Conference Agenda

We use our private feelings of confidence to coordinate our public, social lives. For example, when making joint decisions we can share our uncertainty honestly to reach an accurate consensus, or exaggerate our confidence to exaggerate our influence on the group. Some theories in cognitive neuroscience suggest that we can strategically distort the confidence we express to others while leaving our private feelings unchanged. But other theories suppose that our interactions with others may be a key source of beliefs about uncertainty in our own minds. If the latter is true, communicating about confidence with others could alter introspection on ourselves. We tested this idea in a novel decision-making task, where participants made perceptual choices and gave confidence ratings alone, together with a partner and then alone again. Across studies, participants showed ‘confidence matching’, with the confidence they expressed gravitating towards the confidence of their partners. Most importantly, these shifts in reported confidence persisted even after the interaction terminated, and participants made decisions alone again. These effects can be captured by a new predictive learning model, which assumes that agents use the confidence expressed by themselves and others to predict the confidence they ought to feel. These results reveal how the dynamics of social interaction can shape our sense of confidence and point to a mechanism that explains how communication with others can change our private states of mind.

Accurate global confidence -our overall belief in our abilities across tasks- is essential for effective decision-making and goal-setting. While local confidence reflects certainty in individual decisions, the process by which these judgments aggregate into global confidence remains unclear. We investigated how saliency in local confidence influences the formation of global confidence across two experiments, with the hypothesis that salient trials would disrupt the optimal aggregation of local confidences into global confidence. In both experiments, trials were organized into mini-blocks. Participants made decisions about a perceptual stimulus (dot numerosity task), reported their confidence in the correctness of each decision (local confidence), and at the end of each mini-block, reported their global confidence in their overall performance. In Experiment 1, saliency was manipulated by presenting memorable or forgettable images before the perceptual stimulus. Nevertheless, no significant difference was observed in how local confidence contributed to global confidence between memorable and forgettable trials. In Experiment 2, saliency was introduced by tagging certain trials as “special” and assigning them a post-decision bonus. Participants were informed that accurate decisions and local confidence in special trials could yield a monetary reward, but they only learned a trial’s special status after reporting their decision and confidence, ensuring no influence on task performance or local confidence. We found that special trials enhanced local-global metacognitive sensitivity -alignment between local and global confidence- for subsequent trials while disrupting this for earlier ones. Therefore, salient trials acted like spotlights, improving alignment for future trials but casting a shadow on preceding ones.

Psychosis is characterized by erroneous beliefs and delusions, thought to stem from impaired predictive processes. However, how these deficits originate and implicit and explicit processes’ contribution remains unresolved. Given well-documented ocular abnormalities in psychosis, we focus on the interaction between implicit ocular expectations and explicit decisions in forming aberrant predictions.

We examine probabilistic learning in a cross-sectional sample spanning the psychosis continuum (N = 150), including clinical high-risk for psychosis (CHRp), after a first psychotic episode (FEP), and control groups. In a virtual-reality-based paradigm, explicit decisions and post-decision gaze dynamics, that reflect an ocular confidence-like assessment of explicit decisions (Stern et al., 2024), were compared.

Explicit accuracy was impaired across the psychosis continuum (r = -0.19, p = .03), CHRp’s learning curve plateau was reduced (p = .005), and they switched their response rule more often (p = .01). These behaviors presumably reflect an overestimation of the environment’s volatility. Turning to implicit ocular expectations, CHRp’s gaze increasingly diverged from their explicit decisions (p = .01), reflecting reduced ocular confidence and the “hedging” of their responses in a volatile environment. Nonetheless, despite reduced ocular confidence, CHRp’s ocular metacognition was preserved (p = .57). Metacognition was significantly impaired only in FEP (p = .04), possibly serving as a protective factor in CHRp against full-blown psychosis.

These findings point to the overestimation of the environment’s volatility as driving impaired probabilistic learning in psychosis. Perceived volatility affected primarily explicit decisions, yet also reduced ocular confidence and increased the divergence between explicit and implicit expectations across the psychosis continuum.

The cold control theory of response to imaginative suggestions calling for distortions in veridical experience (including hypnotic suggestions) states that behavioural and cognitive responses are generated intentionally, but are perceived as involuntary due to inaccurate higher order thoughts of intending. Previous research has placed imagination as central to this response, yet imagined scenarios alone do not appear to result in feelings of automaticity or involuntariness. Here we explored whether imagined involuntariness, while imagining not being aware of thoughts to the contrary, will result in a greater sensation of involuntariness. We compared training in imagined involuntariness with simple practice in responding to imaginative suggestions, by comparing scores for subjective response and feelings of involuntariness for six suggestions. The pilot results have been confirmed as reproducible by an independent statistician. These were insensitive, but a post hoc analysis indicated that had the test suggestions been limited to motor suggestions (as the training had been), with the hallucination suggestions eliminated, then it would have found evidence to support the training group resulting in greater subjective effects and sensations of involuntariness than the control group. We are currently conducting a registered study to test training on motor suggestions and, if successful, we will repeat with hallucination suggestions. The results bare on the debate as to whether hypnotic response can be trained.

Voluntary decisions are endogenously driven choices guided by subjective goals and preferences. Although fMRI studies have identified brain regions involved in voluntary decision-making, it remains unclear how voluntary decisions unfold over time. Specifically, do evidence accumulation dynamics observed during perceptual and memory-based decisions also occur during voluntary decisions? Additionally, do motor preparatory processes differ between voluntary and forced (i.e., instructed or stimulus-cued) decisions?

Using electroencephalography, we examined two event-related potential (ERP) components that gradually build-up and peak around response onset: the centro-parietal positivity (CPP), indexing evidence accumulation, and the readiness potential (RP), reflecting motor preparation. Forty-nine participants completed a colour decision task with voluntary (two options) and forced (one option) trials. We compared CPP and RP waveforms across decision types and assessed whether their build-up rates scaled with response times (RTs). Stimulus- and response-aligned ERPs were separately examined using a deconvolution algorithm.

We did not find differences in CPP or RP waveforms between voluntary and forced decisions. However, RTs were positively associated with build-up rates for both components, with faster decisions reflecting steeper accumulation rates. This suggests that the CPP and RP reliably track accumulation dynamics during voluntary decisions.

This study extends key findings from other decision-making domains to voluntary decision-making, revealing similar evidence accumulation and motor preparatory dynamics between voluntary and other forms of decision-making. Notably, this is the first study to show that the CPP tracks evidence accumulation in voluntary decision-making. Furthermore, our results support the notion that motor preparatory processes are similar between voluntary and forced decisions.

A fundamental question in memory research is why some past events are vividly recalled while others are forgotten. The sense of agency (SoA), the feeling of controlling one’s own actions and their effects in the environment (Haggard, 2017), may play a critical role in shaping memory, as both involve self-referential processes (Tulving, 2002; Wen & Haggard, 2018). Here, we investigated whether SoA over the displacement of objects via motor actions modulates long-term memory for these objects in healthy human adults (N =30). Participants pressed a button to move a picture of an object displayed on the screen, where the picture’s movement was either spatially congruent or incongruent with their action. One hour later, memory for the objects (what) and their locations (where) was assessed in a recognition task. Results showed higher agency ratings following congruent versus incongruent trials (difference: 81.45±3.84; p < 0.001, Cohen’s d = 3.87). Crucially, higher SoA during encoding was associated with faster reaction times in the old-new recognition task (difference: -0.14±0.06; p = 0.034, Cohen’s d = -0.41; trial-by-trial analysis: p = 0.067) and improved location memory accuracy (difference: 0.19±0.05; p = 0.001, Cohen’s d = 0.66; trial-by-trial analysis: p < 0.001). These findings highlight a link between SoA and memory, implying that the brain may prioritize the encoding of events with clear cause-and-effect relationships.

Humans have the capacity to feel a subjective sense of regret – a rather sophisticated emotion that involves counterfactual simulation of paths-not-taken (Frith and Metzinger, 2016). Pathological levels of regret, as in states like rumination, are a hallmark of multiple psychiatric conditions (Menzies, 2022; Sijtsema et al. 2022). However, little is understood about how subjective feelings of regret connect to the learning and inference processes we use to make decisions – and how alterations in these processes may explain atypical experiences of regret.

We developed a reward learning task with sporadic probes for post-choice regret. Reinforcement learning modelling revealed that subjective feelings of regret were closely and independently tied to evolving beliefs about action values and trial outcomes. Crucially, by retesting the sample at separate timepoints, we found that the ‘regret weights’ afforded to predictions and outcomes were stable within individuals over time – with some participants consistently relying less on their predictive models of the environment and more on (potentially spurious) outcomes when forming feelings of regret.

In a large follow-up sample, we connected these insights about prediction, learning and regret to transdiagnostic psychiatric symptoms. This revealed how the tendency to downweight internal models and upweight trial-by-trial outcomes related to mental health dimensions through trait rumination.

Our findings provide a mechanistic perspective on how the subjective experience of regret arises through internal monitoring of basic learning and inference processes, and offer insights into how disruptions of this monitoring process become a core feature of pathological experiences.

Moral dilemmas involve choosing between two mutually exclusive options with opposing moral values. In the trolley dilemma, for example, a participant has to choose between saving as many people as possible, or not being involved in the death of a single individual. Multiple studies showed that individual participants experience a sense of agency and responsibility for the outcomes of moral decisions when they are the actor of that decision. In real life, however, moral decisions are rarely made in isolation. When making decisions together with others, people typically do not experience the same level of agency or responsibility for the outcome. The current research aims to investigate how participants’ reactions to the outcomes of moral dilemmas are different for collective vs individual decisions. To investigate this, we designed an experiment in which participants are paired into face-to-face dyads and are presented with a series of trolley-like moral dilemmas. They first report their response privately on an external device (alone trials), then make a collective decision by reaching a consensus (social trials). Simultaneous EEG recordings of the two participants are made during the whole task. Feedback-related negativity (FRN) and the P3 component will be used to quantify to which degree participants process outcomes of individual decisions differently from collective decisions. Preliminary results will be presented at the conference. These results will contribute to explaining how the sense of agency can be modulated by social contexts.

Brain-computer interface (BCI) technology promises to enhance our agency by decoding user intentions, even if the user lacks the capacity for intentional bodily movement. Users have been documented as experiencing a sense of agency, both explicit and implicit, over their BCI actions and outcomes, akin to the sense of agency experienced in the embodied case (Caspar et al.). In contrast, experimental decoding of intentions — such as the much-discussed broad Libet-inspired paradigm — has been taken to threaten agency, exposing the illusion of conscious control, even if the agent has a full range of intentional bodily movement. Intriguingly, both projects rely on the same neurotechnology and brain activity.

