Tomorrow Oct 16 at 10:15AM Pacific.
From @FFiebig:
“Facilitation/augmentation, the pre-synaptic calcium dynamics that are behind it, and hope to touch on how some cortical models have leveraged it as a mechanism to enable/implement cortical working memory.” This presentation is, however part of a larger arc of maybe 3-4 presentations I hope to give in the coming weeks that will also include alternative Bayesian-Hebbian models of working memory that I have proposed in recent years to complement the ongoing paradigm shift in cortical working memory theory, away from the “persistent activity hypothesis” to a “synaptic theory of working memory” (which Earl Miller at MIT promptly re-branded “Working Memory 2.0”, when he took on a PostDoc from our lab and proved some of our model predictions in macaque PFC). I’ll post some relevant papers below, but its all optional reading really. I’ll start this topic slow and steady, because I think might eventually provide a neat cortical mechanism to solve an issue I see with the excessive activity of the actively maintained hypothesis space for object and places during inference in in the cortical learning framework.
I’ll start of with a review of synaptic facilitation/augmentation, direct PDF link here.