The Movie After-Effect: widespread adaptation of human cortex following naturalistic sensory experience
Simony, E.; Yahav, N.; Malach, R.
Show abstract
Neural systems adapt to prolonged sensory input through mechanisms such as gain control or homeostatic plasticity to maintain stable operating ranges. However, this phenomenon has so far been documented under extreme, non-ecological stimuli targeting specific sensory systems. Here, we reveal a widespread adaptation process across diverse human cortical regions following naturalistic movie watching conditions. The effect was evident in 218 out of 251 cortical regions (87%) that exhibit significant stimulus-driven activations. Analyzing the HCP fMRI data set in which 170 participants watched 14 movie clips, followed by rest periods - we found robust evidence for a movie-induced adaptation process, revealed in the post-movies rest periods. Within a region, voxels highly activated at the end of movies subsequently reduced their activity below baseline during rest, with the magnitude of this drop proportional to initial activation levels, manifested as a consistent voxel-population inversion effect. Conversely, persistently movie-inactivated voxels exhibited increased activation above baseline. Importantly, this Movie After-Effect (MvAE) enabled successful decoding of the specific rest periods following individual movie clips. Our findings suggest that under naturalistic conditions, cortical neurons dynamically change their gain to achieve homeostatic balance in a process akin to batch instance normalization in artificial neural networks. Whether this ubiquitous MvAE has additional cognitive and memory-related implications remains to be explored.
Matching journals
The top 4 journals account for 50% of the predicted probability mass.
Similar papers in this journal
- Movement trajectories as a window into the dynamics of emerging neural representations. 94%
- Ramp-shaped neural tuning supports graded population-level representation of the object-to-scene continuum 94%
- Switching on Behavioral and Neural Rhythmicity to Retrieve Memories When the Number of Retained Items Exceeds Four 93%
"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.