Examining Context Effects in Humans and Machine Learning Models
By drawing on insights from cognitive science, this project explores how both humans and machines perceive and compare visual information. We use deep neural networks as models of human behavior to investigate where these models succeed and where they still fall short in capturing human-like understanding of visual relationships. We are particularly interested in how visual judgments are influenced by the broader structure of experience, for example how perception and comparison unfold over time, rather than in isolation. By doing so, we aim to foster the development of models that represent information in more human-aligned and meaningful ways.
