Successor-like representation guides the prediction of future events in human visual cortex and hippocampus

Non-feature-specific elevated responses and feature-specific backward replay in human brain induced by visual sequence exposure

The ability of cortical circuits to adapt in response to experience is a fundamental property of the brain. After exposure to a moving dot sequence, flashing a dot as a cue at the starting point of the sequence can elicit successive elevated responses even in the absence of the sequence. These cue-triggered elevated responses have been shown to play a crucial role in predicting future events in dynamic environments. However, temporal sequences we are exposed to typically contain rich feature information. It remains unknown whether the elevated responses are feature-specific and, more crucially, how the brain organizes sequence information after exposure. To address these questions, participants were exposed to a predefined sequence of four motion directions for about 30 min, followed by the presentation of the start or end motion direction of the sequence as a cue. Surprisingly, we found that cue-triggered elevated responses were not specific to any motion direction. Interestingly, motion direction information was spontaneously reactivated, and the motion sequence was backward replayed in a time-compressed manner. These effects were observed even after brief exposure. Notably, no replay events were observed when the second or third motion direction of the sequence served as a cue. Further analyses revealed that activity in the medial temporal lobe (MTL) preceded the ripple power increase in visual cortex at the onset of replay, implying a coordinated relationship between the activities in the MTL and visual cortex. Together, these findings demonstrate that visual sequence exposure induces twofold brain plasticity that may simultaneously serve for different functional purposes. The non-feature-specific elevated responses may facilitate general processing of upcoming stimuli, whereas the feature-specific backward replay may underpin passive learning of visual sequences.

Noise Resilience of Successor and Predecessor Feature Algorithms in One- and Two-Dimensional Environments

Noisy inputs pose significant challenges for reinforcement learning (RL) agents navigating real-world environments. While animals demonstrate robust spatial learning under dynamic conditions, the mechanisms underlying this resilience remain understudied in RL frameworks. This paper introduces a novel comparative analysis of predecessor feature (PF) and successor feature (SF) algorithms under controlled noise conditions, revealing several insights. Our key innovation lies in demonstrating that SF algorithms achieve superior noise resilience compared to traditional approaches, with cumulative rewards of 2216.88±3.83 (mean ± SEM), even under high noise conditions (σ=0.5) in one-dimensional environments, while Q learning achieves only 19.22±0.57. In two-dimensional environments, we discover an unprecedented nonlinear relationship between noise level and algorithm performance, with SF showing optimal performance at moderate noise levels (σ=0.25), achieving cumulative rewards of 2886.03±1.63 compared to 2798.16±3.54 for Q learning. The λ parameter in PF learning is a significant factor, with λ=0.7 consistently achieving higher λ values under most noise conditions. These findings bridge computational neuroscience and RL, offering practical insights for developing noise-resistant learning systems. Our results have direct applications in robotics, autonomous navigation, and sensor-based AI systems, particularly in environments with inherent observational uncertainty.

Humans rationally balance detailed and temporally abstract world models

How do people model the world's dynamics to guide mental simulation and evaluate choices? One prominent approach, the Successor Representation (SR), takes advantage of temporal abstraction of future states: by aggregating trajectory predictions over multiple timesteps, the brain can avoid the costs of iterative, multi-step mental simulation. Human behavior broadly shows signatures of such temporal abstraction, but finer-grained characterization of individuals' strategies and their dynamic adjustment remains an open question. We developed a task to measure SR usage during dynamic, trial-by-trial learning. Using this approach, we find that participants exhibit a mix of SR and model-based learning strategies that varies across individuals. Further, by dynamically manipulating the task contingencies within-subject to favor or disfavor temporal abstraction, we observe evidence of resource-rational reliance on the SR, which decreases when future states are less predictable. Our work adds to a growing body of research showing that the brain arbitrates between approximate decision strategies. The current study extends these ideas from simple habits into usage of more sophisticated approximate predictive models, and demonstrates that individuals dynamically adapt these in response to the predictability of their environment.

Revisiting the role of computational neuroimaging in the era of integrative neuroscience

Computational models have become integral to human neuroimaging research, providing both mechanistic insights and predictive tools for human cognition and behavior. However, concerns persist regarding the ecological validity of lab-based neuroimaging studies and whether their spatiotemporal resolution is not sufficient for capturing neural dynamics. This review aims to re-examine the utility of computational neuroimaging, particularly in light of the growing prominence of alternative neuroscientific methods and the growing emphasis on more naturalistic behaviors and paradigms. Specifically, we will explore how computational modeling can both enhance the analysis of high-dimensional imaging datasets and, conversely, how neuroimaging, in conjunction with other data modalities, can inform computational models through the lens of neurobiological plausibility. Collectively, this evidence suggests that neuroimaging remains critical for human neuroscience research, and when enhanced by computational models, imaging can serve an important role in bridging levels of analysis and understanding. We conclude by proposing key directions for future research, emphasizing the development of standardized paradigms and the integrative use of computational modeling across neuroimaging techniques.

