Structure learning and the posterior parietal cortex

A mathematical theory of relational generalization in transitive inference

Humans and animals routinely infer relations between different items or events and generalize these relations to novel combinations of items. This allows them to respond appropriately to radically novel circumstances and is fundamental to advanced cognition. However, how learning systems (including the brain) can implement the necessary inductive biases has been unclear. We investigated transitive inference (TI), a classic relational task paradigm in which subjects must learn a relation ([Formula: see text] and [Formula: see text]) and generalize it to new combinations of items ([Formula: see text]). Through mathematical analysis, we found that a broad range of biologically relevant learning models (e.g. gradient flow or ridge regression) perform TI successfully and recapitulate signature behavioral patterns long observed in living subjects. First, we found that models with item-wise additive representations automatically encode transitive relations. Second, for more general representations, a single scalar "conjunctivity factor" determines model behavior on TI and, further, the principle of norm minimization (a standard statistical inductive bias) enables models with fixed, partly conjunctive representations to generalize transitively. Finally, neural networks in the "rich regime," which enables representation learning and improves generalization on many tasks, unexpectedly show poor generalization and anomalous behavior on TI. We find that such networks implement a form of norm minimization (over hidden weights) that yields a local encoding mechanism lacking transitivity. Our findings show how minimal statistical learning principles give rise to a classical relational inductive bias (transitivity), explain empirically observed behaviors, and establish a formal approach to understanding the neural basis of relational abstraction.

The interplay between spatial and non-spatial grouping cues over approximate number perception

Humans and animals share the cognitive ability to quickly extract approximate number information from sets. Main psychophysical models suggest that visual approximate numerosity relies on segmented units, which can be affected by Gestalt rules. Indeed, arrays containing spatial grouping cues, such as connectedness, closure, and even symmetry, are underestimated compared to ungrouped arrays with equal low-level features. Recent evidence suggests that non-spatial cues, such as color-similarity, also trigger numerosity underestimation. However, in natural vision, several grouping cues may coexist in the scene. Notably, conjunction of grouping cues (color and closure) reduces perceived numerosity following an additive rule. To test whether the conjunction-effect holds for other Gestalt cues, we investigated the effect of connectedness and symmetry over numerosity perception both in isolation and, critically, in conjunction with luminance similarity. Participants performed a comparison-task between a reference and a test stimulus varying in numerosity. In Experiment 1, test stimuli contained two isolated groupings (connectedness or luminance), a conjunction (connectedness and luminance), and a neutral condition (no groupings). Results show that point of subjective equality was higher in both isolated grouping conditions compared to the neutral condition. Furthermore, in the conjunction condition, the biases from isolated grouping cues added linearly, resulting in a numerosity underestimation equal to the sum of the isolated biases. In Experiment 2 we found that conjunction of symmetry and luminance followed the same additive rule. These findings strongly suggest that both spatial and non-spatial isolated cues affect numerosity perception. Crucially, we show that their conjunction effect extends to symmetry and connectedness.

The relational bottleneck as an inductive bias for efficient abstraction

A central challenge for cognitive science is to explain how abstract concepts are acquired from limited experience. This has often been framed in terms of a dichotomy between connectionist and symbolic cognitive models. Here, we highlight a recently emerging line of work that suggests a novel reconciliation of these approaches, by exploiting an inductive bias that we term the relational bottleneck. In that approach, neural networks are constrained via their architecture to focus on relations between perceptual inputs, rather than the attributes of individual inputs. We review a family of models that employ this approach to induce abstractions in a data-efficient manner, emphasizing their potential as candidate models for the acquisition of abstract concepts in the human mind and brain.

Spontaneous eye movements reflect the representational geometries of conceptual spaces

Functional neuroimaging studies indicate that the human brain can represent concepts and their relational structure in memory using coding schemes typical of spatial navigation. However, whether we can read out the internal representational geometries of conceptual spaces solely from human behavior remains unclear. Here, we report that the relational structure between concepts in memory might be reflected in spontaneous eye movements during verbal fluency tasks: When we asked participants to randomly generate numbers, their eye movements correlated with distances along the left-to-right one-dimensional geometry of the number space (mental number line), while they scaled with distance along the ring-like two-dimensional geometry of the color space (color wheel) when they randomly generated color names. Moreover, when participants randomly produced animal names, eye movements correlated with low-dimensional similarity in word frequencies. These results suggest that the representational geometries used to internally organize conceptual spaces might be read out from gaze behavior.

