoy hanging out with friends, cooking, playing chess, and messing with synthesizers.