Transfer entropy and transient limits of computation.

Area of Science:

• Thermodynamics
• Information Theory
• Computational Theory

Background:

• Transfer entropy quantifies directed statistical coherence in spatiotemporal processes.
• Its relationship to fundamental computational limits like Landauer's limit was previously unknown.

Purpose of the Study:

• To investigate the connection between transfer entropy and thermodynamic limits of computation.
• To generalize Landauer's limit for non-equilibrium, bidirectional information dynamics.

Main Methods:

• Theoretical analysis of information-theoretic measures and thermodynamic principles.
• Generalization of Landauer's principle to information prediction.

Main Results:

• Increasing transfer entropy by one bit necessitates heat flow at or above Landauer's limit.
• Landauer's limit is extended to cover prediction in non-equilibrium processes.
• Connections established with negentropy, Bremermann's limit, and Bekenstein bound.

Conclusions:

• New computational limits are defined using fundamental physical quantities.
• Transfer entropy is established as a key measure linking information theory, thermodynamics, and computation.