This is a nicely detailed explanation! But when people express frustration with low-D dynamics, I think it is exactly what you say this paper disproves: the notion that variance explained is a proxy for the important parts of neural computation

“the conventional focus on low-dimensional coding subspaces” is tired, and in many cases just an artifact of low dimensional behavior. Excited to see comp models moving past it!

This was such a joy to read! Naive question: what are the D2-msn learning?

Yea, perhaps makes more sense to psychologists than neuroscientists? In practice I think those are best used as 3 styles of investigation, for any topic, rather than as saying anything intrinsic about the subject matter

Thanks Tom! I appreciate the comments, and I agree our evidence is for cued recall in the paper. But I should add: we have no evidence *against* free recall, we just did not analyze it yet. Topic of ongoing experiments. So I’d add that caveat to your comments here until we get results!

flying back from interviews to lab = same vibe

Technically a mountain chickadee, not the black capped chickadee I study. But they all share that ‘tiny dog aggressively yelling at every big dog it sees on the street’ energy