Conference Agenda

Psychology of human learning distinguishes implicit learning, resulting in nonconscious knowledge, and explicit learning, resulting in conscious knowledge. Following Reber’s seminal work, a whole field has focused on sequence and grammar learning. Although behavioral sequences are widespread in animals, few studies have examined the implicit or explicit character of sequence learning . We address this gap by adapting standard paradigms from the human literature to other species. This experiment tested brown-tufted capuchin monkeys (Sapajus sp., N=7) in a sequence completion task. The monkeys first learned to follow a target moving on a touchscreen from left to right. The delay between targets' appearance was progressively increased, eliciting anticipatory responses. Correct anticipatory responses triggered the target's display and food reward delivery. With this training, three capuchins performed the sequence completion task with the left-right rule, and one generalized it to an up-down rule. These results show that a sequence completion task can be learned without verbal instructions, opening new perspectives to the study of implicit/explicit learning in non-verbal or pre-verbal participants. We discuss a range of follow-up experimental designs to assess distinctive properties of conscious knowledge within this task.

A key debate in consciousness and affective sciences concerns whether emotions are necessarily conscious, as argued by higher-order (HO) theories (1). In this paper, I highlight how the kind of phenomenon known as hedonic reversal could inform this debate. Hedonic reversals refer to taking pleasure from experiences that are usually aversive, such as the fear from watching horror movies or the pain from consuming chili peppers. A widely accepted account (2) suggests that it is enabled by a “mind over body” process of enjoying bodily arousal elicited by defensive reactions while one knows there is no actual danger. I point out that this explanation supports HO theories, in the sense that an unconscious, conceptual state is required for engendering defensive reactions but is yet to constitute valence. In particular, hedonic reversal redeems HO theories from defective examples used in the literature (3) by showcasing cases where valence necessarily depends on a third-order state evaluating contextual information using self-schema. This is because valenced processes occurring prior to self-referential appraisal would contradict with danger perception and thereby blocking the interoceptive aspect of the experience. I further discuss how this interpretation is supported by non-fixed views of valence (4) as well as agentive accounts of fear (5). Together, I show that hedonic reversal offers a powerful case against unconscious valence (and thus emotion).

References:

1. LeDoux & Brown (2017). PNAS.

2. Rozin (1999). Well-being, pp. 109-133.

3. Ludwig (2025). Conscious. Cogn.

4. Prinz (2010). Emot. Rev.

5. de Vignemont (2024). Mind Lang.

Counterfactual actions are alternatives that could have been chosen and executed but were not. Investigating their neural bases is challenging due to the difficulty of measuring events that do not occur. To address this, we conducted a meta-analysis of 35 studies examining counterfactuality, aiming to disentangle overlapping and distinct neural mechanisms underlying counterfactual perception and action.

Building on these insights, we developed a novel experimental task. Previous research in our lab revealed a memory bias: participants performing the Tower of London task erroneously claimed to have seen or executed plausible counterfactual configurations more often than entirely new ones. This suggests that planning an action is sufficient to encode it in memory, while execution is not necessary. However, that study could not distinguish between the prospective (planning-related) and retrospective (plausibility-based) contributions of counterfactual actions to these false memories.

To clarify these influences, our current study employs a novel maze task designed to separate prospective and retrospective contributions to counterfactual memory biases. By comparing behavioural responses, we shed light on the cognitive mechanisms driving false memories for actions that were planned but not executed.

This approach provides a novel framework for understanding the neural and cognitive bases of counterfactuality, with implications for decision-making, memory, and goal-directed behavior.

This OSF preregistered study examines how the brain represents the extrinsic and intrinsic value of emotional states and how this influences emotion regulation. Thirty-two participants viewed emotionally evocative video clips while undergoing functional magnetic resonance imaging. Using a within-subjects design, participants were instructed to either allow their emotional response to unfold naturally ("Allow") or to suppress their response ("Suppress"). In each trial, participants could earn either £10 ("High Extrinsic Value") or £1 ("Low Extrinsic Value") for successfully following the instructions. Outside of the scanner, participants rated how pleasant/unpleasant, arousing and effortful it was to watch the same videos when suppressing and allowing their emotional response. We calculated the influence of these ratings on participants decisions to suppress or allow their emotional responses to video stimuli in a separate task, and used these parameters to predict the intrinsic value of the instructed emotional states for each trial of the imaging task.

We present preliminary findings on how the extrinsic and intrinsic value of emotion states influences activity in known value-sensitive regions (ventral striatum, ventral medial prefrontal cortex & orbitofrontal cortex), and how this modulates functional connectivity between regions associated with engaging in emotion regulation (dorsolateral and dorsomedial prefrontal cortex) and those associated with generating emotional responses (insula, amygdala). By linking emotion regulation to value representation, this research offers novel insights into the neural underpinnings of the evaluation of emotional states and affective decision-making.

Peace of mind, as an internal state of peacefulness and harmony, serves as an index of affective well-being. Recent evidence suggests that peace of mind contributes to subjective well-being across cultures. However, the mechanisms underlying this relationship remain underexplored, particularly in university student populations.

Self-determination theory posits that human beings must satisfy their basic psychological needs—autonomy, relatedness, and competence—to function optimally. Research indicates that satisfying these needs enhances positive and reduces negative affect. However, limited studies have examined whether peace of mind mediates the relationship between psychological needs satisfaction and emotions in university students.

This study investigated the relationships among peace of mind, psychological needs satisfaction, and students’ positive and negative emotions. Data were collected from 350 university students at the University of Crete using the Peace of Mind Scale, the Positive and Negative Affect Scale, and the Basic Psychological Needs Satisfaction Scale. Correlation analyses examined associations, while path analysis tested the mediating role of peace of mind in the relationship between psychological needs satisfaction and emotions.

Findings showed a positive association between peace of mind and positive emotions, with competence and autonomy also correlating positively with positive emotions. Conversely, peace of mind negatively correlated with negative emotions. Path analysis confirmed that peace of mind mediated the effects of psychological needs satisfaction on emotions.

Findings highlight the importance of fostering peace of mind to enhance students’ emotional well-being. Universities should implement interventions promoting peace of mind, and future research should explore its role as a protective factor against academic stress.

Emotion recognition plays a vital role in developmental processes, influencing behavior, social interactions, and learning across diverse settings and age groups. Despite its importance, few studies have focused on developing picture databases suitable for investigating emotion recognition across a wide age range, including children as young as five years old. This study addresses this gap by creating a database of emotional pictures tailored for such research. To ensure the appropriateness of the pictures for younger audiences, experts first screened potential pictures. Subsequently, we established normative data by presenting pictures to a large group of adults, who rated each picture on emotional category, confidence, and intensity. The result is a validated set of 566 emotional pictures, providing a unique resource for studying emotion recognition and processing across ages. To evaluate the reliability of this database, we conducted a series of tests. In one analysis, the pictures were divided into 14 subsets. Then we calculated measures of dominance for the participants, for each subset. Correlations between subsets were exceptionally high (r > .9), demonstrating robust consistency in categorization and scoring across the database. This study not only provides a valuable tool for cross-age emotion research but also establishes a reliable methodological framework. These findings invite researchers to explore emotion using a resource that bridges developmental and experimental boundaries.

Theory suggests loneliness is dynamic, manifesting as a stable trait, but also fluctuating over time in everyday life. These trait and state forms of loneliness are argued to have different consequences: trait loneliness is associated with social withdrawal, while state loneliness is associated with social affiliation. Despite its centrality to theory, the time-course of loneliness in daily life is underexplored. We investigated trait and state loneliness and tested their association with a well-established mechanism of social connection: interpersonal emotion regulation. Accordingly, we aim to understand how different forms of loneliness affect seeking and offering support in social interactions.

We conducted an experience sampling study where participants (N=239) completed a baseline survey—assessing trait loneliness—and seven daily surveys over seven days—assessing state loneliness. In daily surveys, participants also reported whether, to what degree, and how (i.e., through which strategies) they tried to change others’ emotions or their own emotions by turning to others.

State loneliness was associated with greater effort invested towards interpersonal emotion regulation. Trait and state loneliness were linked to invalidating other’s feelings or feeling invalidated by others. Moreover, state loneliness was associated with offering less cognitive support to others and receiving cognitive support from others.

Our findings are in line with the social seeking motivation account of state loneliness. Nevertheless, the picture appears more complex than anticipated, in that loneliness predicted engagement in both adaptative and maladaptive social strategies. The experience of loneliness is associated with detrimental social experiences at both trait and state levels.

People often engage in episodic counterfactual thinking, or mentally imagining alternative ways in which past events might have unfolded. For instance, after failing an exam, I might think about how things could have been different if I had studied more. One feature by which these simulations can vary is their plausibility: while some simulations closely mirror the past, others may deviate significantly from what I know could have happened. Extant research has shown that plausibility judgments influences mood, prosocial behavior, and false memories. However, knowledge about the factors that influence this phenomenological feature remains limited. To address this question, we tested three models for what might shape the plausibility of episodic counterfactuals: (1) Whether plausibility is an epiphenomenal feature of how difficult a simulation is to generate; (2) Whether plausibility is derived from the vividness of the simulation or (3) A sampling mechanism that generates more plausible alternatives for the past. Across two experiments, we demonstrated that: (1) Plausibility appears to be an independent property from vividness and difficulty; and (2) When people generate episodic counterfactual thoughts, they tend to produce the most vivid and plausible mental simulations first. Importantly, changes in plausibility and vividness were not accompanied by concurrent changes in difficulty, indicating that this sampling process is not a byproduct of a heuristic that bases metacognitive appraisals on the ease of simulation. Our results support a sampling process that prioritizes the generation of more plausible and vivid counterfactual alternatives.

An “Aha! moment” corresponds to the moment the solution to a problem suddenly pops into consciousness. The phenomenological experience of the “Aha! moment” is characterized by a feeling of suddenness and pleasure and by an increased confidence in the solution. These “Aha!”, also called insights, are defined as the result of restructuration – a representational change of information. With a new paradigm, we asked participants to find 5-letter solution words in six guesses. After each guess, participants were asked to rate their feeling of insight before receiving feedback on which letters of their guess were in the solution word. This feedback provided them with the necessary information to find the solution. To illustrate, the feedback could indicate that the solution contains the letters “I”, “T” and “H” with the position of two letters: _ I _ T_. For some solutions, a representational change of the given feedback was needed to find the word (e.g. separating “T” and “H” in “HINTS”). For other words, there was no need to alter the information representation (e.g. “BIRTH”). We expected the words that needed restructuration to result in more “Aha! moments” but, surprisingly, preliminary data showed the opposite. Participants had fewer “Aha! moments” in the condition for which a change of representation was needed compared to the condition where it was not needed. In contrast to what the definition of an insight suggests, our results indicate that restructuration may not be a prerequisite for the “Aha!” phenomenological experience.

Artificial Intelligence (AI) literacy is a crucial competency in contemporary education, integrating affective, behavioral, cognitive, and ethical dimensions. Understanding AI goes beyond technical proficiency, encompassing motivation, ethical awareness, and responsible use. The present study aims to examine the psychometric properties of the Greek adaptation of the AI Literacy Questionnaire (AILQ), developed by Ng et al. (2024), which assesses AI literacy through a structured framework known as the Affective, Behavioral, Cognitive, and Ethical (ABCE) approach.

The AILQ serves multiple purposes: (1) empowering students by enhancing their AI-related self-efficacy and knowledge, (2) providing educators with a reliable instrument to assess AI literacy levels, and (3) enabling researchers to investigate AI literacy development in diverse educational settings. The adaptation of the questionnaire follows rigorous psychometric validation procedures to ensure conceptual validity and measurement reliability.

