Visual short-term memory capacity predicts the "bandwidth" of visual long-term memory encoding

Decomposing the multiple encoding benefit in visual long-term memory: Primary contributions by the number of encoding opportunities

Although access to the seemingly infinite capacity of our visual long-term memory (VLTM) can be restricted by visual working memory (VWM) capacity at encoding and retrieval, access can be improved with repeated encoding. This leads to the multiple encoding benefit (MEB), the finding that VLTM performance improves as the number of opportunities to encode the same information increases over time. However, as the number of encoding opportunities increases, so do other factors such as the number of identical encoded VWM representations and chances to engage in successful retrieval during each opportunity. Thus, across two experiments, we disentangled the contributions of each of these factors to the MEB by having participants encode a varying number of identical objects across multiple encoding opportunities. Along with behavioural data, we also examined two established EEG correlates that track the number of maintained VWM representations, namely the posterior alpha suppression and the negative slow wave. Here, we identified that the primary mechanism behind the MEB was the number of encoding opportunities. That is, recognition memory performance was higher following an increase in the number of encoding opportunities, and this could not be attributed solely to an increase in the number of encoded VWM representations or successful retrieval. Our results thus contribute to the understanding of the fundamental mechanisms behind the influence of VWM on VLTM encoding.

Perceptual encoding benefit of visual memorability on visual memory formation

Human observers often exhibit remarkable consistency in remembering specific visual details, such as certain face images. This phenomenon is commonly attributed to visual memorability, a collection of stimulus attributes that enhance the long-term retention of visual information. However, the exact contributions of visual memorability to visual memory formation remain elusive as these effects could emerge anywhere from early perceptual encoding to post-perceptual memory consolidation processes. To clarify this, we tested three key predictions from the hypothesis that visual memorability facilitates early perceptual encoding that supports the formation of visual short-term memory (VSTM) and the retention of visual long-term memory (VLTM). First, we examined whether memorability benefits in VSTM encoding manifest early, even within the constraints of a brief stimulus presentation (100-200 ms; Experiment 1). We achieved this by manipulating stimulus presentation duration in a VSTM change detection task using face images with high- or low-memorability while ensuring they were equally familiar to the participants. Second, we assessed whether this early memorability benefit increases the likelihood of VSTM retention, even with post-stimulus masking designed to interrupt post-perceptual VSTM consolidation processes (Experiment 2). Last, we investigated the durability of memorability benefits by manipulating memory retention intervals from seconds to 24 h (Experiment 3). Across experiments, our data suggest that visual memorability has an early impact on VSTM formation, persisting across variable retention intervals and predicting subsequent VLTM overnight. Combined, these findings highlight that visual memorability enhances visual memory within 100-200 ms following stimulus onset, resulting in robust memory traces resistant to post-perceptual interruption and long-term forgetting.

Working memory limitations constrain visual episodic long-term memory at both specific and gist levels of representation

Limitations in one's capacity to encode information in working memory (WM) constrain later access to that information in long-term memory (LTM). The present study examined whether these WM constraints on episodic LTM are limited to specific representations of past episodes or also extend to gist representations. Across three experiments, young adult participants (n = 40 per experiment) studied objects in set sizes of two or six items, either sequentially (Experiments 1a and 1b) or simultaneously (Experiment 2). They then completed old/new recognition tests immediately after each sequence (WM tests). After a long study phase, participants completed LTM conjoint recognition tests, featuring old but untested items from the WM phase, lures that were similar to studied items at gist but not specific levels of representation, and new items unrelated to studied items at both specific and gist levels of representation. Results showed that LTM estimates of specific and gist memory representations from a multinomial-processing-tree model were reduced for items encoded under supra-capacity set sizes (six items) relative to within-capacity set sizes (two items). These results suggest that WM encoding capacity limitations constrain episodic LTM at both specific and gist levels of representation, at least for visual objects. The ability to retrieve from LTM each type of representation for a visual item is contingent on the degree to which the item could be encoded in WM.

