Sleep strengthens successor representations of learned sequences in humans.

Decomposition of the interaction showed a trend for a linear component ( p = 0.050) and a significant cosine-shaped component ( p = 0.040), indicating that the changes across layers are best described as a mixture of monotonic and non-linear (curved) changes.

also in 33,481 other papers

First, higher proportions of SWS during a post-learning nap predicted stronger successor representations, though the association with representational shift to high-level formats only showed a statistical trend ( Fig 4b ).

also in 1,573 other papers