The study is conducted with a sample of university students from higher education institutions. Data collection takes place through an online survey using Google Forms, ensuring voluntary and anonymous participation. Exploratory Factor Analysis (EFA) and Confirmatory Factor Analysis (CFA) will be employed to examine the underlying structure and verify construct validity. Reliability analysis will be conducted to determine the internal consistency of the scale.

Findings are expected to contribute to the validation of the Greek AILQ as a reliable tool for assessing AI literacy. Results will offer valuable insights for educational policy, curriculum development, and AI pedagogy. This research highlights the importance of a holistic AI literacy framework that fosters both technical competence and ethical responsibility in AI use.

External memory, historically conceptualized within the extended mind framework (Clark & Chalmers, 1998), has long been controversial among theorists. Early critiques highlighted their seemingly passive character. Theorists like Michaelian (2012) identified a disconnect between the static nature of external memory and the constructive processes of internal memory. These criticisms led theorists to retreat toward viewing such systems as pragmatic cognitive scaffolds (Heersmink, 2020) rather than genuine memory extensions. However, we propose a fundamental reconceptualization of external memory, shifting focus from system functionality to content dynamics. Drawing on Chalmers's (2025) notion of propositional interpretability, we demonstrate how AI-driven external memory systems now generate personalized propositional attitudes from user data through three key mechanisms: dynamic retrieval-based contextualization, adaptive content reconstruction, and preference-integrated updating. These systems transcend traditional paradigms of passive storage and retrieval, exhibiting properties previously thought exclusive to internal cognition.

To demonstrate this empirically, we collected an unprecedented multimodal dataset including 5-year continuous (24/7) audio recordings and comprehensive life-logging data. This dataset enables exploration of how AI systems can leverage personal data to reconstruct human mental content, potentially transforming our understanding of external memory in the age of AI.

Key References:

Chalmers, D. J. (2025). Propositional Interpretability in Artificial Intelligence. arXiv:2501.15740

Clark, A., & Chalmers, D. J. (1998). The Extended Mind.

Heersmink, R., & Carter, J. A. (2020). The Philosophy of Memory Technologies: Metaphysics, Knowledge, and Values. Memory Studies, 13(4)

Michaelian, K. (2012). Is External Memory Memory? Biological Memory and Extended Mind. Consciousness and Cognition, 21(3)

In visual working memory (VWM), bottom-up attention prioritizes recalling perceptually similar features, a relational property of stimuli over dissimilar ones. Here, we have conducted a behaviour cum EEG-based VWM experiment on human participants (n = 25) to capture the bias in recalling repeated items vs. not-repeated items by evaluating response time and accuracy in responding to relevant probes. We then investigated their neural correlates using EEG to test our hypotheses that band-specific spectral power differences across two conditions reveal the neural changes that can be associated with underlying causes for task-specific behavioural differences and required attentional facilitation. Our behavioural findings demonstrated faster and more accurate VWM recalling of repeated items than non-repeated items. Early desynchronization of the beta band (13-30 Hz) was found in the C3 electrode at around 200 ms shortly after probe presentation for repeated items which acts as an index for the prioritization of repeated items in VWM and is associated with faster motor preparation for response selection supporting our behavioural results. Comparatively high theta power (4-7Hz) in front-medial sensors for responding to the not-repeated items implies more cognitive control for recalling task-relevant information and potentially suppressing responses to task-irrelevant information. We also observed decreased right posterior parietal alpha power (8-13 Hz) for recalling repeated items over not-repeated items acting as a marker for their attentional facilitation. This provides substantial evidence for the stimulus-driven attentional facilitation and prioritization of probe matching for repeated items over not-repeated items.

Working memory (WM) is traditionally linked to persistent neuronal firing, but recent evidence challenged this, suggesting the presence of silent mechanisms such as short-term synaptic plasticity (STSP). Recent neurocomputational models tested STSP as a possible basis for unconscious WM, but their results depend on architectural decisions. One key factor is network connectivity, particularly weight rectification, which may prune connections and influence how STSP supports memory. Here, we investigate how weight rectification affects STSP-driven WM representations in one of the most cited neurocomputational simulation studies on STSP.

We used a biologically constrained recurrent neural network incorporating STSP (Masse et al., 2019). To examine the impact of network connectivity, we tested different weight rectification strategies and analyzed their effects on WM task performance.

Our results show that input and hidden layer connectivity interact with STSP, affecting performance, while output layer connections play a minor role. Input weights modulate information flow into the memory layer, influencing how STSP encodes memory.

In conclusion, network connectivity significantly impacts the role of STSP in WM, particularly in how input weights filter information before reaching the recurrent memory layer. These findings highlight the need for biologically grounded connectivity choices in neurocomputational models.

Stimuli that consistently predict rewards can capture attention, leading to reward-related attentional biases. These biases are often considered independent of task goals or physical salience, driven by Pavlovian learning processes. Recent evidence suggests that may be a necessary condition for such biases, although the mechanisms underlying this relationship remain unclear. One potential mechanism is that explicit awareness biases participants’ selective attention toward the relevant features of the stimuli. This preregistered study tested whether reward-related attentional biases arise primarily as a consequence of selectively attending to reward-predictive features, independent of explicit awareness. Participants performed a visual search task in which one of two potential distractors—one predicting high reward and the other low reward—were presented on some trials. We manipulated selective attention to the reward-predictive feature (distractor color) by assigning participants to one of two groups. In some trials, one group of participants was instructed to report the color of the distractor from the preceding trial, while in the other group, participants were instructed to report an irrelevant distractor feature (its location). As predicted, reward-related attentional biases, as indicated by slower response times for the high-value distractor compared to the low-value distractor, were observed only in the group tasked with reporting the distractor color. Notably, explicit knowledge of the color-reward contingencies did not differ significantly between the two groups. When controlling for individual differences in awareness, both selective attention manipulation and explicit awareness independently influenced reward-related attentional biases.

Understanding the neural mechanisms of conscious experience requires distinguishing brain processes generating awareness from those supporting or following it. A key challenge is identifying electrophysiological markers that track perceptual awareness rather than general cognitive or task-related processes. This study examined visual and auditory awareness, focusing on the effects of reporting and stimulus processing levels on early and late ERP components.

During the EEG recordings, participants completed three experimental blocks: (1) a no-report block with stimuli presented at varying levels of visibility/audibility, (2) a report block with an identification task (with the same stimuli as in block 1), and (3) a block with low-level and high-level objective tasks (designed following the Levels of Processing hypothesis), and subjective Perceptual Awareness Scale (PAS).

Analyzes comparing blocks 1 and 2 showed that early VAN, AAN, and later LP components were present in both blocks, but their amplitudes were more pronounced in the report than in the no-report task. This suggests that the necessity to produce a report affects ERP markers of aware perception. The analysis results for block 3 revealed that both early and late components exhibit gradual changes in amplitude—i.e., the higher the PAS rating, the more pronounced the amplitude of a given ERP. However, VAN and AAN components were differentiated by levels of processing to a lesser extent than the late LP markers. This may indicate that early components correlate with non-specific processes related to the early stages of perception, not specific features of conscious content.

Introduction:

Synaesthesia provides a framework to investigate visual perception and memory in the general population. Synaesthesia has been associated with perceptual and memory advantages, but the mechanisms underlying these enhancements remain unclear.

Methods:

We compared the cognitive profiles of synaesthetes, non-synaesthetic relatives of synaesthetes, and non-synaesthetic non-relative controls. Participants completed visual perception (VP), short-term memory (STM) and long-term memory tasks (LTM-immediate and LTM-delayed), involving colour and location manipulations. Mental imagery, cognitive style, and motivation were assessed using questionnaires.

Results:

Data collection is ongoing (Synaesthetes n = 52, Relatives n = 15, Controls n = 42). Preliminary evidence suggests that synaesthetes outperformed controls on most tasks, with no significant differences from relatives. In the VP task, synaesthetes and relatives demonstrated significantly better colour and location perception compared to controls. Synaesthetes more accurately recalled colours than controls in the STM task at a low cognitive load (load = 1). In the LTM tasks, synaesthetes showed significantly better immediate and delayed colour recall, and better location recall in the delayed task. Synaesthesia subtypes were not significantly related to VP or STM performance. Grapheme-colour synaesthetes showed enhanced colour recall in LTM-immediate task, with combined subtypes influencing accuracy more than individual factors. Synaesthetes reported greater self-reported imagery ability than controls.

Conclusion:

This research explores individual cognitive differences, revealing how synaesthetes and their non-synaesthetic relatives share cognitive enhancements, suggesting a potential endophenotype that contributes to understanding perception, memory, and consciousness.

Introduction:

Culture profoundly shapes who we are, how we perceive the world, and how we process information. Yet, the links between cultural context, cognitive processes, and mental health remain elusive. This study examines how cultural differences impact subjective experiences, cognitive and attentional styles, and mental health, focusing on participants from the UK (Western), Chile (Western-Indigenous mix), and Japan (Eastern), selected for their variations according to Hofstede's cultural dimensions.

Method:

A control study (N=99) identifies suitable stimuli for change detection and change blindness tasks. The main study (N=300), with an equal number of participants from Japan, the UK, and Chile, recruits participants via local online platforms. Participants must be native residents. Tasks developed in the Gorilla Experiment Builder assess visual-spatial attention, including change blindness/detection, landscape drawing, cognitive reflection test, and non-verbal reasoning. Additionally, validated surveys measure thinking styles, cognitive-emotional dimensions, and mental health indicators.

Results:

Preliminary findings from the pilot study indicate significant cultural differences in thinking styles and visual-spatial attention. Japanese participants, who emphasize holistic thinking, favor background details. UK participants prefer foreground details, while Chilean participants fall in between in both attention style and thinking. Furthermore, intuitive cognitive styles were linked to higher anxiety levels, while holistic and reflective styles correlated with lower anxiety susceptibility.

Conclusion:

These findings suggest the critical role of culture in shaping cognitive processes, worldviews, and mental well-being. They have implications for developing culturally tailored mental health interventions, inclusive educational tools, and UX design. Additionally, they inform policymakers to respect cultural diversity in their strategies.

Facial emotion perception is influenced by the expresser’s age. While attitudes and beliefs about certain age groups (e.g., the elderly) may contribute to this effect, direct evidence remains limited. This study examined whether age-related stereotypes about emotional experiences (e.g., “older people frequently feel sad”) influence facial emotion perception. Young adults (aged 18–30) completed a facial emotion perception task and an emotional experience rating task. In the perception task, they identified emotions (anger, disgust, fear, sadness) expressed by younger and older male faces. In the rating task, they viewed neutral facial images of the same individuals and rated how frequently they believed these individuals experienced each emotion. The results revealed that younger faces were more likely to be mistaken for being angry (perception task) and were rated as experiencing anger more frequently (rating task). On the other hand, older faces were more likely to be mistaken for being sad and were rated as experiencing sadness more frequently. This pattern of misattributions and emotional experience stereotypes was largely consistent with previous studies. The results were the same when we conducted the experiment using real human faces and standardized 3D facial images. Thus, while the patterns of errors in facial emotion perception seem to align with age-related stereotypes about emotional experiences, the relationship does not hold at the individual difference level.

