Ramsés J. Sánchez studied theoretical physics at Simon Bolivar University (Venezuela) and the University of Bonn, with a focus on statistical mechanics. In 2017 he wrote his PhD dissertation on anomalous transport and out-of-equilibrium dynamics, and received his doctoral degree from the University of Bonn in October that year. From 2018 on he started working as a postdoctoral researcher in the field of Machine Learning (ML), first in the Competence Center Machine Learning Rhine-Ruhr, and later in the Lamarr Institute. During this time he has worked on problems ranging from artificial reasoning in natural language to the inference of different families of stochastic processes from noisy data. Since 2020 he is the scientific coordinator of the Hybrid ML group of the Lamarr institute. Most recently, he spearheads the Deep Learning for Scientific Discovery (DL4SD) group of Hybrid ML.
Current research topics:
- Defining the notion of foundation models for zero-shot inference of dynamical systems from noisy time series data
- Constructing phenomenological models for formal and commonsense reasoning, that can be used to power large, pretrained language models
- Studying the dynamics of learning in deep neural networks through the lens of stochastic thermodynamics and optimal transport theory
Special interests:
I am interested in the definition of neural network representations from carefully selected, simple algebraic equations. These representations are designed to be composed by neural reasoners (aka artificial scientists) trained to automatically construct novel scientific theories from data.
