BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//pretalx//pretalx.com//euroscipy-2026//speaker//8MTSZC
BEGIN:VTIMEZONE
TZID:CET
BEGIN:STANDARD
DTSTART:20001029T040000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
END:STANDARD
BEGIN:DAYLIGHT
DTSTART:20000326T030000
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
END:DAYLIGHT
END:VTIMEZONE
BEGIN:VEVENT
UID:pretalx-euroscipy-2026-QJRACH@pretalx.com
DTSTART;TZID=CET:20260720T113000
DTEND;TZID=CET:20260720T120000
DESCRIPTION:Many scientific models\, from climate systems to neural circuit
 s\, are defined as simulators: computer programs that generate data from p
 arameters but provide no tractable likelihood function. This makes them in
 visible to probabilistic programming languages (PPLs) like PyMC and NumPyr
 o\, which require explicit likelihoods for Bayesian inference. Practitione
 rs are forced to choose: make simplifying assumptions about the simulator 
 to use a PPL\, or use the real simulator and give up the rich modeling cap
 abilities PPLs offer\, such as prior specification\, uncertainty quantific
 ation and exploitation of hierarchical structures.\n\nWe present `setu` ("
 bridge")\, a JAX-native Python package that closes this gap. `setu` uses g
 enerative neural networks trained on simulated data to learn a neural surr
 ogate of the likelihood. This learned likelihood can then be exported dire
 ctly into PPLs via a simple API: `nle.to_pymc()` or `nle.to_numpyro()`. On
 ce inside a PPL\, the full Bayesian toolbox becomes available: hierarchica
 l models\, custom priors\, posterior predictive checks\, and standard MCMC
  samplers — all running on a simulator that was previously out of reach.
 \n\nThe package follows a clean simulate\, train\, validate\, export workf
 low\, with built-in diagnostics to ensure the learned likelihood is trustw
 orthy before it ever enters a PPL. In this talk\, we walk through the moti
 vation\, design\, and a real-world example showing how a black-box simulat
 or gains full PPL capabilities.
DTSTAMP:20260603T185654Z
LOCATION:Room 1.38 (Ground Floor\, Turing)
SUMMARY:setu: Bridging Simulators to Probabilistic Programming in JAX - Jan
  Boelts (Teusen)\, sethaxen
URL:https://pretalx.com/euroscipy-2026/talk/QJRACH/
END:VEVENT
END:VCALENDAR