Space as a scaffold for rotational generalisation of abstract concepts

Learning invariances allows us to generalise. In the visual modality, invariant representations allow us to recognise objects despite translations or rotations in physical space. However, how we learn the invariances that allow us to generalise abstract patterns of sensory data ('concepts') is a longstanding puzzle. Here, we study how humans generalise relational patterns in stimulation sequences that are defined by either transitions on a nonspatial two-dimensional feature manifold, or by transitions in physical space. We measure rotational generalisation, i.e., the ability to recognise concepts even when their corresponding transition vectors are rotated. We find that humans naturally generalise to rotated exemplars when stimuli are defined in physical space, but not when they are defined as positions on a nonspatial feature manifold. However, if participants are first pre-trained to map auditory or visual features to spatial locations, then rotational generalisation becomes possible even in nonspatial domains. These results imply that space acts as a scaffold for learning more abstract conceptual invariances.

Emergent neural dynamics and geometry for generalization in a transitive inference task

Relational cognition-the ability to infer relationships that generalize to novel combinations of objects-is fundamental to human and animal intelligence. Despite this importance, it remains unclear how relational cognition is implemented in the brain due in part to a lack of hypotheses and predictions at the levels of collective neural activity and behavior. Here we discovered, analyzed, and experimentally tested neural networks (NNs) that perform transitive inference (TI), a classic relational task (if A > B and B > C, then A > C). We found NNs that (i) generalized perfectly, despite lacking overt transitive structure prior to training, (ii) generalized when the task required working memory (WM), a capacity thought to be essential to inference in the brain, (iii) emergently expressed behaviors long observed in living subjects, in addition to a novel order-dependent behavior, and (iv) expressed different task solutions yielding alternative behavioral and neural predictions. Further, in a large-scale experiment, we found that human subjects performing WM-based TI showed behavior inconsistent with a class of NNs that characteristically expressed an intuitive task solution. These findings provide neural insights into a classical relational ability, with wider implications for how the brain realizes relational cognition.

Identifying content-invariant neural signatures of perceptual vividness

Some conscious experiences are more vivid than others. Although perceptual vividness is a key component of human consciousness, how variation in this magnitude property is registered by the human brain is unknown. A striking feature of neural codes for magnitude in other psychological domains, such as number or reward, is that the magnitude property is represented independently of its sensory features. To test whether perceptual vividness also covaries with neural codes that are invariant to sensory content, we reanalyzed existing magnetoencephalography and functional MRI data from two distinct studies which quantified perceptual vividness via subjective ratings of awareness and visibility. Using representational similarity and decoding analyses, we find evidence for content-invariant neural signatures of perceptual vividness distributed across visual, parietal, and frontal cortices. Our findings indicate that the neural correlates of subjective vividness may share similar properties to magnitude codes in other cognitive domains.

Electrophysiological hallmarks for event relations and event roles in working memory

The ability to maintain events (i.e., interactions between/among objects) in working memory is crucial for our everyday cognition, yet the format of this representation is poorly understood. The current ERP study was designed to answer two questions: How is maintaining events (e.g., the tiger hit the lion) neurally different from maintaining item coordinations (e.g., the tiger and the lion)? That is, how is the event relation (present in events but not coordinations) represented? And how is the agent, or initiator of the event encoded differently from the patient, or receiver of the event during maintenance? We used a novel picture-sentence match-across-delay approach in which the working memory representation was "pinged" during the delay, replicated across two ERP experiments with Chinese and English materials. We found that maintenance of events elicited a long-lasting late sustained difference in posterior-occipital electrodes relative to non-events. This effect resembled the negative slow wave reported in previous studies of working memory, suggesting that the maintenance of events in working memory may impose a higher cost compared to coordinations. Although we did not observe significant ERP differences associated with pinging the agent vs. the patient during the delay, we did find that the ping appeared to dampen the ongoing sustained difference, suggesting a shift from sustained activity to activity silent mechanisms. These results suggest a new method by which ERPs can be used to elucidate the format of neural representation for events in working memory.

