Head direction (HD) cells in the anterodorsal thalamic nuclei form the brain's internal compass, and are often modeled as a ring attractor maintaining azimuth coding by leveraging continuous visual and inertial sensory input. Here, we test how the common experimental preparation of head-fixed animals alters this code. Complete head-fixation that creates vestibular conflict disrupts both unit and population encoding of head direction, while selectively constraining head-on-body movements either in real or virtual reality uniquely impairs HD population activity. More specifically, attractor dynamics is altered in head-restrained mice during periods of immobility, but remarkably recover several hundred milliseconds prior to locomotion onset. The rescue preceding movement onset suggests that an efference copy or prediction of a re-afferent signal is necessary to maintain HD network activity during head restraint. A computational model recapitulates these effects by perturbing lateral connectivity among HD neurons. More generally, the results indicate that the HD network is a context- and state-dependent predictive estimator, stabilized by forthcoming self-motion signals. The classic ring-attractor models should be revised to integrate context-dependent dynamics with prospective motor signals, offering a more complete account of how the brain's compass remains stable across both naturalistic and constrained conditions.

Our brains support at least two major spatial representations: an egocentric, short-horizon representation of peripersonal space (PPS), and an allocentric, longer-horizon cognitive map system. Here, we speculate on their relationship from an evolutionary perspective. We argue that an ancient proto-PPS system for contact prediction likely emerged early in vertebrate evolution as a fight-or-flight mechanism for managing imminent threats, supported by evolutionarily conserved midbrain circuits such as the optic tectum (superior colliculus in mammals). The subsequent transition of many species from aquatic to terrestrial environments dramatically expanded sensory ranges and environmental complexity, arguably creating selective pressures for longer-horizon navigation, prediction, and planning. These pressures likely contributed to the emergence of hippocampal and entorhinal systems supporting allocentric spatial representations. Next, as navigation and foraging behaviors became more sophisticated and required dexterous multi-joint movements, it is possible that a cortical PPS system emerged within neocortical networks, including posterior parietal and ventral premotor cortices. Rather than mediating coarse fight-or-flight responses, we suggest this cortical PPS system now supports fine-grained hierarchical sensorimotor control, reference-frame transformations, tool use, and social interactions. In parallel, the allocentric mapping systems seemingly expanded beyond spatial navigation to support relational, conceptual, and abstract cognitive structures. We suggest that PPS and cognitive maps instantiate shared computational principles operating at different spatial and temporal scales and in different reference frames; a view supported by recent reinforcement-learning frameworks linking body-centered and allocentric predictive value. Together, this perspective points to a putative scaffolded evolutionary trajectory in which cognition may have expanded from near-body safety to a multiscale predictive architecture supporting flexible, goal-directed behavior across space and time.

Reliably detecting consciousness in unresponsive patients remains an urgent ethical and clinical challenge, as no behavior-independent marker is currently accepted in clinical practice. We characterize consciousness as linked to a representation of the embodied subject of experience, mediated by multisensory integration within the peripersonal space (PPS) system. We test whether a neural marker of PPS representation could detect consciousness and predict clinical outcome in disorders of consciousness (DoC) patients. Using high-density electroencephalography (EEG) during a task-free audiotactile task, we derive a PPS index based on high-beta oscillations. In healthy participants, the PPS index is present during wakefulness and dreaming, but absent in dreamless sleep. In 72 DoC patients, the PPS index correlates with behavioral measures of consciousness and predicts recovery at discharge. The index is associated with forebrain mesocircuit integrity. These findings highlight a bedside-compatible electrophysiological marker with potential clinical utility for detecting covert consciousness and predicting outcomes in non-responsive patients.

The Simons Collaboration on Ecological Neuroscience (SCENE) seeks to uncover general principles of brain function through an ecological perspective: studying perception, cognition, and action in the context of the affordances available to different agents. Here, we introduce SCENE's goals, hypotheses, and approaches outlining a collaborative vision for the next decade.

