Sadhika Malladi
Sadhika Malladi
Research Affiliate
Sadhika is a final-year PhD student at Princeton University working on technical and theoretical machine learning research. Her work focuses on deriving theoretical insights that explain the successes and failures of language models in a variety of real-world settings, and many of these insights have translated to popular training methodologies. She has recently developed results showing surprising failures of preference learning techniques and is looking to dip her toes into more philosophical research on AI safety. Prior to Princeton, she received a B.S. degree in Math and Computer science as well as in Philosophy from the Massachusetts Institute of Technology.