I am a fourth-year PhD student @ Department of Computer Science and Engineering, UMN. I am member of DMLSys and currently be advised by Ali Anwar.
My research focuses on designing efficient systems for large language models. I am interested in exploring how to optimize model performance while minimizing computational resources and what trade-offs exist between model size, speed, and accuracy for specific tasks. My current work involves improving the inference of LLMs by reusing past reasoning.
Prior to starting my PhD, I worked as research assistant at the Information technology university under Mudassir Shabbir. During this time, I worked on graph learning and its applications in various domains.
