Speaker
Kiyoshi Kanazawa
(Kyoto University)
Description
Stochastic thermodynamics is a powerful framework to formulate various thermodynamic bounds for small systems. However, this framework has largely relied on the Markov assumption for the underlying dynamics, and its application to non-Markov processes with strong memory has been limited, except for a few special classes (like the generalized Langevin equation and the semi-Markov processes). In this talk, we will present stochastic thermodynamics for general non-Markov jump processes. We develop the Fourier embedding as a key mathematical technique to formulate the time-reversal symmetry for general jump processes. Finally, we present several novel non-Markov models that satisfy the second law.