Education · Psychological Research
Research on the Characterization Mechanism and Influencing Factors of Visual Working Memory Precision
SUN Tianyia, BU Yanlua, HOU Xiulib
(a. School of Educational Science; b. School of Law and Sociology, Xinyang Normal University, Xinyang 464000, Henan)
Abstract：Precision is a key factor affecting visual working memory, and the neural mechanism of precision representation is of great significance for improving working memory ability. Previous studies have found that the precision of working memory representation is related to the activity patterns of the entire visual cortex and the parietal sulcus. Detailed memory features are encoded in a distributed activation pattern in the visual cortex and relayed to the parietal cortex. The upper parietal sulcus of the parietal cortex directs attention towards a limited number of memory representations, thereby controlling the representation process of visual information from top to bottom. In addition, factors such as working memory capacity, attention, and negative emotions can all affect the precision of visual working memory representation. Future research should explore the temporal characteristics of visual working memory precision representation and delve deeper into the influencing factors of visual working memory precision representation.
Keywords：visual working memory; precision representation; neural mechanism; influencing factor
1. Introduction
Visual working memory (VWM) refers to the ability of individuals to store and process visual information for a short period after the visual stimuli presented before them disappear[1, 2]. As one of the main mechanisms of human information processing systems, visual working memory provides a mental workspace for higher cognitive activities such as language, learning, and problem-solving. It is generally believed that visual working memory is a resource-limited cognitive system that can temporarily store only 3-4 items for current cognitive processing[3, 4].
Visual working memory is mainly influenced by both the number of items and the precision of representations[5, 6, 7]. The number of items refers to the number of visual representations within a unit time of visual working memory, while representation precision reflects the accuracy of the representations in visual working memory relative to the original memory stimuli[8]. Several studies have shown that as the number of items stored in visual working memory increases, memory precision monotonically decreases[2, 9]. In other words, there is a trade-off in the allocation of limited working memory resources between storage capacity and representation precision. To address this issue, researchers have proposed the capacity-limited model (also known as the slot model), the flexible resource allocation model, and the slot-averaging model that integrates both models[10].
The slot model assumes that working memory resources are quantifiable, and any stored item must be allocated to a limited number of “slots,” which restricts resource allocation. For a large amount of information that needs to be remembered, individuals can only represent a few objects accurately, while they have almost no impression of objects that exceed capacity[2]. The flexible resource allocation model posits that individuals can selectively remember information within limited resources, achieving flexible allocation of resources, meaning that when the number of memory items is high, their precision decreases, and when the precision of memory items is high, the number of items remembered decreases[6, 9]. The slot-averaging model integrates the views of the above two models, suggesting that the number of slots in the memory system is fixed, but unlike the slot model, multiple slots can be used to represent the same object, with each slot storing a representation of that object, and individuals reporting the average of multiple representations during the detection phase. Therefore, when the number of objects is less than the number of slots, representation precision decreases as the number of items increases, while representation precision does not change after exceeding the number of slots[2].
Regarding the two dimensions of visual working memory capacity and precision, due to the significant impact of load levels on memory performance, existing research has focused more on the former, while studies on precision are still in the exploratory stage, especially concerning the neural mechanisms of processing and the influencing factors of precision representation, which require further investigation. Based on this, it is necessary to systematically review and summarize the relevant literature on the precision representation of visual working memory to provide new insights for advancing research in this field. Specifically, this paper aims to first review the research on the precision representation of visual working memory in the field of cognitive neuroscience, including studies using techniques such as functional magnetic resonance imaging (fMRI), transcranial magnetic stimulation (TMS), transcranial direct current stimulation (tDCS), and event-related potentials (ERPs), and based on this, analyze the impact of working memory capacity, attention, negative emotions, and other factors on the precision representation of visual working memory.
2. Neural Mechanisms of Visual Working Memory Precision Representation
The ability to represent working memory precision is one of the factors affecting working memory capacity, and understanding the neural mechanisms of working memory precision representation is crucial for finding training methods to enhance working memory ability. Comprehensive previous research has found that the brain regions associated with visual working memory precision representation mainly involve the visual cortex (VC), parietal cortex (PC), and lateral occipital complex (LOC).
