I'm a final year undergraduate student, interested in learning theory and statistics and the use of information theoretic tools therein. My prior work has focussed on upper bounds on the generalization error of learning algorithms; more recently, I've grown interested in lower bounds for learning/statistical inference. In particular, one of the aspects I'm interested in is the fundamental tradeoff between the number of samples and the amount of memory used by the algorithm. Over the past decade, a series of seminal works have shown strong, unconditional lower bounds for problems like learning parities. This area lies at the intersection of information theory and statistics, often borrowing tools from fields like rate distortion theory and coding theory to get surprisingly strong bounds. Furthemore, I'm also interested in exploring learning under various kinds of privacy constraints: differential privacy offers a concrete, airtight framework to explore privacy and its implications.