Hi, I'm Joey Bose,

About Me

Joey Bose is a Post-Doctoral Fellow at University of Oxford working with Michael Bronstein, a Distinguished Research scientist at Dreamfold, and an Affiliate member of Mila. He completed his PhD at McGill/Mila under the supervision of Will Hamilton, Gauthier Gidel, and Prakash Panagaden. His research interests span Generative Modelling, Differential Geometry for Machine Learning with a current emphasis on geometric generative models for scientific applications. Previously, he completed his Bachelors and Master’s degrees from the University of Toronto working on adversarial attacks against face detection and is the President and CEO of FaceShield Inc an educational platform for digital privacy for facial data. His work has been featured in Forbes, New York Times, CBC, VentureBeat and other media outlets and was generously supported by the IVADO PhD Fellowship, and NSERC Post-doc Fellowship.

Research Interests

Generative Models
Normalizing Flows
Equivariant Networks
Adversarial Attacks
Graph Neural Networks
Pytorch
Pytorch-Geometric
Fall 2022 COMP760 Geometry and Generative Models

Education & Experience

For more information, have a look at my curriculum vitae .

  • Mila Sept 2024 - Present
    Affiliate Member
  • University of Oxford Jan 2024 - Present
    Post-doc
    Geometric Deep Learning Generative Models
  • Dreamfold July 2023 - Present
    Distinguished Research Scientist
    Proteins Generative models and Sampling
  • McGill and Mila Sept 2018 - Sept 2023
    Phd Student
    Geometric Deep Learning Generative Models Equivariant Networks Adversarial Attacks
  • Qualcomm AI Research Oct 2022 - Feb 2023
    Research Intern
    Generative Models Geometry
  • Co-Instructor at McGill Sept 2022 - Dec 2022
    Teaching COMP760
    Generative Models Geometry
  • Facebook AI Research May 2021 - Dec 2021
    Research Intern
    Generative Models Causality
  • FaceShield.ai (acquired) August 2018 - June 2020
    Founder and CEO
    Adversarial Attacks on Face Detection
  • Uber AI May 2019 - Aug 2019
    Research Intern
    Graph Neural Networks Meta-Learning
  • University of Toronto Sept 2017 - Aug 2018
    Masters of Applied Science Computer Engineering
    Adversarial Attacks on Face Detection
  • Borealis AI May 2017 - Jul 2018
    Resarch Intern
    NLP Embeddings Negative Sampling
  • University of Toronto 2012 - 2017
    Bachelors of Applied Science in Computer Engineering and Minor in Mechatronics

Publications and Pre-Prints

The Superposition of Diffusion Models

Preprint 2024

M. Skreta, L. Atanackovic, A. J. Bose, A. Tong, K. Neklyudov

arxiv

Steering Masked Discrete Diffusion Models via Discrete Denoising Posterior Prediction

Preprint 2024

J. Rector-Brooks, M. Hasan, Z. Peng, Z. Quinn, C. Liu, S. Mittal, N. Dziri, M. Bronstein, Y. Bengio, P. Chatterjee, A. Tong*, A. J. Bose*

arxiv

Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Backbone Generation

NeurIPS 2024

G. Huguet*, J. Vuckovic*, K. Fatras, E. Thibodeau-Laufer, P. Lemos, R. Islam, C. Liu, J. Rector-Brooks, T. Akhound-Sadegh, M. Bronstein, A. Tong, A. J. Bose

arxiv

Self-Consuming Generative Models with Curated Data Provably Optimize Human Preferences

NeurIPS 2024 (Spotlight)

D. Ferbach, Q. Bertrand, A. J. Bose, G. Gidel

arxiv

Metric Flow Matching for Smooth Interpolations on the Data Manifold

NeurIPS 2024

K. Kapusniak, P. Potaptchik, T. Reu, L. Zhang, A. Tong, M. Bronstein, A. J. Bose, F. D. Giovanni

arxiv

Fisher Flow Matching for Generative Modeling over Discrete Data

NeurIPS 2024

O. Davis, S. Kessler, M. Petrache, I. I. Ceylan, M. Bronstein, A. J. Bose

arxiv

Iterated Denoising Energy Matching for Sampling from Boltzmann Densities

ICML 2024

T. Akhound-Sadegh*, J. Rector-Brooks*, A.J. Bose*, S. Mittal, P. Lemos, C.H. Liu, M. Sendera, S. Ravanbaksh, G. Gidel, Y. Bengio, N. Malkin, A. Tong

arxiv

A Geometric Perspective to Deep Generative Models

Phd Thesis 2023

A. J. Bose

arxiv

SE(3) Stochastic Flow Matching for Protein For Protein Backbone Generation

ICLR 2024 (Spotlight)

