Benjamin Bloem-Reddy

I am Assistant Professor of Statistics at the University of British Columbia. I work on problems in statistics and machine learning, with an emphasis on probabilistic approaches. Most of my ongoing work is motivated by some form of (scientific) knowledge acquisition, and the interrelated roles of models, data, inference, and prediction. Some recent examples:

• Uses and benefits of symmetry in modeling, inference, and computation.
• Statistical hypothesis tests for symmetry.
• Integrating probability, statistics, and data with scientific models.
• The role of causality, broadly construed, in learning from observations of and interactions with the world.

I also collaborate with researchers in the sciences on statistical problems arising in their research.
My research is supported by funding from NSERC, CANSSI, and UBC, and by computational resources and services provided by Advanced Research Computing at UBC.

I was a PhD student with Peter Orbanz at Columbia and a postdoc with Yee Whye Teh in the CSML group at the University of Oxford. Before moving to statistics and machine learning, I studied physics at Stanford University and Northwestern University.

Contact: benbr at stat dot ubc dot ca
Office: Department of Statistics, Earth Sciences Building, Room 3168

Research group

Current (alphabetical order)

• Boyan Beronov, Computer Science PhD student (with Anne Condon)
• Kenny Chiu, PhD student
• Gian Carlo Diluvi (with Trevor Campbell), PhD student
• Sebastian Gonzalez, Undergraduate student
• Kevin Lam (with Yongjin Park), PhD student
• Alex Sharp (with Chris Maddison), CANSSI Distinguished Postdoctoral Fellow
• Quanhan (Johnny) Xi, PhD student (NSERC CGS-D Scholarship)

Past (reverse chronological order)

• Daniel Kennedy, NSERC USRA
• Liam Gilson, MSc student while on leave from Forestry PhD
• William Laplante (with Daniel J. McDonald), MSc student
• Quanhan (Johnny) Xi, MSc student (2022) → PhD student, UBC Statistics
• Weijia (Grace) Yin, MSc (2022) → BC Centre for Excellence in HIV/AIDS
• Kenny Chiu, MSc (2021) → PhD student, UBC Statistics
• Gian Carlo Diluvi (with Trevor Campbell), MSc (2021) → PhD student, UBC Statistics
• Sean La (with Alexandre Bouchard-Côté), MSc (2021) → Data Scientist, Kabam Games

Papers

Pre-prints

• Non-parametric Hypothesis Tests for Distributional Group Symmetry
  K. Chiu and B. Bloem-Reddy
  [arxiv] [pdf] [code]

• Uncertainty in Neural Processes
  S. Naderiparizi, K. Chiu, B. Bloem-Reddy, F. Wood
  [arxiv]

• On the Benefits of Invariance in Neural Networks
  C. Lyle, M. van der Wilk, M. Kwiatkowska, Y. Gal, B. Bloem-Reddy
  [arxiv]

• Preferential Attachment and Vertex Arrival Times
  B. Bloem-Reddy and P. Orbanz
  [arxiv]
  (Sampling and Inference for Beta Neutral-to-the-Left Models of Sparse Networks develops inference methods and applies these models to data.)

Published and to appear

• Mixed Variational Flows for Discrete Variables
  Gian Carlo Diluvi, B. Bloem-Reddy, and Trevor Campbell
  AISTATS 2024
  [arxiv] [code]

• Indeterminacy in Generative Models: Characterization and Strong Identifiability
  Q. Xi and B. Bloem-Reddy
  AISTATS 2023 (Oral presentation and notable paper)
  [aistats] [arxiv]

• Lossy Compression for Lossless Prediction
  Y. Dubois, B. Bloem-Reddy, K. Ullrich, C. J. Maddison
  NeurIPS 2021 (Spotlight)
  [neurips] [arxiv] [code] [Neural Compression Workshop @ ICLR 2021 version (oral presentation)]

• Probabilistic symmetry and invariant neural networks
  B. Bloem-Reddy and Y. W. Teh
  Journal of Machine Learning Research
  [jmlr] [arxiv] [slides from a talk on this work] [Yee Whye talked about this in his IMS Medallion Lecture]

