Mathias Vigouroux

AI & Cognitive Science Researcher
mathias.vigouroux2@gmail.com
Paris, France
https://github.com/MathiasVigouroux
https://www.linkedin.com/in/mathias-vigouroux/
Download CV (PDF)

Research Topic

Theory of Mind & metacognition, Interpretability & Robustness in AI.

Education

Master in Applied Maths - MVA:

ENS & University Paris-Saclay

2023 - 2024
  • Robusteness of AI
  • Interpretability of AI

The Mathematique, Vision, Apprentissage Master is the most rekown french master for AI.

University Paris-Saclay is ranked 1st in the World for Mathematics by the Shanghai Ranking.

Others topics learned were : Graph neural network, topolological data analysis, optimal transport, RL, NLP

Normalien Science

École Normale Supérieure de Paris - PSL University

2022 - 2025
  • Major in Cognitive Science, Human-Machine Comparison
  • Possible minor : Theater & Cinema
  • Math Lecturer for non-math ENS student

Master 1 CogMaster

École Normale Supérieure de Paris - PSL University

2022 - 2023
  • Major in Modelling for Human Psychology
  • Theory of Machine Learning

The CogMaster is the most rekown french master for Cognitive Science.

Master 2 in Computational Neurosciences

and Neuroengineering

University Paris-Saclay

2021 - 2022
  • Modelling Biological Neurons, dynamical Systems
  • Neuroprostetics
  • Neurophysiology and cognitive neurosciences

Master of Engineering in Statics and Economics

École Nationale de la Statistique et de l'Administration Économique

2021 - 2022
  • Statistical expertise
  • Modelling Rational Human Behavior

First engineer student to become normalien at ENS Paris. I created a novel double diploma. Interview from the the ENSAE.

Professional Experience

Psychophysics Research Intern

LCSP Team - ENS

11/2024 - Present

Research on ability to detect attentional variation in others. Under the Supervision of Jérôme Sackur

AI Research Intern

MILES Team - PRAIRIE - PSL University

04/2024 - 09/2024
  • Developing uncertainty metrics for Large Language Models
  • Direct Supervisor: Jamal Atif and Yann Chevaleyre
  • Master supervisor: Gabriel Peyré

Technical Skills

  • Programming: Python (PyTorch, Transformers, Scikit-learn, MNE), SLURM, R, Matlab
  • Languages: Hungarian (Native), French (Native), English (Proficient), Spanish (Intermediate), Japanese (Beginner)
  • Domains: Machine Learning, Deep Learning, NLP, Cognitive Science
  • Research: Statistical Analysis, Bayesian Modeling, Neural Networks

Selected Projects

  • NLP: Pattern-Exploitation Training for low resource languages
  • Interpretability: Discovering Latent Knowledge in Language Models
  • Deep RL: Teacher algorithms for curriculum learning
  • TDA: Principal Geodesic Analysis of Merge Trees
  • See Projects Page for more details