Rahul Gupta
I work with a team of scientists focusing on Responsible AI. Examples of my research work include building metrics applicable to RAI (see BOLD and Tango datasets), mitigations (see our attribute controlled fine-tuning for LLMs), and analysis articles (see our analysis on fairness evaluation metrics). I organize the TrustNLP series of workshops, and in the past organized the Trustworthy Speech Processing workshop at ICASSP and the Unlearning Challenge at SemEval. Back when I was a graduate student, our team won the Interspeech Paralinguistics Challenge twice (in 2013 and 2015).
My latest publications can be accessed through my Google Scholar Profile and I often post latest research updates on X and LinkedIn.