Biology subjects

Deouell, L. Y.

Publications and source records attributed to Deouell, L. Y..

At least 19 recordsLinked to original sources

bioRxiv2026

EEG-Based Decoding of Color and Visual Category Representations Is Reliable Within and Across Sessions

The human visual system represents stimuli in a rich and detailed manner. Traditional methods of studying visual representations in humans, such as event-related potentials (ERP), revealed numerous distinctions between the brain activity elicited by different categories of stimuli. However, these methods miss the information embedded in the spatial distributions of brain activity, or patterns, and are not always sensitive to study visual representations of different stimuli at the single participant or single trial level. Time-resolved multivariate pattern classification analysis (MVPA), or Decoding, efficiently extracts the visual representations of stimuli from the EEG topography without a-priori assumptions about the location of the effect in time and space at the single participant level. The rich information this method provides has increased its popularity dramatically in recent years. Yet, different participants show variable quality of decoding performance, and it is unclear if the accuracy of decoding is maintained within participants across multiple sessions, tasks, attentional conditions and visual features. In the current study, participants performed three visual tasks, over two sessions (1-7 days apart). We examined the correlation of decoding accuracy: within the cross-validation set, between sessions, between features (color and category) and to different measurements of the ERP signal and behavioral performance. We also examined how models generalized to different tasks and different attention conditions. We found that decoding accuracies varied substantially across participants, and that decoding accuracy was reliable within participant, over sessions, attention condition and task. This suggests the decodability behaves like an individual trait. Moreover, the spatial patterns underlying the decoding (classification weights) generalized across different tasks, attentional conditions and sessions. This suggests minimal representational drift at the resolution allowed by the EEG. We conclude that EEG decoding is a reliable method, and that visual representations are stable.

neuroscience↗

bioRxiv2025

Systematic Classification Differences Across Eye-Movement Detection Algorithms

Eye movement (EM) detection is a critical step in most eye-tracking (ET) research, typically relying on detectors - specialized algorithms designed to segment raw ET data into discrete oculomotor events. However, variability in detection algorithms and the lack of standardized evaluation frameworks hinder transparency and reproducibility across studies. In this work, we introduce pEYES, an open-source toolkit designed to streamline EM detection and enable robust, quantitative comparisons between detectors. The toolkit provides implementations for several widely used threshold-based detectors, along with multiple standardized evaluation procedures for assessing detection performance. Using pEYES, we evaluated seven detection algorithms on two publicly-available human-annotated datasets containing recordings of subjects freely viewing color images. Performance was assessed using metrics such as Cohens Kappa, Relative Timing Offset and Deviation, and a sensitivity index (') for fixation and saccade onsets and offsets. Engberts adaptive velocity-threshold algorithm consistently matched or outperformed the other detectors, occasionally achieving human-level precision. In contrast, several other detectors exhibited substantial variability in performance between datasets. We also found systematic differences in detection scores between fixation and saccade boundaries, with fixation offsets and saccade onsets detected more reliably than their counterparts. These findings highlight the importance of task- and dataset-specific detector selection in EM analysis. The pEYES toolkit is freely available, and its codebase - along with the analyses presented in this report - is accessible at https://github.com/huji-hcnl/pEYES. We invite the research community to use, extend, and contribute to its ongoing development. Through open collaboration, we aim to advance the rigor and reproducibility of EM detection practices.

neuroscience↗

bioRxiv2025

High-level Speech Processing During Mind-Wandering: Evidence from Neural Alignment with Language Models

