Thermodynamics and evolutionary biology through optimal control
We consider a particular instance of the lift of controlled systems recently proposed in the theory of irreversible thermodynamics and show that it leads to a variational principle for an optimal control in the sense of Pontryagin. Then we focus on two important applications: in thermodynamics and in evolutionary biology. In the thermodynamic context, we show that this principle provides a dynamical implementation of the Second Law, which stabilizes the equilibrium manifold of a system. In the evolutionary context, we discuss several interesting features: it provides a robust scheme for the coevolution of the population and its fitness landscape; it has a clear informational interpretation; it recovers Price equation naturally; and finally, it extends standard evolutionary dynamics to include phenomena such as the emergence of cooperation and the coexistence of qualitatively different phases of evolution, which we speculate can be associated with Darwinism and punctuated equilibria.
Code (0)
등록된 구현이 없습니다.
Similar Papers 제목 키워드 기반
A Test of the Thermodynamics of Evolution
Recent research has extended methods from the fields of thermodynamics and statistical mechanics into other disciplines. Most notably, one recent work creates a unified theoretical framework to understand evolutionary bi…
Mega-scale experimental analysis of protein folding stability in biology and design
Advances in DNA sequencing and machine learning are providing insights into protein sequences and structures on an enormous scale1. However, the energetics driving folding are invisible in these structures and remain lar…
Protein FoldingProtein Stability PredictionThermodynamics of Evolution and the Origin of Life
We outline a phenomenological theory of evolution and origin of life by combining the formalism of classical thermodynamics with a statistical description of learning. The maximum entropy principle constrained by the req…
Analysis and Explainability of LLMs Via Evolutionary Methods
Evolutionary methods have long been useful for analysis and explanation in genetics, biology, ecology, and related fields. In this work, we extend these methods to neural networks, specifically large language models (LLM…
Characterizing optimal hierarchical policy inference on graphs via non-equilibrium thermodynamics
Hierarchies are of fundamental interest in both stochastic optimal control and biological control due to their facilitation of a range of desirable computational traits in a control algorithm and the possibility that the…