my education

projects

Are preference-based models for AI value alignment enough?: This study investigates alternative value elicitation techniques through crowd computing, exploring the effects of storytelling and rationale writing in shaping participants’ fairness judgements.

Retrieval Augmented Generation (RAG) for Open Domain Complex Question Answering: In this research project, we have conducted experiments to evaluate what kind of context is valuable to provide the LLM with when answering compositional questions.

Reproducibility of ”EG3D: Efficient Geometry-aware 3D Generative Adversarial Networks”: Recreating the results of ”EG3D: Efficient Geometry-aware 3D Generative Adversarial Networks”. My work primarily focused on getting the model to work on Kaggle and extracting a feature vector which allows us to shift the gender of generated subjects.

projects

Arenas in Rust: encoding and benefits: Exploring Arenas in Rust and how they can be used to enable cyclic (self-referential) types.

How do different environments affect the evolution of robot morphologies?: Honors research project exploring simulated robot evolution via different environments and how environmental conditions guide morphologies.