Interested in memory, learning, and decision making 🧠. Prof at TUD and PI of the AMC group https://amc-tud.github.io. Happy on two wheels 🚲 and four strings 🎸
Bernhard Spitzer
Loading...
Together, 'abstract' (object-independent) and 'concrete' (object-specific) WM info appeared multiplexed within the same areas, and this coexistence of formats was widely distributed.
NB: a distributed 180° representation emerged similarly also in an artificial vision model (CORnet-S) — but the model showed no 'abstract' 360° representation
To our surprise, we found that the neural format of the objects’ orientation in 360° space was fully generalized (object-independent) in all our ROIs, from early visual to parietal cortex.
While we found no gradient of abstraction in terms of generalization, we (like others) did see regional differences in categorization, where cardinal repulsion emerged only in the parietal cortex. Generalization and categorization might be distinct facets of WM abstraction
So, was distributed WM all abstract, everywhere? No: The same areas also showed an orthogonal representation of the object’s orientation in a 180° space, i.e. a line-like format. This format was object-dependent, likely reflecting the stimuli's concrete visual appearance.
More in our preprint www.biorxiv.org/content/10.6... A big shout out to the stellar Or Yizhar and the whole team (@inespontsanc.bsky.social also just joined Bsky)!