Abundant evidence shows that when listening to speech or reading text, we continuously make predictions about upcoming words. Does this process stop when our mind wanders away and subjective experience becomes decoupled from the content of the speech? Conversely, is word prediction sufficient for conscious experience, or can these processes dissociate? Participants (N = 25) listened to audiobook content (> 12,000 words) while providing self-reports of mind-wandering. Mind-wandering was associated with spectral changes in the EEG signal, compared to attentive listening, as well as decreased early response to word onset, in line with current accounts. In contrast, EEG markers of word-level contextual surprise (indexed by the contextual surprise of large language models) remained intact during mind-wandering, alongside significant encoding of semantic content. Last, we measured the neural encoding of predictive context, measured by the correlation between EEG activity and the vectorial embeddings that language models use to predict each word given prior context. This encoding also persisted during mind-wandering, albeit weaker compared to attentive periods. Our findings show that the brain predicts and monitors relevant inputs even when the subjective experience of the external environment is dimmed during mind wandering. This suggests that the shared computational mechanisms between humans and Language Models are insufficient for having a subjective experience of the meaning of speech contents, with implications for theoretical accounts of mind-wandering and predictive processing accounts of consciousness. Significance StatementLarge Language Models (LLMs) achieve human-like conversational abilities by using long range context to predict the most likely next word. Such predictive processing is also common in the brain, and some link the best prediction to the current conscious experience. Using EEG, we show that next-word prediction markers, marked by LLM-brain alignment, persist during periods of mind-wandering while people listen to audiobooks. Specifically, we found that when people report mind-wandering, their brains continue to represent context-based predictions, word-level surprisal, and word-level semantics. These findings dissociate high-level predictions from subjective experience.

neuroscience↗

bioRxiv2025

Computational Mechanisms of Temporal Anticipation in Perception and Action

To interact effectively with our surroundings, we rely on strategies to reduce uncertainty. One important source of information is temporal regularities, which enable us to form predictions about when events will occur, and through this, prepare for them in advance. Such preparation was shown to facilitate motor planning, yet the impact of temporal anticipation on perceptual acuity is unknown, and the cognitive computations underlying this process remain debated. To answer these questions, we administered a new difficult change discrimination task with variable anticipatory periods (N=142). We show that both perceptual sensitivity and motor responses are influenced by temporal structure. Using computational modelling, we show that the cognitive operation underlying this behavior is based on a logarithmic transformation of the event hazard rate (HR) and reveal a crucial role of temporal-estimation noise in shaping this computation, both when temporal information is encoded, and when it is decoded. Together, our results highlight the fundamental contribution of anticipation in directing behavior and advance our understanding of temporal processing in the brain.

neuroscience↗

bioRxiv2024

Beyond stimulus onset: Ongoing fixations within an object do not re-evoke category representations during free-viewing.

Human visual perception in natural conditions involves multiple fixations within single objects. While traditional studies focus on transient neural responses to initial stimuli, this study investigates how object-category representations evolve across sequential fixations on an object. Using electroencephalography (EEG) and eye-tracking, we analyzed fixation-related potentials (FRPs) and applied multivariate pattern analysis (MVPA) to decode neural representations of faces and watches during prolonged viewing. Results revealed robust category-selective responses, including the N170 component, at stimulus onset, with sustained representations persisting throughout object presentation. Temporal signal deconvolution showed that subsequent fixations did not re-evoke the N170 but elicited transient occipital responses, likely reflecting low-level differences. These findings underscore the dynamic interplay between transient and sustained neural processes during naturalistic vision and highlight the importance of disentangling overlapping neural signals during free viewing.

neuroscience↗

bioRxiv2024

Maintenance of Bound or Independent Features in Visual Working Memory is Taskdependent

Conflicting findings exist regarding whether features of an object are stored separately or bound together in visual working memory. This controversy is based on an implicit assumption about a default, or fixed, mode of working memory storage. In contrast, here we asked whether the anticipated task might determine the format in which information is maintained in working memory, consistent with its task-oriented function. To test this flexible maintenance hypothesis, we recorded EEG while subjects performed a delayed (Yes/No) recognition task with different requirements and loads. Across three experiments, we compared event-related potentials (ERPs) in conditions with and without the necessity of maintaining conjunctions between features, while controlling for differences in visual stimulation. In Experiment 1 (N=24), involving color-location conjunctions, we identified a delay-period effect characterized by a positive potential shift in central-parietal channels when conjunction was not required by the task. This pattern, distinct from the effect caused by an increased working memory load, was confirmed in Experiment 2 (N=23) with an independent group of subjects using a similar paradigm, while also controlling for the physical appearance of probes. Finally, the observation of color and location conjunction in Experiments 1 and 2 was extended to Color and Orientation conjunction in Experiment 3 (N=22). Collectively, these three experiments provided reliable evidence demonstrating that the maintenance of feature conjunctions in working memory, whether spatial (location) or non-spatial (non-location), depends on the task goal. HighlightsO_LITask demands determine whether objects are represented as separate features or as conjoined items (maintenance format) in visual working memory. C_LIO_LIEEG recordings show differential activity despite similar stimulation, reflecting task demands. C_LIO_LIThe effect of the task is evident in the early stages of working memory processing. C_LIO_LIFlexibility in maintenance format was observed for both color-location and color-orientation conjunctions. C_LIO_LIThe findings challenge the notion of fixed maintenance format during visual working memory. C_LI

