Episode

• Join this channel to get access to perks:
https://www.patreon.com/c/learnbayesstats

• Proudly sponsored by PyMC Labs! Get in touch at alex.andorra@pymc-labs.com

• Intro to Bayes Course (first 2 lessons free): https://topmate.io/alex_andorra/503302
• Advanced Regression Course (first 2 lessons free): https://topmate.io/alex_andorra/1011122

Our theme music is « Good Bayesian », by Baba Brinkman (feat MC Lars and Mega Ran). Check out his awesome work at https://bababrinkman.com/ !

Takeaways:
* Bayesian neural networks are crucial for uncertainty quantification.
* Scaling Bayesian methods to high dimensions is a significant challenge.
* JAX offers substantial speed improvements for Bayesian sampling.
* Initialization errors can hinder the performance of Bayesian neural networks.
* Microcanonical Langevin sampler enhances sampling efficiency.
* Practical tools are essential for wider adoption of Bayesian methods.
* Understanding neural networks requires better uncertainty quantification.
* Ensemble methods can improve the performance of Bayesian models.
* Computational efficiency must be balanced with posterior fidelity.
* Community-driven tools are vital for advancing Bayesian deep learning. Bayesian deep ensembles provide a more flexible approximation.
* Sampling methods can yield better predictive performance.
* Uncertainty quantification is crucial for practical applications.
* The overhead of Bayesian methods is decreasing.
* Bayesian neural networks outperform standard approaches in many cases.
* Exploration-exploitation trade-offs are important in sampling.
* Future advancements may allow for Bayesian deep learning at scale.
* Community efforts are needed to improve Bayesian inference packages.
* Practical applications of Bayesian methods are expanding.
* Understanding life and probabilistic modeling are key future goals.

Chapters:

00:00 Scaling Bayesian Neural Networks
04:26 Origin Stories of the Researchers
09:46 Research Themes in Bayesian Neural Networks
12:05 Making Bayesian Neural Networks Fast
16:19 Microcanonical Langevin Sampler Explained
22:57 Bottlenecks in Scaling Bayesian Neural Networks
29:09 Practical Tools for Bayesian Neural Networks
36:48 Trade-offs in Computational Efficiency and Posterior Fidelity
40:13 Exploring High Dimensional Gaussians
43:03 Practical Applications of Bayesian Deep Ensembles
45:20 Comparing Bayesian Neural Networks with Standard Approaches
50:03 Identifying Real-World Applications for Bayesian Methods
57:44 Future of Bayesian Deep Learning at Scale
01:05:56 The Evolution of Bayesian Inference Packages
01:10:39 Vision for the Future of Bayesian Statistics

Thank you to my Patrons (https://learnbayesstats.com/#patrons) for making this episode possible!

Come meet Alex at the Field of Play Conference in Manchester, UK, March 27, 2026!
https://www.fieldofplay.co.uk/

Links from the show:

David Rügamer:
* Website: https://www.statistik.uni-muenchen.de/people/professors/rügamer/index.html
* Google Scholar: https://scholar.google.com/citations?user=y1p8VhsAAAAJ&hl=en
* GitHub: https://github.com/compstat-lmu

Emanuel Sommer:
* Website: https://emanuelsommer.github.io/my-yourney/
* GitHub: https://github.com/emanuelsommer
* Google Scholar: https://scholar.google.com/citations?user=qa2P1tYAAAAJ&hl=en

Jakob Robnik:
* Google Scholar: https://scholar.google.com/citations?user=J9E2DxAAAAAJ&hl=en
* GitHub: https://github.com/JakobRobnik
* Microcanonical Langevin paper: https://www.jmlr.org/papers/volume24/22-1450/22-1450.pdf
* LinkedIn: https://www.linkedin.com/in/emanuelsommer/

General references:
* JAX: https://github.com/google/jax
* BlackJAX: https://github.com/blackjax-devs/blackjax
* sklearn-contrib-bde: https://github.com/scikit-learn-contrib/bde (easy to use and fast MILE for tabular data)
* A Beginner's guide to Variational Inference: https://www.youtube.com/watch?v=XECLmgnS6Ng
* posteriors (pytorch+sampling): https://github.com/normal-computing/posteriors
* LBS #142, Bayesian Trees & Deep Learning for Optimization: https://learnbayesstats.com/episode/142-bayesian-trees-deep-learning-optimization-big-data-gabriel-stechschulte
* MILE paper: https://arxiv.org/abs/2502.06335