"Just in Time" World Modeling Supports Human Planning and Reasoning

TL;DR

This paper introduces a 'Just-in-Time' framework for mental simulation that constructs simplified, online representations of complex environments, enabling efficient human-like reasoning and planning with minimal computational effort.

Contribution

The paper proposes a novel online simulation model that dynamically constructs simplified environment representations, supported by empirical evidence from planning and reasoning tasks.

Findings

Model predicts human behavior in grid-world planning tasks

Model outperforms alternative approaches in physical reasoning tasks

Supports efficient mental simulation with minimal encoding of objects

Abstract

Probabilistic mental simulation is thought to play a key role in human reasoning, planning, and prediction, yet the demands of simulation in complex environments exceed realistic human capacity limits. A theory with growing evidence is that people simulate using simplified representations of the environment that abstract away from irrelevant details, but it is unclear how people determine these simplifications efficiently. Here, we present a "Just-in-Time" framework for simulation-based reasoning that demonstrates how such representations can be constructed online with minimal added computation. The model uses a tight interleaving of simulation, visual search, and representation modification, with the current simulation guiding where to look and visual search flagging objects that should be encoded for subsequent simulation. Despite only ever encoding a small subset of objects, the model makes high-utility predictions. We find strong empirical support for this account over alternative models in a grid-world planning task and a physical reasoning task across a range of behavioral measures. Together, these results offer a concrete algorithmic account of how people construct reduced representations to support efficient mental simulation.