The brain hierarchically represents the past and future during multistep anticipation

Memory for temporal structure enables both planning of future events and retrospection of past events. We investigated how the brain flexibly represents extended temporal sequences into the past and future during anticipation. Participants learned sequences of environments in immersive virtual reality. Pairs of sequences had the same environments in a different order, enabling context-specific learning. During fMRI, participants anticipated upcoming environments multiple steps into the future in a given sequence. Temporal structure was represented in the hippocampus and across higher-order visual regions (1) bidirectionally, with graded representations into the past and future and (2) hierarchically, with further events into the past and future represented in successively more anterior brain regions. In hippocampus, these bidirectional representations were context-specific, and suppression of far-away environments predicted response time costs in anticipation. Together, this work sheds light on how we flexibly represent sequential structure to enable planning over multiple timescales.

The brain hierarchically represents the past and future during multistep anticipation

Memory for temporal structure enables both planning of future events and retrospection of past events. We investigated how the brain flexibly represents extended temporal sequences into the past and future during anticipation. Participants learned sequences of environments in immersive virtual reality. Pairs of sequences had the same environments in a different order, enabling context-specific learning. During fMRI, participants anticipated upcoming environments multiple steps into the future in a given sequence. Temporal structure was represented in the hippocampus and across higher-order visual regions (1) bidirectionally, with graded representations into the past and future and (2) hierarchically, with further events into the past and future represented in successively more anterior brain regions. In hippocampus, these bidirectional representations were context-specific, and suppression of far-away environments predicted response time costs in anticipation. Together, this work sheds light on how we flexibly represent sequential structure to enable planning over multiple timescales.

Attending is not enough: Responding to targets is needed for across-trial statistical learning

Recent evidence shows that observers are able to learn across-trial regularities as indicated by faster responses to targets whose location was predicted by the target's location on the preceding trial. The present study investigated whether responding to both targets of the pair, as was the case in studies thus far, was needed for learning to occur. Participants searched for a shape singleton target and responded to the line inside. There were two across-trial predicting-predicted regularities regarding target locations: if the target appeared at one specific location on a given trial, it would appear at another specific location on the next trial. Unlike previous experiments, for one of these regularity pairs a response was only needed on either the first or the second target in the pair. Experiment 1 showed that across-trial learning only occurred when responding was required to both targets of a pair. If the response to one target of a pair had to be withheld, no learning occurred. Experiment 2 showed that the absence of learning cannot be attributed to carry-over inhibition resulting from not having to respond. After learning across-trial contingencies, learning remained in place even when the response to the first target of the pair had to be withheld. Our findings show that the execution of the (arbitrary) simple key-press response for both trials of the pair was needed for across-trial statistical learning to occur, whereas solely attending target locations did not result in any learning.

Predictable and unpredictable deviance detection in the human hippocampus and amygdala

Our brains extract structure from the environment and form predictions given past experience. Predictive circuits have been identified in wide-spread cortical regions. However, the contribution of medial temporal structures in predictions remains under-explored. The hippocampus underlies sequence detection and is sensitive to novel stimuli, sufficient to gain access to memory, while the amygdala to novelty. Yet, their electrophysiological profiles in detecting predictable and unpredictable deviant auditory events remain unknown. Here, we hypothesized that the hippocampus would be sensitive to predictability, while the amygdala to unexpected deviance. We presented epileptic patients undergoing presurgical monitoring with standard and deviant sounds, in predictable or unpredictable contexts. Onsets of auditory responses and unpredictable deviance effects were detected earlier in the temporal cortex compared with the amygdala and hippocampus. Deviance effects in 1-20 Hz local field potentials were detected in the lateral temporal cortex, irrespective of predictability. The amygdala showed stronger deviance in the unpredictable context. Low-frequency deviance responses in the hippocampus (1-8 Hz) were observed in the predictable but not in the unpredictable context. Our results reveal a distributed network underlying the generation of auditory predictions and suggest that the neural basis of sensory predictions and prediction error signals needs to be extended.

