Neural mechanisms of learned suppression uncovered by probing the hidden attentional priority map
- Department of Experimental and Applied Psychology, Vrije Universiteit Amsterdam, Netherlands
- Institute Brain and Behavior, Netherlands
- Department of Psychological and Brain Sciences, Boston University, United States
- William James Center for Research, ISPA – Instituto Universitario, Portugal
Figures
Figure 1
Schematic trial and experimental design.
The experiment commenced with a training phase, during which participants were tasked with retaining the spatial cue’s location in their memory for subsequent testing following a delay. After a 5 min break, the training phase transitioned to the test phase, where a search task was embedded during memory maintenance. Upon search display onset, participants were instructed to search for a unique shape singleton while ignoring the color singleton. Unbeknownst to the participants, the color singleton was presented more frequently at a location referred to as the high-probability location (HPL).
Figure 2 with 1 supplement
Behavioral findings in the search task.
The top panels show (A) the mean reaction times (RTs) and (B) the mean accuracy under conditions where the distractor was absent, presented at either the high-probability location (hpl) or the low-probability locations (lpl). The bottom panels show (C) the mean RTs and (D) the mean accuracy in relation to the relative distance between the distractor and the high-probability location. Specifically, dist0, dist1, dist2, dist3, and dist4 signify instances where the distractor was at the hpl, one position, two positions, three positions, and four positions away from the hpl. Small gray dots show data for individual participants. The presence of ‘***’ denotes statistical significance at the level of p<0.001. Error bars (N = 24) indicate condition-specific, within-subject 95% confidence intervals (Morey, 2008).
Figure 2—figure supplement 1
No evidence of behavioral interaction between spatial working memory and visual search.
(A) The average memory recall deviation as a function of the distance between memory cue and high-probability location. (B) The average reaction times as a function of the spatial overlap between the memory cue location and any of the search items, separating trials by distractor-present (match-target, match-distractor, match-neutral) and distractor-absent (match-target, match-neutral) conditions. Error bars (N = 24) indicate condition-specific, within-subject 95% confidence intervals (Morey, 2008).
Figure 3 with 1 supplement
The performance of the localizer in the training phase.
In the training phase, the average alpha-band channel tuning function (CTF) profiles reliably tracked the location of the memory cue following its onset (A). Although the spatial memory representation returned to baseline during the maintenance period, it was revived by a neutral placeholder display (B) that was irrelevant to the memory task. The lower images in (A and B) depict the responses across channels, while the plots above show CTF slopes, with amplitude indicating spatial selectivity. The light-colored lines in the background represent CTF slopes tuned to different memory cue locations, grouped by their distance from a location that will be the high-probability location in the test phase. Visual inspection reveals no spatial bias. Shaded areas represent bootstrapped standard error of the mean (SEM, N = 24). Time points showing significant differences in CTF slopes, identified through a cluster-based permutation test (p<0.05), are marked with horizontal black insets. (C) illustrates that individual CTF profiles, following the memory cue onset, were finely tuned to each of the eight spatial cue locations with comparable response strengths. (D) shows that individual CTF profiles tuned to each of the eight spatial cue locations were revived by the placeholder to a similar extent, although the effect was somewhat obscured in the fourth row of the panel.
Figure 3—figure supplement 1
The performance of the localizer in the training phase.
(A) The average evoked alpha-band channel tuning function (CTF) profiles time-locked to memory onset. (B) The average total alpha CTF profiles time-locked to placeholder onset. The lower image depicts the responses across channels, while the plot above shows CTF slopes, with amplitude signifying spatial selectivity. Shaded areas reflect bootstrapped SEM (N = 24). Time points exhibiting significant differences in CTF slopes, identified through a cluster-based permutation test (p<0.05), are marked with horizontal black insets. The light color lines in the background indicate the CTF slopes tuned to different memory cue locations, grouped by their distance to the artificial high-probability location. (C) Individual total CTF profiles, synchronized with memory cue onset, finely tuned to each of the eight spatial cue locations, respectively. (D) Individual evoked CTF profiles, synchronized with placeholder onset, finely tuned to each of the eight spatial cue locations, respectively.
Figure 4
Cross-session encoding: alpha channel tuning functions (CTFs) of memory cue display and placeholder display.
The location CTFs obtained by generalizing the trained model to the test set, ensuring strict alignment of time samples between the training and test phases. CTFs were reconstructed from evoked and total alpha power following the onset of (A) the memory cue display and (B) the placeholder display. The spatial location of the memory cue was reliably tracked by the alpha CTFs, which returned to baseline before the presentation of the placeholder display. The spatial information maintained in working memory was then revived upon the presentation of the placeholder display, selectively for evoked power. The lower image depicts the responses across channels, while the plot above shows CTF slopes, with amplitude signifying spatial selectivity. Shaded areas reflect bootstrapped SEM (N = 24). Time points exhibiting significant differences in CTF slopes, identified through a cluster-based permutation test (p<0.05), are marked with horizontal black insets.
Figure 5
The gradient pattern of alpha channel tuning functions (CTFs) following the onset of the placeholder display.
(A) Numerically larger CTFs were observed post-placeholder onset when the spatial memory cue matched the high-probability location compared to the low-probability locations. (B) The location-specific CTFs were categorized based on the relative position distance between the spatial memory cue and the high-probability location. To statistically evaluate the observed gradient pattern, a linear regression model was applied to the data points, followed by a permutation test on the slope of the regression model. Shaded areas depict bootstrapped SEM (N = 24), and time points with significant differences (cluster-based permutation test, p<0.05) in regression model slopes are indicated with horizontal black insets.
Author response image 1
Author response image 2
Additional files
Download links
A two-part list of links to download the article, or parts of the article, in various formats.
Downloads (link to download the article as PDF)
Open citations (links to open the citations from this article in various online reference manager services)
Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)
Neural mechanisms of learned suppression uncovered by probing the hidden attentional priority map
eLife 13:RP98304.
https://doi.org/10.7554/eLife.98304.3
