Computational Analysis of Disorganized Attachment as a Multilevel Epigenetic and Neural Network
摘要
This paper presents a fifth-order adaptive network architecture that formalizes how attachment patterns emerge and adapt across six biologically grounded layers. Our developed model integrates relational context states to simulate caregiving environments over time and embeds a biologically plausible recovery loop in which demethylation enables oxytocin availability. This oxytocin availability in turn restores neural regulation and belief restructuring. Finally, our model captures the temporal unfolding of recovery, showing that molecular repair precedes shifts in belief and behavioural expression.