Supriti Vijay

WORK HISTORY

EDUCATION

ABOUT SUPRITI VIJAY

I\'m currently a Graduate Student at LTI, CMU, where I\'m advised by Dr. Fernando Diaz. Prior to this, I worked as a Member of Technical Staff at Adobe\'s Digital Experience. In brief, my research interests are centered around making NLP systems more accessible and practical for people all around the world (Multi-lingual LLMs, De-biasing, ZSL/FSL, Inference-time Reasoning). This entails diving deep into the intricacies of Transformer models, uncovering biases, applying mechanistic interpretability, and critically evaluating their privacy and fairness threats. I am also interested in the societal and ethical implications of these models (specifically AI Safety, Toxicity-detection & benchmarking, Off-Topic detection, Hallucinations, and their ability of Deception) and the potential harms that emerge in their context.In addition to this, I am a Research Lead and Open-Source Contributor at OpenMined where I dedicate my time researching and evaluating secure and privacy-preserving AI tools.I\'m an LLM enthusiast ready to take on the world, open to ideas and all that life has to offer!