Direct encoding of new visual concepts in early visual cortex
Klinkowski, S.; Seewald, A.; Fath, B.; Kasties, V.; Iliopoulos, P.; Schmidt, S.; Voss, F.; Erb, M.; Scheffler, K.; Gais, S.; Brodt, S.
Show abstract
Neocortical memories are assumed to rely on slow systems consolidation through hippocampal offline reactivation. However, recent studies show rapid neocortical integration of new memories for information associated with prior knowledge. Testing the limits of rapid neocortical learning, we tracked memory formation across 24h with functional and diffusion-weighted MRI in subjects acquiring either conceptual or detailed knowledge of novel objects from identical visual stimulation. Concept learning-induced rapid, 24h-stable changes in strength and pattern of functional early visual cortex responses. These were co-localized with microstructural changes, correlated positively with categorization performance and negatively with context recognition. Detailed item-context learning elicited analogous changes in posterior parietal areas. Results show that the neocortex can rapidly learn without prior knowledge and that concept learning can occur in early sensory regions that process corresponding visual features. Thus, we substantiate the possibility of direct or parallel neocortical encoding without offline abstraction from individual episodes.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.
Similar papers in this journal
Similar papers in this journal
Similar papers in this journal
Similar papers in this journal
- Mapping multidimensional content representations to neural and behavioral expressions of episodic memory 97%
- Neural similarity between overlapping events at learning differentially affects reinstatement across the cortex 96%
- A characterization of the neural representation of confidence during probabilistic learning 96%
"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.