Biological neural networks perform complex computations to predict their environment, far exceeding the capabilities of individual neurons. Here, we argue that understanding these computations requires considering emergent dynamics—dynamics that make the whole system “more than the sum of its parts.” We examine the relationship between prediction performance and emergence by leveraging quantitative metrics of emergence and modeling environmental time-series prediction within a bio-inspired computational framework called reservoir computing. Notably, three key results reveal a robust bidirectional coupling between prediction performance and emergence: (1) optimizing hyperparameters for performance enhances emergent dynamics, and vice versa; (2) emergent dynamics serve as a highly sufficient and often also necessary condition for prediction success in most environments; and (3) training with larger datasets results in stronger emergent dynamics, encoding task-relevant information. These findings emphasize the importance of emergence-based approaches for studying neural networks—biological or artificial—as they enable network-level insights, complementing traditional single-neuron-based analyses.

Stroboscopic light stimulation (SLS) on closed eyes typically induces simple visual hallucinations, characterized by vivid, geometric, and colourful patterns. A dataset of 898 sentences, extracted from 407 open subjective reports, was recently compiled as part of the Dreamachine programme (https://dreamachine.world/) (Collective Act, 2022), an immersive multisensory experience that combines SLS and spatial sound in a collective setting. Although open reports extend the range of reportable phenomenology, their analysis presents significant challenges, particularly in systematically identifying patterns. To address this challenge, we implemented a data-driven approach leveraging large language models and topic modelling to uncover and interpret latent experiential topics directly from the Dreamachine's text-based reports. Our analysis confirmed the presence of simple visual hallucinations typically documented in scientific studies of SLS, while also revealing experiences of altered states of consciousness and complex hallucinations. Building on these findings, our computational approach expands the systematic study of subjective experience by enabling data-driven analyses of open-ended phenomenological reports, capturing experiences not readily identified through standard questionnaires. By revealing rich and multifaceted aspects of experiences, our study broadens our understanding of stroboscopically induced phenomena while highlighting the potential of natural language processing and large language models in the field of computational phenomenology. More generally, this approach provides a practically applicable methodology for uncovering subtle hidden patterns of subjective experience across diverse research domains. Open-source implementation and an interactive web application are provided to facilitate application of this methodology.

Consciousness science faces the challenge of bridging first-person experience with third-person empirical measurements. Neurophenomenology aims to build such `generative passages' connecting the content of experience with behavioural and neuroscientific data. However, the mathematical machinery for such bridges remains underdeveloped. Here we develop a Rosetta Stone hypothesis from predictive processing, where beliefs serve as a central hub connecting phenomenology, behaviour, and neural dynamics. This hinges on a central technical assumption that phenomenology is a function of beliefs. We pursue a conditional approach: if this assumption holds, then certain predictions mathematically follow. We derive predictions for subjective similarity judgements, cognitive metabolic cost, subjective cognitive effort, and time perception. We review the connection between beliefs and neural dynamics to complete the generative passage for neurophenomenology, omitting the connection between beliefs and behaviour as this is already well-documented elsewhere. Testing our predictions will inform the validity of the central assumption connecting beliefs and phenomenology, and advance the neurophenomenology research programme.

Hallucinations are percepts that occur in the absence of corresponding sensory input, reflecting internally generated activity within sensory systems. Among visual hallucinations, geometric formations are common, yet their phenomenological structure remains poorly characterised. They occur across contexts ranging from psychedelic experiences to psychiatric and neurological disorders, likely involving related mechanisms within the early sensory cortices. Despite their relevance across disciplines, existing measures rely primarily on verbal reports or coarse rating scales, which do not capture geometric structure. To address this limitation, we developed and validated image recreation methods in which healthy participants were trained to recreate images of their hallucinations induced by stroboscopic light at frequencies known to elicit visual hallucinations under laboratory conditions. In Experiment 1 (N = 55), participants used a freehand drawing interface; in Experiment 2 (N = 54), a generative image recreation interface was employed. Training procedures mitigated individual differences in expressive ability. Validation trials demonstrated that geometric features were recoverable at the population level when image stimuli were used in place of hallucination-inducing stimuli. Across experiments, visual qualities of induced hallucinations varied systematically with stimulation frequency, consistent with prior reports. The data revealed a broader repertoire of geometric forms than typically described in the literature, including structures not predicted by current computational models of simple hallucinations. Applying these methods to other contexts could clarify the range of geometric hallucinations the visual system tends to produce under controlled perturbation across psychedelic, pathological, and laboratory settings. More precise characterisation of hallucination phenomenology may refine constraints on theoretical models of how visual experience is constructed in normal and altered states.