Daydreaming is a structured form of mind wandering involving imaginative scenarios, while maladaptive daydreaming (MD) is marked by vivid, prolonged, and compulsive fantasies that disrupt daily life. Music is a particularly powerful trigger for these immersive mental states and uniquely suited to studying daydreaming. Our research investigates the effects of music on daydreaming and MD through survey and experimental approaches. An online survey of 137 participants (83 individuals with MD, 53 controls) revealed that music, compared to other media (e.g., movies, books), is the most effective medium for initiating and sustaining daydreams, especially among MDers. Musical qualities such as positive valence, emotional intensity, and pleasure were pivotal in fostering daydreaming, with MDers reporting longer, more intense music-driven fantasies than controls. Additionally, musical absorption strongly predicted MD severity, suggesting potential avenues for targeted interventions. In a second study, 68 participants (34 MDers, 31 controls) completed the Sustained Attention to Response Task (SART) under three conditions: silence, positive music, and negative music. While music did not improve task performance, it reduced both internal and external distractions and increased fantasy-related thoughts for MDers compared to controls. These findings highlight music’s unique role in modulating attention and immersive mental states, providing insights into MD mechanisms and suggesting music-based strategies to support mental well-being. By bridging cognition, emotion, and consciousness, this research advances theoretical frameworks of mind wandering and offers practical implications for mental health interventions.

In this talk, I challenge the format approach of drawing the perception-cognition border by arguing that it fails to accommodate to sensory modalities other than vision, particularly olfaction. To distinguish between cognition and perception, one major kind of approach is the format-based approach. According to the format-based approach, the key differences between perception and cognition lies in the format of their representations: perceptual states are iconic, while cognitive states are discursive or symbolic. One notable version of format-based approach is recently proposed by Ned Block. Block argues that perception is fully iconic, nonconceptual and non-propositional, but cognition does not require these features. I argue against the format-based approaches by questioning the role of iconicity as a defining character of perception. Specifically, I focus on Block’s version of iconicity: Analog Tracking and Mirroring (ATM). ATM holds that differences in representations will track and mirror environmental differences in a degree-sensitive way. I argue that this criterion fails to capture the border of cognition and perception when applied to olfaction. To support this claim, I present three key challenges: 1) the irregular and non-linear mapping of odors onto molecular structures, 2) the tight connection between cognition and olfactory perception, and 3) the multisensory nature of olfaction. Based on these challenges, I further argue that olfactory perception is fundamentally non-analogue and non-iconic. Finally, I examine why olfactory and visual perception are radically different in format and why it is dangerous to treat visual perception as a paradigm case for perception.

A central challenge in consciousness research is distinguishing conscious access from the unconscious sensory processing leading up to it. Retro-perception offers a potential solution: a retrospective cue can trigger conscious access to a stimulus long after its disappearance, separating conscious access from initial sensory processing in time. However, previous studies used weak or masked stimuli, which mostly engage feedforward processing, leaving open the possibility that retrospective cues trigger later, but still unconscious, recurrent sensory activity rather than isolated conscious access. To address this, we use an attentional blink paradigm, where stimuli undergo strong, recurrent sensory processing yet often fail to reach consciousness. Participants view a target word (T2, upper or lower case), followed by a retrospective cue (500 ms later): either congruent (spoken version of T2) or incongruent (pseudoword). Participants report T2’s case and visibility. If retro-perception occurs, congruent cues should enhance case discrimination and visibility for blinked words. Preliminary results indicate that retro-cueing can indeed "bring back" blinked words, supporting the idea that conscious access can be cleanly decoupled from sensory processing. Future MEG studies will use this design to isolate the neural mechanisms of retro-perception, clarifying the distinction between sensory processing and conscious access.

As the global population ages, understanding age-related changes to the information processing dynamics that underlie our conscious experiences becomes increasingly urgent. Recent work has pioneered “cognitive matching,” where given patterns of neural activity are compared to a set of cognitively-distinct functional activity maps from the NeuroSynth meta-analysis of ~14,000 fMRI experiments, and has found that the strength of fit diminishes with unconsciousness (Luppi et al., bioRxiv 2023.11.08.566332; bioRxiv 2024.10.19.619194). In order to enhance our understanding of how resting state (RS) activity transitions through meaningful “canonical brain states” as we age, the current analysis aims to apply cognitive matching to RS fMRI data from the Cam-CAN cohort, a rich dataset of ~600 participants spanning ages 18-88 (Shafto et al., 2014, BMC Neurology). We are investigating the degree to which activity patterns match recognizable cognitive states, which NeuroSynth maps display the highest correlation with RS activity, and how both of these factors change over time. Preliminary results suggest that the best fit of RS fMRI data with NeuroSynth brain maps decreases significantly as a function of age (r^2=0.17, p<0.000001). Furthermore, initial analyses suggest that pairing this emerging cognitive matching technique with established proxies of conscious level (Lempel-Ziv complexity) reveals an interaction between age and Lempel-Ziv such that, under age 48, as Lempel-Ziv increases, so does best NeuroSynth fit, but over age 48, increasing Lempel-Ziv corresponds to decreasing NeuroSynth fit (r^2=.11, p<0.001). Through this analysis, we aim to uncover age-related changes in the interplay between cognition and consciousness.

Intentional binding (IB) refers to the phenomenon where the temporal interval between an action and its outcome is subjectively compressed. Typically, when an action (e.g., key press) causes a stimulus (e.g., auditory tone) to occur in an operant condition, the action is perceived as occurring later and the stimulus as occurring earlier compared to when the same events occur independently in a baseline condition. Recent studies have focused on the hypothesis that this phenomenon is driven by the perception of causality between the action and the subsequent stimulus. Based on this perspective, we hypothesized that a similar temporal compression might occur between a preceding stimulus and a response to it. To test this, we conducted two experiments using the Libet clock task, a common procedure for measuring IB, and psychophysical measurements based on the method of constant stimuli. The results demonstrated a robust compression effect, comparable in magnitude to traditional IB. These findings support the explanation that temporal compression is mediated by causal perception and call for a reconsideration of the interpretation of IB as a reflection of the sense of agency.

Conscious perception is intricately intertwined with expectations and attention, making it challenging to separate their neural basis. Most research has explored interactions in the spatial domain, while the relationship between consciousness, attention, and temporal expectations remains less understood. We hypothesize that unlike bottom-up attention, top-down temporal attention operates sluggishly, requiring precise temporal expectations to effectively enhance conscious perception. By employing a modified version of the temporal orienting task, we demonstrate how increasingly precise temporal predictions differentially guide bottom-up and top-down attention, and how temporal attention in turn facilitates conscious perception. We use the corresponding gradual changes in single-trial EEG dynamics to track the neural effects of expectations and attention during and after the foreperiod. Specifically, participants in our experiment identify the orientation of a threshold-contrast Gabor patch and then rate their perceptual awareness of this target using PAS. On every trial, the target is preceded by a neutral cue and a short or a long foreperiod with equal probability. Participants can therefore learn the precise duration of the two foreperiods only gradually. The ERP results identify the N2/P2 and the P3 complex as the neural correlates of consciousness. With increasingly precise predictions, P3 amplitude decreases over trials whereas N2/P2 amplitude increases. These changes correlate with a gradual increase of CNV amplitude in the foreperiod, even though CNV amplitude does not distinguish between seen and unseen trials. These preliminary findings therefore indicate that temporal expectations may not interact with conscious perception directly, but are mediated by selective attention.

Duration is the only feature obviously shared by qualia and their neural substrates. While perceptual experience integrates multisensory information into a unified whole, it remains bounded by a temporal horizon in which inputs and cognitive states are experienced as co-occurring. Here we propose a model for the phenomenological present as a ‘collective memory horizon’. Specifically, we propose that the duration of the present moment in a conscious system is governed by the decay rate of its collective active information storage (AIS) reflecting how long its past continues to disclose information about the present. The temporal horizon of experience thus corresponds to the timescale over which emergent information persists across system-wide interactions, before dissipating. Using tools from statistical mechanics, we define a landscape of energy levels that dictates the likelihood of each possible configuration of a network of binary spiking-neuron-like units, and then optimise the AIS of trajectories over possible landscapes. By tuning self and pairwise interactions across the network, we create conditions under which the group retains past information for longer than any single unit. In other words, while single units are comparatively ‘forgetful’, their interactions shape an energy landscape over which their emergent dynamics exhibit an extended memory window, as shown by a slower AIS decay. As the entire network history goes beyond the histories of individual elements, this suggests an emergent memory effect, thus offering a formal framework to investigate how a unified sense of the present could feasibly arise from the interactions between many comparatively forgetful components.

William James famously described the stream of thought as alternating between moments of stability and transition. Recent empirical studies support this distinction and highlight individual variability in these dynamics. This parallels clinical observations in psychiatric conditions such as Attention Deficit Hyperactivity Disorder (ADHD), where individuals may experience both sluggish and rushed thought patterns, suggesting altered temporal organization of conscious experience. In this study, we investigate these phenomena by reanalyzing data from four experiments, involving 220 neurotypical participants, who completed a validated clinical attention questionnaire and performed spontaneous word production tasks. Using semantic word embedding models (FastText), we define the semantic speed of thought as the ratio of semantic distance between consecutive words to the time interval between them. This metric integrates the semantic flow of thoughts and their temporal unfolding. We then investigated the relationship between the variability in the speed of thought and scores on the attention questionnaire. We found that participants with higher trait inattention exhibit greater variability in semantic speed, in two guises: First, they show an increased standard deviation of semantic speed. Second, the auto-correlation of semantic speed extends less far back in time, suggesting weaker temporal dependency in their stream of thought or faster fluctuations. Notably, this heightened variability occurs even though trait inattention does not affect mean semantic speed. We discuss these findings in relation to cognitive theories of the stream of consciousness, emphasizing their relevance for understanding individual differences in the temporal organization of conscious thought.

Our conscious experience suggests that our sensory systems continuously track an evolving stream of data in real time, but to some extent this is an illusion. Instead, both our sensory environment and neural processing show considerable rhythmicity that must be coordinated to optimize processing of the world. Research in the auditory domain has shown that neural tracking of environmental rhythms aligns favourable processing timepoints with likely moments of stimulus appearance, to optimise perception. Meanwhile, visual research has largely focused on the influences of neural rhythmic fluctuations that are ‘intrinsic’ and not defined by the external environment. It remains unclear how and whether the brain tracks visual rhythms similarly to audition. To address this, we conducted an MEG study in which participants attended rhythmic visual sequences of rotating Gabor patches and consequently judged the final patch’s orientation or timing. Using a rate-specific phase-locking analysis, we show that the motor system covertly tracks external rhythms, even when they are no longer concurrently present and with task-specific, perceptual effects. We further examine the mechanism of this tracking response and, using time-resolved decoding analyses, show that visual tracking of rhythms is linked to temporally specific mental simulation of the rhythm’s inferred sensory outcomes suggesting a tight coupling between temporal and feature-based anticipation. By showing that flexible tracking of external rhythms occurs similarly in vision as it had previously been shown in audition, our findings highlight the need to, regardless of modality, consider influences of environmental dynamics alongside and in combination with intrinsic constraints.

Self-produced events appear attenuated to us. The classic explanation of sensory

attenuation involves a forward model comparing predicted and perceived sensory

consequences of actions. While initially observed in the tactile domain, multiple

investigations have extended its applicability to the auditory domain. However, recently,

studies have shown that self-generated sounds can also be enhanced. We propose a novel

theory explaining both sensory attenuation and enhancement through adaptation to mean

sound statistics. We tested the perceived loudness of sounds between 40 and 80 dB and

found a regression to the mean sound level. Interestingly, this effect was nearly identical for

self-produced and passively observed but temporally predictable tones, suggesting that

predictability alone is sufficient to induce these perceptual changes. This pattern supports

the idea that perception adapts to the mean sound level rather than differentiating between

self- and externally generated sounds. To experimentally induce a new mean sound level,

we presented sessions where subjects produced in the majority of trials either loud (80 dB)

or quiet (40 dB) tones. When the mean level was loud, rarely presented quiet tones were

massively enhanced. Conversely, when the mean level was quiet, loud tones were

attenuated. Our findings suggest that sensory attenuation and enhancement are not

mechanisms modifying the perception of predicted action consequences but rather reflect

perceptual adjustments to average sound statistics. In conclusion, our brain does not

attempt to cancel out temporally predictable sounds but rather biases them to the average

sound statistics.

