My research in Natural Language Processing focuses on three primary themes: (1) the application of language models to social science-relevant tasks, (2) investigating and mitigating biases in language models, and (3) increasing language model robustness in processing unclean, unstructured text from social media. In the social sciences, I am particularly interested in the application of NLP to suicidality, mental health, and emotions in psychology as well as to political rhetoric and polarization in political science.