In this work, I review the case for BCI actions and outcomes as intentional action and as attributable to the conscious agent. I re-visit classic armchair thought experiments in action theory as well as neuroscience of agency. I outline four candidate characterizations of (active) BCI use, as non-action events, as basic mental action, as non-basic action, and as basic extended action.The standard causalist story of action should cohere well with BCI action under the umbrella of basis mental action and nonbasic intentional action given its compatibility with physical realizers and causal mechanisms. However, I argue that on the standard causalist view, BCI users are only agents in the sense of consciously (or not) intending outcomes; their agency stops there, “in the head”. In response, I argue that the best case for BCI use as agency involves adopting an extended casual mind framework.

In psychology, shared representations are thought to exist beyond individual minds, as theorized by Carl Jung's concept of the collective unconscious—a shared repository of universal experiences, memories, and symbols. Our focus is on how artificial architectures underlying artificial agents can share their representations to build a common sense and collective construct. We will propose a model of agentivity for the assembly of agents that accounts for the agentivity of the parts but cannot be restricted to them. In this model, each agent is an active inferer; it performs a cycle of actions based on preferences and inferences based on observations, with an additional message-passing step where all agents exchange their beliefs to ensure compatibility of their beliefs while keeping their own singular point of view on their environment. Compatibility of beliefs ensures a collective representation that serves as the basis for shared agentivity. Such a model relies on a novel methodological tool for variational inference in the multi-agent setting for active inference, where each agent has its own partial and subjective experience of the environment and must reconcile with others' points of view in order for the whole collection of agents to work in synchrony. In our model, the compatibility of beliefs does not imply consensus but is simply a way to ensure that each agent is referring to the same environment. The beliefs reached by such agents in an exploration task are informative on what particular beliefs arise from shared agentivity.

Agency, the feeling of controlling one's actions and their outcomes, is distorted in several neurodegenerative disorders, disrupting normal consciousness (Haggard, 2017). Here, we study the effects of agency on apathy, a transdiagnostic syndrome of amotivation (Marin, 1991) linked to decreased reward sensitivity and resulting in a reduced acceptance rate during effort-based decision-making (EBDM) tasks (Le Heron et al., 2019). We hypothesize that loss of agency decreases willingness to act, reflecting a state of increased apathy. To test this hypothesis, we ran a study (n=22) in a healthy young sample using a validated EBDM task, where participants receive offers to perform physical effort (i.e., keyboard tapping) for a reward, which they can accept or reject. We modified this task to manipulate agency by adding a visuomotor conflict to the visual tapping feedback. Our study found a marginally significant reduction in trial acceptance for the high visuomotor conflict condition. Moreover, group modeling of the data suggests that low agency specifically affects participants’ reward evaluation of the effort-reward decision, akin to the decreased reward sensitivity found in apathetic patients. These findings suggest agency plays a role in modulating motivation in a way that reduces willingness to exert effort for reward when agency is diminished. Broadly, our results highlight the role of agency in conscious experience, as agency disruptions alter decision-making and motivation. This offers a new path for studying apathy in diseases like Parkinson’s that have been linked to a reduced sense of agency.

Introduction. Higher brain functions, such as perceptual decision-making, are thought to depend on maintaining an optimum balance between excitatory and inhibitory interactions in cortical circuits(1-3). The neurotransmitters setting cortical excitation-inhibition ratio (“E/I-ratio”), GABA and glutamate (via NMDA receptors), are altered in schizophrenia/SCZ(4). We compared behavioural markers of GABAergic and NMDA-mediated interactions between SCZ patients and healthy controls (HCs).

Methods. 41 SCZ patients and 41 matched HCs compared two simultaneous visual circular grating sequences, which fluctuated in contrast over 1 second, reporting which within-trial sequence (left/right) displayed higher average contrast. This task quantified two candidate markers of cortical E/I-interactions: “psychophysical kernels”, the temporal weighting of contrast fluctuations on choice; and “choice history bias” (CHB), the impact of previous choices on current ones.

Results. SCZ placed less emphasis on contrast information occurring early in trials than HCs. There was no overall group difference in CHB, but SCZ showed a reduced tendency to repeat previous correct choices during repetitive, predictable stimulus sequences.

Conclusion. Results identify SCZ-related alterations in GABAergic, NMDA-mediated cortical interactions shaping decision-making at both within-/across-trial timescales. Although somewhat contentious, reduced within-trial weighting of early perceptual information does occur in some circuit models of SCZ under decreased E/I-ratio(5). Additionally, reduced CHB in predictable environments in SCZ is consistent with NMDA hypofunction effects found previously in related task contexts(6). Together, these findings provide multi-scalar mechanistic markers elucidating the neurocomputational underpinnings of aberrantly altered perception.

1.van Vreeswijk, Science(1996);274(5293):1724-6

2.Shadlen, J Neurosci(1998);18(10):3870-96

3.Deneve, Nat Neurosci(2016);19(3):375-82

4.Uhlhaas, Nat Rev Neurosci(2010);11(2):100-13

5.Lam, J Neurosci(2022);42(6):1035-53

6.Stein, Nat Commun(2020);11:4250

To what extent does subjective appraisal shape cognition and behavior? While this question may appear trivial, it becomes essential when subjective experience misrepresents reality. We know that placebos, as well as suggestion in general, can produce genuine physiological effects. However, their underlying mechanisms remain unclear.

Here, we focused on the role of subjective experience in conflict tasks, which involve congruency relationships between prime target and stimuli. In this context, Desender et al. (2014) showed that the sequential congruency effect (SCE; Gratton, 1992) only occurs when people actually experience the conflict. In contrast, van Gaal et al. (2010) argued that adaptation effects can result from unconsciously triggered conflicts.

Here, taking inspiration from Palmeri et al. (1999) and the phenomenal suggestions literature, we aimed to go one step further and explore what happens when there is no conflict whatsoever, but people believe there is. To do so, we asked participants to respond to the orientation of a target arrow also acting as a metacontrast mask for a congruent or incongruent prime. Unknown to participants, the prime was neutral on half the trials, but an auditory cue suggested the presence of a conflict, just as a nocebo pill.

Our behavioral results show that the SCE is triggered after a sham-induced experience of conflict and that this effect is influenced by participants' capacity to modulate their subjective experience (i.e., phenomenological control, see Dienes & Lush, 2023). We conclude that this spontaneous adaptive behavior (SCE) can be shaped solely by people’s subjective experience.

Serial dependencies in perceptual tasks have been suggested to reflect implicit perceptual priors derived from short-term perceptual history. Generally, a reduced weighting of perceptual priors has been proposed as a mechanism underlying psychotic experiences such as delusions and hallucinations. In this study, we examined how implicit and explicit perceptual priors influence auditory decision-making and their modulation by psychosis proneness. Across two online experiments, 197 healthy participants in whom psychosis proneness was measured using standard questionnaires completed an auditory two-alternative forced-choice task. In experiment 1, implicit perceptual priors were assessed by quantifying the influence of preceding on current perceptual choices. We found a significant positive effect of choice history and a negative modulation of this choice history bias by psychosis proneness. In experiment 2, we manipulated explicit perceptual priors by providing probabilistic cues before each trial. We found significant effects of choice history and probabilistic cues, but only the latter was negatively influenced by psychosis proneness. Interestingly, both experiments showed an increased reliance on the available stimulus information with higher psychosis proneness suggesting alternated perceptual awareness. Using Hidden Markov Models, we found no conclusive evidence for a modulation of the temporal dynamics of perceptual decision-making by psychosis proneness. Our findings support the notion of a shift in perceptual inference away from perceptual priors and towards sensory information as a mechanism facilitating psychosis. The degree to which psychosis proneness modulates the effects of different types of perceptual priors seems to depend on the sources of priors available in a given experimental context.

The influence of action-based predictions on perception is fundamental, particularly for understanding the sense of agency, i.e., the subjective experience of control over one’s actions and their outcomes. While the literature reports mixed results, such as sensory attenuation or enhancement, the current research aims to contribute to our understanding of how actions and predictions influence visual perception through two experiments. In Experiment 1, participants viewed a grating and were asked to reproduce its orientation. Before the grating onset, different cues allowed them to predict its global orientation based on their own action (free-choice or forced-choice conditions) or an external cue (passive predicted condition), while in a control condition (passive neutral), no prediction was possible. Results revealed that perceptual precision in the reproduction task improved in both the free-choice and externally cued predictions compared to the passive neutral condition, whereas forced-choice predictions provided no such benefit. Experiment 2 investigated whether participants relied more on their predictions when the stimulus was harder to perceive. This was tested by examining the impact of free-choice action-based predictions under two levels of stimulus ambiguity within the same reproduction task. Results showed enhanced perceptual precision in the predicted condition compared to the unpredicted condition, regardless of task difficulty. Findings suggest that both cue-based and action-based predictions can improve perceptual precision, but crucially, only when action are freely chosen. A lack of intentionality on action-based predictions may reduce attentional allocation to sensory action-effects, potentially limiting the benefits of sensory predictions for perception.

The sense of agency (SoA), the feeling of being the author of one’s actions and outcomes, is critical for decision-making. While prior research has explored its neural correlates, most studies have focused on neutral tasks, overlooking moral decision-making. In addition, previous studies mainly used convenience samples, ignoring that some social environments may influence how authorship in moral decision-making is processed. This study investigated the neural correlates of SoA in civilians and military officer cadets, examining free and coerced choices in both agent and commander roles. Using an fMRI paradigm where participants could either freely choose or follow orders to inflict a mild shock on a victim, we assessed SoA through temporal binding (TB)—a temporal distortion between voluntary and less voluntary decisions. Our findings suggested that SoA is reduced when following orders compared to acting freely in both roles. Several brain regions correlated with TB, notably the occipital lobe, superior/middle/inferior frontal gyrus, precuneus, and lateral occipital cortex. Importantly, no differences emerged between military and civilians at corrected thresholds, suggesting that daily environments have minimal influence on the neural basis of moral decision-making, enhancing the generalizability of the findings.

When we make decisions, we often have a subjective experience of "knowing that we know" - a feeling of confidence in our choice. Researchers have studied this feeling by asking people to rate how confident they feel after making a decision. Traditionally, researchers assumed that asking participants to self-report their subjective confidence wouldn't change the decision-making process. However, recent studies have begun to question this assumption, showing that the very act of thinking about our subjective confidence can either help or hurt our ability to make accurate decisions, a so-called reactivity effect1. In three experiments, we had participants perform blocks of a perceptual discrimination task with or without confidence ratings. We found that confidence ratings elicited contemporaneously with perceptual decisions impaired decision accuracy, while retrospective confidence ratings did not. Furthermore, we found that when we delayed responding to allow for sufficient post-stimulus processing, then eliciting contemporaneous confidence ratings was no longer reactive. These findings suggest that there is a disassociation between the perceptual experience and the metacognitive experience, such that when they occur simultaneously, interference arises that impairs decision accuracy. This implies that subjective feelings of confidence arise from a distinct set of evidence to the primary perceptual decision. In practice, these findings also raise important questions about whether asking people to rate their confidence is an appropriate way to study consciousness, as scientists have long assumed.

