profile_picture
Adhi Saravanan
Ph.D. Candidate (StatML CDT), University of Oxford
adhithya.saravanan@stats.ox.ac.uk

Hey! I am a ML PhD student at the University of Oxford, where I am fortunate to be advised by Yee Whye Teh and Tom Rainforth. I am broadly interested in generative modelling through diffusion and flows, with a focus on efficient inference-time adaptation. I am particularly interested in applying these techniques for modelling sequences including in generative world models and language models.

Previously, I completed my B.A. & M.Eng. in Information Engineering at the University of Cambridge, where I did my disseration with the MLG group. During my undergrad, I spent a summer at Caltech, advised by Anima Anandkumar.

News

  • 15 Jul 2026 Invited talk at the ELLIS Summer School. Presented Flow Maps.
  • 22 Jun 2026 Started as a Student Researcher at DeepMind! Working on generative world models & deep RL.
  • 10 Apr 2026 Discrete Flow Maps Preprint!

Selected Works

Peter Potaptchik*, Adhithya Saravanan*, Abbas Mammadov, Alvaro Prat, Michael Albergo, Yee Whye Teh
Discrete Flow Maps
2026 Preprint
Peter Potaptchik, Jason Yim, Adhithya Saravanan, Peter Holderreith, Eric Vanden-Eijnden, Michael Albergo