Skip to content

Details

🎙️ Speakers: Nathanael Nussbaumer, Markus Böck, Jürgen Cito, Christopher Fonnesbeck, Oriol Abril Pla, Evan Wimpey | ⏰ Time: 15:00 UTC / 8:00 AM PT / 11:00 AM ET / 4:00 PM Berlin

Probabilistic programming enables the formulation of Bayesian models as programs and automates posterior inference. However, identifying and repairing issues with inference is notoriously difficult, requiring deep knowledge and long wait times for MCMC algorithms to finish.

Join PyMC Labs as we host researchers Nathanael Nussbaumer, Markus Böck, and Jürgen Cito to discuss their recent paper, Online and Interactive Bayesian Inference Debugging.

They will introduce INFERLOG HOLMES, an open-source debugger specialized for MCMC workflows. By hooking directly into the inference loop, this tool allows practitioners to analyze the output of a probabilistic program during its execution, rather than waiting for a potentially long inference time to finish.

What you'll learn:

  • How to use live debugging views to monitor MCMC sampling in real-time, tracking not just trace plots and marginals, but also continuously updated diagnostics like $\hat{R}$, ESS, and sampler metrics.
  • How heuristic rules combine computed MCMC diagnostics and sampler stats (like acceptance rates or divergences) to automatically warn you about inference issues.
  • How interactive, online diagnostics can drastically speed up your development cycles by allowing you to cancel bad runs early.

📜 Outline of Talk / Agenda:

  • 5 min: Introduction to PyMC Labs and speakers
  • 40 min: Panel discussion
  • 15 min: Q&A

-------------------------------------------------

💼 About the speakers:

Nathanael Nussbaumer (InferLog Holmes Co-Author)
Nathanael is a master’s student at the Technical University of Vienna. He has been working on probabilistic programming for 4 years with a focus on developer tools for probabilistic programming and Bayesian inference debugging.
🔗 Connect with Nathanael
👉 LinkedIn: https://www.linkedin.com/in/nathanael-nu

Markus Böck (InferLog Holmes Co-Author)
Markus is a PhD student at the Technical University of Vienna, where he sits at the intersection of Computer Science and Mathematics. His research focuses on probabilistic programming, developing sound program analysis methods to enhance debugging and to optimise inference algorithms. Additionally, he explores the use of GPU acceleration to scale the performance of inference algorithms for challenging probabilistic models.

🔗 Connect with Markus
👉 LinkedIn: https://www.linkedin.com/in/markus-boeck-aut/

Jürgen Cito (InferLog Holmes Co-Author)
Dr. Jürgen Cito is an Associate Professor for Computer Science at TU Wien (Vienna, Austria), where he leads the IPA Lab (Interactive Programming and Analysis Lab). His work sits at the intersection of Software Engineering (SE), Programming Languages (PL), and Artificial Intelligence (AI), with a specific focus on making software systems more reliable, explainable, and performant, and on empowering domain experts to leverage computational methods and AI in their own fields.

His research also maintains strong ties to industrial practice through industry engagements, including a Visiting Research Scientist position at Google (DevAI group) and a Software Engineer role at Meta (Probability group).

🔗 Connect with Jürgen
👉 LinkedIn: **https://www.linkedin.com/in/jcito/**

Christopher Fonnesbeck (Principal Data Scientist at PyMC Labs)
Chris is a Principal Quantitative Analyst at PyMC Labs and an Adjoint Associate Professor at the Vanderbilt University Medical Center, with 20 years of experience as a data scientist in academia, industry, and government, including 7 years in pro baseball research with the Philadelphia Phillies, New York Yankees, and Milwaukee Brewers. He is interested in computational statistics, machine learning, Bayesian methods, and applied decision analysis. He hails from Vancouver, Canada and received his Ph.D. from the University of Georgia.
🔗 Connect with Chris:
👉 Linkedin: https://www.linkedin.com/in/christopher-fonnesbeck
👉 GitHub: https://github.com/fonnesbeck

Oriol Abril Pla (Principal Data Scientist at PyMC Labs)
Oriol discovered his passion for computational statistics and open source in 2018 during his MSc in Astrophysics and has been working the topic since then. He started contributing to ArviZ and PyMC in 2019, joining their core teams not long after that. He started in academia but he left after some years in order to be able to work more freely and collaboratively on open source, software and knowledge sharing. His main areas of interest are data visualization, model and inference diagnostics, model comparison, and prior elicitation. Within open source projects, he has also dedicated a large part of his work to documentation, governance and EDIA.

🔗 Connect with Oriol
👉 Linkedin: https://www.linkedin.com/in/oriol-abril-pla-1b9123180/https://www.linkedin.com/in/oriol-abril-pla
👉 GitHub: https://github.com/OriolAbril

💼 About the Host:
Evan Wimpey (Director of Analytics at PyMC Labs)
Evan helps clients design Bayesian solutions tailored to their goals, ensuring they understand both the how and why of inference. With master’s degrees in Economics and Analytics, he focuses on delivering clear value throughout projects and brings a unique twist with his background in data comedy.

🔗 Connect with Evan:
👉 Linkedin: https://www.linkedin.com/in/evan-wimpey/
👉 GitHub: https://github.com/ewimpey

📖 Code of Conduct:
Please note that participants are expected to abide by PyMC's Code of Conduct.

🔗 Connecting with PyMC Labs:
🌐 Website: https://www.pymc-labs.com/
👥 LinkedIn: https://www.linkedin.com/company/pymc-labs/
🐦 Twitter: https://twitter.com/pymc_labs
🎥 YouTube: https://www.youtube.com/c/PyMCLabs
🤝 Meetup: https://www.meetup.com/pymc-labs-online-meetup/
🎮 Discord: https://discord.gg/KJt32Ty8

You may also like