Researcher in Applied Mathematics & Machine Learning
Master MVA (Mathematics, Vision & Learning) - École Normale Supérieure Paris-Saclay
Research at École Normale Supérieure (Ulm) - IBENS
Under review at ICLR 2026
Extending Schrödinger bridges to multiple marginals for reconstructing temporal evolution from static data snapshots. Applications in video generation, cellular differentiation, and disease progression modeling.
- Generative Modeling - VAEs, Diffusion Models, GANs
- Optimal Transport - Schrödinger Bridges, Multi-marginal OT
- Information Geometry - Manifold Learning, Riemannian Optimization
- Applied Probability - Stochastic Processes, MCMC Methods
- Scientific Computing - Physics-Informed ML, Inverse Problems
- 1st Place - EchoCem Data Challenge (Collège de France)
- Highest Honours - Master MVA (17.37/20 average)
- ICLR 2026 - Paper submission on Multi-marginal Schrödinger Bridges
- EDF Excellence Scholarship - Mathematical Sciences
Research Intern - École Normale Supérieure (Ulm) - IBENS (2025)
Multi-marginal Schrödinger Bridges for video generation from unpaired data
Research Engineer - French Space Agency (CNES) (2023-2024)
Inverse problems, Physics-Informed ML, signal processing for aerospace
Research Intern - French Space Agency (CNES) (2022-2023)
Capillary fluid dynamics models tested in microgravity
Master 2 MVA - École Normale Supérieure Paris-Saclay (2024-2025)
Mathematics, Vision & Learning - Highest Honours
Master in Engineering - Arts et Métiers (2020-2023)
Applied Mathematics, Optimization, Modeling & Simulation - Top 10%
- Video Generation - Multi-marginal Schrödinger bridges for temporal dynamics
- Medical Imaging - Generative vs Discriminative robustness (MedMNIST)
- Audio Processing - Wave-U-Net for source separation
- Topological Data Analysis - PersLay neural networks
- Geophysics - EchoCem ultrasonic well image segmentation
- Computational Statistics - MCMC methods and optimization
Languages: Python, MATLAB, C++, R
ML/DL: PyTorch, TensorFlow, scikit-learn, JAX
Scientific: NumPy, SciPy, Matplotlib, Seaborn
Tools: Git, Docker, LaTeX, Jupyter, VS Code
When not diving into mathematics and machine learning, you'll find me:
- Sailing expeditions across Mediterranean waters
- Paragliding over Alpine landscapes
- Mountaineering and technical climbing
- Trekking in remote mountain ranges
Website: tgravier.github.io
Email: thomas.gravier@gadz.org
Google Scholar: Academic Profile
LinkedIn: thomas-gravier
Seeking PhD opportunities in Machine Learning Theory, Stochastic Processes, or Optimal Transport