Since 18th century experiments demonstrated that the effects of mesmerism arose from belief rather than magnetic fluid, there has been evidence that situational cues can drive change in experience. Across four studies (total N = 2,042), we investigated the degree to which experiences in psychological experiments may arise from participants controlling their experience to satisfy their goals, i.e., phenomenological control. Trait phenomenological control predicts some experimental measures of changes in experience when demand characteristics are not controlled. Here, we investigated the reach of phenomenological control as a demand effect by contrasting relationships between Phenomenological Control Scale scores (PCS) and effects presumed to either be highly sensitive to beliefs (visually evoked auditory response or vEAR; tingling in the scalp or ASMR) or less sensitive to beliefs (Müller Lyer and vertical-horizontal illusions). Phenomenological control accounted for much of the effect for the posited belief-sensitive effects but not for the classic visual illusions (Study 1). This pattern of results remained after controlling for context effects (Study 2) and measurement differences across procedures (Study 3). Finally, Study 4 presents causal evidence that vEAR and ASMR can be modulated by high PCS participants in response to demand characteristics delivered by direct suggestion. When an effect is substantially predicted by trait phenomenological control and can be caused by phenomenological control following demand characteristics, this domain-general mechanism should be ruled out before appealing to domain-specific explanations. This presents a previously unrecognised threat to validity in psychology experiments analogous to that of placebo effects in medicine.

Across four studies (total n = 2,042, two preregistered, one partly preregistered), we investigated the degree to which experiences in psychological experiments may arise from participants controlling their experience to satisfy their goals, i.e., phenomenological control (PC). Trait PC has previously been shown to predict some experimental measures of changes in experience when demand characteristics are not controlled. Here, we investigated the reach of PC as a demand effect by contrasting relationships between Phenomenological Control Scale scores (PCS; a test of direct imaginative suggestion) and effects presumed to either be sensitive to beliefs (visually evoked auditory response or vEAR; tingling in the scalp or ASMR) or insensitive to beliefs (the Müller Lyer illusion; the vertical-horizontal illusion). PC accounted for much of the effect for the posited belief-sensitive effects but not for the classic visual illusions (Study 1). This pattern of results remained after controlling for context effects (Study 2) and measurement differences across procedures (Study 3). Finally, we present the first causal evidence that two established psychological effects which are predicted by PC (vEAR and ASMR) can be modulated by participants high in PC in accordance with demand characteristics delivered by direct suggestion (Study 4). PC may account for the effects of psychological experiments when an effect is sensitive to belief, but not where it is not. This presents a previously unrecognised threat to validity in psychology experiments analogous to that of placebo effects in medicine.

AbstractHallucinations arise from disruptions in neural processes that construct perceptual experience from sensory input. Progress in consciousness science may therefore benefit from methods that precisely characterize their phenomenology. We provide an initial validation of the Six-Dimensional Visual Hallucination Questionnaire (6D-VHQ), a brief instrument designed to quantify visual hallucination content across contexts. The 6D-VHQ measures six dimensions of visual experience: geometric and semantic content, detail level, vividness, entropy and focality. We report three studies: an online image-based validation study, a laboratory-based study with stroboscopically induced hallucinations and an online study in which participants reported closed-eye psychedelic hallucinations which occurred before enrolment. Across 962 responses, the questionnaire showed high internal consistency and a factor structure aligned with the intended dimensions, while also exhibiting substantial inter-item redundancy consistent with an over-complete initial implementation. Compared with the 11-Dimensional Altered States of Consciousness Questionnaire, both instruments captured simple and complex hallucinations, despite differing psychometric profiles. Stroboscopic stimulation was associated with simple geometric phenomena, whereas psychedelic reports showed greater semantic content, vividness and detail. These findings provide an initial operationalization of multi-dimensional visual hallucination content. Future work may expand and improve upon this work, including refinements to the instrument and applications to other hallucinatory contexts.

