Back

Stimulus-specific recruitment of human amygdala neurons predicts episodic memory encoding success

Xiao, J.; Daume, J.; Salimpour, Y.; Kurilenko, N.; Mosher, C.; Anderson, W. S.; Valiante, T. A.; Mamelak, A. N.; Rutishauser, U.

2025-10-30 neuroscience
10.1101/2025.10.29.685162 bioRxiv
Show abstract

Controlling whether a given experience is encoded into long-term memory and thus later remembered is a crucial component of our memory system whose failure is often at the root of memory disorders. One brain area that takes part in controlling which experiences are remembered is the amygdala, but the mechanisms by which it does so remain poorly understood. Here we examined single-neuron activity and local field potentials as human participants performed recognition memory tasks with visual stimuli. Category-selective amygdala neurons exhibited elevated firing rates during encoding of later remembered items versus forgotten items. This subsequent memory effect was restricted to images of the preferred category of a given cell, was stronger and appeared earlier in the amygdala compared to the hippocampus, and did not depend on the valence and arousal of the stimuli. In contrast, category selective cells immediately upstream in the ventral temporal cortex did not exhibit a subsequent memory effect, highlighting specificity to the amygdala. Successful memory formation was accompanied by enhanced spike-field coherence between the activity of category cells in the amygdala and hippocampal field potentials. These findings, replicated in two large independent datasets with two different tasks, demonstrate that recruitment of stimulus-specific amygdala representations predicts episodic memory formation, particularly in the right amygdala. This data suggests category cells in the right amygdala as a cellular target for interventions to treat memory disorders in humans.

Matching journals

The top 4 journals account for 50% of the predicted probability mass.

50% of probability mass above

"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.