References

1. Lei, Wei, et al. "Metacognition-related regions modulate the reactivity effect of confidence ratings on perceptual decision-making." Neuropsychologia 144 (2020): 107502.

In the philosophy of mind, there is a view that we experience our own actions in a distinct way, contrasting with the passive experience of other events. This “agentive awareness” is difficult to pinpoint in mental vocabulary without reducing it to mere sensory effects. The representational theory of mind may help clarify this awareness through the concept of directive representations that function in an “output-oriented” manner – such as motor representations (MR). I highlight the challenges of using directive representations as vehicles of conscious content by drawing on examples from cognitive science. First, the motor simulation theory posits that MR become conscious during motor imagery, but there their role is reduced to reproducing the sensory effects of movement. Second, according to the comparator model of the sense of agency (SoA) a copy of the MR is processed and compared with movement effects, leading to SoA, yet the format of SoA is independent of the MR format and thus open to multiple interpretations. Third, motor intentions, the immediate causal antecedents of movement, remain unconscious and do not directly support the experience of initiating movement despite depending on higher-level intentions. In all these cases, the functional role of MR as output does not yield a specific format of mental content. To address this, I propose a process-based account of directive representations, linking agentive awareness to changes in sensory and cognitive content, and offering a novel framework for understanding how actions are experienced. This framework encompasses both metacognitive and motor control.

Metacognition comprises the ability to estimate the accuracy of predictions about the world. A variety of measures for this ability exist. The most established measures (meta-d’, M-ratio) are based on signal detection theory and require assumptions on underlying normal noise. Recently, measures based on Classical Information Theory have been proposed which have the advantage of requiring fewer assumptions. We further develop this approach based on the key quantity: the information transmitted via confidence reports. We show that predictions with a fixed overall accuracy are consistent with a range of possible transmitted information values and we present tight upper and lower bounds for this range. Based on these bounds, we propose a new, normalized measure of metacognition: For a fixed accuracy, this measure assigns a value of 1 when information is transmitted at the most possible rate (corresponding to underlying uniform noise) and 0 when information is transmitted at the least possible rate (corresponding to underlying binary noise). This normalization decouples our measure from the overall accuracy improving upon existing measures, whose ranges depend on the accuracy of the responses — an undesirable confound. We further relate metacognitive ability to the performance in groups: For the case of multiple independent group members combining their predictions about states of the world, we provide upper and lower bounds on the group accuracy and show a strong relation to the individual metacognitive ability. Overall, we present an advanced information-theoretic perspective on measures of metacognition and their connection to group research.

In human evolution, tool use has been one of the most important milestones. Metacognition has been suggested as a critical mental ability for this capability. To investigate its role in tool use under uncertainty, we developed a novel paradigm that involves action selection during tool use scenarios and subjective metacognitive reports.

We adapted a claw machine arcade game into a 2D video game where participants pick and drop balls. To manipulate outcome uncertainty, we varied the balls' weights in terms of their mean and spread, resulting in a 2×2 design with four conditions: lighter or heavier (generally easier or harder to pick up), and certain or uncertain (narrow and wide weight distributions around given mean weights).

After training, participants made trial-wise choices between two balls from different conditions, each associated with a reward. Then, before executing the action, participants reported their metacognitive judgements via prospective confidence ratings: one about the optimality of their choice for their overall reward count, and another about their ability to successfully pick and drop the ball.

Preliminary results from our pilot study indicate that participants have metacognitive insight into both the potential success of their tool use and outcome uncertainty, with a tendency for higher confidence in the former.

We discuss these findings in the context of existing confidence models and suggest potential applications for artificial agents.

Self-evaluations of perceptual decision accuracy are thought to rely on a general metacognitive mechanism that computes confidence as a common currency across tasks. As a subjective variable, it is a challenge to directly manipulate confidence. Rather, the study of confidence has relied on correlational approaches, indirectly manipulating confidence by varying the available decision evidence. Here, we tested whether we can directly manipulate confidence by appropriating the classical psychophysical method of sensory adaptation to confidence adaptation.

We tested 100 observers’ metacognitive sensitivity in a forced-choice confidence paradigm: Participants were asked to choose which of two orientation discrimination decisions was more likely to be correct. This followed a long sequence of numerosity and colour discrimination decisions, in which task difficulty was manipulated to either be varied (baseline), relatively difficult (low-confidence adaptation), or relatively easy (high-confidence adaptation).

The presented stimuli, orientation discrimination sensitivity, and thus the underlying decision evidence relevant for evaluating confidence, was unchanged following each of these prolonged exposure sequences of numerosity/colour discrimination decisions. Yet, we found a systematic change in metacognitive sensitivity following prolonged exposure to high/low confidence decisions compared to baseline, consistent with cross-task confidence adaptation.

The pattern of this change in metacognitive sensitivity disambiguates models of the neural coding scheme of confidence, suggesting a dual-channel coding scheme involving tuning to both increasing confidence and increasing uncertainty. Greater channel responsiveness and more channel overlap predicted better metacognitive sensitivity. This cross-task adaptation provides direct evidence for general metacognitive computations in human observers, and demonstrates adaptive resource allocation for metacognitive processes.

When performing a piece, musicians combine automatic motor control with conscious monitoring. For example, a guitarist may monitor their finger placement, or rely on automatic movements to monitor the sound produced. The Focus of Attention (FOA) effect describes the tradeoff between focusing internally (on finger positioning), or externally (on the resulting sound): focusing externally enhances motor performance compared to focusing internally. While described in movement sciences, the underlying mechanisms of this effect remain poorly understood.

This study examined whether individual differences in metacognitive ability—the capacity to recognize successful cognitive performance— explain FOA effects in music performance. We hypothesized that young adults would perform better under an external focus of attention, and explored whether differences in metacognitive ability to monitor internal (movement-related) versus external (outcome-related) parameters explain this effect. Amateur guitarists played a melody under three FOA conditions (no instruction, internal focus, external focus). Performance was assessed using semi-automated Music Information Retrieval techniques for pitch and rhythm accuracy. Metacognitive ability was measured using m-ratio in a novel paradigm requiring participants to monitor their movements on a guitar neck under visual (internal) and tonal (external) tasks.

Contrary to predictions, FOA conditions did not significantly affect performance, nor did metacognitive ability interact with FOA. However, participants showed higher metacognitive accuracy when monitoring visual (internal) compared to tonal (external) aspects. Psychometric modeling assessed the sources of these differences. These findings challenge external FOA benefits in motor performance, highlight metacognitive accuracy's variability across perceptual domains, and demonstrate novel measurement paradigms for motor skill acquisition insights.

Metacognition can be used to study various non-ordinary states of consciousness like meditation, lucid dreaming and ecstatic epilepsy. Metacognition presents a declarative facet, defined as the declarative knowledge on how cognition works (e.g. epistemic beliefs), and a procedural facet, defined as monitoring and control functions for cognition to modulate itself. Within the procedural facet, the concept of metacognitive feelings (e.g. tip of the tongue feeling) related to cognitive actions (Proust, 2013) is described, giving special attention to “insights”, also defined as metacognitive outcome-oriented feelings, such as the "eureka" or "A-ha Erlebnis” experience. In the context of the modern resurging interest in Psychedelic Assisted Therapy (PAT), metacognitive feelings are linked to psychedelic induced non-ordinary states of consciousness by proposing a neuro-cognitive framework to study their interaction. In particular, a parallel between the “metacognitive feeling of epistemic gain” and the noetic quality of the mystical-type experience potentially occurring in PAT is described. By discussing the interplay between metacognition processes and psychedelic-induced non-ordinary states of consciousness, we can gain understanding into fundamental mechanisms of the nature of consciousness and of self-awareness.

Metacognitive self-monitoring is thought to be largely domain-general, with numerous prior studies providing evidence of a metacognitive g-factor. The observation of shared inter-individual variance across different measures of metacognition does not however preclude the possibility that some aspects may nevertheless be domain-specific. In particular, it is unknown the degree to which explicit metacognitive beliefs regarding one's own abilities may exhibit domain generality. Similarly, little is known about how such prior self-beliefs are maintained and updated in the face of new metacognitive experiences. In this study of 330 healthy individuals, we explored metacognitive belief updating across memory, visual, and general knowledge domains spanning nutritional and socioeconomic facts. We find that across all domains, participants strongly reduced their self-belief (i.e., expressed less confidence in their abilities) after completing a multi-domain metacognition test battery. Using psychological network and cross-correlation analyses, we further found that while metacognitive confidence exhibited strong domain generality, metacognitive belief updating was highly domain-specific, such that participants shifted their confidence specifically according to their performance on each domain. Overall, our findings suggest that metacognitive experiences prompt a shift in self-priors from a more general to a more specific focus.

Recent years have seen a substantial proliferation of computational models of confidence and metacognition. The most widely used model, although mostly implicitly assumed by metacognition researchers without empirical testing, is the Independent Truncated Gaussian model (ITG). ITG is the basis of the popular meta-d′/d′ method used to quantify metacognitive ability. However, previous modelling studies of perceptual confidence have not included ITG in formal model comparisons. The present study compares model fit of ITG to seven different alternative models of confidence and metacognition all derived from signal detection theory in a reanalysis of four previously published experiments and one new experiment, (i) a masked orientation discrimination task, (ii) a random-dot motion discrimination task, (iii) a low contrast orientation discrimination task, (iv) a dot numerosity discrimination task, and (v) a low contrast number discrimination task. I show that in all five experiments, alternative models provide a better fit than ITG: In the dot numerosity discrimination task, the best fit is achieved by the signal detection rating model. In the low contrast orientation discrimination task, the logistic noise model performs best. In the other three experiments, the best fit is achieved by either the weighted evidence and visibility model or the logistic weighted evidence and visibility model, implying that at least two sources of evidence are required to account for perceptual confidence, one related to the discrimination judgment, and one related to the reliability of the perceptual evidence. I discuss implications for the measurement of metacognition.

