Boltzmann State-Dependent Rationality

TL;DR

This paper introduces a state-dependent rationality model that enhances the expressivity of human behavior modeling by replacing a constant with a function over states, supported by mathematical theory and preliminary experiments.

Contribution

It proposes a novel Boltzmann rationality model with a state-dependent suboptimality parameter, expanding the modeling capabilities of human decision-making.

Findings

Mathematical framework for state-dependent rationality

Experimental designs for testing the model

Preliminary results indicating potential improvements

Abstract

This paper expands on existing learned models of human behavior via a measured step in structured irrationality. Specifically, by replacing the suboptimality constant $\beta$ in a Boltzmann rationality model with a function over states $\beta(s)$, we gain natural expressivity in a computationally tractable manner. This paper discusses relevant mathematical theory, sets up several experimental designs, presents limited preliminary results, and proposes future investigations.