Bhuwan Dhingra

Bhuwan Dhingra is an Assistant Professor of Computer Science at Duke University. His research focuses on understanding how language models acquire and store knowledge from dynamic sources and developing epistemic interfaces to communicate this knowledge reliably. He completed his PhD in 2020 from the Language Technologies Institute at Carnegie Mellon University, advised by William Cohen and Russ Salakhutdinov. Prior to his PhD, he earned an Electrical Engineering degree from IIT Kanpur. He has worked at Google Deepmind and Qualcomm Research. His current research is conducted through the DukeNLP Lab, and he maintains an active presence on GitHub and Google Scholar. He also teaches graduate-level courses in NLP and large language models at Duke.

Bhuwan Dhingra is an Assistant Professor in Computer Science at Duke University, specializing in Natural Language Processing. His research focuses on advancing question-answering systems, dialogue agents, and text comprehension through machine learning. Dhingra's work includes developing datasets like PubmedQA and Quasar for biomedical and open-domain question answering, as well as exploring table-to-text generation and robust word recognition. He also investigates reinforcement learning for dialogue systems and neural models for reasoning over multiple mentions. His research emphasizes creating models that can handle ambiguity, adversarial inputs, and complex reasoning tasks, contributing to the fields of information access and natural language understanding.