William Narhi-Martinez¹, Pranava Dhar¹, Kaiki Chiu¹, Anna C. Nobre¹; ¹Yale University

We investigated how internal attention is directed to contents within or between short-term (STM) and long-term memory (LTM) across two in-person behavioral studies. Many contexts in daily life are associated with both LTM and STM content. We wanted to understand how attention operates when directed to one versus both of these internal domains. For LTM, participants learned the identities and locations of two objects within each of 48 scenes. The next day, the same scenes appeared with two new objects in either different (Experiment 1) or occasionally the same (Experiment 2) locations as the LTM objects. Participants encoded these new items into STM. At the end of the trial, an old item and foil appeared, and participants chose the old item. The target could come from either LTM or STM. During the STM delay, retrocues could guide internal attention. In Experiment 1, retrocues were spatial and indicated the side of two objects (both LTM, both STM, or one from each). In Experiment 2, retrocues were either spatial or neutral, indicating all four objects should continue to be maintained. The spatial retrocues could indicate two items with overlapping or non-overlapping locations, one from each memory domain. Our results showed significant performance benefits of internal attention for both LTM and STM, with retrocue benefits being stronger for STM than LTM probes. Surprisingly, spatially overlapping LTM and STM associations had relatively little effect. Interestingly, performance on LTM probes was significantly worse when retrocues prioritized items from both LTM and STM domains rather than the LTM domain alone. Our study paves the way for studying flexible internal attention for contextual associations in different memory domains. The initial findings challenge the standard view that once an LTM is recalled into STM, it is maintained in an equivalent format to new information encoded from sensory stimulation.