A neural-network Maxwell's demon learns cold damping for work extraction

Abstract

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 position of a harmonic trap, is trained to maximize the power extracted under steady-state operation. When the demon is given the cantilever position and trap position as inputs it learns a refined version of an existing hand-designed protocol, yielding a substantial improvement in performance. When the demon receives the oscillator velocity as input it discovers a qualitatively different strategy that extracts substantially more work, close to the theoretical power bound. Analysis of the protocol shows that it implements {\em cold damping}: the trap position is displaced approximately linearly with velocity, producing an effective increase of the oscillator's damping coefficient and a reduction of its effective temperature. Thus a neural-network Maxwell's demon rediscovers a well-known cooling strategy from optomechanics, revealing a simple physical mechanism underlying near-optimal work extraction from thermal fluctuations in an underdamped system.