Behavioral signatures of the rapid recruitment of long-term memory to overcome working memory capacity limits

Working- and long-term memory are often studied in isolation. To better understand the specific limitations of working memory, effort is made to reduce the potential influence of long-term memory on performance in working memory tasks (e.g., asking participants to remember artificial, abstract items rather than familiar real-world objects). However, in everyday life we use working- and long-term memory in tandem. Here, our goal was to characterize how long-term memory can be recruited to circumvent capacity limits in a typical visual working memory task (i.e., remembering colored squares). Prior work has shown that incidental repetitions of working memory arrays often do not improve visual working memory performance - even after dozens of incidental repetitions, working memory performance often shows no improvement for repeated arrays. Here, we used a whole-report working memory task with explicit rather than incidental repetitions of arrays. In contrast to prior work with incidental repetitions, in two behavioral experiments we found that explicit repetitions of arrays yielded robust improvement to working memory performance, even after a single repetition. Participants performed above chance at recognizing repeated arrays in a later long-term memory test, consistent with the idea that long-term memory was used to rapidly improve performance across array repetitions. Finally, we analyzed inter-item response times and we found a response time signature of chunk formation that only emerged after the array was repeated (inter-response time slowing after two to three items); thus, inter-item response times may be useful for examining the coordinated interaction of visual working and long-term memory in future work.

Long-term memory for distractors: Effects of involuntary attention from working memory

In a visual search task, attention to task-irrelevant distractors impedes search performance. However, is it maladaptive to future performance? Here, I showed that attended distractors in a visual search task were better remembered in long-term memory (LTM) in the subsequent surprise recognition task than non-attended distractors. In four experiments, participants performed a visual search task using real-world objects of a single color. They encoded color in working memory (WM) during the task; because each object had a different color, participants directed their attention to the WM-matching colored distractor. Then, in the surprise recognition task, participants were required to indicate whether an object had been shown in the earlier visual search task, regardless of its color. The results showed that attended distractors were remembered better in LTM than non-attended distractors (Experiments 1 and 2). Moreover, the more participants directed their attention to distractors, the better they explicitly remembered them. Participants did not explicitly remember the color of the attended distractors (Experiment 3) but remembered integrated information with object and color (Experiment 4). When the color of the distractors in the recognition task was mismatched with the color in the visual search task, LTM decreased compared to color-matching distractors. These results suggest that attention to distractors impairs search for a target but is helpful in remembering distractors in LTM. When task-irrelevant distractors become task-relevant information in the future, their attention becomes beneficial.

The Relation Between Attention and Memory

The relation between attention and memory has long been deemed important for understanding cognition, and it was heavily researched even in the first experimental psychology laboratory by Wilhelm Wundt and his colleagues. Since then, the importance of the relation between attention and memory has been explored in myriad subdisciplines of psychology, and we incorporate a wide range of these diverse fields. Here, we examine some of the practical consequences of this relation and summarize work with various methodologies relating attention to memory in the fields of working memory, long-term memory, individual differences, life-span development, typical brain function, and neuropsychological conditions. We point out strengths and unanswered questions for our own embedded processes view of information processing, which is used to organize a large body of evidence. Last, we briefly consider the relation of the evidence to a range of other theoretical views before drawing conclusions about the state of the field.

Delayed memory for complex visual stimuli does not benefit from distraction during encoding

The covert retrieval model (McCabe, Journal of Memory and Language 58(2), 480-494, 2008) postulates that delayed memory performance is enhanced when the encoding of memoranda in working memory (WM) is interrupted by distraction. When subjects are asked to remember stimuli for an immediate memory test, they usually remember them better when the items are presented without distraction, compared to a condition in which a distraction occurs following each item. In a delayed memory test, this effect has been shown to be reversed: Memory performance is better for items followed by distraction than without. Yet, this so-called McCabe effect has not been consistently replicated in the past. In an extensive replication attempt of a previous study showing the effect for complex visual stimuli, we investigated five potential boundary conditions of the predictions of the covert retrieval model: (1) Type of Stimuli (doors vs. faces), (2) type of distractor (pictures vs. math equations), (3) expectation about task difficulty (mixed vs. blocked lists), (4) memory load in WM (small vs. large), and (5) expectation about the long-term memory (LTM) test (intentional vs. incidental encoding). Across four experiments we failed to replicate the original findings and show that delayed memory for faces and other complex visual stimuli does not benefit from covert retrieval during encoding - as suggested as being induced by distractors. Our results indicate that the transfer of information from WM to LTM does not seem to be influenced by covert retrieval processes, but rather that a fixed proportion of information is laid down as a more permanent trace.

