Kaitlyn Zhou

Dr. Kaitlyn Zhou is an incoming Assistant Professor in the Department of Information Science at Cornell University's Bowers College of Computing and Information Science, commencing August 2026. Her research focuses on human-language model interaction dynamics, with recognition at premier NLP and HCI conferences.

Education:

• PhD in Computer Science, Stanford University (Advised by Dan Jurafsky)
• B.Sc., B.Se., M.S. in Computer Science and Human Centered Design and Engineering, University of Washington

Research Focus: Her work investigates how language models shape human decision-making through three pillars: (1) identifying model overconfidence risks, (2) developing context-aware evaluation frameworks, and (3) reimagining interactions for marginalized user groups. This spans NLP, HCI, and AI ethics with emphasis on trust calibration and inclusive design.

Publication Trends: Recent work examines human reliance on unreliable language models (NAACL 2025 Best Paper Runner-Up), uncertainty expression failures (ACL 2024), and evaluation metric limitations. Collectively, these advance responsible AI through human-centered methodologies across 12 major publications from 2017-2025.

Scientific Awards:

• NAACL Best Paper Runner-Up (2025)
• MIT EECS Rising Star (2024)
• Stanford Graduate Fellowship
• College of Engineering Dean's Medal
• School of Engineering Dean's Medal of Excellence (UW)

Advising & Grants: Supported by Stanford Graduate Fellowship and research internships at Microsoft Research FATE (hosted by Olteanu/Blodgett) and Allen Institute for AI (hosted by Sap/Hwang/Ren). Will recruit NLP/HCI students at Cornell starting 2026. Appointed by Washington Governor to UW Board of Regents, advocating for educational equity.

Research Ecosystem: Collaborates with Stanford's NLP group, Microsoft's FATE team, and Allen Institute researchers. Features in NYT/WSJ for methods impacting real-world AI deployment.