About

Email: [email protected]

Agrima is a Ph.D. candidate at the University of Michigan School of Information, working with Professor Kentaro Toyama. Her research focuses on culture and computing, where she uses and develops methods in Human-Computer Interaction (HCI) and Natural Language Processing (NLP). She has developed methodologies to study the cultural awareness and extent of cultural biases encoded in LLMs, especially as they relate to the Global South (India) and understudied axes of marginalization like religion and caste. Her work also looks at the online behavioral patterns of users across various cultures to understand how cultural values drive and influence user interactions on social media platforms.

Personal website

CV

Dissertation title 

Provisional Dissertation Title: Cultural Variability and Bias in Online Social Interactions and Large Language Models

Fields of interest

Human-centered AI; Responsible AI; HCI; NLP

Education

Bachelor of Engineering, University of Pune, 2016; 

Masters of Science in Engineering, University of Pennsylvania, 2018

In the news

She has published papers in top-tier HCI and NLP venues, receiving the best paper award at FAccT 2024. Additionally, she was awarded a Google PhD Fellowship [1] and recognized as a finalist for the Snap PhD Fellowship [2]. The Data Nutrition Project has featured her work to highlight inclusive methods for research and dataset creation [3].

[1] https://research.google/programs-and-events/phd-fellowship/recipients/ 

[2] https://research.snap.com/fellowships.html 

[3] https://vimeo.com/1018720653 (https://datanutrition.org/)