Energy Dissipation and Information Flow in Coupled Markovian Systems.

Area of Science:

• Statistical Physics
• Information Theory
• Thermodynamics

Background:

• Stochastic systems interacting with their environment exhibit correlations with present and future environmental states.
• Such systems can be interpreted as implicitly constructing predictive models of their environment.
• Non-predictive model complexity is known to impose a lower bound on thermodynamic dissipation.

Purpose of the Study:

• To investigate the statistical and physical quantities, specifically model complexity and thermodynamic dissipation, in simple stochastic systems at steady state.
• To explore the relationship between model complexity and dissipation under different driving conditions.

Main Methods:

• Analysis of simple stochastic models at steady state.
• Examination of system behavior under quasi-static driving conditions.
• Investigation of system dynamics beyond the quasi-static limit, focusing on weak driving.

Main Results:

• Model complexity saturates thermodynamic dissipation under quasi-static driving.
• A lower bound on the ratio of model complexity to total dissipation is demonstrated.
• This lower bound is achieved in the limit of weak environmental driving.

Conclusions:

• Stochastic systems inherently possess predictive capabilities reflected in their model complexity.
• Model complexity plays a crucial role in determining thermodynamic dissipation, especially under varying driving regimes.
• The findings provide insights into the fundamental relationship between information processing and energy dissipation in physical systems.