• Sampling and Inference for Beta Neutral-to-the-Left Models of Sparse Networks
  B. Bloem-Reddy, A. Foster, E. Mathieu, Y. W. Teh
  UAI 2018
  [uai] [arxiv] [code]

• Discussion of F. Caron and E. B. Fox, "Sparse graphs using exchangeable random measures"
  B. Bloem-Reddy
  Journal of the Royal Statistical Society, Series B (Statistical Methodology), 79(5)
  [jrss b] [pdf] [slides from discussion at RSS meeting]

• Random-walk models of networks formation and sequential Monte Carlo methods for graphs
  B. Bloem-Reddy and P. Orbanz
  Journal of the Royal Statistical Society, Series B (Statistical Methodology), 80(5): 871-898
  [jrss b] [arxiv] [code]
  [A talk on this work] [Another talk on this work, by Peter Orbanz]

• Slice Sampling on Hamiltonian Trajectories
  B. Bloem-Reddy and J. P. Cunningham
  ICML 2016: JMLR W+CP
  [pdf] [icml/jmlr] [A talk on this work]

• Superfluid Phase Stability of ³He in Axially Anisotropic Aerogel
  J. Pollanen, J. P. Davis, B. Reddy, K. R. Shirer, H. Choi, W. P. Halperin
  Journal of Physics: Conference Series, 150(3), 032084
  [iop]

• Stability of the axial phase of superfluid ³He in aerogel with globally anisotropic scattering
  J. P. Davis, J. Pollanen, B. Reddy, K. R. Shirer, H. Choi, W. P. Halperin
  Physical Review B 77, 140502(R)
  [aps] [arxiv]

In progress (comments welcome)

• Beauty in Machine Learning: Fluency and Leaps
  I (Still) Can't Believe It's Not Better! Workshop at NeurIPS 2021 (selected for contributed talk)
  [paper] [poster]

Workshop contributions

(Most of these have a more fully developed counterpart above)

• Multiple Environments Can Reduce Indeterminacy in VAEs
  Q. Xi and B. Bloem-Reddy
  NeurIPS Workshop on Causal Inference & Machine Learning: Why now? (WHY-21)
  [paper] [poster]

• Generalization Bounds for Invariant Neural Networks
  C. Lyle, B. Bloem-Reddy, Y. Gal, M. Kwiatkowska
  NeurIPS 2019 Workshop on Machine Learning with Guarantees

• Neural network models of exchangeable sequences
  B. Bloem-Reddy and Y. W. Teh
  NeurIPS 2018 Workshop on Bayesian Deep Learning
  [pdf] [slides from a talk on this work]
• Sequential sampling of Gaussian process latent variable models
  M. Tegner, B. Bloem-Reddy, S. Roberts
  ICML 2018 Workshop on Tractable Probabilistic Models
  [arxiv]
• Sampling and inference for discrete random probability measures in probabilistic programs
  B. Bloem-Reddy*, E. Mathieu*, A. Foster, T. Rainforth, Y. W. Teh, M. Lomeli, H. Ge, Z. Ghahramani
  NeurIPS 2017 Workshop on Advances in Approximate Bayesian Inference
  [pdf] [poster]
• Random walk models of sparse graphs and networks
  B. Reddy and P. Orbanz
  NeurIPS 2014 Workshop on Networks: From Graphs to Rich Data. Best student poster award.
  

Teaching

Current/upcoming

• Fall 2023: STAT 460/560, Theory of Statistical Inference I

• Spring 2024: STAT 305, Introduction to Statistical Inference
  Course website on Canvas

• STAT 548 qualifying papers
  [guide, expectations, and list of potential papers]

Past courses

• Fall 2019, 2020, 2021, 2022: STAT 547C, Topics in Probability
  [course website: 2021, 2022]

• Summer 2020, Spring 2020, 2021, 2022, 2023: STAT 305, Introduction to Statistical Inference

• Spring 2023: STAT 547S, Topics on Symmetry in Statistics and Machine Learning
  [course outline] [course notes]

Notes

• Exchangeable random partitions and random discrete probability measures: a brief tour guided by the Dirichlet Process
  B. Bloem-Reddy
  Notes for a lecture given to Oxford PhD students (these are a work in progress)
  [pdf]