Scholar
Jose Gallego-Posada
Google Scholar ID: tfKnkRQAAAAJ
Mila - Université de Montréal
Citations & Impact
Citations
249
H-index
8
i10-index
8
Publications
20
Co-authors
10
list available
Publications
Resume (English only)
Academic Achievements
- - Published multiple papers including 'Feasible Learning' (AISTATS 2025) and 'On PI Controllers for Updating Lagrange Multipliers in Constrained Optimization' (ICML 2024)
- - First last-author paper 'Balancing Act: Constraining Disparate Impact in Sparse Models' accepted at ICLR 2024
- - Delivered talks at multiple international conferences such as ICML 2024, ICLR 2024, NeurIPS 2025, etc.
- - Released version 1.0.0 of the Cooper library
- - Served as area chair for workshops at NeurIPS 2025
Research Experience
- - Doctoral researcher at Mila and Université de Montréal
- - One of the lead developers of Cooper, an open-source library for non-convex constrained optimization in PyTorch
- - Speaker or organizer at multiple international conferences such as NeurIPS, ICML, AISTATS
- - Visiting researcher at Meta, working on scalable adaptive optimization methods
Education
- - PhD, Université de Montréal and Mila, supervised by Simon Lacoste-Julien, partially supported by an IVADO PhD Excellence Scholarship
- - MSc in Artificial Intelligence, University of Amsterdam, supervised by Patrick Forré, graduated in 2018
- - BSc in Mathematical Engineering, Universidad EAFIT, Medellin
Background
- Research interests include constrained optimization problems, adaptive optimization, equivariant deep learning, geometric information theory, federated learning, and applications of algebraic geometry to machine learning.
Miscellany
- Dijkstra and Erdős numbers are both 4