neuroscience↗

bioRxiv2023

Task-Dependent Representation of Features within the Same Dimension in Visual Working Memory

Working memory (WM) serves as a neurocognitive system responsible for temporarily storing and manipulating information when its source has disappeared. Previous investigations into whether features within the same dimension were stored separately or conjoined into objects have yielded conflicting findings. Based on increasing evidence on the adaptivity of the working memory system, we conjectured that the format in which objects from the same dimension are stored in Visual Working Memory (VWM) might be contingent on the specific task demands, and subtle distinctions across experiments may account for the disparities in results. In the current study, we introduced modifications to the paradigm used by Luck and Vogel (1997) and Wheeler and Treisman (2002), where the same paradigm led to different conclusions, to assess whether implicit task requirements could influence the storage format of objects in visual working memory. In two conditions, we manipulated the relevance of conjunction between two colors by varying the proportions of Mis-conjunct probes, a probe type depending on conjunction information for accurate responses. The results showed that in both conditions, performance was primarily determined by the number of features rather than the number of objects, aligning with the results of Wheeler and Treisman. Nevertheless, we observed that Mis-conjunct probes, which require conjunction information, exhibited improved performance when tested more frequently. This suggests that the format of retention in working memory, whether in separate or conjoined form, is influenced by the task demands.

animal behavior and cognition↗

bioRxiv2023

Binding in visual working memory is task-dependent

Working memory is a neurocognitive system for maintaining and manipulating information online for a short period after the source of information disappears. The information held in working memory has been shown to flexibly match current functional goals. Considering this, we revisited the question of whether information is held in WM as separate features or as bound objects. We conjectured that rather than having a fixed answer, the format in which information is maintained in working memory is also task dependent. In two separate experiments, we investigated the binding between features when the location was not (Experiment 1, color and orientation binding) or was (Experiment 2, color and location binding) a task-relevant feature in a delayed (yes/no) recognition task by manipulating the relative relevance of conjunctions and separate features. Each experiment included two conditions, Binding dominant (BD), which emphasizes the retention of binding between features, and Feature dominant (FD), which emphasizes the retention of individual features. In both experiments, we found that memory for conjunctions was better in the BD condition and that memory for separate features was better in the FD condition. These patterns suggested that the formats of objects in VWM could be shaped by the tasks they serve. Additionally, we found that the memory of location was impaired when conjunction between location and color was task-irrelevant, while memory of color and orientation was relatively independent of each other. We conclude that the representation format of objects in working memory is influenced by task requirements.

neuroscience↗

bioRxiv2023

Decoding object categories from EEG during free viewing reveals early information evolution compared to passive viewing

