Adaptive Memory Distortions Are Predicted by Feature Representations in Parietal Cortex

Goal-directed attention transforms both working and long-term memory representations in the human parietal cortex

The abundance of distractors in the world poses a major challenge to our brain's limited processing capacity, but little is known about how selective attention modulates stimulus representations in the brain to reduce interference and support durable target memory. Here, we collected functional magnetic resonance imaging (fMRI) data in a selective attention task in which target and distractor pictures of different visual categories were simultaneously presented. Participants were asked to selectively process the target according to the effective cue, either before the encoding period (i.e., perceptual attention) or the maintenance period (i.e., reflective attention). On the next day, participants were asked to perform a memory recognition task in the scanner in which the targets, distractors, and novel items were presented in a pseudorandom order. Behavioral results showed that perceptual attention was better at enhancing target memory and reducing distractor memory than reflective attention, although the overall memory capacity (memory for both target and distractor) was comparable. Using multiple-voxel pattern analysis of the neural data, we found more robust target representation and weaker distractor representation in working memory for perceptual attention than for reflective attention. Interestingly, perceptual attention partially shifted the regions involved in maintaining the target representation from the visual cortex to the parietal cortex. Furthermore, the targets and distractors simultaneously presented in the perceptual attention condition showed reduced pattern similarity in the parietal cortex during retrieval compared to items not presented together. This neural pattern repulsion positively correlated with individuals' recognition of both targets and distractors. These results emphasize the critical role of selective attention in transforming memory representations to reduce interference and improve long-term memory performance.

Differentiation and Integration of Competing Memories: A Neural Network Model

What determines when neural representations of memories move together (integrate) or apart (differentiate)? Classic supervised learning models posit that, when two stimuli predict similar outcomes, their representations should integrate. However, these models have recently been challenged by studies showing that pairing two stimuli with a shared associate can sometimes cause differentiation, depending on the parameters of the study and the brain region being examined. Here, we provide a purely unsupervised neural network model that can explain these and other related findings. The model can exhibit integration or differentiation depending on the amount of activity allowed to spread to competitors - inactive memories are not modified, connections to moderately active competitors are weakened (leading to differentiation), and connections to highly active competitors are strengthened (leading to integration). The model also makes several novel predictions - most importantly, that when differentiation occurs as a result of this unsupervised learning mechanism, it will be rapid and asymmetric, and it will give rise to anticorrelated representations in the region of the brain that is the source of the differentiation. Overall, these modeling results provide a computational explanation for a diverse set of seemingly contradictory empirical findings in the memory literature, as well as new insights into the dynamics at play during learning.

Caricaturing Shapes in Visual Memory

When representing high-level stimuli, such as faces and animals, we tend to emphasize salient features-such as a face's prominent cheekbones or a bird's pointed beak. Such mental caricaturing leaves traces in memory, which exaggerates these distinctive qualities. How broadly does this phenomenon extend? Here, in six experiments (N = 700 adults), we explored how memory automatically caricatures basic units of visual processing-simple geometric shapes-even without task-related demands to do so. Participants saw a novel shape and then immediately adjusted a copy of that shape to match what they had seen. Surprisingly, participants reconstructed shapes in exaggerated form, amplifying curvature, enlarging salient parts, and so on. Follow-up experiments generalized this bias to new parameters, ruled out strategic responding, and amplified the effects in serial transmission. Thus, even the most basic stimuli we encounter are remembered as caricatures of themselves.

Distinct hippocampal mechanisms support concept formation and updating

Learning systems must constantly decide whether to create new representations or update existing ones. For example, a child learning that a bat is a mammal and not a bird would be best served by creating a new representation, whereas updating may be best when encountering a second similar bat. Characterizing the neural dynamics that underlie these complementary memory operations requires identifying the exact moments when each operation occurs. We address this challenge by interrogating fMRI brain activation with a computational learning model that predicts trial-by-trial when memories are created versus updated. We found distinct neural engagement in anterior hippocampus and ventral striatum for model-predicted memory create and update events during early learning. Notably, the degree of this effect in hippocampus, but not ventral striatum, significantly related to learning outcome. Hippocampus additionally showed distinct patterns of functional coactivation with ventromedial prefrontal cortex and angular gyrus during memory creation and premotor cortex during memory updating. These findings suggest that complementary memory functions, as formalized in computational learning models, underlie the rapid formation of novel conceptual knowledge, with the hippocampus and its interactions with frontoparietal circuits playing a crucial role in successful learning.

Significance statement

How do we reconcile new experiences with existing knowledge? Prominent theories suggest that novel information is either captured by creating new memories or leveraged to update existing memories, yet empirical support of how these distinct memory operations unfold during learning is limited. Here, we combine computational modeling of human learning behaviour with functional neuroimaging to identify moments of memory formation and updating and characterize their neural signatures. We find that both hippocampus and ventral striatum are distinctly engaged when memories are created versus updated; however, it is only hippocampus activation that is associated with learning outcomes. Our findings motivate a key theoretical revision that positions hippocampus is a key player in building organized memories from the earliest moments of learning.

Long-term memory interference is resolved via repulsion and precision along diagnostic memory dimensions

When memories share similar features, this can lead to interference, and ultimately forgetting. With experience, however, interference can be resolved. This raises the important question of how memories change, with experience, to minimize interference. Intuitively, interference might be minimized by increasing the precision and accuracy of memories. However, recent evidence suggests a potentially adaptive role for memory distortions. Namely, similarity can trigger exaggerations of subtle differences between memories (repulsion). Here, we tested whether repulsion specifically occurs on feature dimensions along which memories compete and whether repulsion is predictive of reduced memory interference. To test these ideas, we developed synthetic faces in a two-dimensional face space (affect and gender). This allowed us to precisely manipulate similarity between faces and the feature dimension along which faces differed. In three experiments, participants learned to associate faces with unique cue words. Associative memory tests confirmed that when faces were similar (face pairmates), this produced interference. Using a continuous face reconstruction task, we found two changes in face memory that preferentially occurred along the feature dimension that was "diagnostic" of the difference between face pairmates: (1) there was a bias to remember pairmates with exaggerated differences (repulsion) and (2) there was an increase in the precision of feature memory. Critically, repulsion and precision were each associated with reduced associative memory interference, but these were statistically dissociable contributions. Collectively, our findings reveal that similarity between memories triggers dissociable, experience-dependent changes that serve an adaptive role in reducing interference.