My research interest involves quantifying how much of the complexity of natural language is learned by large, transformer-based language models.