Terminology def gets in the way. Can you sent me a citation? Something like Wilson-Cowan E/I population nodes, trained by backprop?

Yeah, but they don’t leak, so they have no meaningful temporal dynamics. And their ‘synapses’ have no temporal dynamics either. Same holds for rate models, ie they leak too. And then there are exc/inh interactions, which also matter

Perhaps learning is proactive and adaptation is reactive, ie we adapt ‘to’ something, whereas learning seems to imply agency in some general sense

Mos def. Some say it's niche. We say there's a gaping hole in the literature. Maybe we should refer to context-dependence as neuro-dark-matter. We can see its effect on behaviour, but our analysis only explains 15% of the variance!

I'm guessing the use of 'context-dependence' is context-dependent. It's part of the lexicon for cognitive control, i.e. do the same task in different contexts (e.g. speed and accuracy emphasis) and see what changes. But I share you wariness of attributing inconsistent findings to 'context'