Real-world choices often require balancing short- and long-term goals. We reasoned that seemingly suboptimal single-trial decisions may reflect strategic planning over longer timescales. We demonstrate that male macaques freely navigating in virtual reality strategically aborted offers, forgoing immediate rewards to maximize session-long returns. This behavior was highly individual-specific, suggesting that macaques account for their own long-run performance. Reinforcement-learning models suggest that this strategy is supported by modular actor-critic networks in which a policy module optimizes long-term value while also incorporating state-action values for rapid policy adjustment. These models predict that policy changes for matched offers should emerge at offer presentation, even when aborts occur much later. Consistent with this prediction, units and population dynamics in dorsolateral prefrontal cortex (dlPFC), but not parietal area 7a or dorsomedial superior temporal area (MSTd), encoded upcoming reward-optimizing aborts at offer onset. These findings cast dlPFC as a specialized policy module within closed-loop behaviors.

This article presents a compiled interview with three researchers in the field of multisensory integration, with an emphasis on how multisensory integration may go awry in psychiatric and neurodevelopmental conditions. Through structured dialogue, the authors reflect on the origins of the field, the development of key empirical and computational frameworks, and the growing relevance of multisensory science for understanding autism, schizophrenia, dyslexia, and related conditions. The discussion highlights the shift from descriptive accounts of multisensory phenomena toward normative models, including maximum likelihood estimation, causal inference, and correlation detection, as well as the need to link these computations to neural circuits and clinical heterogeneity. The authors consider ongoing challenges in the field, including inconsistent findings across clinical studies, limited statistical power, difficulties in defining and measuring multisensory function, and the need for larger, better-characterized cohorts. They also discuss emerging directions, including naturalistic paradigms, adaptive coding, open and team science, precision psychiatry, and the application of artificial intelligence to complex multimodal datasets. The interview concludes by emphasizing the promise of multisensory integration as a framework for understanding perception, cognition, and belief updating in health and disease.

Peripersonal space (PPS) is the region of space near the body, the multisensory interface where interactions with the environment predominantly occur. This space is represented by a specialized neural system that integrates tactile and external stimuli as a function of their distance from the body. Previous studies uncovered plastic and dynamical properties of PPS representation, links between PPS encoding and higher-level cognitive functions (e.g., social cognition), as well as its alterations in neurological and psychiatric disorders. These findings have expanded the definition of PPS and have led to the development of an array of computational models of PPS, addressing the why and how of PPS encoding. Although computational models are crucial for advancing our mechanistic and functional understanding of PPS representation, no prior work has reviewed these models. Here, we address this gap by analysing computational models of PPS, and proposing a taxonomy to classify them based on their level of description, capacity to reproduce empirical findings, and ability to generate novel predictions. This effort leads us to propose that PPS may be best understood as a system that detects spatiotemporal regularities in body-environment interactions, in order to predict potential future interactions. Hence, we suggest re-defining PPS as a unified spatiotemporal field that integrates not only spatial dimensions, but also temporal ones.

A key challenge in neuroscience is understanding how neurons in hundreds of interconnected brain regions integrate sensory inputs with previous expectations to initiate movements and make decisions(1). It is difficult to meet this challenge if different laboratories apply different analyses to different recordings in different regions during different behaviours. Here we report a comprehensive set of recordings from 621,733 neurons recorded with 699 Neuropixels probes across 139 mice in 12 laboratories. The data were obtained from mice performing a decision-making task with sensory, motor and cognitive components. The probes covered 279 brain areas in the left forebrain and midbrain and the right hindbrain and cerebellum. We provide an initial appraisal of this brain-wide map and assess how neural activity encodes key task variables. Representations of visual stimuli transiently appeared in classical visual areas after stimulus onset and then spread to ramp-like activity in a collection of midbrain and hindbrain regions that also encoded choices. Neural responses correlated with impending motor action almost everywhere in the brain. Responses to reward delivery and consumption were also widespread. This publicly available dataset represents a resource for understanding how computations distributed across and within brain areas drive behaviour.