Tracing the emergence of the memorability benefit

Some visual stimuli are consistently better remembered than others across individuals, due to variations in memorability (the stimulus-intrinsic property that determines ease of encoding into visual long-term memory (VLTM)). However, it remains unclear what cognitive processes give rise to this mnemonic benefit. One possibility is that this benefit is imbued within the capacity-limited bottleneck of VLTM encoding, namely visual working memory (VWM). More precisely, memorable stimuli may be preferentially encoded into VLTM because fewer cognitive resources are required to store them in VWM (efficiency hypothesis). Alternatively, memorable stimuli may be more competitive in obtaining cognitive resources than forgettable stimuli, leading to more successful storage in VWM (competitiveness hypothesis). Additionally, the memorability benefit might emerge post-VWM, specifically, if memorable stimuli are less prone to be forgotten (i.e., are "stickier") than forgettable stimuli after they pass through the encoding bottleneck (stickiness hypothesis). To test this, we conducted two experiments to examine how memorability benefits emerge by manipulating the stimulus memorability, set size, and degree of competition among stimuli as participants encoded them in the context of a working memory task. Subsequently, their memory for the encoded stimuli was tested in a VLTM task. In the VWM task, performance was better for memorable stimuli compared to forgettable stimuli, supporting the efficiency hypothesis. In addition, we found that when in direct competition, memorable stimuli were also better at attracting limited VWM resources than forgettable stimuli, supporting the competitiveness hypothesis. However, only the efficiency advantage translated to a performance benefit in VLTM. Lastly, we found that memorable stimuli were less likely to be forgotten after they passed through the encoding bottleneck imposed by VWM, supporting the "stickiness" hypothesis. Thus, our results demonstrate that the memorability benefit develops across multiple cognitive processes.

Working Memory Constrains Long-Term Memory in Children and Adults: Memory of Objects and Bindings

We explored how individual and age-related differences in working memory (WM) capacity affected subsequent long-term memory (LTM) retrieval. Unlike past studies, we tested WM and LTM not only for items, but also for item-color bindings. Our sample included 82 elementary school children and 42 young adults. The participants performed a WM task with images of unique everyday items presented sequentially at varying set sizes in different colors. Later, we tested LTM for items and item-color bindings from the WM task. The WM load during encoding constrained LTM, and participants with a higher WM capacity retrieved more items in the LTM test. Even when accounting for young children's poor item memory by considering only the items that they did remember, they exhibited an exacerbated difficulty with remembering item-color bindings in WM. Their LTM binding performance, however, as a proportion of remembered objects, was comparable to that of older children and adults. The WM binding performance was better during sub-span encoding loads, but with no clear transfer of this benefit to LTM. Overall, LTM item memory performance was constrained by individual and age-related WM limitations, but with mixed consequences for binding. We discuss the theoretical, practical, and developmental implications of this WM-to-LTM bottleneck.

Contralateral delay activity, but not alpha lateralization, indexes prioritization of information for working memory storage

Working memory is inherently limited, which makes it important to select and maintain only task-relevant information and to protect it from distraction. Previous research has suggested the contralateral delay activity (CDA) and lateralized alpha oscillations as neural candidates for such a prioritization process. While most of this work focused on distraction during encoding, we examined the effect of external distraction presented during memory maintenance. Participants memorized the orientations of three lateralized objects. After an initial distraction-free maintenance interval, distractors appeared in the same location as the targets or in the opposite hemifield. This distraction was followed by another distraction-free interval. Our results show that CDA amplitudes were stronger in the interval before compared with the interval after the distraction (i.e., CDA amplitudes were stronger in response to targets compared with distractors). This amplitude reduction in response to distractors was more pronounced in participants with higher memory accuracy, indicating prioritization and maintenance of relevant over irrelevant information. In contrast, alpha lateralization did not change from the interval before distraction compared with the interval after distraction, and we found no correlation between alpha lateralization and memory accuracy. These results suggest that alpha lateralization plays no direct role in either selective maintenance of task-relevant information or inhibition of distractors. Instead, alpha lateralization reflects the current allocation of spatial attention to the most salient information regardless of task-relevance. In contrast, CDA indicates flexible allocation of working memory resources depending on task-relevance.

