Attentional refreshing multiple working memory objects via exogenous retro-cues
Poster Presentation: Tuesday, May 20, 2025, 2:45 – 6:45 pm, Pavilion
Session: Visual Memory: Working memory and attention
Schedule of Events | Search Abstracts | Symposia | Talk Sessions | Poster Sessions
Juyeon Joe1, Min-Shik Kim1; 1Yonsei University
Working memory (WM) is an important cognitive function that enables the temporary storage and manipulation of information necessary for daily life. Attention enhances WM performance and is often studied through internal attention manipulations such as retro-cues. Recent studies suggest that non-predictive retro-cues can selectively refresh WM contents (Souza et al., 2015; van Ede et al., 2020). However, it remains unclear whether nonpredictive exogenous retro-cues can facilitate the maintenance of multiple objects in WM. To address this, we investigated the effect of attentional refreshing on multiple objects. Participants were instructed to memorize four different colored shapes, and after the memory array disappeared, two of the four placeholders flickered to direct internal attention to specific objects in WM. Participants then determined whether a test probe presented at the end of each trial matched any of the memorized objects. For non-matching probes, feature combinations varied across conditions: containing both shape and color, only one feature, or no features from the WM items. Sensitivity was higher for cued compared to uncued objects, demonstrating that non-predictive exogenous retro-cues effectively refresh multiple WM objects. However, the presence of cued features in nonmatching probes did not improve rejection accuracy. Additionally, accuracy decreased as the number of shared features between the test probe and WM objects increased. These findings suggest that while attentional refreshing in WM operates on bound object representations, WM performance is constrained by feature-based processes. Thus, non-predictive exogenous cues can enhance attention to multiple objects, but only when the objects' features are integrated.
Acknowledgements: This work was supported by the National Research Foundation of Korea (NRF) grant funded by the Korea government (MSIT) (NRF-2024S1A5B5A19043380)