Research Talks and Presentations

The CML at SfN 2024

The Complex Memory Lab shared exciting research updates this week at the Society for Neuroscience annual meeting held in Chicago, IL.

Adi Upadhyayula, a postdoctoral scholar in the lab, presented a talk titled “Meaningful moments during film viewing are represented and remembered similarly across participants.” Adi’s work starts off with a simple premise: he asks people first to watch short videos, then to reconstruct the overall story of the video using still frames from the videos. Adi found that participants tended to select the same moments in the movies in their “storyboard” reconstructions, and that these moments were distinct from event boundaries. Next, Adi looked at an open fMRI dataset and found that the pattern of activity in posterior medial cortex (PMC) was strongly aligned across participants at these meaningful moments. Further, he found that when participants were recalling the videos, patterns in the PMC were more strongly related to these “storyboard” moments than to moments not chosen for storyboards. Altogether, Adi’s findings suggest that information that is consistent across observers during initial viewing and reinstated during recall reflects a disproportionately small number of moments – those “storyboard” moments – that are able to convey the essentials of the activity.

Ata Karagoz, a graduate student in the lab, presented a talk titled “Event boundaries and interactive inference shape the content and organization of memory.” To date, much of our understanding of how event boundaries affect episodic memory – e.g., enhancement of item recognition for boundary items – has been based on passive viewing tasks. However, little is known about the relationship between event boundaries and memory during interactive events where inference is necessary. To address this, Ata designed an interactive rule inference task using common words, allowing him to test recall & recognition for words at event boundaries (i.e., rule shifts). In contrast to established passive viewing tasks, he found that participants were less likely to recall & recognize items that were encoded immediately after an event boundary. Ata also found that recall was organized by event structure, with boundary and pre-boundary words serving as anchors for retrieval of individual words. In another study variation, Ata informed participants of the sequence of rule changes and found that this eliminated the post-boundary memory deficit. These results confirm that the post-boundary memory deficit is caused by increased inference demands associated with rule shifts. Together, Ata’s findings demonstrate that event segmentation during interactive events is different than during passive viewing, and that, under some circumstances, event boundaries might inhibit rather than enhance memory for specific items.

Leave a Reply

Your email address will not be published. Required fields are marked *