Hallucinations arise from disruptions in neural processes that construct perceptual experience from sensory input. Progress in consciousness science may therefore benefit from methods that precisely characterise their phenomenology. We introduce and validate the 6-Dimensional Visual Hallucination Questionnaire (6D-VHQ), a brief instrument designed to quantify visual hallucination content across contexts. The 6D-VHQ measures six dimensions: geometric and semantic content, detail level, vividness, entropy, and focality. It focuses on perceptual content rather than broader altered-state features. We report three studies: an image-based validation study, an application to stroboscopically induced visual hallucinations, and an application to psychedelic closed-eye visuals. Across 962 responses, the questionnaire showed high internal consistency, a factor structure aligned with the intended dimensions, and stable performance across settings. High inter-item redundancy indicated an initially over-complete but coherent implementation of the instrument. Compared with the 11-Dimensional Altered States of Consciousness questionnaire (11D-ASC), both instruments similarly captured simple and complex visual hallucinations despite differing psychometric properties. While the 11D-ASC showed low internal consistency, the 6D-VHQ has high internal consistency alongside high inter-dimensional redundancy. While the 11D-ASC correlates to participant traits, and 6D-VHQ is largely invariant to these, capturing the differences in visual phenomenology between different classes of hallucinatory experiences. The 6D-VHQ showed that stroboscopic stimulation predominantly elicited vivid simple geometric hallucinations, whereas psychedelic experiences showed combined geometric content alongside more vivid, detailed, and semantic visuals. The 6D-VHQ provides an operationalisation of a six-dimensional model of visual hallucination content that distinguishes broad classes of hallucinations across induction methods.

Voluntary actions are associated with sense of agency and distortions of perceived duration. This study examined how sensorimotor coupling and visual perspective modulate perceived duration and sense of agency during manual actions in four pre-registered experiments using virtual reality. Participants moved their hands while observing a virtual hand moving (a)synchronously, and evaluated movement duration and their sense of agency. We found that sensorimotor coupling provided by synchronous visual feedback is associated with longer perceived duration relative to action-related time compression, compared with delayed feedback or pre-recorded movements of another person (Experiments 1 and 2). Sensorimotor coupling also modulated sense of agency and may serve as a shared basis for perceived time and agency. Comparable modulations across first- and third-person perspectives (Experiments 1 and 2), anatomical configurations (Experiment 3), and attention on external entities (Experiment 4) suggest a shared predictive system for sensory consequences of self-generated actions and others’ reactions. This modulation was observed irrespective of anatomical configuration and when attention was directed toward objects unrelated to one’s own body, suggesting that the underlying mechanism is not confined to representations specific to the self, but may instead reflect domain-general predictive computations. These findings, together with the virtual reality technique developed, offer new insights into how humans experience passage of time and agency during voluntary action.

The scientific study of consciousness was sanctioned as an orthodox field of study only three decades ago. Since then, a variety of prominent theories have flourished, including integrated information theory, which has been recently accused of being pseudoscience by more than 100 academics. Here we critically assess this charge and offer thoughts to elevate the clash into positive lessons for our field.

