Hello!
I am an Assistant Professor in the Department of Computer Science at Lafayette College in Easton, PA. I received my PhD in Computer Science & Engineering (more specifically, in natural language processing) from the University of Washington, advised by Noah Smith. Before that, I was an undergrad at Carleton College studying math and computer science.
My research mainly focuses on topics related to interpretability and explainability of natural language processing models, although I’m also interested in model evaluation and NLP for social science. I’m also very interested in computer science education and AI literacy more broadly! (If you’re here looking for the primer that I wrote on language models with Zander Brumbaugh and Noah Smith, you can find that here: “Language Models: A Guide for the Perplexed.” I also have a more screen-reader-friendly version of the guide here.)
This summer, I’m excited to be giving a keynote for the PA Regional Digital Collegium conference (June 25). Looking forward to chatting with others thinking about the digital sphere in higher ed!
Research
- “Troubles in Text: Using Natural Language Processing to recognize government rationalizations for human rights abuses.” Sarah K. Dreier, Sofia Serrano, Emily K. Gade, Noah A. Smith. Journal of Politics 2025. Early versions of this work were presented at Politics and Computational Social Science in 2020 and at the American Political Science Association Annual Meeting in 2021. [publish-format pdf]
- “Stubborn Lexical Bias in Data and Models.” Sofia Serrano, Jesse Dodge, Noah A. Smith. ACL Findings 2023. [screen-reader-friendly pdf] [publish-format pdf] [summary video]
- “All That’s ‘Human’ Is Not Gold: Evaluating Human Evaluation of Generated Text.” Elizabeth Clark, Tal August, Sofia Serrano, Nikita Haduong, Suchin Gururangan, Noah A. Smith. ACL 2021. Chosen as an outstanding paper. [screen-reader-friendly pdf] [publish-format pdf] [recorded talk by Elizabeth]
- “Is Attention Interpretable?” Sofia Serrano and Noah A. Smith. ACL 2019. [screen-reader-friendly pdf] [publish-format pdf] [recorded talk]
- “The partisan dimensions of religious rhetoric: Merging qualitative and natural language processing approaches to measure Congressional behavior.” Sarah K. Dreier, Lucy H. Lin, Sofia Serrano, Emily K. Gade, Noah A. Smith. Working paper; presented at Text as Data (2018) & Politics and Computational Social Science (2019).
Teaching
In academic year 2024-25, I have taught (or am teaching) CS420 (Artificial Intelligence), CM151 (Introduction to Computational Science), and a couple sections of CS105 (Digital Media). Here are some other teaching-type things I’ve done:
- Spring 2025 — Gave a talk titled “Promise and Pitfalls of Scale: Working with Machine Learning Models of Text” for the Lafayette Hanson Center’s spring 2025 series of programming themed around AI
- Spring 2024 — Gave an invited talk for UW/Fred Hutch Center for AIDS Research’s Behavioral Innovations Core on large language models [video]
- Spring 2023 — Gave two invited talks for the UW Tacoma Computer Literacy seminar series, the first on neural network fundamentals [video] and the second on large language models [video]
- Spring 2023 — Gave a talk on large language models for UW CSE’s Security Seminar (590Y) [slide deck]
- Winter 2023 — I was the instructor of record for CSE 447: Natural Language Processing
- Autumn 2022 — Gave two guest lectures covering text classification for CSE 447: Natural Language Processing
- Winter 2022 — Gave a guest lecture on interpretability for NLP for CSE 447/517: Natural Language Processing
- Autumn 2021 — Gave a guest lecture on transformers and BERT for POLS 585: Texts as Data (Emory University)
- Winter 2021 — Teaching Assistant for CSE 447/517: Natural Language Processing
- Winter 2020 — Teaching Assistant for CSE 517: Natural Language Processing
Misc
- December 2024 — Together with Lauren Biernacki, I ran an activity representing the Lafayette CS department at Wilson Area Intermediate School’s Family CS Night. If you’re a Lafayette student and you’d be interested in getting involved with future iterations of this event, please reach out, we’d love to have you on board!
- Summer 2024 — Panelist at Lafayette’s AI Literacy Across the Curriculum symposium
- Spring 2023 — I moderated a panel focused on demystifying ChatGPT for academics [video]