Distinct processing of the state prediction error signals in frontal and parietal correlates in learning the environment model

Goal-directed reinforcement learning constructs a model of how the states in the environment are connected and prospectively evaluates action values by simulating experience. State prediction error (SPE) is theorized as a crucial signal for learning the environment model. However, the underlying neural mechanisms remain unclear. Here, using electroencephalogram, we verified in a two-stage Markov task two neural correlates of SPEs: an early negative correlate transferring from frontal to central electrodes and a late positive correlate over parietal regions. Furthermore, by investigating the effects of explicit knowledge about the environment model and rewards in the environment, we found that, for the parietal correlate, rewards enhanced the representation efficiency (beta values of regression coefficient) of SPEs, whereas explicit knowledge elicited a larger SPE representation (event-related potential activity) for rare transitions. However, for the frontal and central correlates, rewards increased activities in a content-independent way and explicit knowledge enhanced activities only for common transitions. Our results suggest that the parietal correlate of SPEs is responsible for the explicit learning of state transition structure, whereas the frontal and central correlates may be involved in cognitive control. Our study provides novel evidence for distinct roles of the frontal and the parietal cortices in processing SPEs.

Brain-imaging evidence for compression of binary sound sequences in human memory

According to the language-of-thought hypothesis, regular sequences are compressed in human memory using recursive loops akin to a mental program that predicts future items. We tested this theory by probing memory for 16-item sequences made of two sounds. We recorded brain activity with functional MRI and magneto-encephalography (MEG) while participants listened to a hierarchy of sequences of variable complexity, whose minimal description required transition probabilities, chunking, or nested structures. Occasional deviant sounds probed the participants' knowledge of the sequence. We predicted that task difficulty and brain activity would be proportional to the complexity derived from the minimal description length in our formal language. Furthermore, activity should increase with complexity for learned sequences, and decrease with complexity for deviants. These predictions were upheld in both fMRI and MEG, indicating that sequence predictions are highly dependent on sequence structure and become weaker and delayed as complexity increases. The proposed language recruited bilateral superior temporal, precentral, anterior intraparietal, and cerebellar cortices. These regions overlapped extensively with a localizer for mathematical calculation, and much less with spoken or written language processing. We propose that these areas collectively encode regular sequences as repetitions with variations and their recursive composition into nested structures.

Propagation of activity through the cortical hierarchy and perception are determined by neural variability

Brains are composed of anatomically and functionally distinct regions performing specialized tasks, but regions do not operate in isolation. Orchestration of complex behaviors requires communication between brain regions, but how neural dynamics are organized to facilitate reliable transmission is not well understood. Here we studied this process directly by generating neural activity that propagates between brain regions and drives behavior, assessing how neural populations in sensory cortex cooperate to transmit information. We achieved this by imaging two densely interconnected regions-the primary and secondary somatosensory cortex (S1 and S2)-in mice while performing two-photon photostimulation of S1 neurons and assigning behavioral salience to the photostimulation. We found that the probability of perception is determined not only by the strength of the photostimulation but also by the variability of S1 neural activity. Therefore, maximizing the signal-to-noise ratio of the stimulus representation in cortex relative to the noise or variability is critical to facilitate activity propagation and perception.

Disentangling Abstraction from Statistical Pattern Matching in Human and Machine Learning

The ability to acquire abstract knowledge is a hallmark of human intelligence and is believed by many to be one of the core differences between humans and neural network models. Agents can be endowed with an inductive bias towards abstraction through meta-learning, where they are trained on a distribution of tasks that share some abstract structure that can be learned and applied. However, because neural networks are hard to interpret, it can be difficult to tell whether agents have learned the underlying abstraction, or alternatively statistical patterns that are characteristic of that abstraction. In this work, we compare the performance of humans and agents in a meta-reinforcement learning paradigm in which tasks are generated from abstract rules. We define a novel methodology for building "task metamers" that closely match the statistics of the abstract tasks but use a different underlying generative process, and evaluate performance on both abstract and metamer tasks. We find that humans perform better at abstract tasks than metamer tasks whereas common neural network architectures typically perform worse on the abstract tasks than the matched metamers. This work provides a foundation for characterizing differences between humans and machine learning that can be used in future work towards developing machines with more human-like behavior.

Goal-oriented representations in the human hippocampus during planning and navigation

Recent work in cognitive and systems neuroscience has suggested that the hippocampus might support planning, imagination, and navigation by forming cognitive maps that capture the abstract structure of physical spaces, tasks, and situations. Navigation involves disambiguating similar contexts, and the planning and execution of a sequence of decisions to reach a goal. Here, we examine hippocampal activity patterns in humans during a goal-directed navigation task to investigate how contextual and goal information are incorporated in the construction and execution of navigational plans. During planning, hippocampal pattern similarity is enhanced across routes that share a context and a goal. During navigation, we observe prospective activation in the hippocampus that reflects the retrieval of pattern information related to a key-decision point. These results suggest that, rather than simply representing overlapping associations or state transitions, hippocampal activity patterns are shaped by context and goals.