Object processing is fundamental to visual perception, and understanding its neural substrates informs many cognitive and computational visual processing models. Thus far, most human studies have used passive viewing paradigms, during which self-driven behavior, such as eye movements, is constrained, and brain activity is evoked by abrupt stimuli onsets. This artificial dissociation of perception and action ignores the natural dynamics of visual processing. Thus, conclusions based on such passive viewing paradigms may not apply to active vision. Here, we study the human neural correlates of category representations during active visual processing by time-locking EEG to self-driven fixations during visual search for natural objects. We combine the deconvolution of overlapping responses to consecutive fixations with multivariate pattern analysis (MVPA) to decode object categories from responses to single fixation. We bridge the active and passive viewing literature by comparing the temporal dynamics of multivariate object representations during free visual search (active viewing) and rapid serial visual presentation (passive viewing), leveraging the high temporal resolution of EEG. We found that categorical information, at different levels of abstraction, can be decoded from single fixations during natural visual processing, and cross-condition decoding revealed that object representations are similar between active and passive viewing conditions. However, representational dynamics emerge significantly earlier in active compared to passive conditions, likely due to the availability of predictive information in free viewing. We highlight methodological considerations for combining MVPA with deconvolution methods. Significance StatementUnderstanding the neural correlates of visual perception is crucial for advancing cognitive and computational models of human vision. This study bridges the gap between passive- and active-vision literature while shedding light on the intricate relationship between perception and action in visual processing. Although eye movements are a fundamental behavior through which visual information is naturally sampled, most neuroimaging studies probe the brain by presenting stimuli abruptly at the center of the screen while participants refrain from moving their eyes. We investigated EEG correlates of visual processing during active visual search and demonstrated that object categories of naturally fixated objects can be decoded from the EEG. We provide novel findings regarding the dynamics of active, compared to passive, visual processing, while contributing to the advancement of EEG analysis methodology.

neuroscience↗

bioRxiv2022

Representation of sustained visual experience by time-invariant distributed neural patterns

Instances of sustained stationary sensory input are ubiquitous. However, previous work focused almost exclusively on the transient onset responses. This presents a critical challenge for neural theories of consciousness, which should account for the full temporal extent of experience. Here we use intracranial recordings from human patients viewing diverse images in multiple durations. In sensory regions, we reveal that despite dramatic changes in activation magnitude, the distributed representation of categories and exemplars remains sustained and stable. In contrast, in frontoparietal regions we find transient content representation at stimulus onset. Our results highlight the connection between the anatomical and temporal correlates of experience. To the extent perception is sustained, it may rely on sensory representations, and to the extent it is discrete, centered on perceptual updating, it may rely on frontoparietal representations.

neuroscience↗

bioRxiv2021

Maintenance of Bound or Independent Features in Visual Working Memory is Task-dependent

Over the last decade, seemingly conflicting results were obtained regarding the question of whether features of an object are stored separately, or bound together, in visual working memory. Many of these studies are based on an implicit assumption about a default, or fixed, mode of working memory storage. In contrast, here we asked whether the anticipated memory probes used in a given experiment might determine the format in which information is maintained in working memory, consistent with its task-oriented function. In order to test this flexible maintenance hypothesis, we recorded EEG while subjects performed a delayed-match-to-sample task with different requirements and load. In Experiment 1, by contrasting conditions with and without the requirement of maintaining bound features, we found significant differences in EEG signals recorded in central-parietal channels while controlling for differences in visual stimulation. Such difference was confirmed by multivariate pattern analysis (MVPA). Moreover, behavior results, topographical distributions and cross-contrast MVPA decoding jointly suggested that this effect is not the same as task load effect. These results were confirmed in Experiment 2 with an independent group of subjects with the same paradigm, providing reliable evidence that the format of object maintenance in visual working memory could be flexibly shaped by the current task.

neuroscience↗

bioRxiv2020

Context Sensitivity Across Multiple Time and Frequency Scales

Everyday auditory streams are complex, including spectro-temporal content that varies at multiple time scales. Using EEG, we investigate the sensitivity of human auditory cortex to the content of past stimulation in unattended sequences of equiprobable tones. In 3 experiments including 82 participants overall, we found that neural responses measured at different latencies after stimulus onset were sensitive to frequency intervals computed over distinct time scales. Importantly, early responses were sensitive to a longer history of stimulation than later responses. To account for these results, we tested a model consisting of neural populations with frequency-specific but broad tuning that undergo adaptation with exponential recovery. We found that the coexistence of neural populations with distinct recovery rates can explain our results. Furthermore, the adaptation bandwidth of these populations depended on spectral context - it was wider when the stimulation sequence had a wider frequency range. Our results provide electrophysiological evidence as well as a possible mechanistic explanation for dynamic and multi-scale context-dependent auditory processing in the human cortex. SIGNIFICANCE STATEMENTIt has become clear that brain processing of sensory stimuli depends on their temporal context, but context can be construed at time scales from the recent millisecond to life-long. How do different contextual time scales affect sensory processing? We show that auditory context is integrated across at least two separate time scales, and that at both of these time scales responses dynamically adapt to a varying frequency stimulation range. Using computational modeling, we develop a rigorous methodology to estimate the time and frequency scales of context integration for separate response components. Our robust results replicated across 3 EEG experiments, and contribute to the understanding of neural mechanisms supporting complex and dynamic context integration.