Perceptual experiences are not merely a reflection of incoming sensory input but are also shaped by prior experiences and expectations. One important source of prior information which has received comparatively less attention in the consciousness literature is the temporal properties of past events. Here we address this gap, by examining the influence of temporal expectations on perceptual acuity and characterizing the underlying cognitive computations. We designed a difficult change discrimination task, where the timing of change was varied on a trial-by-trial basis, drawing from one of three predefined distributions (uniform, exponential and flipped-exponential) and administered it to N=142 participants online. Our results show that both reaction times and perceptual accuracy fluctuate systematically with environmental temporal structure, suggesting that not only motor output, but also perceptual experience is modulated by temporal priors. Using computational modelling, we identify the logarithmic transformation of the event hazard rate (the probability that an event will occur given that it did not occur yet) as the core operation shaping these effects, in contrast with recent work suggesting that the reciprocal of the probability distribution underlies the influence of temporal priors (Grabenhorst et al., 2019, 2021). Moreover, our modelling reveals a key role for temporal estimation noise in shaping the process, both at the encoding stage, when temporal information is learned and represented and at the decoding stage, when it is used to influence perception. These findings suggest that temporal priors shapes perceptual experience, and shed light on the computational mechanisms behind this effect.

Our perceptual experience is continuously shaped by the integration/segregation of information within temporal windows operating at the sub-second timescale. While some properties of temporal windows like the temporal extent, influence of attention and load have been investigated, the effect of ongoing temporal context--particularly when it could facilitate anticipation--has not been systematically studied. To investigate this, we developed a novel paradigm of temporal integration where two halves of a Kanizsa square are integrated or segregated (seen or not seen as a full square) as a function of their ISI (30,100, 300, 800ms). The two halves were preceded and succeeded by rotating-discs of the same Kanizsa square in a temporally correlated fashion. Three context conditions varied the mean rotation increments (15°, 25°, and 40°), along with a no-context baseline. Participants (N=18) reported whether they perceived a full square in the sequence. We found that, at intermediate and longer ISIs, participants were more likely to perceive the integrated square in the presence of a correlated context compared to no context, suggesting that context-induced anticipation extended the temporal window. The integration was also enhanced in the smallest mean compared to the largest mean condition at the longest ISI, indicating that higher density of correlated context allowed for integration over longer gaps. Our study shows that the speed of temporally correlated context systematically affects integration, highlighting the adaptive nature of how information is organized over time.

Introduction: Existing functional imaging studies assessing neural correlates of consciousness (NCC) only used brief stimulus presentations. However, such an approach does not allow the investigation of whether an NCC candidate corresponds to the onset or the maintenance of a conscious percept.

Methods: In the present functional magnetic resonance imaging (fMRI) study, we used a visual stimulation design, which allows for two different prolonged stimulus presentations in both an aware and an unaware condition independent of the task relevance of stimuli. Participants (N=56) performed a visual distractor task with line stimuli presented in the background, sometimes containing an abstract human face for either 500 or 1000 ms. Depending on prior information, faces were either perceived or remained unaware as assessed by unannounced post-experimental reports.

Results: Results indicate that stimulus awareness and duration increased activation in the extrastriate visual cortex but not in frontal areas.

Conclusion: The findings support the hypothesis that activation in visual areas is tightly coupled to the content and duration of conscious visual perception. Using long stimulus presentations, our study takes a first step towards understanding sustained conscious perception and explaining the continuity of consciousness.

A key aim of consciousness research is to identify factors that impact perceptual processing versus subjective visibility. In vision, spatial attention facilitates both, but appears to affect them differently in magnitude and neural mechanisms. However, the effect of temporal attention was mostly studied for perceptual performance, while the impact on subjective visibility remains unknown.

Here we investigated how temporal attention affects perceptual processing and subjective visibility and underlying neural mechanisms. Participants (N=30) performed a discrimination task in which target onset times were predictable based on either rhythmic or interval-based regularities. We hypothesized that rhythmic attention could have a unique impact as it is putatively mediated by low-level oscillatory entrainment. We found that temporal attention improves perceptual performance as well as subjective perception, as indicated through lower perceptual thresholds compared to an irregular condition. However, this improvement was comparable between rhythm- and interval-based regularities. Surprisingly, EEG data revealed a similar increase in delta phase alignment in both predictive conditions. Anticipatory alpha-band power suppression however, which has also been reported in spatial attention paradigms, was observed in all conditions, including the irregular, suggesting that it represents a separate mechanism of attention.

These results highlight the impact of temporal attention on our perception and thus on the unfolding of our stream of consciousness in time. Further, they suggest that aperiodic interval-based temporal attention is mediated through alignment of non-oscillatory delta-band activity and leads to similar behavioral benefits as periodic temporal attention, in contrast to predictions of prominent entrainment models.

Although still fairly novel, the Integrated Information Theory of Consciousness (IIT) has become a relevant alternative in the study of phenomenological experience, especially because of the development of a formal mathematical framework aiming to bridge the gap between philosophical and scientific domains.

In spite of its advances, IIT is still work in progress and, in this sense, a particularly controversial aspect is its treatment of time. As a matter of fact, this sometimes leads to the need of strange or overly complex interpretations in cognitive terms, which naturally casts doubts over the theory as a whole.

To illustrate this problem, we present a set of experiments made using cellular automata (CA) and PyPhi. Our results show how basically any activation pattern of an oscillatory kind yields positive values of integrated information. This applies either to cases where we have considered the whole domain to be the system, or to patterns nested within larger domains.

The fact is, nonetheless, that this is only due to the inherently discrete dynamics underpinning CA systems, rather than some deeper, more interesting phenomena.

Put differently, cellular automata reflect well the world mathematically depicted by the IIT, by enacting the hypothetical isomorphic transformation between physical and subjective time, which is, or so we argue, the main problem of the current framework.

We finally discuss the need for less idealized temporal descriptions for calculating integrated information, as well as some conceptual and mathematical implications stemming from it.

Even in a shared environment, we do not all see the same thing. While the study of individual differences has revolutionized our understanding of personality and various cognitive abilities, consciousness research has focused on universal mechanisms(1). Here, we propose that individual variations in consciousness are not merely noise, but rather reflect stable cognitive traits. Response times in breaking suppression paradigms show remarkable reliability (split-half r>0.9, p<0.01), suggesting a fundamental trait we term nonconscious prioritization speed (NPS, 2). We hypothesized that NPS reflects the rate at which individuals consciously experience their environment, with faster NPS indicating more conscious experiences in a given timeframe. To test this, we developed a self-paced task using Minecraft. Participants (N=49) navigated through virtual urban landscapes while destroying cube-shaped adversaries as a primary task. The environment contained task-irrelevant billboards with embedded target stimuli(3). Then, participants completed a surprise recognition test for these 10 peripheral stimuli. Separately, we measured each participant's NPS through 160 trials of breaking repeated masking suppression(4). Results revealed that faster NPS predicts more accurate incidental memory (r=-0.32, p=0.027), suggesting that individual differences in NPS systematically relate to the breadth of conscious experience. A larger-scale pre-registered replication is underway. These results indicate that NPS represents more than just response speed—this fundamental individual difference in conscious processing has downstream effects on cognitive functions like incidental memory.

1. Gayet, et al., Frontiers in Psychology (2014)

2. Sklar, et al., Cognition (2021).

3. Hirschorn, et al. Behav. Res. Methods (2024).

4. Abir & Hassin, Conscious. Cogn. (2020).

One prominent proposal suggests a separation between the vision for action and vision for perception. It has been proposed that motor responses in priming are based on the rapid feedforward stream of information processing, which extracts only basic, physical properties of the prime stimulus. We tested this hypothesis using a Furrow illusion, in which a veridical, vertical dot movement in the periphery is perceived as tilted when presented against a tilted background grating. In the first experiment, we asked participants to make speeded responses to the target consisting of a dot moving across a tilted path. Preceding the target, prime was presented in the periphery, which consisted of a dot moving vertically, but due to the Furrow illusion was perceived as moving across a tilted path. We asked whether participants’ responses to targets would be affected by the consciously perceived movement along the tilted path or stay unaffected by the veridical, vertical motion of the primes. We observed significant effects of primes on rection times to targets, demonstrating that motor responses were affected by the perceived and not veridical direction of motion. In the second experiment we first estimated the size of the illusion for each participant individually and then used that estimate to eliminate the illusion. With this manipulation, veridical motion was slanted, whereas perceived motion was vertical. Surprisingly, reaction times were not affected by the veridical motion. In summary, our two experiments demonstrate that our visuo-motor responses use consciously perceived information about objects to guide motor responses.

Synesthesia is a cognitive trait where one sensory stimulus simultaneously evokes multiple types of sensations. Although several studies indicate that synesthesia can be acquired through learning, the specific process of transitional acquisition remains unclarified. This study aims to quantify the transitional acquisition of synesthesia by measuring grapheme-color associations observed in synesthetes while learning new graphemes.

We invited four synesthetes to engage in a 10-day online learning program focused on Thai numerals, during which they visited our lab on three occasions to assess their grapheme-color associations. In addition to a standard self-report questionnaire, we utilized eye-tracking methods to measure transient associations. Participants were shown each of the 10 Thai numerals printed in 24 different colors, and we analyzed their gaze patterns on these colored numerals. The findings revealed that, as the learning progressed, the bias in gaze—measured as entropy—increased and gradually converged towards a single color. Furthermore, longitudinal analysis of selected colors indicated that, during the first measurement, grapheme-color associations were predominantly influenced by shape. By the third measurement, however, this determinant had shifted to meaning.

The transition in the factors influencing synesthesia, moving from shape to meaning, can be reframed to highlight how newly learned characters are located in an individual's semantic network. This process transcends synesthesia; when acquiring novel graphemes, placing them within one's semantic framework is essential. Our research has revealed unique and personal experiences associated with graphemes, delving into the qualia connected to these graphemes.

Organisms predict environmental changes moment by moment through adaptive prediction. Changes in the environment that deviate from the context are key features to predict and process efficiently. When deviant sounds interrupt repetitive ones, they reveal neural underpinnings of auditory predictions conceptualised as prediction errors (PE) signals distributed in the brain. In this work, we investigated to what extent event-related potentials (ERP) dynamics encode synergistic and redundant information about PE processing in thalamocortical and corticocortical networks.

We implanted three cats with intracranial electrodes in auditory, prefrontal, motor, parietal-posterior, and visual cortices, and in the geniculate nuclei of the thalamus. Following a Roving auditory oddball paradigm, ERPs were computed for each condition (standard and deviant) during wakefulness, where the Mismatch Negativity was used as neural marker of auditory prediction. Mutual Information (MI) and Co-Information were calculated to reveal the nature of neural dynamics of corticocortical and thalamocortical information.

There was reliable MI between standard and deviant in all the cortices and thalamic locations studied, indicating that neural signals from all locations encode meaningful information about the identity of the oddball stimuli. We further found synergistic information in the ERPs in the thalamus, in the auditory, prefrontal and parietal-posterior cortices, and between auditory cortex and geniculate nucleus, and somatosensory, parietal-posterior and visual cortices.

We conclude that PEs can be encoded in a distributed manner between cortical and subcortical structures. The thalamocortical network, and not only the corticocortical connections, contributes to the integration of information from PEs.

