An Inference-Based Architecture for Intent and Affordance Saturation in Decision-Making

TL;DR

This paper introduces an inference-based hierarchical decision model explaining decision paralysis phenomena like inertia and shutdown, with implications for understanding autism and decision-making processes.

Contribution

It formalizes intent and affordance selection using KL divergence objectives, revealing how different inference biases cause decision saturation and inertia.

Findings

Forward KL inference causes slow, heavy-tailed response times.

Simulations replicate decision inertia and shutdown phenomena.

Model links autism behaviors to extreme decision saturation regimes.

Abstract

Decision paralysis, i.e. hesitation, freezing, or failure to act despite full knowledge and motivation, poses a challenge for choice models that assume options are already specified and readily comparable. Drawing on qualitative reports in autism research that are especially salient, we propose a computational account in which paralysis arises from convergence failure in a hierarchical decision process. We separate intent selection (what to pursue) from affordance selection (how to pursue the goal) and formalize commitment as inference under a mixture of reverse- and forward-Kullback-Leibler (KL) objectives. Reverse KL is mode-seeking and promotes rapid commitment, whereas forward KL is mode-covering and preserves multiple plausible goals or actions. In static and dynamic (drift-diffusion) models, forward-KL-biased inference yields slow, heavy-tailed response times and two distinct failure modes, intent saturation and affordance saturation, when values are similar. Simulations in multi-option tasks reproduce key features of decision inertia and shutdown, treating autism as an extreme regime of a general, inference-based, decision-making continuum.