Learning is a fundamental source of individuality Learning and memory are essential components of our individuality. While it is established that behaviour can vary across genetically identical individuals, it remains unknown how much of this variation stems from momentary experience during learning compared to genetics and its past interactions with the environment. To address this, we measured behaviour in thousands of flies from 90 genetic backgrounds while they performed tasks in conditions that either did or did not require learning. Flies that were genetically identical, raised under the same conditions and tested simultaneously in the same environment persistently modified the extent of expressed individuality when they could learn. This learning-induced residual expression of individuality and its dynamics were subdued or absent in innate, learning-independent behaviours. We could quantify and then recreate this phenomenon in computer simulations. The emergence of in silico behavioural individuality was most consistent with the individuality of real flies once we enabled reinforced learning in simulated agents. Moreover, we showed that minor differences in initial conditions of the experiment can exacerbate the expression of individuality within a genotype in a learning-dependent manner. Our results establish that besides the classical G x E interactions shared between individuals in the past, learning from individual momentary experience further extends the expression of individuality.
From self-organizing systems to subjective temporal extension The self-simulational theory of temporal extension (SST) describes an information-theoretically formalized mechanism by which the “width” of subjective temporality emerges from the architecture of self-modeling. In this paper, the perspective of the free energy principle will be assumed to cast the emergence of subjective temporal extension from first principles of the physics of self-organization and to formalize subjective temporal extension using information geometry. Using active inference, a deep parametric generative model of temporal inference is simulated, which realizes the described dynamics on a computational level. Two “biases” (i.e. variations) of time-perception emerge naturally from the simulated computational model. This concerns the intentional binding effect (i.e. the compression of the temporal interval between voluntarily initiated actions and subsequent sensory consequences) and empirically documented alterations of subjective time experience in deep states of meditative absorption (i.e. in minimal phenomenal experience). Generally, numerous systematic and domain-specific alterations of subjective temporal experience are computationally explained in a unified manner, as enabled by integration with current active inference accounts mapping onto the respective domains. This concerns, next to more general attentional and central tendency effects, the temporality-modulating role of valence, impulsivity, boredom, flow-states, near death-experiences, and various psychopathologies, among others. SST, from the perspective of the free energy principle, explains how the “width” of the subjective temporal moment emerges from first principles, accounting for why sometimes, subjective time seems to fly, and sometimes, moments feel like eternities; with the computational mechanism being readily deployable synthetically.
Perception, Memory, Simulation, and Consciousness: A Convergence of Theories Our theories stemming from perception, memory, and neurology came to similar and complementary conclusions regarding the mechanism of conscious brain processes. We suggest that consciousness is the explicit memory of past events or the general cognitive capacity to simulate events, whether used to consciously remember the past, experience the present, or imagine the future. Perceptual mechanisms may represent an ongoing, editable, “best estimate” of our past, present, and future. In fact, at milliseconds to seconds timescales, there may be no hard boundary between perception and memory. We view conscious perceptions, decisions, and actions as simulations of prior unconscious sensations, decisions, and actions. As consciousness is the simulation/explicit memory of past events, the neural correlates of consciousness may therefore be the neural correlates of simulation/explicit memory. Because the default mode network, along with the frontoparietal control and salience networks, is critical for simulation/explicit memory, it is likely critical for normal consciousness. Each aspect of consciousness (e.g., visual, auditory, decision-making) may have its own neural correlate. Lastly, by combining our three theories, our synthesis can shed light on conscious perceptions, decisions, and actions in timescales ranging from subsecond to seconds, minutes, days, months, and years.
The representational geometry of emotional states in basolateral amygdala Neurons in basolateral amygdala (BLA) encode positive and negative valence. However, many additional variables must be represented to describe all aspects of emotional states. To investigate how BLA encodes these states, we presented mice with conditioned stimuli that elicited two behavioral responses: tremble and ingress into a burrow, reflecting fear and flight to safety, respectively. BLA inactivation eliminated several aspects of differential responses to aversive versus neutral stimuli without eliminating tremble and ingress themselves, consistent with BLA’s encoding valence not motor commands. However, individual neurons rarely represented only valence, exhibiting, instead, mixed selectivity for stimulus identity, stimulus valence, tremble and/or ingress. Despite prevalent mixed selectivity, population activity sometimes realized a representational geometry that conferred two computational properties defining specialized readouts: generalization across conditions and no interference between readouts of different variables. These specialized readouts enable output responses to depend on one specific variable and to remain unaffected by the others.
