These beautiful twisted creatures are probably much simpler than the topologies neuroscientists treat as euclidean spaces when they do silly things like RSA or averaging neural timeseries.

[jonny@neuromatch.social]

These beautiful twisted creatures are probably much simpler than the topologies neuroscientists treat as euclidean spaces when they do silly things like RSA or averaging neural timeseries. Pick any two points on the manifold of the attractor, would a line in the embedding euclidean space be the best way to describe the distance between them?
https://chaos.social/@sml/113460651020035223