A microscopic perspective on stochastic thermodynamics

Abstract

We consider stochastic thermodynamics as a theory of statistical inference for experimentally observed fluctuating time-series. To that end, we introduce a general framework for quantifying the knowledge about the dynamical state of the system on two scales: a fine-grained or microscopic, deterministic and a coarse-grained or mesoscopic, stochastic level of description. For a generic model dynamics, we show how the mathematical expressions for fluctuating entropy changes used in Markovian stochastic thermodynamics emerge naturally. Our ideas are conceptional approaches towards (i) connecting entropy production and its fluctuation relations in deterministic and stochastic systems and (ii) providing a complementary information-theoretic picture to notions of entropy and entropy production in stochastic thermodynamics.