Scholar
Tara Akhound-Sadegh
Google Scholar ID: RHDoTkkAAAAJ
McGill University, Mila
Citations & Impact
Citations
303
H-index
6
i10-index
5
Publications
10
Co-authors
17
list available
Contact
Publications
Resume (English only)
Academic Achievements
- Selected Publications: Progressive Inference-Time Annealing of Diffusion Models for Sampling from Boltzmann Densities, NeurIPS 2025, Spotlight; Feynman-Kac Correctors in Diffusion: Annealing, Guidance, and Product of Experts, ICML 2025; Sequence-Augmented SE(3)-Flow Matching For Conditional Protein Backbone Generation, Neurips 2024; Iterated Denoising Energy Matching for Sampling from Boltzmann Densities, ICML 2024; SE(3)-Stochastic Flow Matching for Protein Backbone Generation, ICLR 2024, Spotlight; Lie Point Symmetry and Physics-Informed Networks, Neurips 2023.
Research Experience
- Visiting Researcher @ Oxford University, Michael Bronstein's group, October 2024 - March 2025; ML Scientist in Residence @ Dreamfold, August 2023 - October 2024.
Education
- PhD (in progress) - McGill University & Mila, Computer Science, January 2022 - present, GPA: 4.0/4.0; MSc - McGill University & Mila, Computer Science, September 2020 - January 2022 (fast-tracked to PhD), GPA: 4.0/4.0; Master 1 - L’Université Paris Diderot, Mathématiques Fondamentales et Appliquées, Sept 2019 - June 2020; BSc - University of British Columbia, Engineering Physics, Minor in Honours Mathematics, September 2014 - May 2019.
Background
- Research Interests: Generative models, sampling, geometric deep learning, and applications to biology.