Hemispherotomy is a neurosurgical procedure for treating refractory epilepsy, which entails disconnecting a significant portion of the cortex, potentially encompassing an entire hemisphere, from its cortical and subcortical connections. While this intervention prevents the spread of seizures, it raises important questions. Given the complete isolation from sensory-motor pathways, it remains unclear whether the disconnected cortex retains any form of inaccessible awareness. More broadly, the activity patterns that large portions of the deafferented cortex can sustain in awake humans remain poorly understood. We address these questions by exploring for the first time the electrophysiological state of the isolated cortex before and after surgery in ten awake pediatric patients. Post-surgery, the isolated cortex exhibited prominent slow oscillations (<2 Hz) and a broad-band shift in power spectral density from high to low frequencies. This resulted in a marked decrease of the spectral exponent, a validated consciousness marker, indicating broad-band slowing characteristic of unconscious states. When compared with a reference pediatric sample across the sleep-wake cycle, the spectral exponent of the contralateral cortex aligned with wakefulness, whereas that of the isolated cortex was consistent with deep NREM sleep. However, spindles did not emerge in the isolated cortex due to the lack of subcortical inputs, constituting a fundamental difference from physiological sleep. These findings demonstrate a unihemispheric sleep-like state during wakefulness, challenging the possibility that hemispherotomy might lead to inaccessible "islands of awareness." Moreover, the persistence of sleep-like patterns years after disconnection provides unique insights into the electrophysiological effects of disconnections in the human brain. ### Competing Interest Statement I have read the journal's policy and the authors of this manuscript have the following competing interests: M.M. is co-founder and shareholder of Intrinsic Powers, Inc., a spin-off of the University of Milan Si.Sa. is advisor of the same company. The remaining co-authors have no conflicts of interest to declare.

Local low-metallicity galaxies with signatures of possible accretion activity are ideal laboratories in which to search for the lowest-mass black holes and study their impact on the host galaxy. Here we present the first JWST NIRSpec IFS observations of SDSS J120122.30+021108.3, a nearby (z = 0.00354) extremely metal-poor dwarf galaxy with no optical signatures of accretion activity but identified by the Wide-field Infrared Survey Explorer to have red mid-infrared (MIR) colors consistent with active galactic nuclei (AGNs). We identify over 100 lines between similar to 1.7 and 5.2 mu m, an unresolved nuclear continuum source with an extremely steep spectral slope consistent with hot dust from an AGN (F-nu approximate to nu(-1.5)), and a plethora of H i, He i, and H2 lines, with no lines from heavier elements, CO or ice absorption features, or polycyclic aromatic hydrocarbons (PAHs). While the bright central unresolved source (less than or similar to 5 pc) is suggestive of an AGN, there are no He ii lines or coronal lines identified in the spectrum, and, importantly, there is no evidence that the radiation field is harder in the nuclear source compared with surrounding regions. The emission-line spectrum can be explained by a young (<5 Myr) nuclear star cluster with stellar mass similar to 3 x 10(4) M-circle dot, but the unresolved continuum source is more typical for a deeply embedded AGN powered by a black hole of minimum mass similar to 1450 M-circle dot. These observations reveal that either a metal-poor stellar population can heat the dust to extremely high temperatures, a result that would have significant impact on the reliability of MIR color selection in AGN surveys and our understanding of the properties of the interstellar medium and young stars in metal-poor galaxies, or an accreting intermediate-mass black hole is deeply embedded and hidden even at near-infrared wavelengths.

The Panchromatic Hubble Andromeda Southern Treasury (PHAST) is a large 195-orbit Hubble Space Telescope program imaging ∼0.45 deg 2 of the southern half of M31's star-forming disk at optical and near-ultraviolet (NUV) wavelengths. The PHAST survey area extends the northern coverage of the Panchromatic Hubble Andromeda Treasury (PHAT) down to the southern half of M31, covering out to a radius of ∼13 kpc along the southern major axis and in total ∼two-thirds of M31's star-forming disk. This new legacy imaging yields stellar photometry of over 90 million resolved stars using the Advanced Camera for Surveys in the optical (F475W and F814W), and the Wide Field Camera 3 (WFC3) in the NUV (F275W and F336W). The photometry is derived using all overlapping exposures across all bands, and achieves a 50% completeness-limited depth of F475W ∼ 27.7 in the lowest surface density regions of the outer disk and F475W ∼ 26.0 in the most crowded, high surface brightness regions near M31's bulge. We provide extensive analysis of the data quality, including artificial star tests to quantify completeness, photometric uncertainties, and flux biases, all of which vary due to the background source density and the number of overlapping exposures. We also present seamless population maps of the entire M31 disk, which show relatively well-mixed distributions for stellar populations older than 1–2 Gyr, and highly structured distributions for younger populations. The combined PHAST + PHAT photometry catalog of ∼0.2 billion stars is the largest ever produced for equidistant sources and is available for public download by the community.