Obsessive-Compulsive Disorder (OCD) is a psychiatric condition characterized by recurrent intrusive thoughts (obsessions) and repetitive, impairing behaviors (compulsions). Another key feature of OCD is pathological doubt. A common hypothesis suggests that patients distrust their own memory and perception, leading them to engage in compulsive checking rituals to make a decision. To investigate these (meta)cognitive processes, we asked participants with OCD (N = 10) and healthy controls (N = 11) to perform a numerosity discrimination task where they indicated which of two boxes contained more dots. Participants reported both their decision and confidence (1 to 4) on a combined scale. Importantly, participants could request to replay the stimulus up to five times before deciding, allowing us to investigate information-seeking behavior similar to checkings in OCD. To assess what performance and confidence would have been when a replay was requested, 25% of these replay requests were randomly rejected, forcing participants to respond without additional information. We found that OCD patients used the replay option less appropriately than healthy participants. Indeed, contrarily to healthy participants, patients with OCD performed similarly regardless of whether they answered or asked to replay the stimulus. Furthermore, during refused replay requests, confidence ratings scaled better with accuracy for participants with OCD than for healthy controls. Our results suggest that checking behaviors may not be truly beneficial to decision-making and confidence in OCD. These results might enhance our understanding of what drives compulsive behaviors.

In standard perceptual metacognition paradigms, participants perform a simple perceptual task and report their subjective confidence in the perceptual decision. Typically, it is assumed that confidence ratings collected under these paradigms reflect the subjective probability of making the correct decision, and many models propose some form of probabilistic inference to underpin the calculation of these probabilities. Here we tested this assumption by asking whether confidence reports are distinguishable from reports of task difficulty, since the latter can be estimated heuristically using cues such as fluency, stimulus visibility (or uncertainty) and response speed.

This study drew on data from the Perception Census, a large-scale citizen science project in which participants (>30K) completed numerous perception tasks and questionnaires online. We concentrated analyses on low-, mid- and high-level perceptual tasks in which confidence or subjective difficulty ratings were reported alongside objective decisions. Using a pre-registered machine learning approach, we tested whether reports of confidence and of subjective difficulty reflect the same underlying information. To the extent that they do, we should be able to “cross-predict” between them: algorithms trained to predict confidence from objective choice and stimulus information should generalise to predict (untrained) subjective difficulty, and vice versa.

Results showed that reports of confidence and subjective difficulty indeed reflected the same information, but only for mid- and high-level tasks. This highlights a potential limit to applying normative models of perceptual metacognition to complex tasks, and underscores the importance of considering heuristic strategies in understanding the formation of confidence across different perceptual domains.

Humans and other creatures seek information to improve their cognition and behaviour. Theories in cognitive neuroscience, developmental psychology and animal cognition tend to assume a strong connection between information seeking behaviour and explicit metacognition – conscious introspection about our mental states and subjective metacognitive feelings like confidence or uncertainty. However, recent developments in computational neuroscience have stressed that metacognition and uncertainty are not equivalent, and many forms of uncertainty may be monitored in the brain without generating subjective metacognitive feelings. Here, across a series of experiments in adult humans, we show that information seeking and subjective confidence are controlled by distinct forms of uncertainty. In particular, information seeking (but not confidence) is controlled by uncertainty in sampled sensory evidence while confidence (but not information seeking) is controlled by uncertainty caused by decision boundaries. This dissociation suggests that separate computations in the mind and brain shape confidence and information seeking: undermining the idea that information seeking behaviour always depends on conscious introspection into our own states of mind.

Blindsight refers to above-chance residual visual discrimination in the blind field of patients with primary visual cortex lesions. While traditionally considered a form of unconscious vision, some patients report subjective visual experiences distinct from normal sight, and others exhibit residual awareness without measurable discrimination abilities. These dissociations challenge existing models of visual consciousness and highlight the need for a taxonomy that integrates subjective experience.

Here, we present an EEG study with multivariate pattern analysis to investigate the neural correlates of residual vision and metacognition in a cohort of cortically blind patients and matched controls. Participants performed motion detection and motion direction discrimination tasks in their blind and intact hemifields while providing both objective (d') and subjective (meta-d') responses. By prompting for confidence rather than visual awareness, our protocol accounts for the often amodal nature of residual sensations and tests whether blindsight reflects disrupted perceptual metacognition rather than strictly unconscious processing. Preliminary results suggest that correctly detected stimuli in the blind field are associated with widespread frontal activation, aligning with predictions from the Global Neuronal Workspace theory.

Our findings aim to refine the characterization of residual perceptual capacities in blindsight and provide insights into the interplay between metacognitive awareness and non-conscious vision. We discuss implications for theories of consciousness, particularly regarding the role of global neural dynamics in shaping subjective experience.

Metacognition plays a crucial role in learning and decision-making across the life span. However, the neurophysiological substrates of metacognitive processes remain poorly understood. Recent but limited evidence suggests that higher activity in the locus coeruleus-noradrenaline (LC-NA) system may subserve post-response monitoring processes such as error awareness. In the present study we aim at investigating whether pupil dilation, as an indirect and non-exclusive measure of LC activation, is a robust biomarker of metacognitive monitoring across tasks and cognitive domains. Thirty-nine young adults performed multiple experimental tasks during eye-tracking recordings. An adapted version of the error awareness task was used to replicate the findings on the link between pupil size and error awareness. Moreover, a novel battery of tasks testing mental rotation, visual perception and working memory was employed to test the association between pupil dilation and confidence judgements across cognitive domains. In line with previous evidence, it was found that pupil size was larger after aware errors as compared to unaware errors. Furthermore, pupil dilation during confidence ratings was found to be greater for high confidence judgements in mental rotation, visual perception and working memory tasks. The study findings suggest that pupil dilation can be used as robust marker of metacognitive processes and that noradrenergic function may support metacognition in a domain-general fashion.

Some of the most profound human experiences involve encounters with invisible others—feeling the presence of the dead, being possessed by a spirit, hearing the voice of God. Our study investigated how the brain enables such experiences and how cultural practices shape this capacity.

Using fMRI, we examined two distinct groups: 22 secular tulpamancers, who cultivate relationships with imaginal companions called tulpas, and 25 charismatic Christians who practice speaking in tongues, a form of prayer in which they release control of their tongue to allow God to speak through them. As predicted by our pre-registered hypothesis, both groups showed deactivation of the pre-supplementary motor area (pre-SMA) when surrendering agency to these invisible others. This region is centrally implicated in voluntary action and the sense of agency.

We then trained a machine-learning classifier on SMA patterns during the tulpa possession condition, in which the tulpa took control of the practitioner’s hand. Remarkably, this tulpa classifier generalized to our prayer group, predicting above chance when charismatic Christians were speaking in tongues compared to engaging in regular prayer. Furthermore, the SMA showed the highest cross-decoding accuracy of any region in the brain. This cross-context decoding, spanning from secular to religious belief systems and from writing to speech output, reveals a consistent neural signature of surrendering agency to an invisible other.

Our findings demonstrate that humans can deliberately modulate core neural mechanisms of agency and suggest that shared brain processes may underlie experiences of invisible presence across religious and secular contexts.

Introduction: The study of altered states of consciousness, particularly minimal phenomenal experience (MPE) – a content-reduced, non-dual state – is essential for advancing our understanding of consciousness. However, MPE research has mainly relied on skilled meditators who are able to reliably access the target state, thereby limiting the pool of participants for empirical investigation. To address this limitation, this pilot study explores Floatation-Restricted Environmental Stimulation Technique (Floatation-REST) as an experimental paradigm to induce MPE under controlled conditions.

Methods: Participants with meditation experience underwent two conditions: (1) Floatation-REST alone and (2) Floatation-REST combined with nondual meditation. Post-intervention assessments involved semi-structured interviews guided by micro-phenomenological methods, while thematic analysis based on the MPE-M92 questionnaire identified key themes related to MPE phenomenology.

Results: In condition 1, participants reported key MPE indicators, such as the dissolution of body boundaries, loss of the sense of time and space, and absence of self-consciousness. In condition 2, despite the unfamiliarity of meditating supine on water, the sensory-reduced environment markedly enhanced MPE induction, deepening and stabilizing the experience.

Conclusion: These findings demonstrate that Floatation-REST effectively induces MPE phenomenology, with nondual meditation significantly amplifying its depth and consistency when combined. This approach offers a practical alternative to traditional methods, potentially broadening the participant pool to include less experienced meditators or non-meditators, while providing a systematic framework for empirical MPE research.

Background: Challenging psychedelic experiences, particularly those characterised by intense anxiety about the loss of self-awareness (known as Dread of Ego Dissolution, or DED), may compromise therapeutic outcomes. These states share similarities with the 'felt sense of anomaly' reported in dissociative experiences. This study examines whether the Anomalous Experience of Self (CEFSA-S) from the Černis Felt Sense of Anomaly scale predisposes individuals to DED, and explores voice analysis as a screening tool.

Methods: Participants (N=26) completed CEFSA-S at baseline, followed by voice journal recordings over two weeks prior to 5-MeO-DMT administration (12mg). DED was assessed within 3 hours after the experience. Voice recordings were analysed using OpenSMILE (eGeMAPSv02), extracting acoustic features. Relationships between CEFSA-S, voice characteristics, and DED were examined using correlations, hierarchical regression modelling and commonality analysis.

Results: CEFSA-S correlated negatively with jitter variability (r = -0.60), F2 bandwidth variability (r = -0.59), and voice intensity (r = -0.65). In hierarchical regression modelling, CEFSA-S alone explained 27.6% of DED adjusted variance, increasing to 52.5% when combined with voice features. Commonality analysis showed CEFSA-S uniquely explained 26.9% of variance, while voice features 35.3%. F2 bandwidth showed positive shared effects (3.61%, 9.52%), while voice intensity showed negative shared effects (-23.1%).

Conclusions: Anomalous self-experience leaves a measurable "acoustic fingerprint" in speech, with voice features capturing immediate emotional states through prosodic variations, while CEFSA-S reflects cognitive evaluation of self-experience. This complementarity between implicit and reflective measures provides a behavioural marker of anomalous self-experience and offers an approach to risk assessment in psychedelic medicine.

The phenomenon of "pure awareness", central to many contemplative traditions, has recently attracted scientific interest for its relevance to the study of consciousness. In this work, we investigate pure awareness through the algorithmic agent model, a computational framework with roots in algorithmic information theory. This framework proposes that agents build compressive models of the world for evolutionary success. Structured experience arises from running such models, thus linking phenomenology with computation and offering insights into the emergence of pure awareness in both natural and artificial systems. We propose that pure awareness, as a stand-alone phenomenon, may correspond to minimally structured experiences achieved through meditation, psychedelics, or other deconstructive practices. A key hypothesis is that the phenomenology of pure awareness arises from a specific model: the agent's model of its own modeling process. Importantly, the agent's recognition of the modeling process can occur alongside other phenomenal content (as in non-dual awareness) and is associated with stable changes in valence computation, potentially reducing suffering. We will explore some empirical hypotheses in terms of neurobiology and neurophenomenology based on whole-brain computational models. Our exploration offers new insights into consciousness science by examining the minimal possible experiences for an agent and helps us better understand the mechanisms and constraints involved in facilitating such transformative experiences.

