I’m a fifth-year PhD student studying computer science at the Center for Language and Speech Processing at Johns Hopkins University, advised by Nicholas Andrews and Kevin Duh. My general research interest is making large language models more reliable and safer. In particular, I’m interested in uncertainty quantification and how it can be used to improve reasoning and detect hallucinations before they reach an end user. Previously, I received my BA in computer science from Smith College, where I worked on controllable music generation.

Publications

2026

• Inducing Artificial Uncertainty in Language Models (arXiv preprint)

  Sophia Hager, Simon Zeng, Nicholas Andrews

• Does Local News Stay Local?: Online Content Shifts in Sinclair-Acquired Stations (NLP-CSS @ ACL 2026)

  Miriam Wanner†, Sophia Hager†, Anjalie Field († equal contribution)

• RuleForge: Automated Generation and Validation for Web Vulnerability Detection at Scale (arXiv preprint)

  Ayush Garg, Sophia Hager, Jacob Montiel, Aditya Tiwari, Michael Gentile, Zach Reavis, David Magnotti, Wayne Fullen

2025

• Hell or High Water: Evaluating Agentic Recovery from External Failures (COLM 2025)

  Andrew Wang, Sophia Hager, Adi Asija, Daniel Khashabi, Nicholas Andrews

• Learning Extrapolative Sequence Transformations from Markov Chains (ICML 2025)

  Sophia Hager, Aleem Khan, Andrew Wang, Nicholas Andrews

• Uncertainty Distillation: Teaching Language Models to Express Semantic Uncertainty (arXiv preprint)

  Sophia Hager, David Mueller, Kevin Duh, Nicholas Andrews

2024

• Learning to Generate Verbalized Confidences (SFLLM @ NeurIPS, 2024)

  Sophia Hager, Nicholas Andrews

• Generating Music with Structure Using Self-Similarity as Attention (arXiv preprint)

  Sophia Hager, Kathleen Hablutzel, Katherine M. Kinnaird

2023

• Learning to Generate Text in Arbitrary Writing Styles (arXiv preprint)

  Aleem Khan, Andrew Wang, Sophia Hager, Nicholas Andrews