Repetition facilitates differentiation of neural representations in the hippocampus
Poster Session C - Sunday, March 30, 2025, 5:00 – 7:00 pm EDT, Back Bay Ballroom/Republic Ballroom
Emily T Cowan1 (ecowan@adelphi.edu), Busra Tanriverdi2, Ingrid Olson2, Vishnu P Murty2,3, Jason Chein2; 1Adelphi University, 2Temple University, 3University of Oregon
While memory seems to benefit from repeated learning opportunities, it remains unclear how repetition affects the organization of memories in the brain. In an fMRI study, 25 participants viewed image-location pairs, each of which was repeated 4 times. To examine how the representational patterns change across repetitions, for each pair, we calculated pattern similarity between consecutive repetitions. In bilateral hippocampus, we found evidence of differentiation in the representational patterns evoked across repetitions, with a significant decrease in pattern similarity between the first two versus last two repetitions (t(24)= 2.73, p=0.01). This pattern of results was particularly evident in the anterior hippocampus (t(24)= 2.38, p=0.03). We next developed several theoretical models of the patterns of voxel-wise changes that could explain the observed differentiation across repetitions. Versions of models tested changes based on novelty, gradual versus abrupt changes across repetitions, and the effects of general representational drift. For each of these models, we simulated patterns of voxels and calculated the similarity between consecutive repetitions (100,000 simulations). The simulated similarity data was then compared to the fMRI data. Comparisons across models will provide theoretical explanations for the differentiation we observed in the fMRI data.
Topic Area: LONG-TERM MEMORY: Episodic