Our ability to understand and control complex systems of many interacting parts remains limited. A key challenge is that we still do not know how best to describe-and quantify-the many-to-many dynamical interactions that characterize their complexity. To address this limitation, we introduce the mathematical framework of Integrated Information Decomposition, or [Formula: see text]ID. [Formula: see text]ID provides a comprehensive framework to disentangle and characterize the information dynamics of complex multivariate systems. On the theoretical side, [Formula: see text]ID reveals the existence of previously unreported modes of collective information flow, providing tools to express well-known measures of information transfer, information storage, and dynamical complexity as aggregates of these modes, thereby overcoming some of their known theoretical shortcomings. On the empirical side, we validate our theoretical results with computational models and examples from over 1,000 biological, social, physical, and synthetic dynamical systems. Altogether, [Formula: see text]ID improves our understanding of the behavior of widely used measures for characterizing complex systems across disciplines and leads to new more refined analyses of dynamical complexity.

Complex neural systems can display structured emergent dynamics. Capturing this structure remains a significant scientific challenge. Using information theory, we apply Dynamical Independence (DI) to uncover the emergent dynamical structure in a minimal 5-node biophysical neural model, shaped by the interplay of two key aspects of brain organisation: integration and segregation. In our study, functional integration within the biophysical neural model is modulated by a global coupling parameter, while functional segregation is influenced by adding dynamical noise, which counteracts global coupling. Leveraging transfer entropy, DI defines a dimensionally-reduced macroscopic variable (e.g., a coarse-graining) as emergent to the extent that it behaves as an independent dynamical process, distinct from the micro-level dynamics. Dynamical dependence (a departure from dynamical independence) is measured by minimising the transfer entropy from microlevel variables to macroscopic variables across spatial scales. Our results indicate that the degree of emergence of macroscopic variables is relatively minimised at balanced points of integration and segregation and maximised at the extremes. Additionally, our method identifies to which degree the macroscopic dynamics are localised across microlevel nodes, thereby elucidating the emergent dynamical structure through the relationship between microscopic and macroscopic processes. We find that deviation from a balanced point between integration and segregation results in a less localised, more distributed emergent dynamical structure as identified by DI. This finding suggests that a balance of functional integration and segregation is associated with lower levels of emergence (higher dynamical dependence), which may be crucial for sustaining coherent, localised emergent macroscopic dynamical structures. This work also provides a complete computational implementation for the identification of emergent neural dynamics that could be applied both in silico and in vivo.

Granger causality (GC) is widely used in neuroimaging to estimate directed statistical dependence between brain regions using time series of brain activity. A known problem is that fMRI measures brain activity indirectly via the blood-oxygen-level-dependent (BOLD) signal, which can distort GC estimates by introducing different time-to-peak responses across brain regions. However, how these distortions affect the validity of inferred connections is not fully understood. Previous studies have shown that false positives are not introduced if the haemodynamic response function (HRF) is minimum-phase; but whether the HRF is actually minimum-phase has remained contentious. Here, we address this issue by studying the transfer functions of three realistic biophysical models. We find that the minimum-phase condition is met for a wide range of physiologically plausible parameter values. Therefore, statistical testing of GC can be viable even if the HRF varies across brain regions, with the following two limitations. First, the minimum-phase condition is violated for parameter combinations that generate an initial dip in the HRF. Second, slow sampling of the BOLD signal (seconds) compared to the timescales of neural signal propagation (milliseconds) may still introduce spurious GC inferences. Beyond GC analysis, the closed-form expressions for the transfer functions of these popular HRF models are valuable for modeling fMRI time series since they balance mathematical tractability with biological plausibility.