(1) Visual Cortex
Existing studies have found that the stimulus features (such as direction, motion direction, and spatial location) that maintain information in visual working memory can be decoded from the visual cortex (e.g., V1-V4, MT) during the delay period, even though these areas do not show enhanced, sustained delay activation[11]. This indicates that visual working memory representation may be encoded in a distributed activation pattern in the visual cortex. To further explore the causal role of the visual sensory cortex in the process of maintaining stimuli in working memory and the spatial range involved, Redemaker et al. used fMRI results for neural navigation, applied TMS to stimulate the primary visual cortex (V1) that transmits retinal information, and required participants to remember various oriented gratings (presented in different quadrants), subsequently calculating the angular deviation between the reported direction and the actual direction in each trial, and analyzing the error distribution of the results through fitting mixed models to exclude guessing rates[12]. The results showed that low-intensity TMS stimulation improved visual sensitivity, leading to improved memory precision (but not guessing rates), while high-intensity TMS stimulation reduced the precision of visual working memory. The researchers interpreted this result as TMS reducing the amount of information at the stimulated location in the visual cortex through local random noise stimulation, thereby negatively impacting the precision of visual working memory. This study clarified the causal relationship between the primary visual cortex and the precision of visual working memory.
However, researchers found that visual working memory precision representation is not influenced by a single brain region but is modulated by a functional network that regulates cognitive processing (broadly interpreted) and the representation of related stimulus feature information from top to bottom[13]. For example, when researchers changed the presence and predictability of distractors during the delay period of visual working memory tasks, they found that when distractors were predictable, V1-V4 showed successful decoding of visual working memory, but when distractors were present and unpredictable, V1-V4 showed weaker visual working memory decoding[14]. This result indicates that the predictability of distractors allows individuals to strategically decide whether to use the occipital visual cortex during visual working memory representation, representing a form of top-down cognitive control.
(2) Parietal Cortex
Multiple studies have found that the parietal cortex exhibits sustained, enhanced activation responses during the delay period of visual working memory[15], and this delay period activation is believed to reflect the general cognitive or attentional demands of the task[16, 17]. Therefore, this may indicate that the parietal cortex has a mechanism for allocating attention, capable of top-down control of attention and directing it towards a limited number of memory representations. Among them, the activation of the posterior parietal cortex (PPC) is considered most likely to affect the quality of sensory representations[14,18, 19]. For example, Wang et al. applied transcranial direct current stimulation to the posterior parietal cortex of participants while requiring them to remember and subsequently recall the direction of visual items (with varying item sizes of 2, 4, or 6), and finally measured visual working memory capacity and precision by modeling response error distributions. To control for sensory and attentional confounds caused by stimulation, participants also participated in sensory memory trials. The results showed that tDCS stimulation over the posterior parietal cortex selectively enhanced visual working memory capacity at the highest load (set size of 6) without significantly improving visual working memory precision[20]. This experimental result suggests that compared to the role of PPC in affecting the quality of sensory representations in visual working memory, it may have a relatively dominant and causal role in supporting the storage capacity of visual working memory.
However, the upper and lower intraparietal sulcus (IPS) contained in the posterior parietal cortex seem to work in parallel with separable neural mechanisms to support visual working memory. Specifically, regardless of the complexity of the objects, the activation of the lower IPS is limited by a fixed number of objects, while the activation of the upper IPS is more influenced by the complexity of the objects and the total amount of visual information encoded, participating in the detailed feature encoding and maintenance process of visual items. This seems to indicate that the activation of the upper intraparietal sulcus is sensitive to the precision representation of visual working memory, while the activation of the lower intraparietal sulcus is positively correlated with the number of items stored in visual working memory[15, 21].
Voxel-based morphometry (VBM) technology can link individual behavioral abilities with neuroanatomical features, as a voxel-based morphological measurement method, it can quantitatively detect the density and volume of brain tissue[22]. VBM technology can measure and compare the entire brain, directly analyzing raw data, which to some extent compensates for the subjectivity of traditional fMRI measurements based on regions of interest (ROI). A VBM study by Konstantinou et al. found that the gray matter volume of the left insula was positively correlated with individuals’ visual working memory object recognition ability, while the gray matter volume of the lower parietal lobe was positively correlated with visual spatial working memory ability, both showing separable neuroanatomical correlations[23]. Similarly, Machizawa et al. using VBM technology found that visual working memory precision representation is also related to the gray matter volume of the parietal lobe[24], specifically, individuals with high precision representation ability tend to have larger gray matter volume in the right parietal lobe, meaning that the gray matter volume of the right parietal lobe is positively correlated with the ability of visual working memory precision representation.
