Sumeet Motwani

I am a Machine Learning PhD student at the University of Oxford, where my research is funded by Eric Schmidt and CAIF. I'm advised by Philip Torr and Christian Schroeder.

My work focuses on RL post-training, multi-agent systems, and AI security. I'm particularly interested in meta-RL, open-endedness, and methods for measuring and improving long-horizon LLM agent capabilities.

During my PhD, I've spent time at Microsoft Research and Google X. At MSR, I was part of AI Frontiers and worked on self-play and RL rewards for open-ended domains. Previously, I was an undergrad at UC Berkeley where I was a member of Berkeley AI Research (advised by Dan Hendrycks) and Cal Boxing. Feel free to get in touch!

Contact

Google Scholar / Twitter / LinkedIn / Email

Selected Papers

Recent Preprints

Conference Publications