Repetition enhances the effects of activated long-term memory

Recent research indicates that visual long-term memory (vLTM) representations directly interface with perception and guide attention. This may be accomplished through a state known as activated LTM, however, little is known about the nature of activated LTM. Is it possible to enhance the attentional effects of these activated representations? And furthermore, is activated LTM discrete (i.e., a representation is either active or not active, but only active representations interact with perception) or continuous (i.e., there are different levels within the active state that all interact with perception)? To answer these questions, in the present study, we measured intrusion effects during a modified Sternberg task. Participants saw two lists of three complex visual objects, were cued that only one list was relevant for the current trial (the other list was, thus, irrelevant), and then their memory for the cued list was probed. Critically, half of the trials contained repeat objects (shown 10 times each), and half of the trials contained non-repeat objects (shown only once each). Results indicated that repetition enhanced activated LTM, as the intrusion effect (i.e., longer reaction times to irrelevant list objects than novel objects) was larger for repeat trials compared with non-repeat trials. These initial findings provide preliminary support that LTM activation is continuous, as the intrusion effect was not the same size for repeat and non-repeat trials. We conclude that researchers should repeat stimuli to increase the size of their effects and enhance how LTM representations interact with perception.

The Relationship between Short- and Long-Term Memory Is Preserved across the Age Range

Both short- and long-term memories decline with healthy ageing. The aims of the current study were twofold: firstly, to build on previous studies and investigate the presence of a relationship between short- and long-term memories and, secondly, to examine cross-sectionally whether there are changes in this relationship with age. In two experiments, participants across the age range were tested on contextual-spatial memories after short and long memory durations. Experimental control in stimulus materials and task demands enabled the analogous encoding and probing for both memory durations, allowing us to examine the relationship between the two memory systems. Across two experiments, in line with previous studies, we found both short-term memory and long-term memory declined from early to late adulthood. Additionally, there was a significant relationship between short- and long-term memory performance, which, interestingly, persisted throughout the age range. Our findings suggest a significant degree of common vulnerability to healthy ageing for short- and long-term memories sharing the same spatial-contextual associations. Furthermore, our tasks provide a sensitive and promising framework for assessing and comparing memory function at different timescales in disorders with memory deficits at their core.

The role of attention control in complex real-world tasks

Working memory capacity is an important psychological construct, and many real-world phenomena are strongly associated with individual differences in working memory functioning. Although working memory and attention are intertwined, several studies have recently shown that individual differences in the general ability to control attention is more strongly predictive of human behavior than working memory capacity. In this review, we argue that researchers would therefore generally be better suited to studying the role of attention control rather than memory-based abilities in explaining real-world behavior and performance in humans. The review begins with a discussion of relevant literature on the nature and measurement of both working memory capacity and attention control, including recent developments in the study of individual differences of attention control. We then selectively review existing literature on the role of both working memory and attention in various applied settings and explain, in each case, why a switch in emphasis to attention control is warranted. Topics covered include psychological testing, cognitive training, education, sports, police decision-making, human factors, and disorders within clinical psychology. The review concludes with general recommendations and best practices for researchers interested in conducting studies of individual differences in attention control.

Linguistic Redundancy and its Effects on Younger and Older Adults' Real-Time Comprehension and Memory

Redundant modifiers can facilitate referential interpretation by narrowing attention to intended referents. This is intriguing because, on traditional accounts, redundancy should impair comprehension. Little is known, however, about the effects of redundancy on older adults' comprehension. Older adults may show different patterns due to age-related decline (e.g., processing speed and memory) or their greater proclivity for linguistic redundancy, as suggested in language production studies. The present study explores the effects of linguistic redundancy on younger and older listeners' incremental referential processing, judgments of informativity, and downstream memory performance. In an eye tracking task, gaze was monitored as listeners followed instructions from a social robot referring to a unique object within a multi-object display. Critical trials were varied in terms of modifier type ("…closed/purple/[NONE] umbrella") and whether displays contained another object matching target properties (closed purple notebook), making modifiers less effective at narrowing attention. Relative to unmodified descriptions, redundant color modifiers facilitated comprehension, particularly when they narrowed attention to a single referent. Descriptions with redundant state modifiers always impaired real-time comprehension. In contrast, memory measures showed faster recognition of objects previously described with redundant state modifiers. Although color and state descriptions had different effects on referential processing and memory, informativity judgments showed participants perceived them as informationally redundant to the same extent relative to unmodified descriptions. Importantly, the patterns did not differ by listener age. Together, the results show that the effects of linguistic redundancy are stable across adulthood but vary as a function of modifier type, visual context, and the measured phenomenon.

