In several real-world scenarios, decision making involves advanced reasoning under uncertainty, i.e., the ability to answer complex probabilistic queries. Typically, it is necessary to compute these answers in a limited amount of time. Moreover, in many domains, such as healthcare and economical decision making, it is crucial that the result of these queries is reliable, i.e. either exact or coming with approximation guarantees. In all these scenarios, tractable probabilistic inference and learning are becoming increasingly important.


t' ("t prime") is an open and inclusive space where researchers and practitioners coming from different fields and backgrounds are engaged to share their experience and ideas on tractable inference. We want participants to get involved on a more practical level while having an informal and entertaining atmosphere. Specifically, we will host contributed hands-on sessions on practical advancements for tractable inference, such as recent software frameworks and support participants to form interest groups for hackathon-like sessions.

We will provide light refreshments in the form of free drinks and snacks for registered attendees and encourage researchers and practitioners to join in fruitful discussions related to research on tractable probabilistic inference and their application in fields outside of the community or in industrial settings.


We welcome researchers and practitioners in all areas of inference and learning to discuss the current state of efficient, flexible and reliable probabilistic reasoning. To this end, we will provide "open mic" sessions to encourage discussion among participants. In addition, we plan to have the following invited spotlight talks:

  • Eli Bingham (UberAI) talking about the challenges of tractable inference in probabilistic programming
  • Eric Nalisnick (DeepMind & Univ. of Cambridge) talking about normalizing flows
  • Hong Ge (Uni. of Cambridge) talking about the prospects of robust and efficient inference in probabilistic programming
  • Molham Aref (Relational AI) on the challenges and opportunities of tractable inference in systems deployed on the real world
  • Pasha Khosravi (UCLA) introducing Juice.jl a Julia library for advanced inference with logical and probabilistic circuits


You can register on eventbrite.


Vancouver Convention Centre, West Level 2, Room 223-224


December 11, 2019 starting from 7pm

A tentative schedule:

  • 7:00 - 7:15 Registration
  • 7:15 - 7:45 Spotlight talks (TBA)
  • 7:45 - 8:30 Open discussions
  • 8:30 - 9:00 Spotlight talks (TBA)
  • 9:00 - 10:00 Open discussions


Martin Trapp (TU Graz)

Antonio Vergari (UCLA)