Bhuwan Dhingra

Bhuwan Dhingra is an Assistant Professor of Computer Science at Duke University's Pratt School of Engineering. He also serves as a Research Scientist at Google Deepmind, where he spends summers working on AI research. His research focuses on natural language processing, knowledge representation, and machine learning.

Dr. Dhingra completed his PhD in 2020 from the Language Technologies Institute at Carnegie Mellon University under the supervision of William Cohen and Russ Salakhutdinov. Prior to that, he earned his degree in Electrical Engineering from IIT Kanpur, where he worked with Amitabha Mukerjee. He also has industry experience from working at Qualcomm Research in San Diego.

His research interests span multiple areas within NLP and AI, with a particular focus on question answering and information retrieval, reasoning over structured and unstructured knowledge, robustness to corrupted text inputs, misinformation and propaganda detection, and model calibration for inherently subjective NLP tasks. His work often bridges the gap between theoretical advances and practical applications, as evidenced by his industry collaborations and grants.

Dr. Dhingra's recent publications indicate a strong focus on large language models, factuality assessment, knowledge attribution, public health content analysis, and efficient model training techniques. His research trajectory shows an evolution from foundational NLP work to more complex applications involving multimodal learning, reasoning, and societal impact.

• Amazon Research Award for 'Long-form QA via collaborative writing'
• NSF Medium grant with Jun Yang and Lavanya Vasudevan
• Learning Engineering Virtual Institute grant (joint with University of Florida)
• Gift from P&G for exploring search interfaces based on concept graphs
• Gift from Google for exploring NLP tools to combat misinformation

Dr. Dhingra actively mentors students, with seven PhD students and four undergraduate/masters students listed on his page. Many of his former students have gone on to prestigious positions at Stanford, Nvidia, McKinsey, and Amazon. He teaches graduate-level courses in NLP and generative models at Duke University. His lab works on machine learning for natural language processing and knowledge representation, with current topics including question answering, reasoning over knowledge, robustness to corrupted inputs, misinformation detection, and model calibration.