POPL 2017
Sun 15 - Sat 21 January 2017

Workshop on probabilistic programming semantics

Probabilistic programming is the idea of expressing probabilistic models and inference methods as programs, to ease use and reuse. The recent rise of practical implementations as well as research activity in probabilistic programming has renewed the need for semantics to help us share insights and innovations.

This workshop aims to bring programming-language and machine-learning researchers together to advance the semantic foundations of probabilistic programming. Topics include but are not limited to:

  • the denotational semantics of probabilistic functions, open universe, loops, and conditioning;
  • the operational semantics of sampling, exact inference, and MCMC transitions;
  • axiomatic and equational reasoning;
  • types and polymorphism;
  • and last but not least, how semantics informs any aspect of probabilistic programming, be it design, theory, implementation, or applications.

Accepted Presentations

Title
A weakest pre-expectation semantics for mixed-sign expectations
PPS
An application of computable distributions to the semantics of probabilistic programs: part 2
PPS
An exponential family basis for probabilistic programming
PPS
Building inference algorithms from monad transformers
PPS
Commutativity logic for probabilistic trace equivalence: complete or not?
PPS
Efficient exact inference in discrete Anglican programs
PPS
Encapsulating models and approximate inference programs in probabilistic modules
PPS
Exchangeable random process and data abstraction
PPS
GraPPa: spanning the expressivity vs. efficiency continuum
PPS
Mathematical structures of probabilistic programming
PPS
Metropolis-Hastings for mixtures of conditional distributions
PPS
On computable representations of exchangeable data
PPS
ProbLog and applicative probabilistic programming
PPS
Probabilistic logic programs: unifying program trace and possible world semantics
PPS
Probabilistic programming and a domain theoretic approach to Skorohod's theorem
PPS
Reasoning about inference in probabilistic programs
PPS
Reducing probabilistic choice to nondeterministic choice
PPS
Support and influence analysis for visualizing posteriors of probabilistic programs
PPS
Synthetic topology in homotopy type theory for probabilistic programming
PPS
The extended semantics for probabilistic programming languages
PPS
The semantics of subroutines and iteration in the Bayesian programming language ProBT
PPS
Towards a metric semantics for probabilistic programming (invited talk)
PPS

Call for extended abstracts

We expect this workshop to be informal, and our goal is to foster collaboration and establish common ground. Thus, the proceedings will not be a formal or archival publication, and we expect to spend only a portion of the workshop day on traditional research talks. Nevertheless, as a concrete basis for fruitful discussions, we call for extended abstracts describing specific and ideally ongoing work on probabilistic programming semantics.

Extended abstracts are up to 2 pages in PDF format. Please submit them by October 31 using EasyChair: https://easychair.org/conferences/?conf=pps2017

You're viewing the program in a time zone which is different from your device's time zone - change time zone

Tue 17 Jan
Times are displayed in time zone: Amsterdam, Berlin, Bern, Rome, Stockholm, Vienna change

