
Thomas Möllenhoff
Postdoctoral Researcher
Approximate Bayesian Inference Team
RIKEN Center for Advanced Intelligence Project
Tokyo, Japan
thomas.moellenhoff (at) riken (dot) jp
CV – Scholar – github – Twitter
About
I’m a researcher at the RIKEN Center for Advanced Intelligence Project and a member of the Approximate Bayesian Inference team where I work with Emtiyaz Khan. Before that, I completed my PhD in the Computer Vision & Artificial Intelligence group at TU Munich under the guidance of Daniel Cremers.
My current research focuses on the design and analysis of new data-driven algorithms to make machine-learning systems more robust and adaptable in a dynamic and uncertain world.
News
- Dec 15, 2022. I gave a talk on SAM as an Optimal Relaxation of Bayes at the Optimization, Systems and Control Group at KU Leuven.
- Oct 31, 2022. I’ll be giving talks on Nov 1st, 10:50 - 11:40 JST in the ABI Team Seminar and on Dec 8th, 23:00 - 24:00 JST at the Mathematical Optimization Group in Tübingen. More talks to be announced soon!
- Oct 5, 2022. New preprint available: SAM as an Optimal Relaxation of Bayes!
- Feb 11, 2022. Outstanding reviewer award (top 10%) for AISTATS ‘22.
- Dec 5, 2021. We won both tracks of the NeurIPS competition on Approximate Inference in Bayesian Deep Learning! I will give a talk at the NeurIPS ‘21 competition track (Dec. 9th, 6pm GMT) and at the BDL workshop (Dec. 14th, 4:40pm GMT) on our solution.
- Nov 8, 2021. I have finally relocated to Tokyo! Please reach out if you are interested in collaborations or want to chat about machine learning.
- Oct 15, 2021. Got outstanding reviewer award (top 8%) for NeurIPS ‘21!
- Oct 1, 2021. Excited to be part of the Bayes-Duality project!
- Sep 11, 2021. Two new papers available.
- Oct 20, 2020. New homepage is online.
- Aug 13, 2020. I successfully defended my thesis and completed my PhD at TU Munich. Moving on, I am joining the Approximate Bayesian Inference Team in Tokyo as a postdoc!