ESPCI web site
Foundational models for materials chemistry
A new computational task has been defined and solved over the past 15 years for extended material systems : the analytic fitting of the Born-Oppenheimer potential energy surface as a function of nuclear coordinates. The resulting potentials ("force fields") are reactive, many-body, with evaluation costs that are currently on the order of 0.1-10 ms/atom/cpu core (or about 1-10ms on a powerful GPU), and reach accuracies of a few meV/atom when trained specifically for a given system using iterative or active learning methods. The latest and most successful architectures leverage many-body symmetric descriptions of local geometry and equivariant message passing networks. Perhaps the most surprising recent result is the stability of models trained on very diverse training sets across the whole periodic table. Our recent discovery is that the MACE-MP-0 model that was trained on just 150,000 real and hypothetical small inorganic crystals (90% of training set < 70 atoms), is capable of stable molecular dynamics at ambient conditions on any system tested so far - this includes crystals, liquids, surfaces, clusters, molecules, and combinations of all of these. The astounding generalisation performance of such foundation models open the possibility to creating a universally applicable interatomic potential with useful accuracy for materials (especially when fine-tuned with a little bit of domain-specific data), and democratise quantum-accurate large scale molecular simulations by lowering the barrier to entry into the field. Similarly, in the domain of organic chemistry, training just on small molecules and small clusters allows accurate simulation of condensed phase systems, and first principles prediction of quantities such as hydration free energies for the first time.
Memory and environment sensing in active systems
The behaviour of active systems, such as living cells, is usually modeled by self-propelled particles driven by internal forces and noise. However, these models often assume memoryless dynamics and no coupling of internal active forces to the environment. Here, we introduce a general theoretical framework that goes beyond this paradigm by incorporating internal state dynamics and environmental sensing into active particle models.
We show that when the self-propulsion of a particle depends on internal variables with their own complex dynamics - modulated by local environmental cues - new classes of behaviours emerge. These include memory-induced responses, controllable localization in complex landscapes, suppression of motility-induced phase separation, and enhanced jamming transitions. Our results demonstrate how minimal information processing capabilities, intrinsic to non-equilibrium systems with internal states like living cells, can profoundly influence both individual and collective behaviours. This framework bridges cell-scale activity and large-scale intelligent motion in active agents, and opens the way to the analysis or design of systems ranging from synthetic colloids to biological collectives and robotic swarms.