"Connectivism = (Network + Stimuli + Interaction) = Trained Reaction"
I view stimuli as a facet of interaction which is a facet of a network. Interaction if the simultaneous processing of input (i.e., stimuli) and output (i.e., behavior). A network is a bunch of nodes interacting one another. Or stated another way, the change of one node causes the change of another node or nodes. According the Downes, a network has the following characteristics: open, diverse, autonomous, and interactive. So, instead of looking at stimuli, interaction, and network separately, I just see each as a subset of the other.
I tend to avoid words like "trained", "learned", "competencies", etc. because they all are dichotomous. If I say,
"I learned the present tense"
"I'm a trained professional"
"I'm technologically competent"
they all insinuate that I either have it or I don't. If we say we have a "trained reaction", then we return to a more behaviorist perspective that implies that certain behavior (i.e., output) will result from specific stimuli (i.e., input).
I would argue that a network is more complex in that it requires the skill of sifting through vast amounts of input and at the same time producing a much quality output as possible. Through reflection, the output thus becomes input as well so through this interaction, the person (i.e., node) moves through the network in ways that best suits the learner. Understandings, learnings, capacities, etc. grow, are cultivated, and flow as if along a continuum. Learners never start at zero percent and end at 100 percent.