The Hippocampus in Pigeons Contributes to the Model-Based Valuation and the Relationship between Temporal Context States

Model-based decision-making guides organism behavior by the representation of the relationships between different states. Previous studies have shown that the mammalian hippocampus (Hp) plays a key role in learning the structure of relationships among experiences. However, the hippocampal neural mechanisms of birds for model-based learning have rarely been reported. Here, we trained six pigeons to perform a two-step task and explore whether their Hp contributes to model-based learning. Behavioral performance and hippocampal multi-channel local field potentials (LFPs) were recorded during the task. We estimated the subjective values using a reinforcement learning model dynamically fitted to the pigeon's choice of behavior. The results show that the model-based learner can capture the behavioral choices of pigeons well throughout the learning process. Neural analysis indicated that high-frequency (12-100 Hz) power in Hp represented the temporal context states. Moreover, dynamic correlation and decoding results provided further support for the high-frequency dependence of model-based valuations. In addition, we observed a significant increase in hippocampal neural similarity at the low-frequency band (1-12 Hz) for common temporal context states after learning. Overall, our findings suggest that pigeons use model-based inferences to learn multi-step tasks, and multiple LFP frequency bands collaboratively contribute to model-based learning. Specifically, the high-frequency (12-100 Hz) oscillations represent model-based valuations, while the low-frequency (1-12 Hz) neural similarity is influenced by the relationship between temporal context states. These results contribute to our understanding of the neural mechanisms underlying model-based learning and broaden the scope of hippocampal contributions to avian behavior.

Traveling waves shape neural population dynamics enabling predictions and internal model updating

The brain generates predictions based on statistical regularities in our environment. However, it is unclear how predictions are optimized through iterative interactions with the environment. Because traveling waves (TWs) propagate across the cortex shaping neural excitability, they can carry information to serve predictive processing. Using human intracranial recordings, we show that anterior-to-posterior alpha TWs correlated with prediction strength. Learning about priors altered neural state space trajectories, and how much it altered correlated with trial-by-trial prediction strength. Learning involved mismatches between predictions and sensory evidence triggering alpha-phase resets in lateral temporal cortex, accompanied by stronger alpha phase-high gamma amplitude coupling and high-gamma power. The mismatch initiated posterior-to-anterior alpha TWs and change in the subsequent trial's state space trajectory, facilitating model updating. Our findings suggest a vital role of alpha TWs carrying both predictions to sensory cortex and mismatch signals to frontal cortex for trial-by-trial fine-tuning of predictive models.

Neural patterns of conscious visual awareness in the Riddoch syndrome

The Riddoch syndrome is one in which patients blinded by lesions to their primary visual cortex can consciously perceive visual motion in their blind field, an ability that correlates with activity in motion area V5. Our assessment of the characteristics of this syndrome in patient ST, using multimodal MRI, showed that: 1. ST's V5 is intact, receives direct subcortical input, and decodable neural patterns emerge in it only during the conscious perception of visual motion; 2. moving stimuli activate medial visual areas but, unless associated with decodable V5 activity, they remain unperceived; 3. ST's high confidence ratings when discriminating motion at chance levels, is associated with inferior frontal gyrus activity. Finally, we report that ST's Riddoch Syndrome results in hallucinatory motion with hippocampal activity as a correlate. Our results shed new light on perceptual experiences associated with this syndrome and on the neural determinants of conscious visual experience.

Gamma oscillations and episodic memory

Enhanced gamma oscillatory activity (30-80 Hz) accompanies the successful formation and retrieval of episodic memories. While this co-occurrence is well documented, the mechanistic contributions of gamma oscillatory activity to episodic memory remain unclear. Here, we review how gamma oscillatory activity may facilitate spike timing-dependent plasticity, neural communication, and sequence encoding/retrieval, thereby ensuring the successful formation and/or retrieval of an episodic memory. Based on the evidence reviewed, we propose that multiple, distinct forms of gamma oscillation can be found within the canonical gamma band, each of which has a complementary role in the neural processes listed above. Further exploration of these theories using causal manipulations may be key to elucidating the relevance of gamma oscillatory activity to episodic memory.

Prior knowledge about events depicted in scenes decreases oculomotor exploration

The visual input that the eyes receive usually contains temporally continuous information about unfolding events. Therefore, humans can accumulate knowledge about their current environment. Typical studies on scene perception, however, involve presenting multiple unrelated images and thereby render this accumulation unnecessary. Our study, instead, facilitated it and explored its effects. Specifically, we investigated how recently-accumulated prior knowledge affects gaze behavior. Participants viewed sequences of static film frames that contained several 'context frames' followed by a 'critical frame'. The context frames showed either events from which the situation depicted in the critical frame naturally followed, or events unrelated to this situation. Therefore, participants viewed identical critical frames while possessing prior knowledge that was either relevant or irrelevant to the frames' content. In the former case, participants' gaze behavior was slightly more exploratory, as revealed by seven gaze characteristics we analyzed. This result demonstrates that recently-gained prior knowledge reduces exploratory eye movements.