The Impact of Prior Beliefs about Volatility on Adaptive Behavior Humans adapt to environmental changes by balancing empirical observations with prior beliefs and evaluating if unexpected events indicate a true change. The specific factors that govern updating behavior in dynamic environments remain to be elucidated. We here examined how prior beliefs about environmental volatility affect updating of cue-target contingencies, particularly when observations violate these beliefs. Thirty-two participants completed two versions of a probabilistic reversal-learning task, in which auditory cues signaled the location of a subsequent visual target stimulus. In a reactive task version, participants indicated the target location after its appearance; in a predictive task version, they predicted the target’s location based on the cue information. Cue-target contingencies either remained stable or reversed once within a block, thereby creating a stable and a reversal environment. Before each block, participants received either true or false information about volatility, i.e., about whether the cue-target contingency would remain stable or change. We analyzed reaction times (reactive task) and choices (predictive task) with model-free measures and a Rescorla-Wagner learning model. Participants generally adapted to the contingency changes in both tasks. In the reactive task, prior beliefs had no significant effect. In the predictive task, believing that a reversal environment was stable reduced learning rates. In stable environments, falsely believing the environment contained a reversal increased decision noise, reduced accuracy and increased choice variability. These findings demonstrate that prior beliefs about volatility shape updating in response to task demands and environmental structure.
Controlling Spatio-Temporal Sequences of Neural Activity by Local Synaptic Changes The neural basis of behavior is believed to consist of sequential patterns of neural activity in the relevant brain regions. Behavioral flexibility also requires neural circuit mechanisms that support dynamic control of sequential activity. However, mechanisms to control and reconfigure sequential activity have received little attention. Here, we show that recurrently connected networks with heterogeneous connectivity and a smooth spatial in-degree landscape (which may arise due to asymmetric neuron morphologies) provide a robust mechanism to evoke and control sequential activity. By modulating the synaptic strength of only a few neurons in local neighborhoods, we uncovered high-impact locations that can start, stop, extend, gate, and redirect sequences. Interestingly, high-impact locations coincide with mid in-degree regions. We demonstrate that these motifs can flexibly reconfigure sequential activity, and hence, provide a framework for fast and flexible computations on behavioral timescales, while the individual parts of the pathways remain rigid and reliable.
Hippocampal place cells map terrain geometry independently of behavior How does the brain map uneven terrain? While spatial neurons such as hippocampal place cells and entorhinal grid cells have been extensively studied in flat, horizontal environments, the natural world is hilly and irregular. To investigate how place cells represent irregular terrain, we recorded from the hippocampus of rats foraging across either flat or ridged terrain. Place cell activity reflected terrain shape, consistent with a surface-bound cognitive map rather than a volumetric one. Place fields were elongated parallel to terrain contours, which contrasted with the movement biases of the rats, and are inconsistent with a predictive coding model. Reflecting the importance of terrain information, a third of the place cells exhibited repeating fields on each ridge. A boundary vector cell model of place cell firing replicated these results. These findings demonstrate that the cognitive map is sensitive to topography, bringing our understanding of spatial cognition closer to the real world.
When less humanity is more: A single target compensation phenomenon between human nature and human uniqueness This research examines the dynamic relationship between the two dimensions of humanity perception – human nature (HN) and human uniqueness (HU) – as defined in Haslam’s (2006) Dual Model of Dehumanization. While these dimensions have long been relied on to study perceptions of humanity, the relationship between them remains unclear. Building on prior research showing compensation effects in social judgment, as well as our own prior unpredicted results, we hypothesized that a similar pattern might occur between HN and HU. Across four studies (N = 660), we independently manipulated vignette-based targets of social perception to appear either low or high on one dimension, using a 2 × 2 factorial design. On a subsequent measure of humanity perception, a strong compensation pattern between HU and HN emerged. Targets described as low in HN were perceived as higher in HU than those described as high in HN. Conversely, targets low in HU were seen as higher in HN than those described as high in HU. These findings support the existence of a dimensional compensation mechanism in humanity perception. Results are discussed in terms of potential cognitive mechanisms behind compensation, the role of social comparison, and the potential implication of the self in this process.
“Nothing” Really Matters: What Omission Responses Reveal About the Predictive Brain Understanding the brain’s predictive machinery requires isolating endogenous activity from responses to external stimulation. Omission paradigms, which study neural responses when expected stimuli are absent, provide this unique window into internal computations. This review synthesizes omission research across species, modalities, and paradigms to reveal how the brain anticipates the world. Across diverse findings, the brain uses explicit learned models to generate detailed, feature-specific representations of expected content. This is evidenced by anticipatory signals in associative areas that peak at omission times and by cortical responses carrying decodable information about missing stimulus features, creating “sensory ghosts” of absent events. These sophisticated computations are complemented by foundational responses emerging rapidly in isolated preparations, reflecting simpler local computations. We propose a framework where omission responses emerge from two cooperative computational styles spanning a mechanistic spectrum. Local Regularity Encoding (LRE) generates fast, automatic signals through intrinsic circuit dynamics like adaptation and rebound, operating within constrained temporal windows. Model-Based Inference (MBI) produces slower, flexible predictions via distributed networks that learn specific “what” and “when” expectations, often requiring attention and top-down control. We organize these findings using empirical signatures including timing constraints, attention dependence, and content-specificity. Our synthesis includes a functional “Omission Atlas” mapping computations onto brain networks, from cerebellar timing scaffolds to hippocampal content predictions. Clinical applications reveal differential vulnerability of these systems in schizophrenia, autism, and neurodevelopmental disorders. This framework provides a unified account of mismatch negativity and related phenomena, offering both theoretical advancement and practical tools for future research into predictive processing mechanisms.