The European Upper Palaeolithic (40,000–11,500 BP) witnessed a surge in artistic expression, notably marked by the emergence of figurative art—a development in symbolic thought which completely changed the nature of mental life. Understanding this cognitive shift should have a fundamental effect on our understanding of the development of human consciousness. To investigate this, we employed eye-tracking methodologies to examine behavioural responses to Palaeolithic imagery, focusing on ‘Venus Figurines’—small 3D female carvings (31,000–23,000 BP) representing the earliest traditions in the depiction of the human form.

Our study explored the influence of participant sex/gender on the perception of these figurines, expecting to find significant differences in line with known perceptual biases toward female bodies. Participants (n = 53, 27 female) viewed 21 Venus Figurines for five-seconds in a free-viewing task. Each stimulus was divided into five regions of interest (Head, Upper Torso, Lower Torso, Upper Legs, Lower Legs). ANOVAs and Bayes factor analyses revealed no significant effect of sex/gender on dwell time (p = .087, BF₀₁ = 172.58) or first fixations (p = .156, BF₀₁ = 5.414) to the ROIs.

These findings suggest that by the Upper Palaeolithic, symbolic cognition had advanced enough for artists to override baseline perceptual tendencies, allowing them to place greater emphasis on symbolic or cultural meanings. This challenges the application of modern intuitive beliefs to prehistoric art and highlights the complexity of early human mental life.

This research project investigates how processing fluency, influenced by affective and perceptual factors, impacts intuitive semantic judgments, thus offering insights into confidence and decision-making mechanisms. Two experiments employed eye-tracking methodology while participants completed the Dyads of Triads task, identifying semantically coherent word triads. Study 1 (n = 39) examined the impact of emotional valence (positive, neutral, negative) on judgment speed, confidence, and accuracy. Study 2 (n = 47) introduced perceptual fluency manipulation through visual contrast alongside emotional valence. Across both studies, participants identified solvable triads above chance, with high fluency conditions enhancing subjective confidence and response speed but not accuracy. Positive affective content led to faster responses and higher confidence ratings, while high visual contrast similarly accelerated judgments and reinforced confidence without improving accuracy. Eye-tracking data revealed reduced fixation durations for solvable triads under high fluency conditions, indicating more efficient processing. Longer inspection times correlated with decreased confidence across affect conditions, suggesting that extended processing signals uncertainty in intuitive judgments. These findings refine our understanding of processing fluency by demonstrating that fluency - whether derived from affective or perceptual cues - modulates subjective experience without affecting objective judgment accuracy. This research provides empirical evidence for the dissociation between confidence and correctness in intuitive decision-making, contributing to metacognitive models and emphasizing the role of eye-tracking in elucidating cognitive mechanisms underlying semantic coherence judgments.

It is commonly agreed that attitudes towards objects, ideas, and people are learned rather than innate. One of the mechanisms of attitude acquisition is evaluative conditioning. Evaluative conditioning (EC) involves pairing a neutral stimulus with an emotionally charged stimulus, leading to a change in the individual's response to the neutral stimulus over time. Cognitive mechanisms of EC include the processes of associative learning, where individuals form connections between stimuli based on experience and emotional responses—additionally, higher-order cognitive processes, such as forming propositions about environmental relations, influence EC. Our research aims to disentangle these processes and track their development. We conducted two experimental studies to examine the developmental trajectory of EC mechanisms. In Experiment 1, we tested N = 47 infants (9-10 months old) using an eye-tracking paradigm to assess visual preferences for conditioned stimuli. In Experiment 2, 74 children aged 3-4 and 6-7 years participated in an evaluative conditioning task measuring changes in preference before and after conditioning. The results indicate that 9-10-month-old infants did not show EC effects, challenging the assumption that associative learning alone drives early attitude formation. Older children exhibited higher preferences for positively conditioned stimuli, which was not the case for 3-4-year-olds. These findings suggest that propositional learning mechanisms, which require higher-order cognitive processing and potentially conscious access, become more dominant with age. Our study contributes to understanding the development of cognitive control in attitude acquisition and its implications for consciousness research.

Synesthesia is a cognitive phenomenon in which one sensory stimulus evokes multiple sensations, e.g., seeing colors when looking at graphemes or attributing personality traits to numbers (Ward, 2013; Simner et al., 2007). Although several studies indicate that synesthesia can be acquired through learning, the specific process of transitional acquisition remains unclarified. This study aims to elucidate the evolution of grapheme-color synesthesia during early learning. We recruited children from the general population, regardless of whether they exhibit synesthesia, since our target age group is just beginning their literacy study, making it difficult to predict which children may develop synesthesia in the future.

We asked children to select colors and personality traits that best represented each number from 0 to 9. We assured them there were no right or wrong answers, allowing for responses such as "none" or "don't know." Surprisingly, every child provided meaningful answers, often offering more nuanced descriptions than the options given (e.g., “’7’ and ‘8’ are twin old men, with ‘8’ being mean but kind only to ‘7’”). We continued our longitudinal interviews to assess the temporal consistency of their responses, a factor commonly used as an indicator of synesthesia. Additionally, we conducted qualitative analyses to understand the elements that contribute to the expression of synesthesia in children and how this differs from adult experiences of the phenomenon. Synesthesia relates to individual differences in subjective experiences. Through this research, we aim to uncover aspects of children's previously hidden subjective worlds.

Attention-Deficit / Hyperactivity Disorder (ADHD) is known to be associated with racing thoughts. Christoff et al. (2016) posit that the main determinant of the dynamics of spontaneous thoughts is the presence of constraints on cognition, be it automatic or deliberate. In the present project, we operationalized the unfolding of spontaneous thoughts with a word generation paradigm, building on the idea that word associations reflect cognitive processes involved in thought generation. We contrasted two levels of constraint on word associations to test the differential impact of subclinical ADHD-like symptomatology on the dynamics of the resulting “trains of thoughts”. In the "strong constraint" conditions, each word had to be "close" to the previous one, in terms of semantics or of context of utilization; in the “weak constraint” condition, participants had to say the words that came spontaneously to their mind. In the two conditions, participants generated series of 10-30 words aloud, following a metronome, after reading an initial seed word presented on the screen.

We first validate the notion that the "weak constraint" condition mirrors free thoughts by comparing the word produced in the task with participants’ self-reported real-life thought topics. Then, using reaction times and semantic metrics, we show that the participants who scored higher on an ADHD diagnostic questionnaire produced more dynamic series of words, but only in the "weak constraint" condition - akin to free thoughts. Our results have both methodological and theoretical implications for the emerging field of spontaneous thought research.

Background: Mind-wandering (MW) refers to the spontaneous experience of thoughts unrelated to the task at hand and has been linked to mental health conditions. However, the relationship between mental health traits and the temporal focus of MW remains underexplored. Aim: This study investigated how mental health traits (i.e., depression, anxiety, rumination, and flourishing) are associated with temporal orientation of MW in adults. Method: Sixty-five online participants completed a slowed digit Go/No-Go task across five blocks. After each block, participants recorded their MW thoughts and rated their temporal orientation (past, present, future) on a 1-7 Likert scale (1 = not at all, 7 = completely). Subsequently, validated psychological scales were administered to assess depressive, anxious, ruminative, and flourishing traits. Result: Multilevel modelling revealed distinct patterns of MW temporal focus across mental health traits. Participants with higher anxiety exhibited a significant tendency toward future-oriented thoughts. By contrast, those with higher depressive traits were significantly less likely to engage in future-oriented MW. Additionally, higher brooding rumination traits correlated with increased past-oriented MW, while reflective pondering, another type of rumination, was negatively associated with present-focused thoughts but positively linked to future-oriented MW. Finally, flourishing traits correlated significantly with greater present-focused MW. Implication: These findings highlight how different mental health traits shape the temporal focus on inner experience. They provide potential applications for personalized psychological interventions, such as cognitive-behavioral therapy and mindfulness-based practices.

Mind-wandering, defined as spontaneous shifts in attention away from a primary task toward unrelated thoughts, is a ubiquitous phenomenon in human cognition. It has been associated with Attention Deficit Hyperactivity Disorder (ADHD), a condition for which methylphenidate—a drug targeting dopaminergic and noradrenergic systems—is the most widely prescribed treatment. Despite its extensive use, it remains uncertain whether methylphenidate’s therapeutic effects in ADHD are mediated by its modulation of mind-wandering. To address this question, we conducted a double-blind, placebo-controlled, within-subject study examining the impact of methylphenidate on mind-wandering. Participants completed a Sustained Attention to Response Task (SART) featuring embedded thought probes that prompted self-reports of mind-wandering episodes. Concurrently, we recorded behavioral responses and electroencephalography (EEG) data, with all analyses preregistered and performed under blinded conditions. Our results indicate that reaction times (RTs) become more variable during mind-wandering compared to on-task periods. In addition, electrophysiological analyses revealed significant reductions in P1 and P300 event-related potentials (ERPs) during episodes of mind-wandering, which is consistent with the perceptual decoupling hypothesis. These findings provide an initial yet crucial step toward a deeper understanding of the neural bases of mind-wandering and elucidate how methylphenidate may modulate this phenomenon. Ultimately, this research could potentially contribute to refining pharmacological strategies for treating ADHD by highlighting potential mechanisms underlying the drug’s efficacy.

Our stream of consciousness is in constant flux. Our attention shifts continuously from one topic to another, and research has shown that individuals with a higher level of trait inattention experience more frequent and pronounced shifts. Similarly, the external world is ever-changing. To what extent do external events influence the dynamics of our stream of consciousness, and how does this influence depend on the internal attentional capacity of each individual? We address these questions through a spontaneous word generation task (free verbal fluency), enabling us to track the semantic content of participants' stream of consciousness. We implemented this task in two experiments (N = 60 in each), incorporating external distractors, either non-semantic (brief dots) or semantic (words of concrete objects) flashed on the computer screen. Across both experiments, we found that distracting events influenced the dynamics of the stream of consciousness, but with notable differences depending on their semantic content. Non-semantic distractors created temporary disruptions in the timing of word generation and the associated semantic distances, with stronger effects in participants with higher trait inattention. In contrast, semantic distractors had a more prolonged derailment effect on both the timing and semantic dimensions of the stream of consciousness, irrespective of participants’ trait inattention. We discuss our findings in the context of the emerging study of the stream of consciousness and their implications for research on consciousness in Attention Deficit Hyperactivity Disorder (ADHD).

Mind wandering (MW) is an ubiquitous phenomenon that despite its negative effects on cognitive processes like working memory, sustained attention, or response inhibition, has been suggested to play an important role in creative cognition, namely during the incubation period of a creative problem-solving task. However it remains unclear whether MW during the course of a creative task benefits the real-time expression of creative behaviour. Musical improvisation provides an ecologically useful framework for studying the real-time effects of MW on creativity. Indeed, in a preliminary study, the present authors showed that MW during a jazz improvisation task enhanced the creativity of musical improvisation in expert pianists, compared with states of on-task attention.

Here, we aim to replicate these findings with a bigger sample size and a more nuanced assessment of conscious experience that also captures other off-task thought phenomena like mind blanking and task-related interference.

In a repeated-measures design, 52 jazz musicians with varying expertise completed musical improvisation tasks interleaved with random thought probes. Musical creativity was independently rated by 2 musical judges using the Consensual Assessment Technique.

We expect to replicate previous findings of MW-associated enhancement of musical creativity. We also expect that this effect will be more prevalent in proficient musicians, compared to novices.

Should we obtain our expected results, this investigation will sediment our previous claim that mind wandering during improvisation is associated with a significant increase in musical creativity. It will also provide the first account of the real-time impact of different attentional states on musical performance.