Visual Information Is Predictively Encoded in Occipital Alpha/Low-Beta Oscillations

Hierarchical predictive coding networks are a general model of sensory processing in the brain. Under neural delays, these networks have been suggested to naturally generate oscillatory activity in approximately the α frequency range (∼8-12 Hz). This suggests that α oscillations, a prominent feature of EEG recordings, may be a spectral "fingerprint" of predictive sensory processing. Here, we probed this possibility by investigating whether oscillations over the visual cortex predictively encode visual information. Specifically, we examined whether their power carries information about the position of a moving stimulus, in a temporally predictive fashion. In two experiments (N = 32, 18 female; N = 34, 17 female), participants viewed an apparent-motion stimulus moving along a circular path while EEG was recorded. To investigate the encoding of stimulus-position information, we developed a method of deriving probabilistic spatial maps from oscillatory power estimates. With this method, we demonstrate that it is possible to reconstruct the trajectory of a moving stimulus from α/low-β oscillations, tracking its position even across unexpected motion reversals. We also show that future position representations are activated in the absence of direct visual input, demonstrating that temporally predictive mechanisms manifest in α/β band oscillations. In a second experiment, we replicate these findings and show that the encoding of information in this range is not driven by visual entrainment. By demonstrating that occipital α/β oscillations carry stimulus-related information, in a temporally predictive fashion, we provide empirical evidence of these rhythms as a spectral "fingerprint" of hierarchical predictive processing in the human visual system.SIGNIFICANCE STATEMENT "Hierarchical predictive coding" is a general model of sensory information processing in the brain. When in silico predictive coding models are constrained by neural transmission delays, their activity naturally oscillates in roughly the α range (∼8-12 Hz). Using time-resolved EEG decoding, we show that neural rhythms in this approximate range (α/low-β) over the human visual cortex predictively encode the position of a moving stimulus. From the amplitude of these oscillations, we are able to reconstruct the stimulus' trajectory, revealing signatures of temporally predictive processing. This provides direct neural evidence linking occipital α/β rhythms to predictive visual processing, supporting the emerging view of such oscillations as a potential spectral "fingerprint" of hierarchical predictive processing in the human visual system.

Predictive neural representations of naturalistic dynamic input

Adaptive behavior such as social interaction requires our brain to predict unfolding external dynamics. While theories assume such dynamic prediction, empirical evidence is limited to static snapshots and indirect consequences of predictions. We present a dynamic extension to representational similarity analysis that uses temporally variable models to capture neural representations of unfolding events. We applied this approach to source-reconstructed magnetoencephalography (MEG) data of healthy human subjects and demonstrate both lagged and predictive neural representations of observed actions. Predictive representations exhibit a hierarchical pattern, such that high-level abstract stimulus features are predicted earlier in time, while low-level visual features are predicted closer in time to the actual sensory input. By quantifying the temporal forecast window of the brain, this approach allows investigating predictive processing of our dynamic world. It can be applied to other naturalistic stimuli (e.g., film, soundscapes, music, motor planning/execution, social interaction) and any biosignal with high temporal resolution.

Optogenetic manipulation of inhibitory interneurons can be used to validate a model of spatiotemporal sequence learning

The brain uses temporal information to link discrete events into memory structures supporting recognition, prediction, and a wide variety of complex behaviors. It is still an open question how experience-dependent synaptic plasticity creates memories including temporal and ordinal information. Various models have been proposed to explain how this could work, but these are often difficult to validate in a living brain. A recent model developed to explain sequence learning in the visual cortex encodes intervals in recurrent excitatory synapses and uses a learned offset between excitation and inhibition to generate precisely timed "messenger" cells that signal the end of an instance of time. This mechanism suggests that the recall of stored temporal intervals should be particularly sensitive to the activity of inhibitory interneurons that can be easily targeted in vivo with standard optogenetic tools. In this work we examined how simulated optogenetic manipulations of inhibitory cells modifies temporal learning and recall based on these mechanisms. We show that disinhibition and excess inhibition during learning or testing cause characteristic errors in recalled timing that could be used to validate the model in vivo using either physiological or behavioral measurements.