Optimal Work Extraction and the Minimum Description Length Principle

Abstract

We discuss work extraction from classical information engines (e.g., Szilárd) with $N$-particles, $q$ partitions, and initial arbitrary non-equilibrium states. In particular, we focus on their {\em optimal} behaviour, which includes the measurement of a set of quantities $\Phi$ with a feedback protocol that extracts the maximal average amount of work. We show that the optimal non-equilibrium state to which the engine should be driven before the measurement is given by the normalised maximum-likelihood probability distribution of a statistical model that admits $\Phi$ as sufficient statistics. Furthermore, we show that the minimax universal code redundancy ${\mathcal{R}}^{\ast }$ associated to this model, provides an upper bound to the work that the demon can extract on average from the cycle, in units of $k_{\mathrm{B}}T$. We also find that, in the limit of $N$ large, the maximum average extracted work cannot exceed $H[\Phi ]/2$, i.e. one half times the Shannon entropy of the measurement. Our results establish a connection between optimal work extraction in stochastic thermodynamics and optimal universal data compression, providing design principles for optimal information engines. In particular, they suggest that: (i) optimal coding is thermodynamically efficient, and (ii) it is essential to drive the system into a critical state in order to achieve optimal performance.

26 pages, 5 figures. To appear in JSTAT

References in corpus (47)

• Stochastic thermodynamics, fluctuation theorems, and molecular machines
• Cavity-based quantum networks with single atoms and optical photons
• Information heat engine: converting information to energy by feedback control
• Strongly Interacting Polaritons in Coupled Arrays of Cavities
• Are biological systems poised at criticality?
• Generalized Jarzynski Equality under Nonequilibrium Feedback Control
• High-precision test of Landauer's principle in a feedback trap
• Measuring thermodynamic length
• Minimal Energy Cost for Thermodynamic Information Processing: Measurement and Information Erasure
• Optimal finite-time processes in stochastic thermodynamics
• Second law and Landauer principle far from equilibrium
• Thermodynamic metrics and optimal paths
• Work and information processing in a solvable model of Maxwell's demon
• Experimental study of mutual information in a Maxwell Demon
• Nonequilibrium Thermodynamics of Feedback Control
• Quantum Szilard Engine
• Truly work-like work extraction
• Estimating dissipation from single stationary trajectories
• Universal features in the energetics of symmetry breaking
• Entropy production and Kullback-Leibler divergence between stationary trajectories of discrete systems
• Complementarity relation for irreversible process derived from stochastic energetics
• Thermodynamic Reversibility in Feedback Processes
• Unifying Three Perspectives on Information Processing in Stochastic Thermodynamics
• Large work extraction and the Landauer limit in a continuous Maxwell demon
• Designing optimal discrete-feedback thermodynamic engines
• Optimal driving of isothermal processes close to equilibrium
• Lossless Brownian information engine
• Erasure without work in an asymmetric, double-well potential
• Sensory capacity: an information theoretical measure of the performance of a sensor
• Computing the optimal protocol for finite-time processes in stochastic thermodynamics
• Learning force fields from stochastic trajectories
• Direct measurement of nonequilibrium system entropy is consistent with Gibbs-Shannon form
• Phase transition in protocols minimizing work fluctuations
• Introducing one-shot work into fluctuation relations
• Optimal probabilistic work extraction beyond the free energy difference with a single-electron device
• Far-from-Equilibrium Measurements of Thermodynamic Length
• Statistical Criticality arises in Most Informative Representations
• Thermodynamic cost and benefit of memory
• Information thermodynamics for interacting stochastic systems without bipartite structure
• Optimal processes for probabilistic work extraction beyond the second law
• Minimum Description Length codes are critical
• Role of sufficient statistics in stochastic thermodynamics and its implication to sensory adaptation
• Szilard engine revisited; information from time forward and backward process
• The Stochastic complexity of spin models: Are pairwise models really simple?
• Optimal work extraction and mutual information in a generalized Szilárd engine
• When memory pays: Discord in hidden Markov models
• Reversible feedback confinement

Cited by in corpus (4)

• Driving rapidly while remaining in control: classical shortcuts from Hamiltonian to stochastic dynamics
• Quantifying Relevance in Learning and Inference
• Optimal work extraction and mutual information in a generalized Szilárd engine
• Gibbs Distribution From Sequentially Predictive Form of the Second Law