Research Project Title:
Learning a Representation
abstract:In AI methods, we frequently employ representations of information. These representations can take many forms -- an embedding function with transforms a high-dimension image to a dense vector, a genotype which describes the construction of a phenotype, or an ensemble of networks which collaborate to solve a task. What makes a representation useful? What frameworks allow us to analyze representations and understand them? This project will involve examining commonly learned representations in various domains, such as images or simulated games, and creating a framework for discovering common patterns and themes.
I am participating in SuperUROP because I am interested in doing research, and the program provides a structured way of doing so. I have acquired various background knowledge about AI, open-endedness, design, etc., and I am excited to apply it to unsolved problems.