neuroscience↗

bioRxiv2019

Face Selective Neural Activity: comparison between fixed and free viewing

Event Related Potentials (ERPs) are widely used to study category-selective EEG responses to visual stimuli, such as the face-selective N170 component. Typically, this is done by flashing stimuli abruptly at the point of static gaze fixation. While allowing for good experimental control, these paradigms ignore the dynamic role of eye-movements in natural vision. Fixation-related potentials (FRPs) obtained using simultaneous EEG and eye-tracking, overcome this limitation. Various studies have used FRPs to study processes such as lexical processing, target detection and attention allocation. The goal of this study was to compare face-sensitive activity evoked by stimulus abrupt appearance with that evoked by self-controlled gaze fixations on a stimulus. Twelve participants were studied in three experimental conditions: Free-viewing (FRPs), Cued-viewing (FRPs) and Control (ERPs). We used a multiple regression approach to disentangle overlapping activity components. Our results show that the N170 face effect (face vs. non-face) is evident for the first fixation on a stimulus, whether it follows a self-generated saccade or stimulus appearance at fixation point. Furthermore, the N170 exhibits category-specific adaptation in free viewing. The N170 face effect had similar topography across viewing conditions, but there were major differences within each stimulus category. We ascribe these differences to an overlap of the fixation-related Lambda response and the N170. We tested the plausibility of this account using dipole simulations. This study establishes the use of the N170 face effect as a signature of face detection in free-viewing experiments while highlighting the importance of accounting for eye-movement related effects.

neuroscience↗

bioRxiv2018

Widespread suppression of high-order visual cortex during blinks and external predictable visual interruptions

Spontaneous eye blinks generate frequent potent interruptions to the retinal input and yet go unnoticed. As such, they provide an attractive approach to the study of the neural correlates of visual awareness. Here, we tested the potential role of predictability in generating blink-related effects using fMRI. While participants attentively watched still images of faces and houses, we monitored naturally occurring spontaneous blinks and introduced three kinds of matched visual interruptions: cued voluntary blinks, self-initiated (and hence, predictable) external darkenings, and physically similar but unpredictable external darkenings. These events impact was inspected using fMRI across the visual hierarchy. In early visual cortex, both spontaneous and voluntary blinks, as well as predictable and unpredictable external darkenings, led to largely similar positive responses in peripheral representations. In mid- and high-level visual cortex, all predictable conditions (spontaneous blinks, voluntary blinks, and self-initiated external darkenings) were associated with signal decreases. In contrast, unpredictable darkenings were associated with signal increases. These findings suggest that general-purpose prediction-related mechanisms are involved in producing a small but widespread suppression of mid- and high-order visual regions during blinks. Such suppression may down-regulate responses to predictable transients in the human visual hierarchy.

neuroscience↗

bioRxiv2018

Human posterior parietal cortex responds to visual stimuli as early as peristriate occipital cortex

Much of what is known about the timing of visual processing in the brain is inferred from intracranial studies in monkeys, with human data limited to mainly non-invasive methods with lower spatial resolution. Here, we estimated visual onset latencies from electrocorticographic (ECoG) recordings in a patient who was implanted with 112 sub-dural electrodes, distributed across the posterior cortex of the right hemisphere, for pre-surgical evaluation of intractable epilepsy. Functional MRI prior to surgery was used to determine boundaries of visual areas. The patient was presented with images of objects from several categories. Event Related Potentials (ERPs) were calculated across all categories excluding targets, and statistically reliable onset latencies were determined using a bootstrapping procedure over the single trial baseline activity in individual electrodes. The distribution of onset latencies broadly reflected the known hierarchy of visual areas, with the earliest cortical responses in primary visual cortex, and higher areas showing later responses. A clear exception to this pattern was robust, statistically reliable and spatially localized, very early responses on the bank of the posterior intra-parietal sulcus (IPS). The response in the IPS started nearly simultaneously with responses detected in peristriate visual areas, around 60 milliseconds post-stimulus onset. Our results support the notion of early visual processing in the posterior parietal lobe, not respecting traditional hierarchies, and give direct evidence for the upper limit of onset times of visual responses across the human cortex.

