Quantum causal inference banner image

Quantum causal inference initiative

The Perimeter Institute’s Quantum Causal Inference Lab establishes Canada as a world leader in this emerging, high-potential field. Quantum causal inference transforms machine learning, moving artificial intelligence beyond pattern-finding to revealing and predicting complex networks of cause and effect.

The physicists working in Perimeter’s Quantum Foundations group have driven major recent progress in causal inference. Their success demonstrated the viability for a standalone lab with an audacious agenda, involving:

  • developing an industry-relevant “unified theory of causal inference”
  • building a Canadian powerhouse for research in quantum causal inference
  • positioning Canada as global leader in this emerging science
  • creating a nexus between real-world challenges and foundational research through collaboration with private industry partners

As global competition increases in this promising research area, we are making Canada a beacon for the world’s top talent, capable of pushing quantum causal inference research further and faster, and fostering deeper collaborations between Perimeter scientists and artificial intelligence researchers at other institutions.

Quantum causal inference can lead to algorithms that extract cause-and-effect insights from large statistical datasets, more powerfully and authoritatively than any existing technology.

Three men standing in a library having a conversation together
Tobias Fritz, Rob Spekkens, and Elie Wolfe

The Quantum Causal Inference Lab will lead this area by:

  • recruiting and retaining the best researchers in the world
  • integrating AI research to accelerate applied use of quantum and quantum inspired causal inference algorithms
  • contributing to Canada’s overall capacity to address industrial and societal challenges
  • providing scientists with direct opportunities to maximize research productivity and to apply their discoveries in partnership with governments, companies, and other organizations

 

Researchers

 
Faculty:
Research Associate:
Graduate Students:
 

 

Collaborators

  • Jonathan Barrett, Oxford University, Oxford, UK
  • Rafael Chaves, International Institute of Physics, Natal, Brazil
  • Xavier Coiteux-Roy, University of Lugano, Switzerland
  • Noam Finkelstein, Department of Computer Science, John Hopkins University, Baltimore, MD
  • Tobias Fritz, Department of Mathematics, University of Innsbruck, Austria
  • Ciaran Lee, Senior Research Scientist at Spotify, Honorary Associate Professor at UCL
  • Alex Posaz, Department of Mathematics, University of Madrid, Spain
  • Marc-Olivier Renou, ICFO, Barcelona, Spain
  • Kevin Resch, Department of Physics and Astronomy, University of Waterloo and Institute for Quantum Computing, Waterloo, Canada
  • John Selby, University of Gdansk, Gdansk, Poland
  • Ilya Shpitser, Department of Computer Science, John Hopkins University, Baltimore, MD

Entropic Inequality Constraints from e-separation Relations in Directed Acyclic Graphs with Hidden Variables
Noam Finkelstein, Beata Zjawin, Elie Wolfe, Ilya Shpitser, Robert W. Spekkens
Proc. UAI 2021
 
Unscrambling the omelette of causation and inference: The framework of causal-inferential theories
David Schmid, John H. Selby, and Robert W. Spekkens
[arXiv:2009.03297 (quant-ph)]
 
Quantum Inflation: A General Approach to Quantum Causal Compatibility
Elie Wolfe, Alejandro Pozas-Kerstjens, Matan Grinberg, Denis Rosset, Antonio Acín, Miguel Navascues
 
The Inflation Technique Solves Completely the Classical Inference Problem
Miguel Navascues and Elie Wolfe
 
The Inflation Technique for Causal Inference with Latent Variables
Elie Wolfe, Robert W. Spekkens, and Tobias Fritz
J. Causal Inference 7(2), (2019)
[arXiv:1609.00672 (quant-ph)]
 
Quantum common causes and quantum causal models
John-Mark A. Allen, Jonathan Barrett, Dominic C. Horsman, Ciaran M. Lee, and Robert W. Spekkens
 
The lesson of causal discovery algorithms for quantum correlations: Causal explanations of Bell-inequality violations require fine-tuning
Christopher J. Wood and Robert W. Spekkens
New J. Phys. 17, 033002 (2015)
[arXiv:1208.4119 (quant-ph)]

A quantum advantage for inferring causal structure
Katja Ried, Megan Agnew, Lydia Vermeyden, Dominik Janzing, Robert W. Spekkens, Kevin J. Resch
 
 

Lectures and talks