
PLoS ONE, Год журнала: 2024, Номер 19(12), С. e0313951 - e0313951
Опубликована: Дек. 17, 2024
Previous explanations for the persistence of polarization opinions have typically included modelling assumptions that predispose possibility (i.e., allowing a pair agents to drift apart in their opinion such as repulsive interactions or bounded confidence). An exception is recent simulation study showing persistent when form using social reinforcement learning. Our goal highlight usefulness learning context modeling dynamics, but caution required selecting tools used model. We show observed model cannot persist indefinitely, and exhibits consensus asymptotically with probability one. By constructing link between voter model, we argue metastable. Finally, slight modification process changes from being non-ergodic ergodic. results may be powerful method (objects stationary distribution, time absorption example) appropriate analysing models crucially depend on properties (such ergodicity, transience). These are determined by details difficult identify based solely simulations.
Язык: Английский