A simulation-heuristics dual-process model for intuitive physics
Keywords: Intuitive physics, physical reasoning, mental simulation, heuristic model
Abstract: The role of mental simulation in human behavior for various physical tasks is widely acknowledged, attributed to the generality of Intuitive Physics Engine (IPE). However, it remains unclear whether mental simulation is consistently employed across scenarios of different simulation costs and where its boundary is. Moreover, cognitive strategies beyond these boundaries have not been thoroughly investigated. Here, we adopted a pouring-marble task containing various conditions to study IPE's limits and strategies beyond. A human study revealed two distinct error patterns in predicting the pouring angle, differentiated by the simulation time using a boundary. This suggests a possible switching of the underlying reasoning strategies. Our initial experiment on IPE showed that its correlation with human judgments diminished in scenarios requiring extended time of simulation. This observation prompted the exploration of an alternative mechanism based on heuristics for intuitive physics. We uncovered that a linear heuristic model, relying exclusively on empirical data, replicated human prediction more accurately when the simulation time exceeded a certain boundary. Motivated by these observations, we propose a new framework, Simulation-Heuristics Model (SHM), which conceptualizes intuitive physics as a dual process: IPE is predominant only in short-time simulation, whereas a heuristics-based approach is applied as IPE's simulation time extends beyond the simulation boundary. The SHM model aligns more precisely with human behavior across various scenarios and demonstrates superior generalization capabilities under different conditions. Crucially, SHM integrates computational methods previously viewed as separate into a unified model, quantitatively studying their switching mechanism.
Primary Area: applications to neuroscience & cognitive science
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics.
Submission Guidelines: I certify that this submission complies with the submission instructions as described on https://iclr.cc/Conferences/2025/AuthorGuide.
Reciprocal Reviewing: I understand the reciprocal reviewing requirement as described on https://iclr.cc/Conferences/2025/CallForPapers. If none of the authors are registered as a reviewer, it may result in a desk rejection at the discretion of the program chairs. To request an exception, please complete this form at https://forms.gle/Huojr6VjkFxiQsUp6.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Submission Number: 634
Loading