As the field of consciousness science matures, the research agenda has expanded from an initial focus on the neural correlates of consciousness, to developing and testing theories of consciousness. Several theories have been put forward, each aiming to elucidate the relationship between consciousness and brain function. However, there is an ongoing, intense debate regarding whether these theories examine the same phenomenon. And, despite ongoing research efforts, it seems like the field has so far failed to converge around any single theory, and instead exhibits significant polarization. To advance this discussion, proponents of five prominent theories of consciousness—Global Neuronal Workspace Theory (GNWT), Higher-Order Theories (HOT), Integrated Information Theory (IIT), Recurrent Processing Theory (RPT), and Predictive Processing (PP)—engaged in a public debate in 2022, as part of the annual meeting of the Association for the Scientific Study of Consciousness (ASSC). They were invited to clarify the explananda of their theories, articulate the core mechanisms underpinning the corresponding explanations, and outline their foundational premises. This was followed by an open discussion that delved into the testability of these theories, potential evidence that could refute them, and areas of consensus and disagreement. Most importantly, the debate demonstrated that at this stage, there is more controversy than agreement between the theories, pertaining to the most basic questions of what consciousness is, how to identify conscious states, and what is required from any theory of consciousness. Addressing these core questions is crucial for advancing the field towards a deeper understanding and comparison of competing theories.

Exposure to rapid and bright stroboscopic light has long been reported to induce vivid visual hallucinations of colour and geometric formations. This phenomenon was first documented by Purkinje over 200 years ago. Since then, significant progress has been made in understanding the effects of stroboscopic light and the experiences it induces through multiple waves of interest from the scientific, therapeutic, and broader cultural communities. Despite these advances, fundamental questions remain unanswered, including comprehensive characterisations of its phenomenology, its precise physiological origins, under which conditions it may lead to altered states of consciousness phenomena, and potential clinical or therapeutic applications. This narrative review provides a historical summary of research into stroboscopic light stimulation alongside its use in recreation and lay-therapeutic contexts. It also discusses the phenomenology of these experiences, current perspectives on the potential neural mechanisms of stroboscopically induced experiences, and provides an outlook for future research in this field.

Psychedelics can profoundly alter consciousness by reorganising brain connectivity; however, their effects are context-sensitive. To understand how this reorganisation depends on the context, we collected and comprehensively analysed the largest psychedelic neuroimaging dataset to date. Sixty-two adults were scanned with fMRI and EEG during rest and naturalistic stimuli (meditation, music, and visual), before and after ingesting 19 mg of psilocybin. Half of the participants ranked the experience among the five most meaningful of their lives. Under psilocybin, fMRI and EEG signals recorded during eyes-closed conditions became similar to those recorded during an eyes-open condition. This change manifested as an increase in global functional connectivity in associative regions and a decrease in sensory areas. For the first time, we used machine learning to directly link the subjective effects of psychedelics to neural activity patterns characterised by low-dimensional embeddings. We show that psilocybin reorganised these low-dimensional trajectories into cohesive patterns of brain activity that were structured by context and quality of subjective experience, with stronger self- and boundary-related effects - which were linked to day-after mindset changes - leading to more structured and distinct neural representations. This reorganisation induces a state of 'embeddedness' - a coherent integration of brain networks that normally segregate internal and external processing - dissolving perceptual boundaries in a way that aligns neural dynamics with context. Beyond its transient expression, embeddedness serves as a construct for understanding the subjective and therapeutic effects of psychedelics. These findings provide a new account of the large-scale neurocognitive effects of psychedelics and demonstrate the utility of using machine learning methods in assessing state- and context-dependent neural dynamics and their association with psychological outcomes. ### Competing Interest Statement The authors have declared no competing interest.