I graduated from the Indian Institute of Technology Jodhpur, majoring in Computer Science. I'm currently most invested in studying misalignment science and interpretability for AI safety and corrigibility. I love to monitor and study model behaviors, training quirks, and understand misalignment risks. I lead
the technical research wing at Safety and Alignment
Research (SAAR-India).
During my undergraduate studies, I largely worked on mechanistic interpretability for privacy preservation, under the amazing mentorhsip of my advisors,
Dr. Mayank Vatsa,
Dr. Richa
Singh, and
Dr. Shivang
Agarwal,
at the Image Analytics and Biometrics Lab, IITJ. I also had the privilege to work at some
really cool places: (i)
Thoughtworks AI Labs (TAILS), under the mentorship of
Shayan
Mohanty
and team, on Fine-grained Incompleteness Evaluation of
Summaries from Language Models; (ii)
Dr.
Tiago Bresolin, at the Center for Digital Agriculture (UIUC), on
Foundation Models for Livestock Images; and (iii)
Dr. Deepak Mishra, on flow-based models for particle physics simulations (HEP) with CERN.
I'm a strong believer of Bruce Lee's
10000-hour to mastery rule; and am forever grateful to
all my mentors & peers who shape my research.
/
/
/
/
/
"Whereof one cannot speak, thereof one must be silent." ~
Ludwig Wittgenstein
News
September, 2026: Machine Learning Research Intern @ Realm Labs.
July, 2026: A recipient of the IndiaAI Fellowship 2026.
June, 2026: Paper accepted at ICML Mech Interp Workshop, 2026!