Which familiar systems emerge from the learned dynamics?

Resting recordings from 96 NYU control participants describe spontaneous functional organization.

The clearest anatomical findings concern the systems represented by each network’s opposing poles. The summaries below use seven reference systems and the 12-network fit.

Network 2 has a prominent default-mode pole. Its positive pole resembles default mode (similarity 0.71; 52% of that pole’s weight). Its negative pole is more distributed, with weaker salience/ventral-attention and somatomotor associations.
Visual organization appears in two different patterns. Network 3’s positive pole (similarity 0.81; 32% of that pole’s weight) and Network 4’s negative pole (similarity 0.79; 62% of that pole’s weight) both resemble visual regions. Similar anatomy alone does not establish that these are duplicate dynamical components.
Network 5 links two distinct anatomical poles. Its positive pole resembles dorsal attention (similarity 0.54; 38% of that pole’s weight), while its negative pole resembles somatomotor organization (similarity 0.55; 28% of that pole’s weight). This is an opposing spatial pattern, not evidence of direct physiological antagonism.
A later pattern also has a clear association. Network 9’s negative pole resembles somatomotor (similarity 0.51; 30% of that pole’s weight). Increasing rank can add useful patterns, but a later network is not automatically more functionally specific.

Explore the functional-system atlas · Inspect network matches and regional connections

From anatomy to expression

After identifying the spatial patterns, inspect their consistency across people and how their contributions vary through resting recordings. Shared coordinates support comparison, but do not demonstrate independent network recovery in every participant.

Spatial associations are descriptive, not validated functional labels. Network numbers refer to this fit; they cannot be equated across studies.