Hi, I'm Marvin.
I like using deep neural networks to solve hard problems in probabilistic machine learning. I am a Computer Scientist and a Psychologist.
I am a final-year PhD student at the European Laboratory for Learning and Intelligent Systems (ELLIS) supervised by Paul Bürkner, Stefan Radev, and Aki Vehtari. My main affiliation is the Cluster of Excellence SimTech at the University of Stuttgart, Germany.
JOB MARKET: I am looking for machine learning positions in Europe (or remote) starting in early 2026. My current interests revolve around deep learning, probabilistic ML, and ML4Science. I'm curious about new topics as well. I spend a considerable amount of my time programming in Python (JAX, PyTorch, TensorFlow, or Keras). Check out my CV (Link) and drop me a message if there might be a fit: mail.marvinschmitt@gmail.com
News
November 25, 2024: Over the past week I migrated to Bluesky and will mostly engage there instead of X/Twitter. The community is growing extremely quickly, the atmosphere is friendly, and I strongly recommend checking it out. You can find me under the handle @marvinschmitt.com
November 9, 2024: My work just reached 100 citations on Google Scholar. It might be small for many others, but it means a lot to me. I started my PhD with no research experience, and I’ve learned so much ever since. Thank you to my mentors, collaborators, and everyone who engaged with my work 🧡
September 25, 2024: Our paper on consistency models for simulation-based inference (Link) has been accepted at NeurIPS 2024!
July 8, 2024: I started a YouTube channel! The videos will broadly fall into these categories: Early career research based on my experience, technical machine learning topics, applied programming in R and Python. In my first video, I talk about my background and my plans for the channel:
June 6, 2024: My guest episode on the Learning Bayesian Statistics podcast is live. We had a great conversation about my research on amortized Bayesian inference with neural networks. Here’s the announcement post on Twitter and LinkedIn.
April 3, 2024: Just finished recording an episode of the Learning Bayesian Statistics podcast. The episode will air later this summer. Thanks Alex Andorra for having me as a guest!
March 1, 2024: I’m spending 6 months as a visiting researcher in the Probabilistic ML group at Aalto University in Finland, hosted by Aki Vehtari.
January 23, 2024: I will be at Bayes on the Beach 2024, Australia, in February. I will give a talk about reliable amortized Bayesian inference with neural networks and co-lead a workshop on amortized Bayesian workflows.
December 2023: We launched the BayesFlow Forums! The BayesFlow Forums provide a community for asking and answering questions about all aspects of BayesFlow and amortized Bayesian workflows in general. Hop over and join the discussion!
October 11, 2023: In our recent preprint, we make simulation-based inference more data-efficient by leveraging self-consistency properties of the Bayesian joint model. You find the preprint on arXiv. Update: Published in the NeurIPS UniReps workshop.
September 22, 2023: Our paper “Detecting Model Misspecification in Amortized Bayesian Inference with Neural Networks” has been awarded the DAGM GCPR Honorable Mention at this year’s German Conference on Pattern Recognition. Huge thanks to my great co-authors!
September 6, 2023: Our paper “Detecting Model Misspecification in Amortized Bayesian Inference with Neural Networks” has been accepted with oral presentation at the German Conference for Pattern Recognition 2023. Learn more at the project website.
June 21, 2023: I will be at the ELLIS Doctoral Symposium 2023 in Helsinki from August 28 to September 1, 2023. Absolutely thrilled to meet other PhD students and researchers and talk about machine learning research in beautiful Helsinki!
May 9, 2023: Our paper “JANA: Jointly Amortized Neural Approximation of Complex Bayesian Models” has been accepted to UAI 2023 (Uncertainty in AI). Check out the project page to find out more!
May 1, 2023: Our paper “Association between vitamin D status and eryptosis–results from the German National Cohort Study” has been published in Annals of Hematology. A great collaboration among twelve researchers from Martin-Luther-University Halle-Wittenberg, Cluster of Excellence SimTech (University of Stuttgart) and University of Hohenheim. Here’s a link to the paper.
April 12, 2023: I have recorded a 15min presentation about our paper on Meta-Uncertainty in Bayesian Model Comparison (AISTATS 2023). Check it out:
March 29, 2023: I’ll be presenting our paper on Meta-Uncertainty in Bayesian Model Comparison at AISTATS 2023 from April 25-27 in Valencia, Spain. You find links to the paper and code on my Projects page or on the dedicated paper website. If you see me around at AISTATS, let’s chat!