(3) Lateral Occipital Complex
Some studies have pointed out that the occipital cortex, as a functional area for perceiving and processing visual information, is involved in many complex visual perception processes and is significantly affected by additional stimuli during the encoding of visual working memory, seeming to prefer receiving visual stimuli rather than memory representations[14]. This cognition may stem from previous studies limiting ROIs to areas sensitive to visual working memory load, treating LOC as part of the occipital cortex ROI[11, 25], as it is modulated by visual working memory load during the encoding phase. However, their analysis results did not distinguish the specific role of regions predicting visual working memory precision (i.e., V1-V4, LOC), so it remains unclear whether LOC has a greater (or lesser) impact on visual working memory precision representation than other visual cortex regions. Whole-brain analysis studies have found that during the maintenance phase of visual working memory, the activation of the parietal lobe is related to load, while the activation of LOC is sensitive to visual working memory precision representation and can enhance its sensitivity through information transfer between the inferior frontal gyrus and the frontal parietal cortex[26]. This study did not distinguish the differential effects of the upper and lower intraparietal sulcus on precise representation, which may be due to the complexity and large number of memory items, requiring further research. Additionally, whether visual working memory load and precision can be functionally or anatomically isolated needs more research to determine.
It should be noted that event-related potentials (ERP) have high temporal resolution and are highly advantageous in revealing the temporal processes of cognition. In studies of the trade-off between the number and precision of visual working memory, CDA can be used to measure individuals’ visual working memory capacity limits[5, 27], but in the field of working memory precision, this technology has only been used to address the contaminating issues of behavioral data. For example, the amplitude of CDA is often sensitive to the number of stored visual items[23], and measuring this neural feature during the maintenance phase of working memory can monitor whether participants are affected by interference; if there is distracting information, CDA tends to show an upward trend. Therefore, it can ensure the objectivity of behavioral research results to some extent, assisting in interpreting experiments related to visual working memory precision, but there is no clear EEG component corresponding to visual working memory precision representation.
In summary, visual working memory precision representation is related to the activity patterns of the entire visual cortex and IPS, forming a functional network. Specifically, detailed memory features are encoded in a distributed activation pattern in the visual cortex and relayed to the parietal cortex, where there is an attention allocation mechanism, and the activation of the upper IPS directs attention towards a limited number of memory representations, thereby controlling the representation process from top to bottom. Among them, the higher visual cortex LOC can establish functional connections between the inferior frontal gyrus and the frontal parietal cortex. Currently, there has been some progress in exploring the neural mechanisms of visual working memory precision representation, but the focus has mainly been on the localization of brain regions involved in the trade-off processing of visual working memory quantity and precision, without achieving functional or anatomical separation of the two. Furthermore, little is known about the temporal characteristics of visual working memory precision representation. For example, in ERP studies, only the CDA component has been used to assist in interpreting research related to visual working memory precision[5, 23, 28], but whether CDA directly affects the process of visual working memory precision representation, or whether there are other ERP components related to precision representation, requires further exploration.
3. Influencing Factors of Visual Working Memory Precision Representation
Visual working memory precision representation is influenced by various factors, such as working memory capacity, attention, and negative emotions. Here, we summarize the influencing factors of visual working memory precision, which can provide new research perspectives for exploring the neural mechanisms of precision representation.
(1) The Impact of Working Memory Capacity on Precision Representation
The slot model posits that individuals have a unique visual working memory capacity limited by a certain number of discrete memory “slots,” with an average visual working memory capacity of about 3 items. Objects within this capacity can be accurately represented, while there is almost no impression of objects that exceed this capacity, leading to no significant differences in precision representation among these objects[6]. Additionally, research by Zhang and Luck has also demonstrated that when the number of items individuals need to remember increases from 3 to 6, there is no significant difference in the precision of memory representations[29]. However, visual working memory capacity exhibits significant individual differences, with low-capacity individuals recalling only 1 to 2 items, while high-capacity individuals can recall 6 to 7 items[30, 31]. Studies have also found that individuals with higher visual working memory capacity tend to have better precision representation abilities. Vellage et al. compared the precision representation abilities of high and low visual working memory capacity participants using a continuous delay estimation task, setting task difficulty across three dimensions (set size, cue position, consistency). The results showed that high-capacity participants did not exhibit a significant decline in precision when completing difficult tasks[32].