Understanding brain function relies on the collective work of many labs generating reproducible results. However, reproducibility has not been systematically assessed within the context of electrophysiological recordings during cognitive behaviors. To address this, we formed a multi-lab collaboration using a shared, open-source behavioral task and experimental apparatus. Experimenters in 10 laboratories repeatedly targeted Neuropixels probes to the same location (spanning secondary visual areas, hippocampus, and thalamus) in mice making decisions; this generated a total of 121 experimental replicates, a unique dataset for evaluating reproducibility of electrophysiology experiments. Despite standardizing both behavioral and electrophysiological procedures, some experimental outcomes were highly variable. A closer analysis uncovered that variability in electrode targeting hindered reproducibility, as did the limited statistical power of some routinely used electrophysiological analyses, such as single-neuron tests of modulation by individual task parameters. Reproducibility was enhanced by histological and electrophysiological quality-control criteria. Our observations suggest that data from systems neuroscience is vulnerable to a lack of reproducibility, but that across-lab standardization, including metrics we propose, can serve to mitigate this.

Self-initiated behavior is accompanied by the experience of intending our actions. Here, we leverage the unique opportunity to examine the full intentional chain-from intention to action to environmental effects-in a tetraplegic person outfitted with a primary motor cortex (M1) brain-machine interface (BMI) generating real hand movements via neuromuscular electrical stimulation (NMES). This combined BMI-NMES approach allowed us to selectively manipulate each element of the intentional chain (intention, action, effect) while probing subjective experience and performing extra-cellular recordings in human M1. Behaviorally, we reveal a novel form of intentional binding: motor intentions are reflected in a perceived temporal attraction between the onset of intentions and that of actions. Neurally, we demonstrate that evoked spiking activity in M1 largely coincides in time with the onset of the experience of intention and that M1 spike counts and the onset of subjective intention may co-vary on a trial-by-trial basis. Further, population-level dynamics, as indexed by a decoder instantiating movement, reflect intention-action temporal binding. The results fill a significant knowledge gap by relating human spiking activity in M1 with the onset of subjective intention and complement prior human intracranial work examining pre-motor and parietal areas.

Our sense of agency, the subjective experience of controlling our actions, is a crucial component of self-awareness and motor control. It is thought to originate from the comparison between intentions and actions across broad cortical networks. However, the underlying neural mechanisms are still not fully understood. We hypothesized that oscillations in the theta-alpha range, thought to orchestrate long-range neural connectivity, may mediate sensorimotor comparisons. To test this, we manipulated the relation between intentions and actions in a tetraplegic user of a brain machine interface (BMI), decoding primary motor cortex (M1) activity to restore hand functionality. We found that the pre-movement phase of low-alpha oscillations in M1 predicted the participant's agency judgements. Further, using EEG-BMI in healthy participants, we found that pre-movement alpha oscillations in M1 and supplementary motor area (SMA) correlated with agency ratings, and with changes in their functional connectivity with parietal, temporal and prefrontal areas. These findings argue for phase-driven gating as a key mechanism for sensorimotor integration and sense of agency.

The neural representations of prior information about the state of the world are poorly understood1. Here, to investigate them, we examined brain-wide Neuropixels recordings and widefield calcium imaging collected by the International Brain Laboratory. Mice were trained to indicate the location of a visual grating stimulus, which appeared on the left or right with a prior probability alternating between 0.2 and 0.8 in blocks of variable length. We found that mice estimate this prior probability and thereby improve their decision accuracy. Furthermore, we report that this subjective prior is encoded in at least 20% to 30% of brain regions that, notably, span all levels of processing, from early sensory areas (the lateral geniculate nucleus and primary visual cortex) to motor regions (secondary and primary motor cortex and gigantocellular reticular nucleus) and high-level cortical regions (the dorsal anterior cingulate area and ventrolateral orbitofrontal cortex). This widespread representation of the prior is consistent with a neural model of Bayesian inference involving loops between areas, as opposed to a model in which the prior is incorporated only in decision-making areas. This study offers a brain-wide perspective on prior encoding at cellular resolution, underscoring the importance of using large-scale recordings on a single standardized task.