Neural knowledge assembly in humans and neural networks

Human understanding of the world can change rapidly when new information comes to light, such as when a plot twist occurs in a work of fiction. This flexible "knowledge assembly" requires few-shot reorganization of neural codes for relations among objects and events. However, existing computational theories are largely silent about how this could occur. Here, participants learned a transitive ordering among novel objects within two distinct contexts before exposure to new knowledge that revealed how they were linked. Blood-oxygen-level-dependent (BOLD) signals in dorsal frontoparietal cortical areas revealed that objects were rapidly and dramatically rearranged on the neural manifold after minimal exposure to linking information. We then adapt online stochastic gradient descent to permit similar rapid knowledge assembly in a neural network model.

Neurocomputational mechanisms underlying fear-biased adaptation learning in changing environments

Humans are able to adapt to the fast-changing world by estimating statistical regularities of the environment. Although fear can profoundly impact adaptive behaviors, the computational and neural mechanisms underlying this phenomenon remain elusive. Here, we conducted a behavioral experiment (n = 21) and a functional magnetic resonance imaging experiment (n = 37) with a novel cue-biased adaptation learning task, during which we simultaneously manipulated emotional valence (fearful/neutral expressions of the cue) and environmental volatility (frequent/infrequent reversals of reward probabilities). Across 2 experiments, computational modeling consistently revealed a higher learning rate for the environment with frequent versus infrequent reversals following neutral cues. In contrast, this flexible adjustment was absent in the environment with fearful cues, suggesting a suppressive role of fear in adaptation to environmental volatility. This suppressive effect was underpinned by activity of the ventral striatum, hippocampus, and dorsal anterior cingulate cortex (dACC) as well as increased functional connectivity between the dACC and temporal-parietal junction (TPJ) for fear with environmental volatility. Dynamic causal modeling identified that the driving effect was located in the TPJ and was associated with dACC activation, suggesting that the suppression of fear on adaptive behaviors occurs at the early stage of bottom-up processing. These findings provide a neuro-computational account of how fear interferes with adaptation to volatility during dynamic environments.

Evolution of neural activity in circuits bridging sensory and abstract knowledge

The ability to associate sensory stimuli with abstract classes is critical for survival. How are these associations implemented in brain circuits? And what governs how neural activity evolves during abstract knowledge acquisition? To investigate these questions, we consider a circuit model that learns to map sensory input to abstract classes via gradient-descent synaptic plasticity. We focus on typical neuroscience tasks (simple, and context-dependent, categorization), and study how both synaptic connectivity and neural activity evolve during learning. To make contact with the current generation of experiments, we analyze activity via standard measures such as selectivity, correlations, and tuning symmetry. We find that the model is able to recapitulate experimental observations, including seemingly disparate ones. We determine how, in the model, the behaviour of these measures depends on details of the circuit and the task. These dependencies make experimentally testable predictions about the circuitry supporting abstract knowledge acquisition in the brain.

From cognitive maps to spatial schemas

A schema refers to a structured body of prior knowledge that captures common patterns across related experiences. Schemas have been studied separately in the realms of episodic memory and spatial navigation across different species and have been grounded in theories of memory consolidation, but there has been little attempt to integrate our understanding across domains, particularly in humans. We propose that experiences during navigation with many similarly structured environments give rise to the formation of spatial schemas (for example, the expected layout of modern cities) that share properties with but are distinct from cognitive maps (for example, the memory of a modern city) and event schemas (such as expected events in a modern city) at both cognitive and neural levels. We describe earlier theoretical frameworks and empirical findings relevant to spatial schemas, along with more targeted investigations of spatial schemas in human and non-human animals. Consideration of architecture and urban analytics, including the influence of scale and regionalization, on different properties of spatial schemas may provide a powerful approach to advance our understanding of spatial schemas.

The precuneus as a central node in declarative memory retrieval

Both, the hippocampal formation and the neocortex are contributing to declarative memory, but their functional specialization remains unclear. We investigated the differential contribution of both memory systems during free recall of word lists. In total, 21 women and 17 men studied the same list but with the help of different encoding associations. Participants associated the words either sequentially with the previous word on the list, with spatial locations on a well-known path, or with unique autobiographical events. After intensive rehearsal, subjects recalled the words during functional magnetic resonance imaging (fMRI). Common activity to all three types of encoding associations was identified in the posterior parietal cortex, in particular in the precuneus. Additionally, when associating spatial or autobiographical material, retrosplenial cortex activity was elicited during word list recall, while hippocampal activity emerged only for autobiographically associated words. These findings support a general, critical function of the precuneus in episodic memory storage and retrieval. The encoding-retrieval repetitions during learning seem to have accelerated hippocampus-independence and lead to direct neocortical integration in the sequentially associated and spatially associated word list tasks. During recall of words associated with autobiographical memories, the hippocampus might add spatiotemporal information supporting detailed scenic and contextual memories.

