Sunny Rai

Sunny Rai is a postdoctoral researcher at the Department of Computer and Information Science at the University of Pennsylvania, where he works under advisors Dr. Lyle H. Ungar and Dr. Sharath C. Guntuku. He holds a Ph.D. from the University of Delhi, India, with a thesis on Computational Metaphor Processing for English Language.

• Current Research: Focuses on culturally aware NLP, AI in health, cross-cultural variations in depression markers, chatbots for vaccine counseling, and bias/risk analysis in LLMs.
• Grants: Awarded Leonard Davis Institute grant (USD 19,500) and PMCRI grant (USD 50,000) in 2024 for mental health and vaccine counseling projects.

Key Research Themes include culturally aligned AI, behavioral health assessment, social emotions across cultures, and LLM bias detection. His work intersects NLP, computational social science, and public health, with a focus on cross-cultural applicability of models.

Publication Trends highlight applications of NLP in healthcare (depression markers, pain analysis), cross-cultural AI (cultural alignment lexica, emotion modeling), and societal impact studies (pandemic needs, radicalization vulnerability). Methodological approaches include RoBERTa fine-tuning, VOSviewer analysis, and metaphorical semantic shifts via diachronic embeddings.

• Scientific Awards: Best Paper Award at WASSA'23, grants from Leonard Davis Institute and PMCRI.
• Conference Participation: Presented at SPSP 2025, NIDA 2024, ICWSM 2023, and NAACL/ACL/WWW conferences.

At UPenn, Sunny contributes to interdisciplinary research groups combining computer science and medical perspectives, with collaborations extending to institutions like The Alan Turing Institute and Aalto University. His technical expertise spans machine learning frameworks like RoBERTa, word2vec, and jekyll development, with prior experience in Facebook AI's RoBERTa and HuggingFace transformers.