The flexible resource allocation model suggests that limited resources can be flexibly allocated to a large number of lower-precision items or a few higher-precision items, without being constrained by working memory capacity[9]. To explore whether individuals change the precision of visual working memory representations due to limitations in working memory capacity, Machizawa et al. built upon the research of Zhang and Luck (2011) using a new direction discrimination paradigm that not only varied the number of memory items but also the expected precision of the items participants needed to remember. The results showed that individuals could improve the precision of visual working memory, but only for a few items[33]. Furthermore, Liu Zhiying and Ku Yixuan set low, medium, and high memory loads in a continuous delay estimation task and increased distracting items, finding that at low memory load levels, distractions significantly affected working memory precision[34]. In other words, when the number of memory items is within the range of working memory capacity, distractions reduce the precision of the represented items in memory. This seems to imply that individuals with higher working memory capacity have stronger resistance to distractions, being able to suppress the impact of distractions on precision representation over a larger range.
The slot-averaging model also posits that the memory system contains a fixed number of slots, i.e., working memory capacity. When the number of memory items is within the range of working memory capacity, precision decreases as the number of items increases, and after exceeding the capacity range, precision no longer changes significantly[33]. This model differs from the slot model only in the way memory item features are represented.
(2) The Impact of Attention on Precision Representation
The impact of attention on visual working memory precision is achieved by enhancing the priority and selectively increasing attention to certain items, thereby improving the precision of that item’s representation[35, 36]. In visual working memory, the representation of memory items is not entirely precise but contains noise[9, 37, 38, 39], and selectively attending to relevant representation information can effectively reduce the impact of noise[40, 41], a phenomenon known as retrospective attention[42, 43].
Previous studies have shown that using retrospective attention to change the priority of memory items can enhance the perceptual precision of those items[44]. For example, a study in the auditory domain found that retrospective attention indeed improved the representation precision of auditory items in memory[45], but this evidence only applied under low-load conditions with two syllables. Lim et al. employed psychophysical modeling methods to further increase memory load and tested whether the changes in representation precision due to auditory retrospective attention could generalize to tasks with different cognitive demands. The results showed that retrospective attention indeed contributed to precision enhancement, and this facilitative effect was positively correlated with individuals’ overall working memory performance under different memory loads[44]. Specifically, as memory load increased, individuals’ overall working memory performance was positively correlated with the degree of precision improvement (under high-load conditions, individuals with lower working memory capacity showed the least improvement in precision).
Similar conclusions have been drawn from studies on visual retrospective attention, indicating that the extent to which effective retrospective cues promote precision improvement depends on visual working memory load[46]. Additionally, Curtis et al. expanded the manipulation of priority beyond simple cue prompts, designing two psychophysical experiments to manipulate item priority, namely testing probability and monetary incentive differences, while measuring the visual working memory precision of multiple items involved in priority ranking. The results showed that precision monotonically increased with higher priority. These results suggest that visual working memory resources can be flexibly allocated across multiple items[47]. Therefore, how attention resources are allocated may represent a shared mechanism, under which observable changes in visual working memory precision occur.
(3) The Impact of Negative Emotions on Precision Representation
Existing research shows that negative emotions can increase the priority of detail storage in visual working memory, exhibiting a regulatory effect on both the quality and quantity of visual working memory[19, 48, 49]. Spachtholz et al. induced participants’ neutral or negative emotions through autobiographical tasks and then required them to complete a visual working memory task (remembering a set of colored dots) and reproduce their colors using a continuous color wheel. In the data analysis phase, they modeled the data and calculated the correct retrieval rate and the precision value of the information. The results showed that compared to the neutral emotional state group, participants in the negative emotional state group had a lower correct retrieval rate, but once the color was retrieved correctly, they could reproduce it with higher precision[49]. Xie and Zhang also found that inducing participants’ emotions using emotional images from the International Affective Picture System (IAPS) showed that negative emotions only promoted the representation of visual working memory precision but did not affect the quantity of memory information[40, 29]. Furthermore, some researchers considered the impact of time intervals, finding that when the presentation time of test items was very short (200ms), inducing negative or neutral emotional states had no effect on the retrieval rate and precision of visual working memory, while at longer times (500ms), the same results as previous studies appeared[19, 50].