Learning the contingencies of a task is difficult. Individuals learn in an idiosyncratic manner, revising their approach multiple times as they explore and adapt. Quantitative characterization of these learning curves requires a model that can capture both new behaviors and slow changes in existing ones. Here we suggest a dynamic infinite hidden semi-Markov model, whose latent states are associated with specific components of behavior. This model can describe new behaviors by introducing new states and capture more modest adaptations through dynamics in existing states. We tested the model by fitting it to behavioral data of >100 mice learning a contrast-detection task. Although animals showed large interindividual differences while learning this task, most mice progressed through three stages of task understanding, new behavior often arose at session onset, and early response biases did not predict later ones. We thus provide a new tool for comprehensively capturing behavior during learning.

Computational psychiatry studies suggest that individuals with autism spectrum disorder (ASD) inflexibly update their expectations. Here we leveraged high-yield rodent psychophysics, extensive behavioral modeling and brain-wide single-cell extracellular recordings to assess whether mice with different genetic perturbations associated with ASD show this same computational anomaly, and if so, what neurophysiological features are shared across genotypes. Mice harboring mutations in Fmr1, Cntnap2 or Shank3B show a blunted update of priors during decision-making. Compared with mice that flexibly updated their priors, inflexible updating of priors was associated with a shift in the weighting of prior encoding from sensory to frontal cortices. Furthermore, frontal areas in mouse models of ASD showed more units encoding deviations from the animals' long-run prior, and sensory responses did not differentiate between expected and unexpected observations. These findings suggest that distinct genetic instantiations of ASD may yield common neurophysiological and behavioral phenotypes.

Methodological advances in neuroscience have enabled the collection of massive datasets which demand innovative approaches for scientific communication. Existing platforms for data storage lack intuitive tools for data exploration, limiting our ability to interact effectively with these brain-wide datasets. We introduce two public websites: Data and Atlas developed for the International Brain Laboratory which provide access to millions of behavioral trials and hundreds of thousands of individual neurons. These interfaces allow users to discover both the raw and processed brain-wide data released by the IBL at the scale of the whole brain, individual sessions, trials, and neurons. By hosting these data interfaces as websites they are available cross-platform with no installation. By releasing each site's code as a modular open-source framework, other researchers can easily develop their own web interfaces and explore their own data. As neuroscience datasets continue to expand, customizable web interfaces offer a glimpse into a future of streamlined data exploration and act as blueprints for future tools.

Real world choices often involve balancing decisions that are optimized for the short-vs. long-term. Here, we reason that apparently sub-optimal single trial decisions in macaques may in fact reflect long-term, strategic planning. We demonstrate that macaques freely navigating in VR for sequentially presented targets will strategically abort offers, forgoing more immediate rewards on individual trials to maximize session-long returns. This behavior is highly specific to the individual, demonstrating that macaques reason about their own long-run performance. Reinforcement-learning (RL) models suggest this behavior is algorithmically supported by modular actor-critic networks with a policy module not only optimizing long-term value functions, but also informed of specific state-action values allowing for rapid policy optimization. The behavior of artificial networks suggests that changes in policy for a matched offer ought to be evident as soon as offers are made, even if the aborting behavior occurs much later. We confirm this prediction by demonstrating that single units and population dynamics in macaque dorsolateral prefrontal cortex (dlPFC), but not parietal area 7a or dorsomedial superior temporal area (MSTd), reflect the upcoming reward-maximizing aborting behavior upon offer presentation. These results cast dlPFC as a specialized policy module, and stand in contrast to recent work demonstrating the distributed and recurrent nature of belief-networks.

Contemporary pose estimation methods enable precise measurements of behavior via supervised deep learning with hand-labeled video frames. Although effective in many cases, the supervised approach requires extensive labeling and often produces outputs that are unreliable for downstream analyses. Here, we introduce "Lightning Pose," an efficient pose estimation package with three algorithmic contributions. First, in addition to training on a few labeled video frames, we use many unlabeled videos and penalize the network whenever its predictions violate motion continuity, multiple-view geometry, and posture plausibility (semi-supervised learning). Second, we introduce a network architecture that resolves occlusions by predicting pose on any given frame using surrounding unlabeled frames. Third, we refine the pose predictions post-hoc by combining ensembling and Kalman smoothing. Together, these components render pose trajectories more accurate and scientifically usable. We release a cloud application that allows users to label data, train networks, and predict new videos directly from the browser.