Aditya Deshmukh

Background

About Aditya Deshmukh

My research interests span the areas of LLMs, deep learning, reinforcement learning, statistical inference and information theory. I am driven by a quest to uncover the fundamental boundaries of learning problems through theoretical analysis while pioneering novel algorithms which achieve these limits with provable guarantees. I am deeply committed to developing principled solutions that address complex challenges across various domains. My portfolio encompasses projects in LLMs, deep compression, multi-armed bandits, robust inference, network optimization and packet scheduling. While the objectives of these endeavors differ, they are unified by a common pursuit of achieving optimality, efficiency, and robustness. Personal Website: https://adityadeshmukh.github.io/ Google scholar profile: & hl=en