Two pathways to resolve relational inconsistencies

TL;DR

This paper investigates why people and neural networks sometimes stick to prior expectations despite conflicting observations, revealing two pathways—adjustment or bypass—based on violation magnitude, and explaining the stability of expectations.

Contribution

The study demonstrates that large expectation violations lead to a bypass mechanism in neural networks, explaining expectation stability without extra assumptions.

Findings

Large violations trigger a bypass in neural networks.

Small violations lead to expectation adjustment.

Expectation stability arises naturally from learning dynamics.

Abstract

When individuals encounter observations that violate their expectations, when will they adjust their expectations and when will they maintain them despite these observations? For example, when individuals expect objects of type A to be smaller than objects B, but observe the opposite, when will they adjust their expectation about the relationship between the two objects (to A being larger than B)? Naively, one would predict that the larger the violation, the greater the adaptation. However, experiments reveal that when violations are extreme, individuals are more likely to hold on to their prior expectations rather than adjust them. To address this puzzle, we tested the adaptation of artificial neural networks (ANNs) capable of relational learning and found a similar phenomenon: Standard learning dynamics dictates that small violations would lead to adjustments of expected relations while larger ones would be resolved using a different mechanism -- a change in object representation that bypasses the need for adaptation of the relational expectations. These results suggest that the experimentally-observed stability of prior expectations when facing large expectation violations is a natural consequence of learning dynamics and does not require any additional mechanisms. We conclude by discussing the effect of intermediate adaptation steps on this stability.