Working in the optimization team on:• LLM inference efficiency• Optimization-based methods for compute resource allocation• Reinforcement learning for recommender systems

• Conducted foundational research on randomized matrix algorithms and optimization methods with applications to machine learning.• Developed novel algorithms and theoretical frameworks to address challenges in high-dimensional data analysis and large-scale machine learning.• Published extensively in top-tier venues such as NeurIPS, ICLR, JMLR, and ICML, with full details available at gregorydexter.com/publications.

• Conducted experimental studies to analyze the behavior of model sharpness across various optimizers and its correlation with test performance.• Developed a fine-grained theoretical framework linking sharpness to SGD stability; results published at ICLR 2024.

• Applied machine learning algorithms like random forest and generative adversarial networks to healthcare datasets.• Contributed to 5 publications, including 3 first author publications.• Presented research results in spotlight presentations at IMIA 2019 and AMIA 2020.

• Coauthored a journal article in IEEE Transactions on Information Theory• Developed strategies for empirical evaluation of novel approximation algorithms• Implemented experiments and created visualizations for final publication