neuroscience↗

bioRxiv2018

Direct evidence for prediction signals in frontal cortex independent of prediction error

Predictive coding (PC) has been suggested as one of the main mechanisms used by brains to interact with complex environments. PC theories posit top-down prediction signals, which are compared with actual outcomes, yielding in turn prediction-error signals, which are used, bottom-up, to modify the ensuing predictions. However, disentangling prediction from prediction-error signals has been challenging. Critically, while many studies found indirect evidence for predictive coding in the form of prediction-error signals, direct evidence for the prediction signal is mostly lacking. Here we provide clear evidence, obtained from intracranial cortical recordings in human surgical patients, that the human lateral prefrontal cortex generates prediction signals while anticipating an event. Patients listened to task-irrelevant sequences of repetitive tones including infrequent predictable or unpredictable pitch deviants. The amplitude of high frequency broadband (HFB) neural activity was decreased prior to the onset of expected relative to unexpected deviants in the frontal cortex only, and its amplitude was sensitive to the increasing likelihood of deviants following longer trains of standards in the unpredictable condition. Single trial HFB amplitudes predicted deviations and correlated with post-stimulus response to deviations. These results provide direct evidence for frontal cortex prediction signals independent of prediction-error signals.

neuroscience↗

bioRxiv2018

Sustained High- and Low-Frequency Neural Responses to Visual Stimuli in Scalp EEG

What are the neurophysiological correlates of sustained visual processing in the scalp EEG signal? In a previous study using intracranial recordings in humans, we found that presentation of visual stimuli for prolonged durations (up to 1.5 seconds) was associated with two kinds of sustained neural activity patterns: a high-frequency broadband (>30 Hz) response that tracked the duration of the stimulus with high precision in early visual cortex (EVC), and with lesser temporal precision in downstream category-selective areas; and a sustained low-frequency potential shift appearing in a small subset of EVC sites. Using a similar approach of presenting images for variable durations to identify sustained activity, we provide the first comprehensive characterization of the manifestation of sustained visual responses as recorded with EEG. In a series of four experiments, we found that both high- and low-frequency sustained responses can be detected on the scalp. The high frequency activity could be detected with high signal to noise ratio only in a subset of individual subjects, in whom it was unequivocal and highly localized. The low frequency sustained response was sensitive to the size and position of the stimulus in the visual field. Both response types showed strong lateralization for stimuli on the left vs. right visual field, suggesting a retinotopic visual cortical source. However, different scalp topographies and different modulation by stimulus properties suggest that the two types of sustained responses are likely driven by distinct sources, and reflect different aspects of sustained processing in the visual cortex.

neuroscience↗

bioRxiv2017

Increasing Suppression Of Saccade-Related Transients Along The Human Visual Hierarchy

A key hallmark of visual perceptual awareness is robustness to instabilities arising from unnoticeable eye and eyelid movements. In previous human intracranial (iEEG) work (Golan et al., 2016) we found that excitatory broadband high-frequency activity transients, driven by eye blinks, are suppressed in higher-level but not early visual cortex. Here, we utilized the broad anatomical coverage of iEEG recordings in 12 eye-tracked neurosurgical patients to test whether a similar stabilizing mechanism operates following small saccades. We compared saccades (1.3{degrees}-3.7{degrees}) initiated during inspection of large individual visual objects with similarly-sized external stimulus displacements. Early visual cortex sites responded with positive transients to both conditions. In contrast, in both dorsal and ventral higher-level sites the response to saccades (but not to external displacements) was suppressed. These findings indicate that early visual cortex is highly unstable compared to higher-level visual regions which apparently constitute the main target of stabilizing extra-retinal oculomotor influences.

neuroscience↗