Rémi Monasson

Rémi Monasson is a Directeur de Recherche at CNRS and Professor at Ecole Polytechnique, based at the Laboratoire de Physique Théorique of École Normale Supérieure (ENS) in Paris. His work bridges statistical physics, machine learning, and biological data analysis, with particular emphasis on protein modeling and neural network theory.

Research Interests:

• High-Dimensional Inference and Applications to Biological Data (Genomics, Neurobiology)
• Statistical Physics of Learning in Neural Networks
• Analysis of Algorithms and Combinatorial Optimization Problems
• Protein Sequence Analysis and Structure Prediction
• Statistical Mechanics of Biological Systems

Monasson's research program integrates principles from statistical physics to develop novel computational approaches for understanding complex biological systems. His work on Restricted Boltzmann Machines for protein family modeling has opened new avenues for protein design and evolutionary analysis. In neural network theory, he explores how physical principles can explain learning and memory mechanisms in both artificial and biological systems. His approach combines rigorous theoretical frameworks with practical applications in molecular biology and neuroscience.

Recent Publication Trends:

Monasson's recent publications (2023-2024) demonstrate a deep integration of statistical physics with cutting-edge problems in biology and machine learning. Key themes include protein sequence analysis using probabilistic models, neural network dynamics and learning mechanisms, and applications to immunology and molecular recognition. His work frequently develops novel computational methods grounded in statistical physics principles, with applications ranging from RNA switch design to T-cell receptor specificity prediction. The interdisciplinary nature of his research is evident in collaborations spanning physics, biology, computer science, and medicine.

Current Research Projects:

• ANR project RBMPro on theory and application of Restricted Boltzmann Machines to protein family modeling (2017-2021)
• HFSP on Analog Computation Underlying Language Mechanisms (2016-2020)
• PSL* project Data Science (2017-2019)
• Workshop on Tumors and Immune Systems: From Theory to Therapy (2019)
• ANR project Coevstat on coevolution in proteins (2013-2016)

Teaching Activities:

• Model-Guided Data Science (2019): Lectures on "Unsupervised learning of features from data: a statistical physics approach"
• Ecole Polytechnique: Physics of Biological Systems (PHY 552B) since 2012
• Netadis Summer School on Complex Systems (2013): lectures on random graphs and maps
• ICFP Master program: Collective phenomena: from phase transitions to statistical field theory (2011-2015)