wow, this paper totally destroys the current mechanistic interpretations of circular dynamics in neural population data: https://biorxiv.org/content/10.1101/2023.09.11.557230v1
One thing they don't talk about is that any state space representation of discrete movements will also produce rotations...
@de594dec@e8501a5d
This is an interesting perspective (from both of you).
I'm not sure though I follow the criticism wrt systems ID vs reverse engineering. In engineering, systems ID always cares about the the overall systems behavior. This is in fact the critical goal, e.g. in systems ID for control. But you want to know something about the components that enable the overall behavior, e.g. in engineering bc you might be able to control some components to achieve a desired overall behavior.
So really to me SID is one powerful way to RE... and I think their definition is appropriate in the presented context.
That being said, it could have been clearer that this is not just a blue brain project of C. elegans... but that the overarching goal is indeed to reverse engineer the overall system's (i.e. animals) behavior as a function of its constituents. This is at least partially mentioned when they talk about wanting to understand sub-system function and building blocks of behavior. So maybe that's where some of the conceptual confusion comes from?
Oddbean new post about login | logout
Notes by ed818a4c | export