6 papers
A neural-network Maxwell's demon learns cold damping for work extraction
Stephen Whitelam, Sergio Ciliberto, Ludovic Bellon
We train a neural-network Maxwell's demon to extract work from a model of an underdamped micromechanical cantilever subject to thermal noise. The demon, which periodically adjusts…
The Future of Artificial Intelligence and the Mathematical and Physical Sciences (AI+MPS)
Andrew Ferguson, Marisa LaFleur, Lars Ruthotto +97
This community paper developed out of the NSF Workshop on the Future of Artificial Intelligence (AI) and the Mathematical and Physics Sciences (MPS), which was held in March 2025 w…
Nonlinear thermodynamic computing out of equilibrium
Stephen Whitelam, Corneel Casert
We present the design for a thermodynamic computer that can perform arbitrary nonlinear calculations in or out of equilibrium. Simple thermodynamic circuits, fluctuating degrees of…
Learning to shine: Neuroevolution enables optical control of phase transitions
Sraddha Agrawal, Stephen Whitelam, Pierre Darancet
We address the problem of active optical steering of structural phase transitions in solids. We demonstrate that existing reinforcement learning approaches can derive optimal time-…
Benchmark control problems in nonequilibrium statistical mechanics
Stephen Whitelam, Corneel Casert, Megan Engel +1
We present a set of computer codes designed to test methods for optimizing time-dependent control protocols in fluctuating nonequilibrium systems. Each problem consists of a stocha…
Learning efficient erasure protocols for an underdamped memory
Nicolas Barros, Stephen Whitelam, Sergio Ciliberto +1
We apply evolutionary reinforcement learning to a simulation model in order to identify efficient time-dependent erasure protocols for a physical realization of a one-bit memory by…