This study used a mirror-gazing paradigm to systematically investigate anomalous self-experiences akin to those seen in dissociative disorders, their relations to psychological traits and physiological underpinnings. Thirty-five participants, aged 20 to 46 , completed two 10-minute mirror gazing tasks (MGT) in dimly lit and well-lit conditions, randomly. During the MGT we measured participants' ECG and EEG. We also measured dissociative states, perceptual anomalies experienced during MGT, and psychological traits.

Results showed that mirror-gazing in dim-lighting significantly increased dissociative symptoms and experiences of perceptual anomalies, with more pronounced effects linked to higher baseline depression and anxiety. Furthermore, when investigating the EEG, we focused on the theta or beta frequency bands. The most significant correlations were global theta reduction associated with fewer symptoms in dim lighting, and increased theta linked to more perceptual anomalies and dissociation in well-lit conditions. Alpha global increase in dim lighting but decreased during mirror-gazing. Decreased alpha in dim lighting was correlated with more perceptual anomalies and dissociation, while decreased alpha in well-lit conditions was associated with fewer anomalies and dissociation. When analysing the ECG, we were able to find that increased heart rate was associated with reduced anomalies and dissociation. Lastly, this study was able to replicate a lot of findings done by previous research, moreover it demonstrated that the MGT is also able to produce dissociative states and anomalous experiences in fully lit condition, thus demonstrating that the dissociative states are due to the MGT and not necessarily dependant on external factors such as sensorial deprivation.

Chronic pain conditions, which affect approximately one in five individuals worldwide, frequently cause visuo-proprioceptive distortion of affected body parts. Virtual reality (VR) is emerging as a promising digital therapeutic due to its capacity to induce immersive bodily illusions and modulate sensory processes related to body perception. By targeting the distorted body perceptions characteristic of certain chronic pain conditions, VR interventions can elicit meaningful analgesia. In parallel, psychoplastogens such as 5-MeO-DMT are receiving increasing attention for their ability to promote neuroplasticity by reshaping maladaptive functional connectivity, suggesting considerable potential in treating chronic pain.

We hypothesize that psychoplastogen-enhanced neuroplasticity will amplify VR’s ability to reorganize self-perception and reduce pain. In the first phase of our proposed experiments, healthy participants will complete VR-based psychophysics tasks concerning their physical and virtual bodies, allowing us to examine changes in body perception as a correlate of acute pain ratings. Building on these findings, the second phase will assess the separate and combined analgesic effects of VR and 5-MeO-DMT in healthy controls and patients with chronic pain, capturing both experiential and physiological markers through subjective measures and neuroimaging techniques in a neurophenomenological framework. By systematically investigating this synergy, the project aims to inform fundamental theories of bodily self-perception and pain modulation. Moreover, the ability of immersive bodily illusions to transiently reconfigure the phenomenal self-model can expand our understanding of consciousness by illuminating the dynamic interplay between neural plasticity, perception, and the sense of corporeal self.

Previous work showed that the Joint Sense of Agency (JSoA) –the sense of conscious control experienced by humans when acting with others– depends on the type of embodied agent we are acting with. Yet, the effect of interacting with human versus artificial bodies on the human body remains an open question. Here we investigate the effect of Depersonalisation (DP) – a condition that makes people feel detached from their self and body - on embodied Joint Sense of Agency in Human/Human versus Human/Artificial other dyads. We have designed a novel Joint Simon Task in Virtual Reality where 100 participants with High versus Low Levels of DP can embody either a Human avatar or a Social Humanoid Robot ‘Pepper’ avatar, then performing the Joint Simon task either together with a Human avatar or a ‘Pepper’ avatar as co-agent.

We hypothesize that participants with High DP will show higher Joint Simon Effect when embodying a robotic ‘Pepper’ avatar. This is because people who feel less connected to their bodies and feel as ‘machines’, or ‘automata’ may develop a higher JSoA while doing a task with another robotic body as opposed to a human body. Our study investigates for the first time the effect of human embodiment on Joint Sense of Agency in Human versus Robotic ‘Pepper’ avatar in Virtual Reality. A better understanding of how feelings of being (dis)connected from one’s body impacts the way people feel (dis)connected from human and artificial others may help us better design human/artificial agents interactions.

Dissociative symptoms are increasingly recognized as transdiagnostic across mental disorders and may result from trauma-related disruptions in interoceptive processing and multisensory integration. However, empirical studies in this area remain scarce. This project addresses a critical gap by investigating the relationship between interoceptive-exteroceptive integration and dissociative symptoms (e.g., depersonalization, derealization, amnesia, absorption, identity fragmentation, and somatoform dissociative symptoms) in the general population.

The study consists of two phases. First, an online screening survey assesses dissociative symptoms using self-report measures, including the Dissociative Experiences Scale, Somatoform Dissociation Questionnaire, Cambridge Depersonalization Scale, and Dissociative Symptoms Scale, along with additional measures of trauma-related symptoms and interoception. Participants with varying levels of dissociative symptoms are invited to the second phase.

In the second phase, participants complete two versions of the simultaneity judgment (SJ) task—cardio-auditory and audio-visual—to compare intero-exteroceptive and extero-exteroceptive integration (i.e., temporal binding window [TBW] width) between individuals with varying levels of dissociative symptoms. We hypothesize that TBW width varies with dissociative symptom type and severity. The study is in the data collection stage, and preliminary results will be presented.

This research has theoretical and practical implications, providing insights into the mechanisms underlying altered states of consciousness associated with dissociative symptoms and informing therapeutic interventions targeting multisensory integration in dissociative states.

The experience of home – a silent, transparent background feeling of familiarity, belonging, and stability that shapes our everyday life - is grounded in our embodiment. This relatedness is most profoundly revealed when it is lost. Depersonalization (DP) is a condition characterized by a distressing sense of detachment from the self, body, and world, often accompanied by disruptions in bodily self-awareness and multisensory integration. As a result, individuals with DP report that experiences are permeated by a sense of unreality, alienation and uncanniness. Embodiment and home appear altered.

This study synthesizes research on DP and healthy individuals to develop an embodied intervention aimed at restoring this lost sense of embodiment. By synchronizing bodily movements (cardiac rhythm) and bodily actions (gait) with real-time cardiac feedback, we investigate whether such alignment enhances embodiment, as measured by a Graphesthesia task, through improved multisensory integration. This approach builds on the understanding that bodily states and regulatory processes are foundational to selfhood and the experience of home.

The findings have implications for both theory and practice, offering a novel perspective on the role of embodiment in psychopathology, suggesting dynamic, embodied therapeutic approches. This work aligns with the conference theme by contributing to contemporary discussions on consciousness, selfhood, and the bodily foundations of experience, with potential significance for both national and international research on DP and related disorders.

A tingling in the body, a sense of calm, relaxation and euphoria, are all sensori-emotional experiences often evoked by a phenomenon called Autonomous Sensory Meridian Response (ASMR) (Barratt & Davis, 2015; Engelbregt et al., 2022; Fredborg et al., 2021; Poerio et al., 2022). ASMR is reported to be an involuntary and atypical physiological response to certain audio, visual and interpersonal triggers often found in ASMR videos (Barratt & Davis, 2015). This study developed and validated a bank of Autonomous Sensory Meridian Response (ASMR) stimuli to address inconsistencies in ASMR research methods. ASMR videos and matched control videos were selected, and participants (N=553) rated them in an online survey. ASMR videos received significantly higher ratings than control videos on the ASMR-15 scale (Roberts et al., 2018), confirming the stimuli's effectiveness in eliciting ASMR responses. Significant differences were found across all subscales, with the sensation subscale showing the largest effect. The study also revealed potential expectancy effects, as control videos presented first received higher ratings, suggesting that prior expectations may influence ASMR experiences. Acoustic analyses of both ASMR and control videos are ongoing, with preliminary results indicating that the ASMR stimuli align with known sound characteristics. This study provides a reliable set of ASMR stimuli for future experimental research, offering a valuable resource to explore ASMR's physiological and psychological effects, as well as its potential applications in mental health interventions.

The full-body illusion (FBI), in which individuals perceive a virtual avatar as their own body, is a key phenomenon in self-recognition research. However, no standardized quantitative index exists to measure its strength. This study explores the validity of postural sway, or changes in the center of pressure (CoP) of the participant’s body, as an objective measure of FBI strength. Previous studies on the rubber hand illusion, a related phenomenon, demonstrated that movements of a fake hand’s fingers can elicit corresponding movements in the participant’s real fingers. Based on this, we hypothesized that changes in an avatar’s posture during the FBI would induce corresponding shifts in the participant’s CoP. In this study, the illusion was induced by simultaneously stroking the backs of the participant and the virtual avatar. We prepared two conditions: one in which the illusion was induced by avatar tracing the participant’s standing body movement, and another in which the illusion was disrupted by the avatar’s movements unrelated to the participant's intentions. In both conditions, the avatar's posture was suddenly and unexpectedly leaned forward six minutes after the illusion induction. We measured the CoP using a stabilometer. The result showed that in the illusion-inducing condition, the avatar's forward lean caused a significant forward shift in the participant's CoP. Thus, the results suggest that changes in the CoP could be an objective index of the FBI strength, reflecting the subjective experience of embodiment.

Full-body illusion (FBI) is the phenomenon of experiencing a virtual avatar as one’s own body. Some studies have shown that avatar users modify their behavior corresponding to their avatars’ features and that the degree of the FBI is related to these modifications. While previous research has focused on the effect of avatar’ s features of appearance, little is known about the effect of the avatar’s narrativity or internal characteristics. This study examines the effect of an avatar’s narrativity induced from contexts on the FBI.

In this experiment, 32 healthy participants experienced an avatar of the strong artificial lifeform in VR. Prior to the VR experience, half of the participants listened to a positive narrative, in which the avatar used its power to protect people, and the other half listened to a negative narrative, in which it used its power inappropriately. The degree of personal familiarity, social desirability, and activism were measured using a questionnaire to evaluate the impressions of the avatar. Sense of Ownership (SoO) and Agency (SoA) were assessed through questionnaires.

A Man-Whitney U test showed that the participants who heard the positive narrative experienced a greater SoA compared to those who heard the negative narrative (p=0.02). Moreover, correlation analysis showed that the factor of the personal familiarity in the positive narrative was related to the greater SoA (r=0.51, p=0.04).

These results suggest that the positivity of the avatar’s narrativity may influence user's SoA while experiencing the avatar.