However, some researchers have obtained inconsistent results. Souza et al. induced participants’ negative, neutral, and positive emotions using emotional images and a combination of emotional images and music, finding that compared to the neutral emotional state, both negative and positive emotions improved precision and the quantity of memory information, but negative emotions did not exhibit a regulatory effect on the quality and quantity of visual working memory[51]. Therefore, the impact of negative emotions on visual working memory precision remains unclear, but emotions, as information about individual states, often become a distraction when unrelated to task goals[52, 53, 54]. From this perspective, exploring the impact of emotions on visual working memory precision is meaningful.
In summary, the importance of visual working memory is reflected not only in daily life and learning tasks but also in individuals’ cognitive thinking processes. The above summarizes the existing research on the neural mechanisms of visual working memory precision representation, finding that the associated brain regions mainly include the visual cortex, parietal cortex, and LOC. Specifically, detailed memory features are encoded in a distributed activation pattern in the visual cortex and relayed to the parietal cortex, where there is an attention allocation mechanism, and the activation of the upper IPS directs attention towards a limited number of memory representations, thereby controlling the representation process of visual memory from top to bottom, while the activation of the lower IPS is sensitive to visual working memory load. The upper and lower IPS support visual working memory with separable parallel processing. Interestingly, whole-brain analysis studies using VBM have also found a correlation between the parietal lobe and visual working memory precision representation, indicating that individuals with high precision representation abilities have larger gray matter volumes in the right parietal lobe. Additionally, the higher visual cortex LOC can establish functional connections between the inferior frontal gyrus and the frontal parietal cortex, enhancing its sensitivity to visual working memory precision representation through information transfer between the two. Previous research has mostly treated sensitivity to visual working memory load as a prerequisite for sensitivity to memory precision, limiting the scope to areas sensitive to visual working memory load. Future research should focus on separating the two and examining the specific characteristics of the processing of visual working memory precision representation. Although whole-brain analysis studies have confirmed the specific sensitivity of LOC to visual working memory precision representation, whether visual working memory load and precision can be functionally or anatomically isolated still requires more research to determine.
4. Limitations and Prospects
Overall, research on the neural mechanisms of visual working memory precision representation has made some progress, but current studies mostly focus on spatial localization of processing brain regions, with little known about the temporal characteristics of visual working memory precision representation. The high temporal resolution technology of ERP has not yet been fully utilized, and the influencing factors of visual working memory, such as emotions, have clear specific components associated with them in existing research[55, 56]. Future research should not only continue to explore the use of technology but also consider working memory capacity, attention, negative emotions, and other influencing factors as new starting points for a deeper investigation into the neural mechanisms of visual working memory precision representation.
First, in the research on the impact of working memory capacity and precision, most studies use the two-component mixture model proposed by Zhang and Luck to achieve the separation of working memory precision and capacity indicators[2, 57], but some researchers have questioned the validity of this indicator[58]. Therefore, a new indicator has been proposed based on guessing rates and standard deviations to quantify individuals’ trade-off abilities between working memory quantity and precision[59], namely the general voluntary trade-off index (GT). The calculation process of this indicator can minimize individual differences among participants, reflecting the degree to which participants spontaneously trade off due to task demands, as indicated by the extent of change in the indicator under high precision conditions compared to low precision conditions. However, it cannot simultaneously account for changes in quantity and precision, and future research should further improve this indicator to explore individual trade-off abilities or investigate the relationship between individual trade-off abilities and attentional filtering capabilities based on attention allocation. Additionally, existing research has focused on the impact of high and low visual working memory capacity on precision, but there is little research on whether object complexity or precision demands can alter visual working memory capacity. Although some studies have indicated that the capacity of visual working memory decreases with increasing object complexity, Awh et al. argue that this decrease is due to increased similarity between memory stimuli and probe stimuli, and that individuals’ memory representation precision is limited, leading to more errors during reporting, which manifests as a decrease in capacity, while the actual number of stored items has not decreased[60]. Therefore, the hypothesis that increased complexity leads to decreased capacity still requires more evidence for verification.
Second, regarding the influencing factors of attention, Zhang and Luck’s method of changing item priority through monetary incentives did not find evidence that priority affects visual working memory precision[30], but Curtis et al.’s use of two psychophysical experiments (testing probability and monetary incentive differences) to manipulate item priority observed changes in precision. In other words, different manipulations of attentional priority led to differing results[47]. Therefore, future research should focus on evaluating the effectiveness of methods for manipulating attentional priority or exploring different ways to manipulate attentional priority, combined with studies on their neural mechanisms, to clarify the role of individual attention in the process of visual working memory precision representation.