Mind wandering (MW), as an important type of spontaneous thought, is a pervasive conscious mental phenomenon that occupies much of our waking life. In this talk, I will address the question: Why is mind wandering phenomenologically effortless?

The influential cognitive control theories of MW (Shepherd, 2019; Smallwood & Schooler, 2006) claim that MW involves cognitive control. A paradox thus arises: Cognitive control typically incurs mental effort, yet MW is experienced as effortless. I resolve this by proposing a new cognitive control account that integrates two key ideas: (1) MW consists of brief, semantically connected thought segments separated by abrupt topic shifts (the dynamic segmented structure of MW) (Sripada & Taxali, 2020); and (2) the costs incurred by cognitive control engaged in a single task accumulate over time (Kurzban et al., 2013; Shenhav et al., 2013).

Cognitive control incurs intrinsic and opportunity costs. These costs accumulate during prolonged engagement with a single task or topic. However, MW’s rapid transitions between various segments/topics disrupt sustained investment in any single segment/topic, resetting costs before they accumulate to a level that would induce phenomenological effortfulness. According to the proposed view, cognitive control should thus be understood as operating locally within each segment rather than globally across the entire MW episode, enabling fluid exploration of multiple current concerns without the effort tied to prolonged single-goal pursuit.

This account thus explains the phenomenological effortlessness of mind wandering by reconsidering the role of cognitive control in terms of its dynamic structure.

Mind blanking (MB) refers to a mental state in which people experience an empty mind or cannot report their immediate mental experience. So far, MB has been behaviorally quantified using probes, meaning that participants were choosing among variant options about mental states after a specific cue (e.g., after hearing an auditory tone). This methodology further allowed to characterize the MB’s neural correlates in terms of specific time-varying fMRI connectivity patterns (Mortaheb et al, PNAS 2022). What remains unknown is whether self-caught MB shares similar attributes with those MB reports that are probed. Behavioral data were collected from 22 healthy participants (age: 28yo ± 4; females: 45%, n = 10) who were instructed to press a button every time they became aware of experiencing MB during a 8-12 minute 3T fMRI scanning protocol. MB episodes were reported on average 4 times across the acquisition period (min: 0, max: 16), with 8 participants (36%) not reporting any MB instance. Self-caught MB were reported uniformly across acquisition time (χ2 = 78.36, p = .50, φ = .30). Clustering analysis of MB-related fMRI volumes revealed brain patterns resembling those using the probing approach. Together, these results are comparable with what has been shown for MB reports with probing in the absence of a background task (Mortaheb et al, PNAS 2022). Overall, we show that the self-caught methodology allows for a representative quantification of MB which can assist future research protocols.

Our minds often drift away from the task at hand. What happens to perception when the mind wanders? A common answer to this question is that perceptual processing diminishes (e.g. slow responses, more mistakes, weaker memory encoding), as in many cases of inattention. Unlike selective attention, however, the external environment remains relevant during mind wandering, suggesting meaningful levels of processing may be beneficial. Here, we use EEG and a Large Language Model (LLM) to ask what characterizes naturalistic speech processing (N=25) during self-reported mind-wandering periods. Using LLM-derived markers of contextual surprise, we found that the significant responses evoked by contextually unexpected words remain unchanged during mind-wandering, compared to attentive listening. Crucially, attention shifts and various acoustic covariates could not explain this response. Next, we used the LLM embeddings (hidden layer activations) of speech contents to predict EEG responses using cross-validated encoding models. By capturing the hidden representation, this method tracks the degree of encoding of contextual-semantic information beyond mere surprisal. This analysis showed that both mind-wandering and attentive listening are associated with significant encoding of the LLM embeddings, suggesting that EEG responses are influenced by the contextual meaning of speech when participants report mind-wandering. The results demonstrate the power of LLM to provide theoretical insights using naturalistic task designs, proposing that mind-wandering does not merely result in diminished processing and that some levels of semantic and predictive processing are still intact.

Although the consciousness-altering effects of drugs such as anesthetics and psychedelics are well-established, their effects on cortex-wide activity at high spatiotemporal resolution is less understood.

Utilizing wide-field optical imaging, we recorded fluorescent signals reflecting membrane potentials across both cortical hemispheres in mice that genetically encode voltage indicators in upper layer pyramidal neurons. Experiments were conducted under three conditions: resting waking state, anesthetized, and post injection of a hallucinogenic 5-HT2A agonist.

As many hypotheses of consciousness rely on brain-wide coherence dynamics, we investigated the holistic nature of ongoing activity. During the resting waking state we observe a stable fragmented distribution of phases. In contrast, in both conditions of altered consciousness, we revealed cortex-wide patterns of activity which oscillated between strong coherence and decoherence. These findings are consistent with prominent hypotheses, which describe metastability or entropy to play a crucial role in understanding consciousness.

To further analyze the underlying changes in brain states, we calculated neural manifolds from the time-resolved activity dynamics. We found relatively compact manifolds with sharp curvature during the wake state, suggesting attractor dynamics with metastable brain-wide neural activity. In contrast, application of a psychedelic induced long sweeping manifolds suggesting a diverse range neural states with high degrees of freedom, indicating a less constrained state of neuronal processing. Anesthesia, on the other hand, generated constrained manifolds, which could infer a less flexible and stable brain state.

Overall, the observed changes in cortex-wide dynamics during states of altered consciousness point towards neural mechanisms that involve massive changes in brain-wide coherence.

Mind-wandering reflects a complex interplay between focused attention and off-task mental states. Various studies have investigated the psychological and systematic mechanisms underlying these shifts. However, previous models have not yet provided an account for the underlying neural mechanisms for autonomous shifts between the two states.

Recent works investigated mind-wandering mechanisms using the Predictive Variational Recurrent Neural Network (PV-RNN), a hierarchically organized model rooted in the Free Energy Principle (FEP). The PV-RNN’s dynamic behavior is governed by a meta-level parameter, the meta-prior w, which balances the complexity term against the accuracy term in free energy minimization. While these studies provide critical insights into macroscopic neural mechanisms, the transition from FS to MW was generated by manual resetting of the meta-prior from a low to a high setting, leaving the mechanism for autonomous FS-MW shifts unexplored.

Motivated by the above, we propose an online adaptive mechanism for w, wherein w is modulated depending on the reconstruction error accumulated over a past period. A simulation experiment is presented to showcase the proposed framework. In particular, using PV-RNN, we trained the model to predict sensory patterns generated by probabilistic transitions among multiple cyclic patterns. Simulation results demonstrate that autonomous shifts between FS and MW emerged as w switched dynamically: high w enhanced top-down predictions, promoting MW, while low w emphasized bottom-up sensory perception, favoring FS. Finally, we speculate that agents could become consciously aware of MW when the accumulated error exceeds a certain threshold.

Decoding spontaneous thought from brain activity remains a fundamental goal in cognitive neuroscience. Traditional psychometric assessments often oversimplify the complex content of thought, while free speech analysis provides richer insights into cognitive and emotional processes. In this study, we propose an innovative method that employs large language models (LLMs) as automated raters of retrospective verbal reports. Twenty-seven participants engaged in a free-thinking task, closing their eyes for 30 seconds while their EEG activity was recorded. After each trial, participants described their thoughts, which were transcribed and analyzed. Multiple independent LLM agents evaluated these transcriptions along several phenomenological dimensions such as valence, arousal, temporal orientation, and “imaginariness”. We first, benchmarked these results comparing against human raters, obtaining similar agreement coefficients between human raters and LLMs. To decode neural markers of spontaneous cognition, we computed spectral, information theory and connectivity measures from the EEG data and converted the LLM ratings into binary categories. A random forest classifier, trained using leave-one-subject-out cross-validation, produced promising results with AUCs of 0.66 for positive versus negative, 0.75 for neutral versus affective, 0.65 for low versus high arousal, 0.62 for past versus future orientation, 0.73 for present versus non-present, and 0.75 for plausible versus imaginary thought. These findings demonstrate that LLMs can provide objective, scalable assessments of thought phenomenology and bridge neural data with nuanced cognitive evaluations. This approach has significant implications for advancing theoretical models and practical applications in neuroscience, aligning with current trends in brain-computer interfacing and neuropsychiatric assessment.

Introduction: Mind-wandering episodes often occur during everyday activities, such as listening to music. While a few previous studies have highlighted the impact of music-evoked emotions on thought contents (Taruffi et al., 2017), the mechanisms through which mind-wandering originates from music remain poorly understood. This study examines whether there is a temporal alignment between the perceived structural dynamics of the music and mind-wandering reports during music listening.

Methods: Self-caught instances of mind-wandering will be measured while 80 participants listen to one of two 5-minute instrumental music pieces (either in a familiar or unfamiliar genre). Additionally, to identify which structural elements may induce mind-wandering, qualitative data on listeners’ specific thought triggers and their associated mind-wandering episodes will be collected. A separate group of 50 participants (both non-musicians and musicians) will segment the music pieces to indicate perceived event changes in the music structure.

Expected Results: While data collection is currently ongoing and will be completed in early spring, it is expected that perceived event changes in the music structure will negatively predict the occurrence of mind-wandering, with a stronger effect for unfamiliar (versus familiar) music.

Conclusion/Significance: This study has important implications for understanding the complex relationship between environmental changes and the modulation of attention in a naturalistic context, such as music listening.

References

Taruffi, L., Pehrs, C., Skouras, S., & Koelsch, S. (2017). Effects of sad and happy music on mind-wandering and the default mode network. Scientific Reports, 7(1), 14396.

Depression is an invalidating psychiatric disorder characterized by a high risk of recurrence. Mind-wandering (MW) is a pervasive cognitive phenomenon that can become maladaptive in psychiatric conditions such as major depressive disorder. In particular, MW in depression is often characterized by increased rumination and negatively biased thought patterns. However, its underlying neurophysiological correlates remain insufficiently understood. We here aimed to identify the electrophysiological markers of MW in participants with a prior history of depression, hence at risk of depressive recurrence (ARDR) compared to healthy controls. To investigate these mechanisms, we employed electroencephalography (EEG) during a Sustained Attention to Response Task (SART) with intermittent thought probes to assess MW episodes. Additionally, a social exclusion manipulation using the Cyberball paradigm allowed to examine whether this intervention influenced MW dynamics and its subtending electrophysiological correlates. Event-related potential (ERP) analysis of Go-stimuli during the SART revealed that ‘on-task’ reports elicited significantly higher P300 amplitudes compared to ‘off-task’ episodes, suggesting reduced attentional engagement during MW. Spectral analyses further indicated increased theta power over frontal regions during on-task episodes, while MW episodes were associated with elevated alpha power over posterior electrodes. These findings align with previous research on MW and attentional disengagement. While these results contribute to our understanding of MW-related neurophysiological markers in depression, further research is needed to refine the characterization of MW subtypes and their role in depression symptomatology. Investigating the variability in MW signatures across different stages pre/post depression may provide valuable insights into cognitive vulnerability and potential targets for intervention.

Background:

It is unclear how the progression of Alzheimer’s disease may impact a person’s conscious experience of their environment. It can be very challenging to know what someone with advanced dementia is experiencing, mainly because they cannot report their thoughts. To circumvent this problem, we studied whether electrophysiological and fMRI signatures of perceptual consciousness found in healthy older adults remained present in people with mild-moderate and severe AD.

Methods:

In two ERP and fMRI experiments, we used a no-report masked visual perception paradigm with healthy older adults (n= 27, patients with mild-moderate AD n = 17 and severe AD n = 7). We contrasted cortical responses to pictures presented for 200ms vs 33ms that are characteristically associated with conscious perception. In the ERP study we examined the visual awareness negativity (VAN) and P400 electrophysiological responses, and in the fMRI study we examined activation in visual cortex and fronto-parietal regions.