The role of visual working memory capacity in attention capture among video game players

It is well established that attention can be captured by salient distractors. Some studies have found that action video game players were less susceptible to attention capture by irrelevant distractors than non-players. Other studies have also found that individuals with greater visual working memory capacity are less susceptible to capture by irrelevant distractors than individuals with lower visual working memory capacity. The present study examined whether action video game players were less susceptible to be captured by salient distractors and, if so, whether that relationship was due to greater visual working memory capacity. Participants completed a questionnaire reporting their video game playing experience, followed by a color change detection task assessing their visual working memory capacity. They then performed an attention capture task, where they determined the orientation of a bar within a shape singleton while attempting to ignore a color singleton distractor that appeared in 50% of the trials. Results showed that action video game players did not produce less capture effect than non-action video game players. However, high visual working memory capacity individuals produced less capture effect than low visual working memory capacity individuals regardless of their video game experience. These results suggest that the ability to resist capture by irrelevant distractors may be better explained by individual differences in visual working memory capacity than by action video game experience.

Is long-term memory used in a visuo-spatial change-detection paradigm?

In tests of working memory with verbal or spatial materials, repeating the same memory sets across trials leads to improved memory performance. This well-established "Hebb repetition effect" could not be shown for visual materials in previous research. The absence of the Hebb effect can be explained in two ways: Either persons fail to acquire a long-term memory representation of the repeated memory sets, or they acquire such long-term memory representations, but fail to use them during the working memory task. In two experiments (N1 = 18 and N2 = 30), we aimed to decide between these two possibilities by manipulating the long-term memory knowledge of some of the memory sets used in a change-detection task. Before the change-detection test, participants learned three arrays of colors to criterion. The subsequent change-detection test contained both previously learned and new color arrays. Change detection performance was better on previously learned compared with new arrays, showing that long-term memory is used in change detection.

Visual Short-Term Memory for Coherent and Sequential Motion: A rTMS Investigation

We investigated the role of the human medio-temporal complex (hMT+) in the memory encoding and storage of a sequence of four coherently moving random dot kinematograms (RDKs), by applying repetitive transcranial magnetic stimulation (rTMS) during an early or late phase of the retention interval. Moreover, in a second experiment, we also tested whether disrupting the functional integrity of hMT+ during the early phase impaired the precision of the encoded motion directions. Overall, results showed that both recognition accuracy and precision were worse in middle serial positions, suggesting the occurrence of primacy and recency effects. We found that rTMS delivered during the early (but not the late) phase of the retention interval was able to impair not only recognition of RDKs, but also the precision of the retained motion direction. However, such impairment occurred only for RDKs presented in middle positions along the presented sequence, where performance was already closer to chance level. Altogether these findings suggest an involvement of hMT+ in the memory encoding of visual motion direction. Given that both position sequence and rTMS modulated not only recognition but also the precision of the stored information, these findings are in support of a model of visual short-term memory with a variable resolution of each stored item, consistent with the assigned amount of memory resources, and that such item-specific memory resolution is supported by the functional integrity of area hMT+.

ResMem-Net: memory based deep CNN for image memorability estimation

Image memorability is a very hard problem in image processing due to its subjective nature. But due to the introduction of Deep Learning and the large availability of data and GPUs, great strides have been made in predicting the memorability of an image. In this paper, we propose a novel deep learning architecture called ResMem-Net that is a hybrid of LSTM and CNN that uses information from the hidden layers of the CNN to compute the memorability score of an image. The intermediate layers are important for predicting the output because they contain information about the intrinsic properties of the image. The proposed architecture automatically learns visual emotions and saliency, shown by the heatmaps generated using the GradRAM technique. We have also used the heatmaps and results to analyze and answer one of the most important questions in image memorability: "What makes an image memorable?". The model is trained and evaluated using the publicly available Large-scale Image Memorability dataset (LaMem) from MIT. The results show that the model achieves a rank correlation of 0.679 and a mean squared error of 0.011, which is better than the current state-of-the-art models and is close to human consistency (p = 0.68). The proposed architecture also has a significantly low number of parameters compared to the state-of-the-art architecture, making it memory efficient and suitable for production.

Sustained Attention and Spatial Attention Distinctly Influence Long-term Memory Encoding

Our attention is critically important for what we remember. Prior measures of the relationship between attention and memory, however, have largely treated "attention" as a monolith. Here, across three experiments, we provide evidence for two dissociable aspects of attention that influence encoding into long-term memory. Using spatial cues together with a sensitive continuous report procedure, we find that long-term memory response error is affected by both trial-by-trial fluctuations of sustained attention and prioritization via covert spatial attention. Furthermore, using multivariate analyses of EEG, we track both sustained attention and spatial attention before stimulus onset. Intriguingly, even during moments of low sustained attention, there is no decline in the representation of the spatially attended location, showing that these two aspects of attention have robust but independent effects on long-term memory encoding. Finally, sustained and spatial attention predicted distinct variance in long-term memory performance across individuals. That is, the relationship between attention and long-term memory suggests a composite model, wherein distinct attentional subcomponents influence encoding into long-term memory. These results point toward a taxonomy of the distinct attentional processes that constrain our memories.

