Lazar Atanackovic

I am a Ph.D. Candidate at the University of Toronto in The Department of Electrical & Computer Engineering and the Vector Institute.

laz_pic.jpg

I am a member of the Probabilistic and Statistical Inference Group supervised by Brendan Frey and co-supervised by Bo Wang. Previously, I was a machine learning research intern at Recursion Pharmaceuticals / Valence Labs and a research intern at MILA.

My research focus is in developing machine learning methods for deciphering and modelling cell dynamics from data. I am interested in generative modelling, dynamical systems, structure learning / causal discovery, and computational biology.

selected publications

  1. mfm.png
    Meta Flow Matching: Integrating Vector Fields on the Wasserstein Manifold
    Lazar Atanackovic*, Xi Zhang* , Brandon Amos , Mathieu Blanchette , Leo J Lee , Yoshua Bengio , Alexander Tong , and Kirill Neklyudov
    In ICML Workshop on Geometry-grounded Representation Learning and Generative Modeling , 2024
  2. gflownet_gen_preview.png
    Investigating Generalization Behaviours of Generative Flow Networks
    Lazar Atanackovic, and Emmanuel Bengio
    In ICML Workshop on Structured Probabilistic Inference & Generative Modeling [Oral] , 2024
  3. wlfs.png
    A Computational Framework for Solving Wasserstein Lagrangian Flows
    Kirill Neklyudov* , Rob Brekelmans* , Alexander Tong , Lazar Atanackovic, Qiang Liu , and Alireza Makhzani
    In International Conference on Machine Learning (ICML) , 2024
  4. sf2m.png
    Simulation-free Schrodinger Bridges via Score and Flow Matching
    Alexander Tong* , Nikolay Malkin* , Kilian Fatras* , Lazar Atanackovic, Yanlei Zhang , Guillaume Huguet , Guy Wolf , and Yoshua Bengio
    In Artificial Intelligence and Statistics (AISTATS) , 2024
  5. dyngfn.png
    DynGFN: Towards Bayesian Inference of Gene Regulatory Networks with GFlowNets
    Lazar Atanackovic*, Alexander Tong* , Bo Wang , Leo J Lee , Yoshua Bengio , and Jason Hartford
    Advances in Neural Information Processing Systems (NeurIPS), 2023