Results:

In healthy controls we found characteristic cortical responses in both EEG and fMRI modalities, with VAN and P400 markers and widespread occipital, fusiform face area and fronto-parietal activation. In people with mild AD, there was significantly reduced VAN and P400 markers, and reduced fronto-parietal activation. In people with severe AD, who are behaviourally minimally responsive, there was limited evidence for any markers of conscious perception.

Conclusion:

These results demonstrate that the brain mechanisms associated with conscious perception become increasingly impaired with progression of AD, illustrating how AD should be considered a disorder of consciousness.

This study explores the intersection of time perspective (TP) and sleep/rest-activity rhythm (RAR) disruptions within the broader framework of consciousness studies, particularly focusing on individuals diagnosed with schizophrenia spectrum disorders (SSD). Utilizing phenomenological psychopathology, we investigate how temporal disintegration reflects broader disturbances in the conscious experience of SSD patients.

We employed the Zimbardo Time Perspective Inventory (ZTPI, a 5-dimensions questionnaire, i.e. Past-Positive (PP), Past-Negative (PN), etc.) and actigraphy to assess TP and RAR in a cohort of 230 subjects (87 females; median age:43, IQR:17). The sample included 122 individuals with SSD (54 outpatients, 68 residential patients) and 108 healthy controls.

Our analysis revealed significant correlations indicative of how temporal orientation impacts sleep as well as activity patterns among SSD patients. Key findings include a positive correlation between PP and Activity Onset (Pearson’s rho (ρ):0.32, p:0.02) and a negative correlation between PN and total sleep time (ρ:-0.29, p:0.03) for outpatients, and a negative correlation between PP and the mean level of activity during the 5 least active hours (Spearman’s rho:-0.27, p:0.02) for residential patients. Importantly, we found more pronounced differences in residential settings, suggesting that environmental factors significantly influence the manifestation of TP disruptions.

In this work emerged the critical role of TP in shaping the conscious experience of time and its impact on RAR. This may be crucial for developing targeted interventions considering both psychological and environmental factors affecting SSD individuals. This work highlights the need to integrate consciousness studies into clinical practices, significantly impacting mental health policies and therapies.

Introduction

Human conscious experience spans a diverse range of thoughts extending beyond the immediate present, such as mind-wandering or mental time travel. This study investigates how depression shapes thought patterns in everyday life and how ongoing experience responds to emotional stimuli in daily contexts. Our study builds on the importance of thought content and context for mental health, offering insights into the experience of depression through naturalistic and experimental approaches.

Methods

We used multilevel models to assess how depression relates to ongoing thought in 140 participants. Multidimensional experience sampling captured thought characteristics- such as valence, intrusiveness, distractibility- six times daily for seven days. Experience sampling was also used while participants completed positive and negative mood induction tasks in the lab using emotional stimuli encountered in everyday life.

Results

Individuals with higher depression levels reported more negative, intrusive, and distracting thoughts in everyday life. Contrary to intuitive beliefs, negative stimuli did not worsen maladaptive thought patterns in people with higher depression. Instead, they showed smaller reductions in negative thoughts after negative stimuli and fewer positive thoughts after positive stimuli compared to those with lower depression levels.

Conclusions

Our results suggest that depression anchors individuals in a persistently negative thought state, where negative stimuli have little effect as baseline experience mirrors negative conditions. We interpret this imbalance as a maladaptive “attractor state” that reinforces depressive symptoms and hinders shifts away from negative states. These findings offer insight into mechanisms maintaining depressive symptoms and highlight the influence of affective traits on conscious experience.

There is an increasing interest in neurodiversity which is often centred around a personalised understanding of one’s conscious experience, as well as strategies to better manage mental distress. Despite research into the brain’s functional landscape in conditions like autism (Hong et al., 2019), there remains little understanding of the structural substrate of inter-individual differences. We are investigating structural coordination in the brains of those with two often co-occurring conditions: synaesthesia and autism.

We present an approach to graph the partial correlations between structural properties of the cortex as parcellated by the human connectome project pipeline (Glasser et al., 2016). These structural covariance networks (SCNs) provide insight into group differences in connectivity and developmental processes (Alexander-Bloch et al., 2013; Sebenius et al., 2025) and yet can be derived from tractable T1-weighted images. We use clustering to estimate high dimensional partial correlations, and employ copulae to facilitate statistical comparison with group-aggregated networks.

Comparing surface area data from a recent study of 100 synaesthetes (Racey et al., 2024) and control data, we find increased network clustering and efficiency in synesthete SCNs, in line with results using cortical thickness (Hanggi et al., 2011). There is a lowered tendency, in synaesthetes, for neighbouring regions to covary in their structural similarity, which drives a reduction in small-world propensity (Muldoon et al., 2016). We will further develop our analysis by examining network complexity. This investigation contributes to our understanding of how perturbation of global brain organisation is related to fundamental shifts in the content of conscious experience.

Damage to the primary visual cortex does not entirely abolish visual processing. Some patients with visual cortical lesions exhibit residual visual performance in their blind field despite lacking conscious awareness—a phenomenon known as blindsight. This study investigated four patients with visual cortical lesions performing a spatial attention task while intracranial EEG and peripheral physiological signals were recorded. Patients identified the presence and orientation of gratings with different spatial frequencies in either their intact or blind visual field. Patients with right-hemisphere lesions, including the primary visual cortex, exhibited classic blindsight, performing above chance in orientation judgments, particularly for low-spatial-frequency stimuli. Patients with left-hemisphere lesions showed no signs of blindsight. Given evidence that interoceptive signals influence perception, in addition to neural activity, we examined the link between respiration, cardiac signals, and residual vision, including the heartbeat-evoked response (HER), previously associated with residual consciousness in comatose patients. In blindsight patients, target visibility and accuracy varied with the phase of the cardiac cycle at stimulus onset, showing opposite patterns for stimuli in the sighted and blind visual fields. Additionally, HER and pre- and post-target respiration amplitudes varied with both accuracy and visibility, while high-gamma oscillatory activity in occipito-temporal electrodes correlated with conscious perception. These findings suggest a causal role of the right primary visual cortex in blindsight, highlight the influence of interoceptive signals on conscious perception, and identify HER and high-gamma activity as potential neurophysiological markers of consciousness. This study advances our understanding of the neural and physiological mechanisms underlying conscious experience.

Inherited Retinal Dystrophies (IRDs) are progressive genetic disorders that lead to severe visual impairment due to photoreceptor degeneration. As vision declines, making the interaction with the environment increasingly challenging, adaptive neural mechanisms come into play to compensate for sensory loss. However, the extent to which IRDs reshape visual information processing remains unclear. This study explores neurophysiological adaptation in IRD patients through a multimodal approach, shedding light on the brain’s capacity for reorganization in late blindness.

A group of IRD patients and age- and gender-matched controls were assessed using Steady-State Visual Evoked Potentials (SSVEPs) in response to 12 Hz flickering checkerboards to evaluate residual visual processing. Additionally, Transcranial Magnetic Stimulation combined with 64-channel Electroencephalography (TMS-EEG) targeted occipital (O1, O2) and motor (M1) regions to examine cortical excitability. Results revealed for the first time preserved yet topographically altered neurophysiological responses and cortical excitability patterns in IRD patients even as the visual impairment was severe, thus suggesting retained but less spatially organized visual cortical function.

These findings contribute to the understanding of the neural correlates of visual awareness by demonstrating how residual sensory input and neuroplasticity shape cortical function in progressive blindness. They provide valuable insights into mind-brain interactions, reinforcing the need for integrated neuroscientific approaches to vision loss. With implications for vision restoration, this research holds international significance, informing both theoretical models of consciousness and clinical interventions for individuals with degenerative visual impairments.

Face recognition is a fundamental social function impaired in Prosopagnosia (i.e., disrupted conscious face identification despite preserved covert recognition). Interestingly, in patients with cortical blindness (i.e., Blindsight) unconscious emotional face processing still occurs despite the absence of visual awareness. However, it is unknown whether covert recognition in prosopagnosia arises from preserved unconscious face familiarity processing or face awareness is required. We investigated this question in a patient affected by Acquired Prosopagnosia and Cortical Blindness. MRI Lesion Mapping and Disconnectome analysis revealed extensive left-hemisphere lesions affecting occipito-parieto-temporal cortices, damaging white matter tracts crucial for cortical face processing. Using the Redundant Target Effect paradigm (bilateral visual stimulation in the intact and blind visual field of same/different stimuli) with familiar and novel faces, combined with the Eye-tracker, we examined whether the patient had blindsight-like visual skills, and if face familiarity processing was preserved irrespectively of the visual field stimulation (blind/intact). Results revealed that familiar faces presented in the intact field speeded the responses compared to novel faces, i.e., covert face familiarity processing. Crucially, responses to novel faces presented in the intact field were faster when a familiar face was presented in the blind field. This shows that unconscious face familiarity processing is preserved despite cortical blindness and Prosopagnosia, suggesting subcortical contributions in unconscious face perception independently from the patient’ brain damages. We suggest that subcortical nuclei may distinguish familiar from novel faces, assigning emotional significance to familiar faces without cortical inputs, promoting covert face processing in the absence of awareness and in Prosopagnosia.

Anorexia Nervosa (AN) is a severe psychiatric disorder characterized by maladaptive eating behaviors, including the avoidance of high-calorie foods and adherence to restrictive diets. While cognitive and emotional disturbances in AN are well-documented, the role of altered visual processing in shaping maladaptive food-related behaviors remains underexplored. This study investigates early-stage visual processing of high- and low-calorie food images in 32 AN-restricted female patients, focusing on their access to visual awareness, dominance within awareness, and implicit evaluative responses. We employed three paradigms: breaking Continuous Flash Suppression (bCFS) to examine initial access to visual awareness, Binocular Rivalry (BR) to assess perceptual dominance, and the Automatic Approach-Avoidance Task (FP-AAT) to probe implicit associations. Results revealed that high-calorie food dominated visual perception longer yet were consistently avoided in implicit tasks. Notably, perceptual dominance was positively correlated with body-focused scales e.g., body listening, while high-calorie avoidance was correlated with body dissatisfaction and interoceptive awareness measures. We hypothesize that disorder-relevant stimuli, such as high-calorie food, act as potent arousal triggers that assign excessive visual salience to aversive categories, reinforcing implicit avoidance tendencies. Given the profound preoccupation with food and distorted eating patterns in AN, our findings highlight the critical role of the earliest visual processing (unconscious), in sustaining food avoidance behaviors. This, in turn, indicates that targeting such stages of visual perception could be a transformative approach in AN treatment. Given the high mortality rate and limited treatment options, this research underscores the urgent need to develop interventions addressing preconscious perceptual biases to promote recovery.

Background: Hyposmia is a common non-motor symptom in Parkinson's disease (PD), linked to olfactory loss and impaired higher-order cognitive functioning. While structural and functional changes in PD with hyposmia are well-documented, studying the dynamic nature of brain states and information processing may offer insights into mechanism of olfactory loss in these patients.

Methods: Structural and functional MRI of PD patients 15 with severe hyposmia (PD-SH), 15 with cognitive normal ability (PD-CN) and 15 healthy controls (HC) were selected. We assessed the dynamic brain state, which characterizes brain's spontaneous spatiotemporal network alterations, and synergy and redundancy measure that capture brain’s capacity for higher-order information exchange. Group differences were analyzed using t-tests with FDR correction.