The mirror self-recognition test is a standard for assessing self-awareness, but it only examines the visual aspect of the Self. The aim of our study is to concurrently explore the development of visual, bodily, and temporal aspects of self-awareness in the second and third years of life. We also investigate whether children with greater exposure to their Self-image (through mirrors, photos, and videos) perform better in behavioral self-awareness tests, as reported by their parents.

Three self-awareness tests completed in the second year of life (first measurement): 1) the Mirror Self-Recognition Test (visual aspect), 2) the Body-as-an-Obstacle Test (bodily aspect), and 3) the Delayed Self-Recognition Test (temporal aspect).

Three self-awareness tests completed in the third year of life (second measurement): 1) the Shadow-Recognition Test (visual aspect), 2) the Door-Choice Test (bodily aspect), and 3) the Delayed Self-Recognition Test (temporal aspect).

We studied 60 children in their second year and found that 73% passed the mirror self-recognition test, and 42% passed the body-as-an-obstacle test. Unlike Povinelli et al. (1996), 27% of 2-year-olds passed the delayed self-recognition test, which is intriguing from a chronosystem perspective. Exposure to Self-image via mirrors, photos, and videos did not correlate with self-awareness, indicating other factors may influence its development. After one year, the same children will be tested again. Findings will be discussed in the context of the chronosystem, considering how self-awareness evolves over time, influenced by the increasing contemporary exposure to Self-image.

Depersonalization is a condition that makes people feel detached from one’s self, body and others. The representation of one’s own face is a salient bodily aspect of self-awareness and identity, and empirical evidence suggests that individuals with depersonalization disorder experience disrupted perception of their faces when viewing themselves in photographs or in the mirror, which has been corroborated by first-person reports. However, no study had yet explored the state of long-term self-face representations stored in visual memory in the context of depersonalization. By visualising how individuals saw themselves “in the mind’s eye”, this study provides the first empirical evidence for a relationship between depersonalization symptoms and impairments in self-face representation. Individuals reporting more frequent and intense depersonalization symptoms had lower self-face representation accuracy, but somewhat counterintuitively, also higher precision and informational content of this representation. These results suggested that individuals with high depersonalization were representing a distinct, but inaccurate, facial identity as themselves. The self-face representations of high-depersonalization participants were rated as visibly more emotionless and younger in appearance than those of low-depersonalisation participants, according to independent raters. These features were found to be specifically related to aspects of depersonalization symptomatology related to anomalous memory experiences. Finally, an intriguing role of interoceptive sensibility was revealed in both self-face representational accuracy and in depersonalization symptoms. These novel results highlight the link between interoceptive and exteroceptive bodily self-awareness and memory processes as important in those individuals who experience distressing feelings of being detached from one’s self, body and the world.

The ability to perceive one's body by integrating internal (interoception) and external (exteroception) percepts is fundamental to bodily self-consciousness (BSC). In schizophrenia spectrum disorders (SSD), disruptions in BSC, termed self-disorders, are prevalent and associated with various symptoms. While the exteroceptive alterations have been extensively investigated, the interoceptive domains remain underexplored. We hypothesize an altered interoception on three levels in SSD compared to healthy controls (HC): self-reported interoceptive sensibility, objective interoceptive accuracy, and neural correlates of interoception. We are recruiting individuals with SSD and age- and gender-matched HC (18-65 years). Participants complete an EEG experiment (with ECG and respiration monitoring) that includes resting-state recordings and a heartbeat counting task (HCT), followed by interoceptive sensibility questionnaires. Patients additionally undergo functional and structural MRI scanning. The HCT measures cardiac interoceptive accuracy, while EEG-derived Heartbeat Evoked Potentials (HEPs) give insight into the neural correlates of interoception.The preliminary analysis (nSSD = 27, nHC = 12) revealed no group differences regarding interoceptive sensibility, interoceptive accuracy, or HEP amplitudes. HEP amplitudes were not modulated by interoceptive attention during the HCT condition. Data collection is ongoing, and the small sample size, particularly for HC, limits statistical power. While HCT has methodological limitations (e.g., reliance on heart rate beliefs), strict instructions were implemented to mitigate this. Notably, the HCT also evaluates the impact of interoceptive attention on HEPs. Future analyses will explore symptom correlations, respiratory interoception, and MRI data focusing on insula. This research enhances the understanding of BSC in SSD and carries potential implications for body-oriented clinical interventions targeting interoception.

Human bodies are highly dynamic systems, constantly moving both inside (e.g., heartbeats) and in the outside world (e.g., footsteps, walking) to secure survival. Bodily self-consciousness crucially depends on the ability to process and couple this dynamic sensory information. Depersonalisation (DP henceforth) is a common phenomenon that makes individuals feel detached from their bodies and as though they do not fully exist, significantly altering their self-consciousness. Previous work has shown that bodily self-consciousness is not rigid; rather, it is constantly updated through dynamic self-related sensory feedback.

We conducted a study investigating the dynamic coupling between bodily movements from inside the body (i.e., cardiac signals) and bodily actions in the world (e.g., walking) in 60 participants with high and low occurrences of DP. Participants were invited to walk while wearing headphones displaying their natural footstep auditory feedback across frequency bands in three conditions (control, high frequency, low frequency), following a procedure from Tajadura-Jiménez and colleagues (2015). In parallel, we recorded participants’ cardiac signals in real time, as well as gait biomechanics, which were used as an implicit measure of changes in perceived body weight across conditions.

We found that in typical controls walking pace is significantly coupled with the systolic cardiac phase, whereas in people detached from their bodies (high DP) this coupling is absent. Our study reveals, for the first time, that real-time cardiolocomotor coupling is altered in DP individuals, with important implications for dynamic body-based potential therapy.

Background:

Voluntary action is linked fundamentally to core aspects of human conscious experience, including selfhood. Recent evidence highlights the crucial role of interoception in shaping the sense of self, with pathophysiological implications across neurological and psychiatric disorders. However, the association between interoception and voluntary action has remained insufficiently characterised.

Methods:

Thirty non-clinical participants engaged in a series of tasks to elucidate this relationship. Interoception was assessed at a perceptual level using convergent methods that included heartbeat detection (tracking and discrimination) tasks, and a self-rated questionnaire. Specifically, participants were ascribed derived measures of interoceptive prediction error, as conceptualised within a predictive coding framework. Electroencephalograms (EEGs) were recorded during the Libet task, in which the participants voluntarily chose to make actions (button press), while noting the time when they first became aware of their intention/decision to act (W-judgment). A Tapping task served as a control, and both tasks evoked EEG readiness potentials.

Results:

During the Libet task, notably over the period preceding the W-judgment time, participants who scored lower on measures of interoceptive prediction error showed greater readiness potential amplitude. This association was more pronounced in comparison to the Tapping task and remained after controlling for potential confounding factors.

Conclusions:

Our findings suggest the mitigation of interoceptive prediction error augments voluntary action and the accompanying sense of voluntariness. Conversely, a greater tendency for interoceptive prediction errors suggesting unresolved interoceptive noise, may hinder the development of voluntary action and experience of voluntariness.

This research explores the experience of embodiment and emotions in healthy adults (HC), long-term meditators (LTM) and people with depersonalisation (DP) experiences. Both LTM and DP subjects usually report the feeling of being detached from one’s experiences. However, while in LTM self detachment experiences are typically positively valenced, in DP they are usually negatively valenced. In order to assess these experiences we utilize the emBODY task (Nummenmaa et al., 2014), a computer-based topographical self-report tool for monitoring emotion-triggered bodily sensations; as well as three questionnaires: Multidimensional Assessment of Interoceptive Awareness (MAIA) (Mehling et al., 2018); Five Facet Mindfulness Questionnaire (FFMQ-15) (Baer et al., 2008); Emotion Regulation Questionnaire (ERQ) (Gross & John, 2003). The questionnaires will provide a rich comprehension of the dimensions of embodiment and emotional processing of these populations. The hypothesis of this research is that the bodily sensations deriving from the triggered emotions will be represented differently across the groups, with lower activation in body maps of DP participants, higher activation in LTM, across emotions, compared to HC. It's anticipated lower activation of body maps for positive emotions in DP participants compared to HC and LTM. Finally, it's expected in the related questionnaires (MAIA, FFMQ-15, ERQ) lower scores in DP participants and higher results in LTM, compared to HC. This research can have significant implications worldwide in clinical interventions for DP by affirming the importance of utilizing mindfulness-based and integrative body-focused therapies; and enhance the theoretical comprehension about consciousness, especially the integration between embodiment, emotions and interoception.

Introduction: Emotional stimuli trigger changes in affective states and conscious experience. This study links variability in emotional responses with self-defining traits such as personality, anxiety and interoceptive awareness - i.e. beliefs, attitudes, and self-reported experiences of bodily sensations

Methods: 297 participants (126 females, aged 18–35 years) viewed 25 video clips featuring negative, positive, disgusting, appetizing, or neutral content. After each clip, participants rated their affect (valence, arousal, disgust, and appetite) and marked body sensations on human silhouettes. Additionally, they completed questionnaires assessing interoceptive awareness, anxiety, disgust sensitivity, mood, and personality.

Results: Hierarchical clustering of questionnaire responses revealed three distinct participant groups. The “Body-Unaware” group, with low interoceptive awareness, showed similar affect ratings across emotions, suggesting reduced emotional granularity. The “Body-Aware” group, characterized by high interoceptive awareness and low anxiety, reported more extensive bodily sensations, indicating heightened emotional engagement. The “High Anxiety” group, marked by elevated anxiety and neuroticism, exhibited stereotypical bodily sensation patterns. Each group also showed unique relationships between affect ratings and bodily sensations, moderated by mood.

Discussion: Our findings underscore the embodied nature of emotions and illustrate how conscious reports of affect vary with individual differences in bodily awareness, emotional processing, and traits that contribute to one’s sense of self. This suggests that awareness of bodily sensations shapes subjective emotional experience and contributes to how individuals construct and interpret their self-identity. These insights promise to advance studies on the interplay between interoception, self-related processes, and emotion regulation in both healthy and clinical populations.

The ability to distinguish between our body and the external world is crucial for our sense of self and interaction with our environment. While research suggests that the perceptual clarity of this boundary can vary, precise methods for measuring body boundary perception remain underexplored. In this study, we developed a psychophysical protocol to directly assess how accurately individuals perceive their body boundaries without visual input. 3D scans of the tested regions were used to determine the true body outline, allowing psychometric functions to be fitted. Additionally, we collected self-reported measures related to mental imagery, embodiment, and the frequency of third-person dreams and memories to examine their influence on body boundary perception.

