Abstract representations of events arise from mental errors in learning and memory

TL;DR

This paper proposes that humans develop abstract representations of events not through complex learning but as a result of natural errors in learning and memory, supported by a model combining information theory and reinforcement learning.

Contribution

It introduces a maximum entropy model explaining how mental errors shape human internal representations of stimulus transitions, with analytical simplicity and predictive power.

Findings

Model explains effects of transition network structure on expectations

Predicts human reaction times in probabilistic tasks

Supports errors as a basis for abstract representations

Abstract

Humans are adept at uncovering abstract associations in the world around them, yet the underlying mechanisms remain poorly understood. Intuitively, learning the higher-order structure of statistical relationships should involve complex mental processes. Here we propose an alternative perspective: that higher-order associations instead arise from natural errors in learning and memory. Combining ideas from information theory and reinforcement learning, we derive a maximum entropy (or minimum complexity) model of people's internal representations of the transitions between stimuli. Importantly, our model (i) affords a concise analytic form, (ii) qualitatively explains the effects of transition network structure on human expectations, and (iii) quantitatively predicts human reaction times in probabilistic sequential motor tasks. Together, these results suggest that mental errors influence our abstract representations of the world in significant and predictable ways, with direct implications for the study and design of optimally learnable information sources.