Third, the impact of negative emotions on visual working memory precision remains controversial. The reasons for this can be roughly divided into several aspects. First, the emotional evaluation methods currently used are overly simplistic, while emotions are diverse; for example, negative emotions include fear, disgust, sadness, and other styles, which may have different effects on visual working memory precision representation. Future research could consider the impact of different conceptualized emotions to enrich existing studies. Second, the duration of emotional manipulation is very short, making it difficult to have a lasting impact on participants, thus failing to ensure the effectiveness of the manipulation, which is unhelpful for evaluating and observing experimental results. Future research should adopt more effective emotional manipulations to enhance the observability of results, such as watching videos. Additionally, a recent meta-analysis on the role of emotions in visual working memory found that emotional information continuously activates brain regions related to processing visual working memory information, but the behavioral impact is minimal, seemingly due to individuals’ effective control over negative emotions, and this control effect is significantly reduced in populations with mental health issues[61]. Future research could utilize ERP technology to investigate the temporal mechanisms of this control effect to further clarify the neural mechanisms of working memory precision representation.
References：
[1] ZHANG L, YAMADA Y. Alternating capture of attention by multiple visual working memory representations[J]. Sci Rep, 2023, 13(1):130-29.
[2] ZHANG W, LUCK S J. Discrete fixed-resolution representations in visual working memory[J]. Nature, 2008, 453:233-235.
[3] PRIETO A, PEINADO V, MAYAS J. Does perceptual grouping improve visuospatial working memory? Optimized processing or encoding bias[J]. Psychol Res, 2022, 86(4):1297-1309.
[4] LEONARD C J, KAISER S T, ROBINSON B M, et al. Toward the neural mechanisms of reduced working memory capacity in schizophrenia[J]. Cereb Cortex, 2013, 23(7):1582-1592.
[5] VOGEL E K, MACHIZAWA M G. Neural activity predicts individual differences in visual working memory capacity[J]. Nature, 2004, 428(6984):748-751.
[6] LUCK S J, VOGEL E K. The capacity of visual working memory for features and conjunctions[J]. Nature, 1997, 390(6657):279-281.
[7] Zhang Z, Zhang L, Gong R. The filtering efficiency of visual working memory[J]. Psychological Science Progress, 2021, 29(4):635-651.
[8] FUKUDA K, AWH E, VOGEL E K. Discrete capacity limits in visual working memory[J]. Current Opinion in Neurobiology, 2010, 20(2):177-182.
[9] BAYS P M, HUSAIN M. Dynamic shifts of limited working memory resources in human vision[J]. Science (New York, N.Y.), 2008, 321(5890):851-854.
[10] He X, Guo C. The capacity and resource allocation of visual working memory[J]. Psychological Science Progress, 2013, 21(10):1741-1748.
[11] EMRICH S M, RIGGALL A C, LAROCQUE J J, et al. Distributed patterns of activity in sensory cortex reflect the precision of multiple items maintained in visual short-term memory[J]. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 2013, 33(15):6516-6523.
[12] RADEMAKER R L, VAN DE VEN V G, TONG F, et al. The impact of early visual cortex transcranial magnetic stimulation on visual working memory precision and guess rate[J]. PlOS ONE, 2017, 12(4):e175230.
[13] Zhang B, Hu C, Chen Y, et al. The modulation of working memory and perceptual load on guiding attention in working memory representation[J]. Acta Psychologica Sinica, 2017, 49(8):1009-1021.
[14] BETTENCOURT K C, XU Y. Decoding the content of visual short-term memory under distraction in occipital and parietal areas[J]. Nature Neuroscience, 2016, 19(1):150-157.
[15] ANDERSON D E, VOGEL E K, AWH E. Precision in visual working memory reaches a stable plateau when individual item limits are exceeded[J]. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 2011, 31(3):1128-1138.
[16] LI X, O’SULLIVAN M J, MATTINGLEY J B. Delay activity during visual working memory: A meta-analysis of 30 fMRI experiments[J]. Neuroimage, 2022, 255:119204.
[17] MITCHELL D J, CUSACK R. Flexible, capacity-limited activity of posterior parietal cortex in perceptual as well as visual short-term memory tasks[Z]. 2008(18):1788-1798.
[18] ZHAO, DI, ZHOU, et al. The causal role of the prefrontal cortex and somatosensory cortex in tactile working memory[J]. Cerebral Cortex, 2018, 28(10):3468-3477.