Our results showed that participants demonstrated millimeter-level precision in identifying their body boundaries, even in regions that are mostly flat, such as the palm, or rarely observed, such as the ankle. However, accuracy varied by body region and individual differences in self-reported experiences. Notably, participants who frequently experience third-person dreams performed significantly worse at the task, whereas those who reported more out-of-body experiences showed greater accuracy in judging their ankle boundaries compared to their hand boundaries—a pattern not observed in other participants. Overall, our findings suggest that the dominant perspective one adopts in the mind’s eye plays a significant role in body boundary perception, influencing both its precision and the regions where this accuracy is altered.

In rubber hand illusion (RHI) experiments, participants report experiences of ownership of a fake hand after it is brushed in synchrony with their own hidden hand, with lower ratings for an asynchronous control. Historically, RHI effects have been proposed to arise from the disruption of multisensory mechanisms. An alternative theory proposes that RHI effects are constructed by participants in accordance with their beliefs about the experimental situation. i.e., attributable to phenomenological control (PC; the ability to change experience to meet situational goals). To date, correlational evidence has provided support for this theory. Here, we report the first causal evidence for and against these competing theories. Fifteen high PC participants (top 10% of PC ability) underwent synchronous and asynchronous RHI brushing procedures in two conditions: first with no information about the expected direction of effects and second with imaginative suggestions for RHI experience in asynchronous and not synchronous procedures. Suggestion reversed the pattern of both ownership ratings and of ‘proprioceptive drift’ (a shift in perceived hand position, which was not directly suggested), with greater effects for asynchrony than synchrony. While a second rated experience of touch referred to the fake hand was modulated by suggestion, there was no evidence for or against reversal. PC dominates over multisensory mechanisms in RHI reports of ownership and proprioceptive drift. That phenomenological control may account for reports of these experiences has important implications for theories of the sense of body ownership which draw on multisensory integration accounts of the RHI.

Recent evidence suggests a role of interoceptive processes in the emergence of a minimal, bodily form of self. Albeit often neglected, such a self is a necessary component of any subjective experience, including the conscious perception of our external environment (exteroception). Still, interoceptive and exteroceptive processing are also frequently assumed to compete with each other. These two seemingly opposing views of the interoception/exteroception dynamic, self-related Facilitation and Competition, have guided largely independent lines of research within the field, yet had never been empirically contrasted to this day. In an EEG experiment on heartbeat-evoked potentials (HEPs), the neural response to heartbeats, we manipulated the self-relevance of a simple audio-tactile stimulus by placing it inside or outside of peripersonal space. The design allows the self-related Facilitation and Competition accounts to produce orthogonal predictions. On one hand, the amplitude of pre-stimulus HEPs over the somatosensory cortex slowed down reaction times and affected the stimulus-evoked response in the same area, indicating Competition for shared neural resources. On the other hand, HEP amplitudes over integrative somatomotor and default-mode network regions predicted the effect of self-relevance on reaction times and the stimulus-evoked response, as predicted by the Facilitation account. Importantly, Facilitation and Competition effects were independent: they were driven by distinct neural sources and at different latencies, and were uncorrelated to each other. We show that spontaneous fluctuations in internal bodily monitoring not only reflect shifts in resource allocation, but also independently index the moment-to-moment integration of ongoing processes with a minimal form of self.

Full-body illusion (FBI) demonstrates the malleability of body ownership through multisensory integration. Although bottom-up processes in the FBI are well-documented, the role of top-down factors remain unclear. This study examined how participants’ interpretations of the relationship between their observing self and a virtual body (considered as their self-body) influenced the third-person perspective FBI. Following the predictive coding theory, we conducted a virtual reality-based FBI experiment with 66 participants, who were randomly assigned to either the integration (viewing a virtual body as their own) or separation (viewing a virtual body as a separate self-body) conditions. Both conditions involved synchronous and asynchronous visuo-tactile stimulation. FBI strength was measured using skin conductance responses (SCR) to threat stimuli and an illusion questionnaire. Results demonstrated significant FBI occurrence in the integration condition, with higher SCR during synchronous versus asynchronous stimulation (Z = 2.662, p = .004, r = 0.463), while the separation condition revealed no significant difference. Comparisons across conditions further highlighted stronger illusions in the integration condition (Z = 1.860, p = .031, r = 0.229), supporting the idea that top-down processes influence participants’ experiences of the illusion. These findings extend the predictive coding theory, demonstrating that body ownership emerges from the interplay between sensory inputs and cognitive predictions. These results enhance our understanding of the cognitive processes underlying embodiment, offering insights into the mechanisms of consciousness as they relate to self-perception and multisensory integration.

Overuse injuries, common in endurance and aesthetic sports, occur from repeated physical stress over time (Clarsen et al., 2013). Preventing these injuries is an emerging research area due to its impact on athletes’ health and performance (Mihalko et al., 2021). While the physical development of overuse injuries is well understood, why athletes continue to train despite pain is unclear. Cavallerio et al. (2016) suggest that elite athletes believe they must endure pain to succeed. Though anecdotal reports claim that athletes perceive pain differently, scientific evidence remains contradictory. Whether athletes feel less pain or simply have a different attitude towards it is not yet understood, and more research is needed to improve pain recognition and communication.

Contextual factors can alter pain perception (Keltner et al., 2006). To explore this, we conducted two electrophysiological (EEG) studies to assess how athletes and non-athletes process pain. Study 1 compared brain responses to pain in 20 athletes and 20 controls, showing no differences in pain-related brain activity (N100 and P200 components). Study 2, still in progress, examines how expectations about pain influence brain responses, with 10 athletes and 5 controls exposed to varying pain intensities and expectations.

Initial findings suggest that athletes perceive pain similarly to non-athletes, implying that interventions should address athletes’ attitudes toward pain. Understanding pain processing and improving pain communication can aid in preventing overuse injuries and benefit athletes’ mental health. The prevailing sports culture that stigmatizes pain disclosure needs to change to promote better mental and physical well-being.

The first-person perspective (1PP), referred to the position and orientation from where “I perceive the world”, is a fundamental aspect of bodily self-consciousness. Typically anchored to the physical body, 1PP is altered in an out-of-body experience (OBE), which can be experimentally induced in virtual reality using conflicting multisensory stimulation (Wu et al., 2024). Because objective measures of changes in 1PP remain limited, we here tested if peripersonal Space (PPS) — a multisensory interface surrounding the body — could index changes in 1PP. Earlier data showed that PPS expands against gravity’s direction (Bufacchi & Iannetti, 2016) and is referenced to the experienced self-location (Noel et al., 2015). Here we tested whether virtual graviceptive cues simulating a down-looking perspective would modulate PPS similarly to physical gravitational cues.

In two experiments with the same 30 subjects, we first confirmed that back PPS, measured in audio-tactile task, extended further backward in prone versus supine position (consistent with Bufacchi & Iannetti’s findings). Next, we examined participants’ PPS while exposing them to visuo-vestibular cues designed to induce an OBE-like state. In line with Wu et al., (2024) showing that congruent visuo-vestibular (vs. visual-only) stimulation elicits stronger disembodied feelings, we found that, despite lying supine, participants in the congruent visuo-vestibular condition exhibited a prone-like PPS. These findings suggest that virtual graviceptive cues modulate PPS, as if their bodies had physically turned into prone position, highlight the importance of vestibular cues in 1PP, and suggest a novel implicit way to quantify 1PP changes.

The sense of body ownership is the feeling that our body belongs to us, and it is a central aspect of our sense of self. It is often studied through bodily illusions such as the rubber hand illusion (RHI), which allows for its manipulation in a non-invasive way. Numerous studies have shown that body ownership relies on the integration of visual, tactile, and proprioceptive signals and that brain regions involved in multisensory perception play a key role in this process. Recent research suggests that Body Scan Meditation (BSM), a Mindfulness meditation practice that involves focusing on one’s bodily sensations, may enhance the ability to sense body ownership. In this study, 37 participants were randomly assigned to either follow recorded BSM instructions or perform an active listening task (control group). We recorded their resting-state electroencephalography (EEG) and had them complete an RHI-based psychophysical task designed to objectively assess body ownership sensitivity while also measuring EEG activity. Participants then engaged in daily ~30-minute BSM or active listening training for 14 days before returning to the lab for a second session, where we again recorded resting-state EEG and repeated the RHI-based task. We found higher theta-band connectivity in the BSM group compared to the control group, suggesting enhanced cognitive control. Additionally, alpha-band connectivity negatively correlated with body ownership sensitivity, while theta-band connectivity showed a positive correlation. Overall, our results suggest that BSM training may strengthen top-down modulation processes relevant to body ownership perception.

In this study, we focused on disembodiment using the avatar’s hand in VR environment. Participants were instructed to move their real hands (real hands; RH) during seeing avatar’s hands, which synchronize with the movement of the RH, in an immersive VR scene (virtual hands; VH). They moved their right RH (rRH) slightly up and moved toward their left RH (lRH). In the first four trials, participants stopped their rRH when the rVH reached to the position above the left VH (lVH). Participants lowered their rRH vertically and felt touch between the rRH and the lRH during seeing that the rVH touch the lVH (simple touch). Following three trials, participants performed almost the identical action to the previous four trials in the different experimental blocks. In a condition, the rRH was displaced 7cm to the right of rVH, so that participants had no touch sensation when they saw that the rVH touch the lVH (loss-of-touch). As control, participants lowered their rVH at the position apart from the lVH with/without 7cm-displacement between the rRH and the rVH (simple swing condition with/without displacement). Afterwards, a set of 2 simple touch condition trials and 3 experimental condition (loss-of-touch/simple swing/simple swing with displacement) trials were repeated 5 times. Participants felt stronger "sense of body loss" in the loss-of-touch condition than the two control conditions. Proprioceptive drift occurred when there was displacement between rRH and rVH regardless of conditions. These results suggest the potential to evoke the sense of body loss through visuotactile incongruence in VR.

It is widely known that self-tickling elicits a reduction of ticklish sensation compared to being tickled by the other person. According to the reports in cognitive science, the sense of agency is considered to be a factor to cause this phenomenon. During self-tickling, the ticklish sensation can be induced when a temporal delay is introduced between stimulation and perception that makes temporal prediction impossible. In addition to this, the present study focuses on haptic prediction, assuming that the accordance of tactile stimuli applying to the body and perceived on the body may attenuate ticklish sensations. In order to test the hypothesis, we designed and developed a force-controllable leader-follower system with a force reflecting bilateral control. Using this system, a self-tickling experiment was performed with ten participants under three conditions: (1) with haptic feedback, (2) without haptic feedback, and (3) with the introduction of a temporal delay. The experimental task involved operating the device to poke one’s own underarm five times. Subjective evaluation was conducted using a questionnaire, assessing ticklishness and sense of agency. The experimental results showed no significant difference in ticklishness scores between the conditions with and without haptic feedback. In contrast, a significant difference in ticklishness scores was observed between the condition with a temporal delay and the condition without a delay. These findings suggest that, in self-tickling, the consistency of stimulation timing plays more dominant to cause ticklish sensations rather than that of haptic information.

