At the Principles of Programming Languages (POPL) conference in January, Facebook launched the Probability and Programming request for proposals and invited the academic community to respond. We were interested in proposals that addressed fundamental problems at the intersection of machine learning, programming languages, and software engineering, including differentiable programming, probabilistic programming, languages and tools for data science, programming tools built using “big code,” and applications of machine learning to optimize systems and human workflows.

We received 66 submissions, and the selection committee was composed of programming language experts throughout Facebook Research. It was a challenge to determine winners among so many high-quality proposals. Thank you to all the researchers who took the time to submit a proposal, and congratulations to the 10 winners.

Research award winners

Differentiable Probabilistic Programming Semantics
Samuel Staton (University of Oxford), Jesse Sigal (University of Oxford), Luke Ong (University of Oxford), Maria Gorinova (University of Edinburgh), Matthijs Vákár (Columbia University), Ohad Kammar (University of Edinburgh)

Opening Up the Black Box of Probabilistic Program Inference
Todd Millstein (UCLA), Guy Van den Broeck (UCLA)

A Probabilistic Domain-Specific Language for Common-Sense Data Cleaning
Vikash Mansinghka (MIT), Alex Lew (MIT)

Zipper Code Embeddings
Aws Albarghouthi (University of Wisconsin-Madison), Somesh Jha (University of Wisconsin-Madison)

Programs as Differentiable Data Objects
Thomas W. Reps (University of Wisconsin-Madison), Jordan Henkel (University of Wisconsin-Madison)

Contextual Ensemble Learning for Software Productivity and Reliability
Lin Tan (Purdue University)

CODA Deep RL Framework for Code Assistant
Mayur Naik (University of Pennsylvania)

Code Embeddings for Bug Finding
Aditya Thakur (UC Davis), Cindy Rubio González (UC Davis)

Higher-Order Differentiable Programming with Delimited Continuations
Tiark Rompf (Purdue University)

Scalable Variational Inference for Probabilistic Programs
Erik B. Sudderth (UC Irvine)

Runners-up

Structured Neural Code Generation
Eran Yahav (Technion-Israel Institute of Technology)

A Metalanguage for Accelerating Human-in-the-Loop Machine Learning
Aditya Parameswaran (UC Berkeley)

A Strategy Language to Optimize ML Using Machine-Learned Optimization Strategies
Michel Steuwer (University of Glasgow)

Augmenting Deep Learning with Program Synthesis
Madhusudan Parthasarathy (University of Illinois at Urbana-Champaign)

Automatic Test Program Generation for Compilers/Optimizers
Xiaokang Qiu (Purdue University)

Debugging Probabilistic Programming Using Automated Model Criticism
Robert Zinkov (University of Oxford)

Deep Probabilistic Programming for Ocaml
Frank Wood (University of British Columbia)

Differentiable Probabilistic Logic Programming
Fabrizio Riguzzi (University of Ferrara)

Differentiable Probabilistic Programming for Data-Driven Precision Medicine
Alan Edelman (MIT)

Differentiable Programming with Scientific Software, and Beyond
Pontus Stenetorp (University College London)

Differentiating Programs with Differential Linear Logic
Neel Krishnaswami (University of Cambridge)

Fixing Neural Networks with Solver-Aided Languages
Karim Ali (University of Alberta)

Geometric Layers for Deep Neural Networks
Dan Raviv (Tel Aviv University)

Machine-Level Primitives for Probabilistic Programming
Adam Smith (UC Santa Cruz)

Meeting the Social Challenge in Static Analysis Tools
Premkumar T. Devanbu (UC Davis)

Migrating to Pluggable Types with Big Code
Manu Sridharan (UC Riverside)

Piperade – Probabilistic Performance-Aware Deployment of FB Services
Antonio Brogi (University of Pisa)

Probabilistic Learning of Code Semantics
Jeff Huang (Texas A&M University)

Programming Abstractions for Inspecting Neural Networks
Eugene Wu (Columbia University)

Property-Based Testing of Probabilistic Programs
Willard Thor Rafnsson (IT University of Copenhagen)

Protoss: A Tool for Testing and Transforming Probabilistic Programs
Sasa Misailovic (University of Illinois at Urbana-Champaign)

Robust and Reliable Probabilistic Programming
David Blei (Columbia University)

Statistical Synthesis of Novel Programs Using “Big Code”
Swarat Chaudhuri (William Marsh Rice University)

Stochastic Programming: Native Support in Deep Learning
Wray Lindsay Buntine (Monash University)

Stochastic Variational Inference for Probabilistic Programs
Hongseok Yang (KAIST)

System-Wide Debugging Assistant Powered by NLP
Ravi Netravali (UCLA)

XDeep: Detecting and Localizing Cross-Implementation Bugs in Machine Learning Software
Lin Tan (Purdue University)

To view our currently open research awards and to subscribe to our email list, visit our Research Awards page.