10:30 - 12:00
Session 2 PPS at Salle 105, Barre 44-54
Chair(s): Chad ScherrerGalois, Inc.
10:30
20m
Talk
An application of computable distributions to the semantics of probabilistic programs: part 2
PPS
Daniel HuangHarvard University, Greg MorrisettCornell University
10:50
10m
Meeting
Discussion 1
PPS
11:00
20m
Talk
Probabilistic programming and a domain theoretic approach to Skorohod's theorem
PPS
11:20
10m
Meeting
Discussion 2
PPS
11:30
20m
Talk
Building inference algorithms from monad transformers
PPS
Adam ŚcibiorUniversity of Cambridge, Yufei CaiUniversity of Tübingen, Germany, Klaus OstermannUniversity of Tübingen, Germany, Zoubin GhahramaniUniversity of Cambridge
11:50
10m
Meeting
Discussion 3
PPS
14:00 - 15:30
Session 3 PPS at Salle 105, Barre 44-54
Chair(s): Sam StatonUniversity of Oxford
14:00
20m
Talk
Commutativity logic for probabilistic trace equivalence: complete or not?
PPS
Paul Blain Levy, Nathan BowlerUniversität Hamburg
14:20
10m
Meeting
Discussion 4
PPS
14:30
20m
Talk
Mathematical structures of probabilistic programming
PPS
Ilias GarnierUniversity of Edinburgh, Fredrik DahlqvistUniversity College London, Florence ClercMcGill University, Vincent DanosENS Paris/CNRS
14:50
10m
Meeting
Discussion 5
PPS
15:00
20m
Talk
A weakest pre-expectation semantics for mixed-sign expectations
PPS
Benjamin Lucien KaminskiRWTH Aachen University, Joost-Pieter KatoenRWTH Aachen University
15:20
10m
Meeting
Discussion 6
PPS
15:30 - 16:30
Poster Session PPS at Salle 105, Barre 44-54
15:30
60m
Meeting
ProbLog and applicative probabilistic programming
PPS
Alexander VandenbrouckeKU Leuven, Belgium, Tom SchrijversKU Leuven
15:30
60m
Meeting
Encapsulating models and approximate inference programs in probabilistic modules
PPS
Marco Cusumano-TownerMIT-CSAIL, Vikash MansinghkaMassachusetts Institute of Technology
15:30
60m
Meeting
The extended semantics for probabilistic programming languages
PPS
Siddharth SrivastavaUTRC Berkeley, Nicholas HayVicarious, Yi WUUC Berkeley, Stuart RussellUniversity of California, Berkeley
15:30
60m
Meeting
Synthetic topology in homotopy type theory for probabilistic programming
PPS
15:30
60m
Meeting
Reasoning about inference in probabilistic programs
PPS
Chandrakana NandiUniversity of Washington, USA, Adrian SampsonCornell University, Dan GrossmanUniversity of Washington, Todd Mytkowicz, Kathryn S McKinleyMicrosoft Research
15:30
60m
Meeting
On computable representations of exchangeable data
PPS
Nathanael L. AckermanHarvard University, Jeremy AvigadCarnegie Mellon University, Cameron FreerGamalon and Borelian, Daniel Roy, Jason M. RutePennsylvania State University
15:30
60m
Meeting
Probabilistic logic programs: unifying program trace and possible world semantics
PPS
Angelika KimmigKU Leuven, Luc De RaedtKU Leuven
15:30
60m
Meeting
Metropolis-Hastings for mixtures of conditional distributions
PPS
15:30
60m
Meeting
Support and influence analysis for visualizing posteriors of probabilistic programs
PPS
Long OuyangStanford University
15:30
60m
Meeting
Efficient exact inference in discrete Anglican programs
PPS
Robert CornishUniversity of Oxford, Frank WoodUniversity of Oxford, Hongseok YangUniversity of Oxford
16:30 - 18:00
Session 5 PPS at Salle 105, Barre 44-54
Chair(s): Chung-chieh ShanIndiana University, USA
16:30
20m
Talk
An exponential family basis for probabilistic programming
PPS
Chad ScherrerGalois, Inc.
16:50
10m
Meeting
Discussion 7
PPS
17:00
20m
Talk
The semantics of subroutines and iteration in the Bayesian programming language ProBT
PPS
17:20
10m
Meeting
Discussion 8
PPS
17:30
20m
Talk
Exchangeable random process and data abstraction
PPS
Sam StatonUniversity of Oxford, Hongseok YangUniversity of Oxford, Nathanael L. AckermanHarvard University, Cameron FreerGamalon and Borelian, Daniel Roy
17:50
10m
Meeting
Discussion 9
PPS
18:15 - 19:15
Session 6 PPS at Salle 105, Barre 44-54
Chair(s): Hongseok YangUniversity of Oxford
18:15
20m
Talk
Reducing probabilistic choice to nondeterministic choice
PPS
Ernie CohenAmazon Web Services
18:35
10m
Meeting
Discussion 10
PPS
18:45
20m
Talk
GraPPa: spanning the expressivity vs. efficiency continuum
PPS
Edwin WestbrookGalois, Inc., Chad ScherrerGalois, Inc., Nathan CollinsGalois, Inc., Eric MertensGalois, Inc.
19:05
10m
Meeting
Discussion 11
PPS