Issam Tafech¹, Karla Matic^1,2,4, Polina Iamshchinina³, Thomas B. Christophel^1,2; ¹Humboldt University of Berlin, Germany, ²Bernstein Center for Computational Neuroscience Berlin, Germany, ³Princeton Neuroscience Institute, USA, ⁴Max Planck School of Cognition, Germany

Attentional prioritization has consistently been shown to improve information maintenance and recall. However, it has remained an open question whether attentional prioritization affects visual working memory and perceptual encoding in similar ways. Prior work has used pre-stimulus and post-stimulus cues to prioritize perceptual input and working memory content respectively, allowing for their comparison. However, experimental variations such as cue predictability and the sensitivity of the report task have led to mixed findings on the relative strength of these effects. Here, we aimed to quantify and compare the effects of prioritization in perception and in working memory on recall error using pre-cues and post-cues, respectively. In our design, participants had to do a continuous recall task, reporting one of two oriented gabor patches. Two separate experiments were run: in Experiment 1, the cues were probabilistic (75% report cued item, 25% report uncued item); in Experiment 2, the cues were deterministic (100% report cued item) but participants were cued twice per trial, possibly requiring an previously uncued item to be reported later. We showed that in both experiments, both prioritization in perception and in working memory have significant effects on performance. By comparing performance for cued and uncued items to a neutral condition with no pre-response prioritization, we found improved recall accuracy for cued items (prioritization effect) and reduced recall accuracy for initially uncued items (deprioritization effects). In both experiments, prioritization and deprioritization effects were stronger in perception than in working memory. The initial prioritization effect on cued items was similar in both tasks and participants reported similar levels of subjective prioritization. Furthermore, subsequent effects of prioritization and deprioritization on the second reported items were weaker than initial effects. These results show that attention can modulate perceptual input more strongly than mnemonic representations and suggest that prioritization effects are subject to decay.

Acknowledgements: This work was supported by DFG Emmy Noether Research Group Grant CH 1674/2-1