[19] SPRAGUE T C, ESTER E F, SERENCES J T. Reconstructions of information in visual spatial working memory degrade with memory load[J]. Current Biology: CB, 2014, 24(18):2174-2180.
[20] WANG S, ITTHIPURIPAT S, KU Y. Electrical stimulation over human posterior parietal cortex selectively enhances the capacity of visual short-term memory[J]. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 2019, 39(3):528-536.
[21] XU Y, CHUN M M. Dissociable neural mechanisms supporting visual short-term memory for objects[J]. Nature, 2006, 440(7080):91-95.
[22] KAMASAK B, ULCAY T, NISARI M, et al. Effects of COVID-19 on brain and cerebellum: A voxel-based morphometrical analysis[J]. Bratisl Lek Listy, 2023, 124(6):442-448.
[23] KONSTANTINOU N, CONSTANTINIDOU F, KANAI R. Discrete capacity limits and neuroanatomical correlates of visual short-term memory for objects and spatial locations[J]. Hum Brain Mapp, 2017, 38:767-778.
[24] MACHIZAWA M G, DRIVER J, WATANABE T. Gray matter volume in different cortical structures dissociably relates to individual differences in capacity and precision of visual working memory[J]. Cerebral Cortex (New York, N.Y.: 1991), 2020, 30(9):4759-4770.
[25] GOSSERIES O, YU Q, LAROCQUE J J, et al. Parietal-occipital interactions underlying control and representation-related processes in working memory for nonspatial visual features[J]. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 2018, 38(18):4357-4366.
[26] ZHAO Y, KUAI S, ZANTO T P, et al. Neural correlates underlying the precision of visual working memory[J]. Neuroscience, 2020, 425:301-311.
[27] HAO R, BECKER M W, YE C, et al. The bandwidth of VWM consolidation varies with the stimulus feature: Evidence from event-related potentials[J]. J Exp Psychol Hum Percept Perform, 2018, 44(5):767-777.
[28] ROY Y, FAUBERT J. Is the contralateral delay activity (CDA) a robust neural correlate for visual working memory (VWM) tasks? A reproducibility study[J]. Psychophysiology, 2023, 60(2):e14180.
[29] ZHANG W, LUCK S J. The number and quality of representations in working memory[Z]. 2011(22):1434-1441.
[30] Wang L, Lü Kuangdi, Zhang Qian, et al. The effect of proximity on the similarity effect in visual working memory under different spatial configurations[J]. Psychological Science, 2021, 44(2):258-265.
[31] YE C, XU Q, LIU Q, et al. The impact of visual working memory capacity on the filtering efficiency of emotional face distractors[J]. Biological Psychology, 2018, 138:63-72.
[32] VELLAGE A, MULLER P, SCHMICKER M, et al. High working memory capacity at the cost of precision[J]. Brain Sciences, 2019, 9(9):210.
[33] MACHIZAWA M G, GOH C C W, DRIVER J. Human visual short-term memory precision can be varied at will when the number of retained items is low[J]. Psychological Science, 2012, 23(6):554-559.
[34] Liu Zhiying, Ku Yixuan. The impact of perceptual representation precision on the ability to inhibit interference in working memory[J]. Acta Psychologica Sinica, 2017, 49(10):1247-1255.
[35] Song Jiaru, Liu Yuanyuan, Wang Xiuli, et al. The guidance of visual and verbal working memory representations on visual attention[J]. Psychological Development and Education, 2022, 38(5):609-617.
[36] Che Xiaowei, Xu Huiyun, Wang Kaixuan, et al. The impact of processing demands of working memory representation precision on attention guidance[J]. Acta Psychologica Sinica, 2021, 53(7):694-713.
[37] BAYS P M. Spikes not slots: Noise in neural populations limits working memory[J]. Trends in Cognitive Sciences, 2015, 19(8):431-438.
[38] MA W J, HUSAIN M, BAYS P M. Changing concepts of working memory.[J]. Nature Neuroscience, 2014, 17(3):347-356.
[39] Che Xiaowei, Wang Kaixuan, Shangguan Mengqi, et al. The representation of attention templates in visual working memory: Evidence from EROS[J]. Psychological and Behavioral Research, 2020, 18(3): 297-303.
[40] XIE K, JIN Z, NI X, et al. Distinct neural substrates underlying target facilitation and distractor suppression: A combined voxel-based morphometry and resting-state functional connectivity study[J]. NeuroImage, 2020, 221:117149.