Growing evidence suggests a bidirectional relationship between cognitive processes and cardiac activity, with embodied theories of consciousness proposing that this interaction may contribute to, or form the basis of, subjective experience. Intersubject correlation of heart rate (ISC-HR) studies suggest that synchronisation of heart rate across individuals reflects shared cognitive processing. Specifically, ISC-HR is thought to require (or reflect) conscious and attentive engagement with a common stimulus, such as listening to the same story or watching the same movie. However, in one study, we observed significant ISC-HR even when participants (n=54) listened to a scrambled story, suggesting that comprehension of a narrative is not necessary for ISC-HR. Here we report a large-scale follow-up ISC-HR study to further delineate the relationship between ISC-HR and conscious experience. For this, we calculated ISC-HR from pulse oximetry data, provided by the Cam-CAN database, collected whilst healthy participants (n=461) watched the 8-minute Alfred Hitchcock movie: “Bang! You’re Dead”. Additionally, we collected suspense ratings from an additional sample of 100 participants, and measured engagement using an auditory Sustained Attention to Response Task (SART) performed by another 100 participants while watching the movie. By combining these measures of subjective experience with our heart-rate data, we form a clearer picture of the role of brain-heart interactions in conscious experience. Furthermore, we describe the implications for ISC-HR as a simple yet valuable physiological marker of residual consciousness in behaviourally unresponsive patients with disorders of consciousness following severe brain injury.

This study investigated the relationship between autonomic nervous system fluctuations and thought state transitions, focusing on individual differences in cardiac interoception. Participants completed an auditory attention task while their cardiac activities were continuously monitored. Throughout the task, participants responded to random thought probes that assessed their immediate preceding thought content. Participants selected one of eight categories to classify their immediately preceding thought content and then rated various aspects, including task focus, arousal level, and deliberateness on a scale of 0 to 100. Trial-by-trial thought states were estimated using a hidden Markov model, and the temporal dynamics of thought state transitions were analyzed using multinomial logistic regression. This analysis examined how the interaction between current thought states, mean RR intervals, and individual cardiac interoceptive accuracy predicted subsequent thought state transitions. Results demonstrated that, compared to task focus and mind-blanking thoughts, the maintenance of self-related thoughts (self-related episodic and future thoughts) across consecutive trials was significantly predicted by the interaction between trial-wise mean RR intervals and cardiac interoceptive accuracy. Notably, individuals with higher cardiac interoceptive accuracy showed increased probability of maintaining self-related thoughts during periods of accelerated heart rate. Moreover, individuals with higher cardiac interoceptive accuracy showed stronger influence of concurrent autonomic activity than individual thought tendencies on thought state transitions. These findings suggest that cardiac interoception serves as a regulatory mechanism for thought adjustment in response to real-time cardiovascular changes, with particular specificity for self-referential thought. The study provides novel insights into the physiological mechanisms underlying spontaneous thought dynamics.

Near-death experiences are reported by up to 20% of patients who suffer in-hospital cardiac arrest and are often linked to a surge of high-frequency neurophysiologic oscillations in the brain. However, whether this surge enables the capacity to report the experience or influences survival outcomes remains unclear.

Previous studies suggest that brain-heart electrophysiological interactions reflect both prognosis during the initial comatose period and the level of consciousness in the post-comatose phase. Building on these findings, we hypothesized that brain-heart interactions are critical for life-sustaining mechanisms and that the previously observed electrophysiological surge during near-death is linked to these interactions.

Using a rodent model of reversible anoxia, we examined the relationship between heart and brain responses to asphyxia to evaluate whether brain-heart interactions could predict resuscitation outcomes. Continuous electrocardiography and multi-site local field potential (LFP) recordings were performed on sedated, curarized rats under artificial ventilation (n=29). Anoxia was induced by halting ventilation, and resuscitation was initiated by restoring oxygen.

While 45% of the rats recovered cortical activity comparable to baseline, the others succumbed to cardiac arrest. Rats with positive outcomes showed a reduced surge in high-frequency somatosensory responses and an increased surge in cardiac vagal activity. Notably, the degree of coordination between increased cardiac vagal activity and reduced brain activity in the thalamus, hippocampus and somatosensory cortex strongly predicted survival.

These findings highlight the potential of brain-heart interactions to understand the neurobiology of near-death, and their connection to the brain's ability to sustain consciousness after critical episodes.

Space and time perception are fundamentally inescapable features of all our conscious perceptions. Depersonalization (DP) is a very common condition characterized by distressing feelings of estrangement from the self and the external world. Previous work showed that peripersonal space perception is disrupted in schizophrenia but not in DP, on the other hand, time perception is disrupted both in DP and schizophrenia. Why this asymmetry? Is time perception more central to the sense of self than space perception? To better understand the specificities of DP, a comprehensive picture of time and space perception disruptions in this population is needed.

To address this gap, we have conducted an online study (N = 1034) investigating the relation between HIGH DP traits measured by the Cambridge Depersonalization Scale (CDS > 50) and altered subjective experiences of body, time, and space perception. The results demonstrated significant relationships between HIGH DP scores and distorted experiences of time, body and space. Looking at the shared variance between CDS facets “Anomalous Body Experience”, we found that altered body perception remains the most important predictor of spatiotemporal disruption experiences, with a slowing of subjective perception of time.

Altogether, these results suggest that the distorted spatiotemporal experiences observed in DP may mainly come from experiences of estrangement from the bodily self. Thus, the body sets the pace of inner felt subjectivity. Our work calls for further investigations linking DP to disruption of internal/external clock and also in relation to active movements in the world.

Uploading identity refers to the thesis that a biological agent at Time 0 can continue its personal identity in a computational agent at Time 1 (Cerullo, 2015). Though it originates in metaphysical discussions, the concept has far-reaching practical implications (Chalmers, 2010). This work examines the minimal empirical conditions for branching identity to be possible. We begin by considering the spatiotemporal constraints, emphasizing that uploading must be understood as a case of branching, in which any computational agent at Time 1 continues from the biological agent at Time 0 but not beyond that point. We then discuss the concept of personal identity as it pertains to uploading, proposing that the classical notion of numerical identity can be reduced to the sense of personal identity within conscious experience (Klein & Nichols, 2012). Furthermore, we assess different empirical predictions regarding the mechanistic nature of consciousness, showing that uploading identity is not possible under the predictions of all major accounts except for a strong version of computational functionalism (Piccinini, 2010). This version maintains that conscious mental states are equivalent to computational states without relying on emergent properties (Butlin & Long, 2023; Graziano, 2024), suggesting that functional continuity achieved through a fine-grained simulation of the computations underlying the biological agent's sense of personal identity could support uploading. After critically evaluating opposing computationalist perspectives, including implementationalism and mortal computation views (Shiller, 2024; Kleiner, 2024; Dung & Kersten, 2024), we conclude that while uploading identity faces significant empirical constraints, current evidence does not rule out its possibility.

The full-body illusion refers to the phenomenon where users perceive the avatar’s body as their own. Some studies indicated that the illusion leads to a transfer of the avatar’s traits to the users and has an effect on changes in the user’s cognitive performance and behavior such as prosocial behavior. However, it is still unclear whether an avatar can modulate physical performance. This study examined whether embodying a heroic avatar, associated with strength and bravery, enhances physical abilities. Moreover, we investigated the relationship between the physical modulation and changes in subjective experiences.

Twelve male participants (21.0 ± 1.32 years) used a heroic avatar and fought against monsters in VR. To evaluate the changes in physical performance and psychological traits, grip strength, physical perseverance, agility, a sense of balance, and the Courage Measure (CM-J) were measured before and after the VR experience. To examine subjective changes, interviews were conducted during the experience, and post-experience questionnaires were administered. Phenomenological control (PCS-J) was also measured to assess individual traits related to generating illusory experiences. The results showed significant increases in physical perseverance and CM-J scores following the VR experience. Additionally, PCS-J scores were positively correlated with perceived physical changes, and a marginal correlation was found between perceived physical change and CM-J score. Subjective reports also indicated notable changes in the sense of embodying a hero. These findings suggest that embodying a heroic avatar may enhance both physical abilities and courage, influenced by individual phenomenological control.

Human faces contain multiple attributes, including information about configuration, emotion, and intention. While the visual system is specially attuned to detect and recognise human faces, the order in which it extracts meaningful attributes—and the extent to which this extraction occurs in the absence of awareness—is largely unknown. Here, we used a novel LCD tachistoscope that enables visual presentations as brief as 0.002 ms to determine the minimal exposures required for the visual system to extract information about facial configuration and emotional expression by combining psychophysical, electroencephalographic (EEG), and electromyographic (EMG) markers of emotional processing, as well as markers of conscious awareness. Importantly, we combined EEG measures of perceptual emotion processing (i.e., markers that distinguish between emotional and non-emotional stimuli) with EMG measures of embodied emotion processing (i.e., facial mimicry of emotional expressions). We found that both perceptual and embodied markers of emotional processing emerged with 4 ms of exposure—a duration that was sufficiently long to also elicit psychophysical and EEG markers of conscious awareness. Our findings suggest that the processing of facial emotion does not occur in the absence of awareness but rather arises alongside it. Furthermore, our findings challenge the hypothesis that facial mimicry—a phenomenon believed to be critical for empathy and social communication—occurs unconsciously by showing that facial mimicry unfolds as emotional information enters awareness.

Central results in the neuroscience of volition demonstrated that the readiness potential (RP) begins before people consciously decide to move, leading to claims that all voluntary action is initiated unconsciously. However, those original studies focused on arbitrary decisions, and it remains contested whether their results generalize to deliberate choices, which are at the heart of the free-will debate. In particular, some studies found RP differences between arbitrary and deliberate actions while others did not. Yet, critically, prior studies tested the influence of arbitrary and deliberate decisions on the RP using divergent tasks, and no unifying study accounted for these fundamental variations among tasks.

Here we present a comprehensive meta-analysis of EEG studies that measure the RP in volitional tasks. Based on predefined philosophical and methodological criteria, studies were categorized on a continuum: arbitrary, near-arbitrary, ambiguous, and deliberate. RP waveforms were compared across studies at RP onset, peak negativity, and movement time. We found clear RP differences between deliberate tasks and the other categories, demonstrating that RP dynamics differ depending on the nature of the decision.

Our study highlights the urgent need to refine the classification of volitional tasks in RP research. A more precise task-categorization framework will help compare future studies and will contribute to a more nuanced understanding of the relation between conscious deliberation, neural precursors of action, and the emergence of volitional choice.