Results: A dynamic brain state (A), characterized by complex long-range global connections, was significantly decreased in PD-SH and PD-CN compared to HC. In contrast, brain state C, featuring modular-local clusters in sensorimotor and frontal areas, had a higher occurrence probability in PD-SH compared to PD-CN. Higher-order information flow (synergy) was significantly reduced in the bilateral superior temporal, parahippocampal, and cerebellar areas in both PD groups compared to HC. PD-SH showed a larger reduction in synergetic information flow in bilateral frontal, insula, and left sensory-motor areas than PD-CN. Brain state C and whole-brain synergy correlated with odor identification abilities, as assessed by the Japanese Odor Stick Identification Test.

Conclusion: PD patients with olfactory loss exhibit greater disruption in spatiotemporal brain connectivity and a reduction in higher-order information exchange, especially within the prefrontal and limbic regions.

Introduction. Individuals suffering from schizophrenia experience an altered perception of the world, distorting reality and shifting conscious experience. The exact neuronal underpinnings for this are only poorly understood. In schizophrenia widespread synaptic dysfunction has been described (Howes&Onwordi, Mol Psychiatry 2023) which is prone to cause dysregulation of excitatory-inhibitory loops. This might be more problematic in epochs of intense (meaningful) vs moderate (noise-level) neuronal firing. Intense neuronal firing might be less efficiently regulated by these loops leading to extreme intrinsic firing events being contagious between brain regions. This kind of dynamics is well captured by the co-kurtosis (CK) measure (Fry-McKibbin&Hsiao, Econometr Rev 2016). We hypothesize that contagion across brain regions is increased in schizophrenia due to altered excitatory-inhibitory neuronal connections.

Methods. We used resting-state fMRI data of the Human Connectome Project for Early Psychosis (HCP-EP) parcellated by ICA. CK and functional connectivity (FC) between all the time courses were compared across groups. CK results were referenced to phase randomized data.

Results. 17% of the connections exhibited significantly increased CK in the patient group as compared to only 10% altered connections found from FC analysis (4% increase, 6% decrease). CK changes were markedly centered on the salience network and correlated with PANSS sub-scores.

Conclusion. The results indicate that events of intense regional neuronal firing are more contagious to other brain regions in schizophrenia than in healthy controls. This contributes to disorganized incoherent thoughts. Increased contagion due to altered excitatory-inhibitory connections might contribute to the distorted conscious experience in schizophrenia.

Background:

People with advanced Alzheimer’s disease (AD) may appear to have reduced or fluctuating consciousness. However it remains unclear whether the capacity for conscious experience reduces with disease progression. The calculation of the perturbational complexity index state-transition (PCIst) from concurrent transcranial magnetic stimulation and encephalography (TMS-EEG) data could provide a no-report marker of consciousness in people with AD.

Methods

In a TMS-EEG study, single-pulse, monophasic TMS (>150 trials) at a stimulation intensity of 120% resting motor threshold, was delivered to the vertex of healthy older adults (n= 29), patients with mild-moderate AD (n = 17) and severe AD (n= 7). EEG was recorded from 63 electrodes, a TMS-evoked potential between 0 and 300ms post-TMS was obtained and the PCIst was computed using publicly available open-source code (Comolatti et al 2019).

Results

There were no significant differences in PCIst values between the groups F (2,51) = 0.213, P = 0.809. The mean (SD) PCIst values for the control participants were 28.5 (11.5); range = 13.4-51.8; for mild-moderate AD participants = 27.1 (11.9); range = 5.4-48.6; and severe AD participants = 30.7 (17.4); range = 14.1 -64.3. There was variability in PCI values between individuals in all groups, with no significant correlation with severity of AD.

Conclusion

PCIst variability is in keeping with previous studies in healthy awake adults and people with locked-in syndrome (Comolatti et al 2019). If capacity for consciousness remains intact even in severe AD, this highlights the importance of understanding how the contents of consciousness may be impacted by advanced dementia to ensure meaningful person-centred care.

Alzheimer’s Disease (AD), a degenerative disorder of the central nervous system, is increasingly being viewed as a disorder of consciousness (Huntley et al., 2021), as it may drastically alter how stimuli are represented and experienced. To explore how information processing is disrupted with AD and mild cognitive impairment (MCI), we applied an extension of the partial information decomposition method, called integrated information decomposition (Mediano et al., 2021), to examine the presence of different types of qualitative information processing in resting state fMRI. We used a large dataset from the Alzheimer’s Disease Neuroimaging Initiative to construct a hierarchy of brain regions from the most synergistic (working together to encode more information) to the most redundant (encoding duplicate copies of information), across healthy controls, individuals with AD, and individuals with MCI. Strikingly, compared to controls, the MCI and AD groups exhibited a near reversal in this information processing hierarchy, with greatest changes in the visual and somatomotor networks (cognitively normal group against AD: p = 1.23e-13, Cohen’s d = 0.843). This effect was even more pronounced when considering cognitive decline directly by comparing MMSE, a test for basic cognitive ability. Intriguingly, we found that almost all subcortical brain regions showed significant alteration in processing (FDR corrected p = 0.05) in almost all types of information. These results show the promise of information-based diagnostic tools and highlight a hitherto unknown potential cognitive-focused mechanism distinguishing AD from controls. Future work may confirm the relationship between departure from typical information processing markers and altered cognition.

Introduction: We and others have postulated that Alzheimer’s disease (AD) and other dementias are disorders of consciousness. The Perturbation Complexity Index - State Transitions (PCI-ST) measures the complexity of the brain’s response to transcranial magnetic stimulation (TMS) using electroencephalography (EEG) and has been shown to be sensitive to levels of consciousness, such as minimally conscious states. We sought to test the hypothesis that PCI-ST would be reduced in AD compared to healthy aging.

Methods: We assessed 28 participants with AD and 28 healthy controls, using TMS-EEG from left M1 and left IPL-angular gyrus, to calculate the PCI-ST. We also measured cognition with the Montreal Cognitive Assessment (MoCA) and disease severity with the Clinical Dementia Rating scale – Sum of Boxes (CDR-SB).

Results: We found that cerebral complexity, as measured by the PCI-ST collapsed across both stimulation sites, was lower in the AD group (M = 20.1) compared to controls (M = 28.9). Moreover, higher complexity across both motor cortex (M1) and inferior parietal lobule (IPL) sites was associated with better MoCA and CDR-SB scores. These results suggest that the PCI-ST is a cognitively relevant marker of conscious cognitive and functional capacity.

Conclusion: These findings support the hypothesis that cortical dementias are disorders of consciousness. This research opens the avenue for future studies on the fundamental nature of consciousness and its neuroanatomical correlates, in addition to enhancing our understanding of dementia and possible therapeutic strategies.

Despite extensive research on the neural correlates of consciousness (NCCs), mechanisms underlying inattentional blindness (IB) remain unclear. Here, we leveraged a three-phase no-report paradigm to investigate the electrophysiological NCCs in IB, minimizing motor and decision-making confounds. Participants performed a peripheral attention task while simultaneously being presented centrally with faces, houses, or visual noise. When probed about these stimuli at the end of the block, approximately 45% of participants failed to perceive the stimuli and exhibited IB. In phase 2, they performed the same task, after being informed about the presence of the stimuli. In phase 3, they were instructed to ignore the peripheral task and instead identify the central stimuli in a three-alternative forced-choice task.

Non-parametric cluster analysis of event-related potentials (ERPs) contrasting phase 2 with phase 1, controlling for noise trials, identified two neural components: the Visual Awareness Negativity (VAN, 180–220 ms) and a novel signal characterized by bilateral-posterior positive and frontal-central negative differences (250-400 ms). The P3b/P300, a traditional NCC marker, was only present during phase 3 confirming its association with task relevance rather than perceptual awareness. Multivariate pattern analysis (MVPA) assessing temporal generalization of decoders showed stable above-chance decoding of seen vs. unseen trials, during the 250–400 ms post-stimulus window. Time-frequency cluster analysis revealed significant differences in theta, alpha, and beta bands, implicating these rhythms in visual consciousness. The results emphasize the need to consider multiple processing stages for perceptual awareness, with two ERPs, temporal-generalization decoding, and EEG rhythms all linked to visual awareness without task-related confounds.

Brain injuries are known to disrupt brain network activity beyond the damaged area (von Monakow, 1914). A recent hypothesis suggests that this disruption arises from sleep-like slow waves intruding into the awake state, impairing cognition and behavior (Massimini et al., 2024).

Here, we tested whether regional intrusions of slow-waves contribute to visuospatial neglect, a condition in which, following brain injury, patients ignore the contralesional visual space. Six patients with visuospatial neglect were assessed in the subacute phase (T0) and upon discharge from rehabilitation, approximately one month later (T1). During each session, they performed a visual exploration task ("starry night") while EEG and eye-tracking data were recorded. We measured their left-right exploration bias and correlated it with regional EEG slowing, analyzing both within- and between-session relationships.

Preliminary results show that recovery from neglect (reduced exploration bias) between T0 and T1 coincides with decreased slow-wave activity in a widespread area encompassing contralateral frontal and parietal regions. Crucially, within-session analysis reveals that higher delta power correlates with stronger exploration bias, while transient reductions in slow waves align with gaze shifts toward the neglected space.

These findings suggest that post-lesional slow waves dynamically influence spatial awareness, and their reduction may support recovery.

Arousal and attention are key requirements for conscious access. Arousal and conscious perception typically exhibit a quadratic (U-shaped) relationship, with optimal performance during mid-level arousal. While arousal’s role in sustained attention is well studied, its effect on the dynamic orienting of attention remains unclear.

We measured spontaneous arousal fluctuations via pre-stimulus pupil size in 50 participants during an exogenous cueing task. Participants identified the orientation of a target grating preceded by a non-predictive onset cue. Attentional orienting, quantified as the cueing effect (validly vs. invalidly cued targets), was analyzed using linear and quadratic models across five bins of increasing pre-stimulus pupil size (reflecting increasing arousal).

The relationship between pupil-linked arousal and both overall performance as well as the cueing effect was well captured with quadratic (U-shaped) models, and model fits for the quadratic models were better than for linear models. Overall performance peaked at moderate arousal levels. Similarly, cueing effects were smallest at moderate arousal levels and larger at low and high arousal levels. This indicates that at mid-level arousal, attentional settings optimized target perception, effectively filtering out the non-predictive exogenous cues.

These findings provide a first characterization of how arousal modulates transient shifts of attention, advancing our understanding of how arousal and attention interact in enabling conscious access.

Creative thinking has traditionally been defined as thought that is both "useful" and "unique." However, this definition overlooks a crucial component: the cognitive and emotional work of meaning-making that lies at the heart of creative thought. This theoretical paper argues that creative thinking should be reconceptualized as an active process of meaning construction, drawing on evidence from psychology and aesthetics. In my paper I theorize that this meaning-making process involves five key components: openness to experience; curiosity driving novel connections; reduced self-referential thinking (similar to "ego dissolution" in psychedelic research); increased connectedness (heightened feeling of oneness with the world around us like when in an awe-state); and positive affect.

This understanding helps explain why mind-wandering often spurs creative thoughts. Research demonstrates that the inward turn away from immediate perceptual input during mind-wandering enables the mental space for constructing personal narratives and meanings (Smallwood & Schooler, 2015). It is also during this inward turn of attention that we generate creative thoughts as has been empirically validated through studies showing improved performance on divergent thinking tasks (Baird et al., 2012). Drawing from Noë’s sentiment that understanding requires thoughtful engagement (2023), I arguing that there is an active undertaking in order to make sense of stimuli. This undertaking – the “thoughtful seeing” or sensing - is at the center of the meaning making, and thus creative thinking process.This understanding of creative thinking as meaning-making reveals an overlooked dimension in the cognitive literature and offers new directions for creativity research.