[41] GETZMANN S, SCHNEIDER D, WASCHER E. The role of inhibition for working memory processes: ERP evidence from a short-term storage task[J]. Psychophysiology, 2018, 55(5):e13026.
[42] BAYS P M, TAYLOR R. A neural model of retrospective attention in visual working memory[J]. Cognitive Psychology, 2018, 100:43-52.
[43] Ye Chaoxiong, Hu Zhonghua, Liang Tengfei, et al. The mechanism of the retrospective cue effect in visual working memory: Cognitive stage separation[J]. Acta Psychologica Sinica, 2020, 52(4):399-413.
[44] LIM S, WOSTMANN M, GEWEKE F, et al. The benefit of attention-to-memory depends on the interplay of memory capacity and memory load[J]. Frontiers in Psychology, 2018, 9:184.
[45] LIM S, WOSTMANN M, OBLESER J. Selective attention to auditory memory neurally enhances perceptual precision[J]. The Journal of Neuroscience: The Official Journal of the Society for Neuroscience, 2015, 35(49):16094-16104.
[46] ASTLE D E, SUMMERFIELD J, GRIFFIN I, et al. Orienting attention to locations in mental representations[J]. Attention, Perception & Psychophysics, 2012, 74(1):146-162.
[47] CURTIS, CLAYTON E, RAHMATI, et al. Attentional priority determines working memory precision[J]. Vision Research: An International Journal in Visual Science, 2014, 105:70-76.
[48] XIE W, ZHANG W. Negative emotion boosts quality of visual working memory representation[J]. Emotion (Washington, D.C.), 2016, 16(5):760-774.
[49] SPACHTHOLZ P, KUHBANDNER C, PEKRUN R. Negative affect improves the quality of memories: Trading capacity for precision in sensory and working memory[J]. J Exp Psychol Gen, 2014, 143(4):1450-1456.
[50] LONG F, YE C, LI Z, et al. Negative emotional state modulates visual working memory in the late consolidation phase[J]. Cognition & Emotion, 2020, 34(8):1646-1663.
[51] SOUZA A S, THALER T, LIESEFELD H R, et al. No evidence that self-rated negative emotion boosts visual working memory precision[J]. Journal of Experimental Psychology: Human Perception and Performance, 2021, 47(2):282-307.
[52] LIESEFELD H R, LIESEFELD A M, SAUSENG P, et al. How visual working memory handles distraction: Cognitive mechanisms and electrophysiological correlates[J]. Visual Cognition, 2020.
[53] Huang Yuesheng, Zhang Bao, Fan Xinghua, et al. Can irrelevant emotional information capture visual attention in working memory? An eye movement study[J]. Acta Psychologica Sinica, 2021, 53(1):26-37.
[54] Zhang Yu, Li Hong, Zhao Shouying, et al. The impact of task-irrelevant emotional stimuli on the updating of working memory information: Evidence from ERP[J]. Psychological Science, 2016, 39(1):2-7.
[55] Sun Bo, Zeng Xianqing, Xu Kaiyu, et al. The neural correlates of conscious processing of emotional faces and their unconscious automatic processing: Evidence from event-related potentials[J]. Acta Psychologica Sinica, 2022, 54(8):867-880.
[56] Zheng Yan, Chen Wei, Wang Hongye. ERP study on the impact of implicit reappraisal on emotional regulation[J]. Psychological Science, 2022, 45(2):268-276.
[57] He Guanghao, Liu Xinyang, Guo Lijing, et al. Is the trade-off relationship between quantity and precision in visual working memory subject to individual spontaneous control? Psychological Science Progress, 2021, 11(10):9.
[58] MA W J. Problematic usage of the Zhang and Luck mixture model[J]. BioRxiv, 2018:1.
[59] YE C, SUN H J, XU Q, et al. Working memory capacity affects trade-off between quality and quantity only when stimulus exposure duration is sufficient: Evidence for the two-phase model[J]. Scientific Reports, 2019, 9(1):8727.
[60] AWH E, BARTON B, VOGEL E K. Visual working memory represents a fixed number of items regardless of complexity[J]. Psychological Science, 2007, 18(7):622-628.
[61] SCHWEIZER S, SATPUTE A B, ATZIL S, et al. The impact of affective information on working memory: A pair of meta-analytic reviews of behavioral and neuroimaging evidence[J]. Psychological Bulletin, 2019, 145:566-609.