A. J. Bose*, T. Akhound-Sadegh*, K. Fatras, G. Huguet, J. Rector-Brooks, C.H. Liu, A.C. Nica, M. Korablyov, M. Bronstein, A. Tong

arxiv

On the Stability of Iterative Retraining of Generative Models on their own Data

ICLR 2024 (Spotlight)

Q. Bertrand, A. J. Bose, A. Duplessis, M. Jiralerspong, G. Gidel

arxiv

EDGI, Equivariant diffusion for planning with embodied agents

NeurIPS 2023

J. Bremer*, A. J. Bose*, P. de Haan, T. Cohen

arxiv
Generative Models Equivariance

Feature Likelihood Score, Evaluating Generalization of Generative Models Using Samples

NeurIPS 2023

M. Jiralerspong, A. J. Bose, G. Gidel

arxiv
Generative Models

A General Framework For Proving The Equivariant Strong Lottery Ticket Hypothesis

ICLR 2023

D. Ferbach*, C. Tsirigotis*, G. Gidel, A.J. Bose

arxiv
Strong Lottery Tickets Equivariant Networks Pruning

Riemannian Diffusion Models

NeurIPS 2022

C.W. Huang*, M. Aghajohari*, A.J. Bose, P. Panangaden, A. Courville

arxiv
Diffusion Models Riemmanian Geometry Generative Models

Controllable Generative Modeling via Causal Reasoning

TMLR 2022

A.J. Bose, R. Monti, A. Grover

arxiv
Causality Generative Models

Equivariant Finite Normalizing Flows

Preprint 2022

A.J. Bose, I. Kobyzev, M. Brubaker

Equivariant Normalizing Flows Generative Models

Matching Normalizing Flows and Probablity Paths on Manifolds

ICML 2022

H. Ben-Hamu*, S. Cohen*, A.J. Bose, B. Amos, M. Nickel, A. Grover, R. Chen, Y. Lipman

arxiv
Continuous Normalizing Flows Generative Models

Online Adversarial Attacks

ICLR 2022

A. Mladenović*, A.J. Bose*, H. Berard*, W.L. Hamilton, S. Lacoste-Julien, P. Vincent, G.Gidel

arxiv
Adversarial Attacks Online Algorithms k-secretary

Neural Path Hunter, Reducing Hallucination in Dialogue Sytems via Path Grounding

EMNLP 2021

N. Dziri, A. Madotto, O. Zaiane, A.J. Bose

arxiv
Hallucination Dialogue Systems Knowledge Graphs

Adversarial Example Games

NeurIPS 2020

A.J. Bose*, G. Gidel*, H. Berard*, A. Cianflone, P. Vincent, S. Lacoste-Julien, W.L. Hamilton

arxiv git blog
Adversarial Attacks Game Theory

Structure Aware Negative Sampling in Knowledge Graphs

EMNLP 2020

K. Ahrabian, A. Feizi, Y. Salehi, W. L. Hamilton, A. J. Bose

arxiv git
Knowledge Graphs Negative Sampling

Latent Variable Modeling with Hyperbolic Normalizing Flows

ICML 2020

A. J. Bose, A. Smofsky, R. Liao, P. Panangaden, W. L. Hamilton

arxiv git blog
Normalizing Flows Hyperbolic Geometry

Meta-Graph Few Shot Link Prediction Via Meta-Learning

Preprint (Unpublished)

A. J. Bose, A. Jain, P. Molino, W. L. Hamilton

arxiv git
Link Prediction Graph Neural Networks Meta-Learning

Compositional Fairness Constraints for Graph Embeddings

ICML 2019

A. J. Bose, W. L. Hamilton

arxiv git
Graph Neural Networks Fairness Node Embeddings

Discrete off-policy policy gradient using continuous relaxations

RLDM 2019

A. Cianflone, Z. Ahmed, R. Islam, A. J. Bose, W. L. Hamilton

arxiv
Policy Gradient off-policy actor-critic continuous relaxation

A Cross-Domain Transferable Neural Coherence Model

ACL 2019

X. Peng, H. Saghir, J. Kang, T. Long, A. J. Bose, Y. Cao, J. Cheung

arxiv
Coherence Modelling NLP

Adversarial Attacks on Face Detectors using Neural Net based Constrained Optimization

IEEE MMSP (Best Paper Runner-up)

A. J. Bose, P. Aarabi

arxiv
Adversarial Attacks Face Detection

Adversarial Contrastive Estimation

ACL 2018 (ORAL)

A. J. Bose*, Y. Cao*, H. Ling*

arxiv
Negative Sampling Embeddings Adversarial Sampling

Contact

If you like what I do and want to potentially collaborate feel free to reach out via email. Note for prospective students interested in joining Mila/McGill/Oxford I cannot assess your candidacy and will not be able to help you get an interview with my advisor. But, I can offer general advice about applying to grad school and the culture of our lab.