People are simple machines, somewhat
It will probably be several years before I can map the language I used for my college work, to the language of the triple network model in contemporary neuroscience. After all, they are both infirm theories. Roughly from my initial reading of the latter :
SMN ... SENSORY-motor ( sensorimotor / somatomotor ) network, covers both exteroception and interoception. ( Which together form a subset of proprioception ... as proprioception broadly involves imagined sensory data as well, not just live sourced sensory data ). Of course this divides into input/output i.e. sense/motor subnets. Various sense modalities have their own subnets.
CEN ... central executive network, mostly concerned with PHYSICS : modelling, simulation, and execution. Bridges the SMN, to other key networks. Since DEDUCTIVE reasoning is a variety of physics ( not a commonly understood or acceptable hypothesis ), perhaps this is also the brain's deductive processor, equivalent to ATPs in maths software.
DMN ... default mode network, generally said to be involved in free association, pattern matching, and INDUCTIVE hypothesis generation ( first stage of inductive reasoning, where the second stage is deductive coherence checking ).
SN ... salience network, tends to handle PRIORITISATION ( probably value judgements in general, as it links up with emotions, and socioeconomic valuation ).
DAN ... dorsal attention network, seems to be closely related with the shortest-term memory buffer, getting bridged alternately to the CEN or DMN. Maybe the seat of CONSCIOUS memory - need to read up on how this works with blind people, as initial pointers talk about visual data ( we do know blind people can use visual imagination internally, to a degree ).