Generating statistical distributions without maximizing the entropy

TL;DR

This paper introduces a thermodynamics-inspired method to generate probability distributions for generalized ensembles by constraining microstate occupation changes using the first law, extending the traditional entropy maximization approach.

Contribution

It presents a novel approach that leverages thermodynamic principles to derive distributions without relying solely on entropy maximization.

Findings

Provides a framework for generating distributions through thermodynamic constraints

Extends the concept of heat to more general microstate occupation changes

Offers a new perspective on ensemble generation beyond traditional methods

Abstract

We show here how to use pieces of thermodynamics' first law to generate probability distributions for generalized ensembles when only level-population changes are involved. Such microstate occupation modifications, if properly constrained via first law ingredients, can be associated not exclusively to heat and acquire a more general meaning.