Children's long-term retention is directly constrained by their working memory capacity limitations

We explored the causal role of individual and age-related differences in working memory (WM) capacity in long-term memory (LTM) retrieval. Our sample of 160 participants included 120 children (6-13-years old) and 40 young adults (18-24 years). Participants performed a WM task with images of unique everyday items, presented at varying set sizes. Subsequently, we tested participants' LTM for items from the WM task. Using these measures, we estimated the ratio at which items successfully held in WM were recognized in LTM. While WM and LTM generally improved with age, the ability to transfer information from WM to LTM appeared consistent between age groups. Moreover, individual differences in WM capacity appeared to predict LTM encoding. Overall, these results suggested that LTM performance was constrained by experimental, individual, and age-related WM limitations. We discuss the theoretical and practical implications of this WM-to-LTM bottleneck.

Working memory limits severely constrain long-term retention

There has been considerable controversy in recent years as to whether information held in working memory (WM) is rapidly forgotten or automatically transferred to long-term memory (LTM). Although visual WM capacity is very limited, we appear able to store a virtually infinite amount of information in visual LTM. Still, LTM retrieval often fails. Some view visual WM as a mental sketchpad that is wiped clean when new information enters, but not a consistent precursor of LTM. Others view the WM and LTM systems as inherently linked. Distinguishing between these possibilities has been difficult, as attempts to directly manipulate the active holding of information in visual WM has typically introduced various confounds. Here, we capitalized on the WM system's capacity limitation to control the likelihood that visual information was actively held in WM. Our young-adult participants (N = 103) performed a WM task with unique everyday items, presented in groups of two, four, six, or eight items. Presentation time was adjusted according to the number of items. Subsequently, we tested participants' LTM for items from the WM task. LTM was better for items presented originally within smaller WM set sizes, indicating that WM limitations contribute to subsequent LTM failures, and that holding items in WM enhances LTM encoding. Our results suggest that a limit in WM capacity contributes to an LTM encoding bottleneck for trial-unique familiar objects, with a relatively large effect size.

Insights into human cognition from intracranial EEG: A review of audition, memory, internal cognition, and causality

By recording neural activity directly from the human brain, researchers gain unprecedented insight into how neurocognitive processes unfold in real time. We first briefly discuss how intracranial electroencephalography (iEEG) recordings, performed for clinical practice, are used to study human cognition with the spatiotemporal and single-trial precision traditionally limited to non-human animal research. We then delineate how studies using iEEG have informed our understanding of issues fundamental to human cognition: auditory prediction, working and episodic memory, and internal cognition. We also discuss the potential of iEEG to infer causality through the manipulation or 'engineering' of neurocognitive processes via spatiotemporally precise electrical stimulation. We close by highlighting limitations of iEEG, potential of burgeoning techniques to further increase spatiotemporal precision, and implications for future research using intracranial approaches to understand, restore, and enhance human cognition.

Perturbing Neural Representations of Working Memory with Task-irrelevant Interruption

Working memory maintains information so that it can be used in complex cognitive tasks. A key challenge for this system is to maintain relevant information in the face of task-irrelevant perturbations. Across two experiments, we investigated the impact of task-irrelevant interruptions on neural representations of working memory. We recorded EEG activity in humans while they performed a working memory task. On a subset of trials, we interrupted participants with salient but task-irrelevant objects. To track the impact of these task-irrelevant interruptions on neural representations of working memory, we measured two well-characterized, temporally sensitive EEG markers that reflect active, prioritized working memory representations: the contralateral delay activity and lateralized alpha power (8-12 Hz). After interruption, we found that contralateral delay activity amplitude momentarily sustained but was gone by the end of the trial. Lateralized alpha power was immediately influenced by the interrupters but recovered by the end of the trial. This suggests that dissociable neural processes contribute to the maintenance of working memory information and that brief irrelevant onsets disrupt two distinct online aspects of working memory. In addition, we found that task expectancy modulated the timing and magnitude of how these two neural signals responded to task-irrelevant interruptions, suggesting that the brain's response to task-irrelevant interruption is shaped by task context.