Reactivating ordinal position information from auditory sequence memory in human brains

Retaining a sequence of events in their order is a core ability of many cognitive functions, such as speech recognition, movement control, and episodic memory. Although content representations have been widely studied in working memory (WM), little is known about how ordinal position information of an auditory sequence is retained in the human brain as well as its coding characteristics. In fact, there is still a lack of an efficient approach to directly accessing the stored ordinal position code during WM retention. Here, 31 participants performed an auditory sequence WM task with their brain activities recorded using electroencephalography (EEG). We developed new triggering events that could successfully reactivate neural representations of ordinal position during the delay period. Importantly, the ordinal position reactivation is further related to recognition behavior, confirming its indexing of WM storage. Furthermore, the ordinal position code displays an intriguing “stable-dynamic” format, i.e. undergoing the same dynamic neutral trajectory in the multivariate neural space during both encoding and retention (whenever reactivated). Overall, our results provide an effective approach to accessing the behaviorally-relevant ordinal position information in auditory sequence WM and reveal its new temporal characteristics.

Intelligence Correlates with the Temporal Variability of Brain Networks

Intelligence is the ability to recognize and understand objective things, and use knowledge and experience to solve problems. Highly intelligent people show the ability to switch between different thought patterns and shift their mental focus. This suggests a link between intelligence and the dynamic interaction of brain networks. Thus, we investigated the relationships between resting-state dynamic brain network remodeling (temporal variability) and scores on the Wechsler Adult Intelligent Scale using a large dataset comprising 606 individuals. We found that performance intelligence was associated with greater temporal variability in the functional connectivity patterns of the dorsal attention network. High variability in these areas indicates flexible connectivity patterns, which may contribute to cognitive processes such as attention selection. In addition, performance intelligence was related to greater temporal variability in the functional connectivity patterns of the salience network. Thus, this study revealed a close relationship between performance intelligence and high variability in brain networks involved in attentional choice, spatial orientation, and cognitive control.

Mnemonic construction and representation of temporal structure in the hippocampal formation

The hippocampal-entorhinal region supports memory for episodic details, such as temporal relations of sequential events, and mnemonic constructions combining experiences for inferential reasoning. However, it is unclear whether hippocampal event memories reflect temporal relations derived from mnemonic constructions, event order, or elapsing time, and whether these sequence representations generalize temporal relations across similar sequences. Here, participants mnemonically constructed times of events from multiple sequences using infrequent cues and their experience of passing time. After learning, event representations in the anterior hippocampus reflected temporal relations based on constructed times. Temporal relations were generalized across sequences, revealing distinct representational formats for events from the same or different sequences. Structural knowledge about time patterns, abstracted from different sequences, biased the construction of specific event times. These findings demonstrate that mnemonic construction and the generalization of relational knowledge combine in the hippocampus, consistent with the simulation of scenarios from episodic details and structural knowledge.

Activity patterns in the hippocampus resemble temporal relations of learned event sequences. Here, the authors show that these relational memories arise through mnemonic construction and are generalized to reflect the temporal event structure.

Effects of Online Single Pulse Transcranial Magnetic Stimulation on Prefrontal and Parietal Cortices in Deceptive Processing: A Preliminary Study

Transcranial magnetic stimulation (TMS) was used to test the functional role of parietal and prefrontal cortical regions activated during a playing card Guilty Knowledge Task (GKT). Single-pulse TMS was applied to 15 healthy volunteers at each of three target sites: left and right dorsolateral prefrontal cortex and midline parietal cortex. TMS pulses were applied at each of five latencies (from 0 to 480 ms) after the onset of a card stimulus. TMS applied to the parietal cortex exerted a latency-specific increase in inverse efficiency score and in reaction time when subjects were instructed to lie relative to when asked to respond with the truth, and this effect was specific to when TMS was applied at 240 ms after stimulus onset. No effects of TMS were detected at left or right DLPFC sites. This manipulation with TMS of performance in a deception task appears to support a critical role for the parietal cortex in intentional false responding, particularly in stimulus selection processes needed to execute a deceptive response in the context of a GKT. However, this interpretation is only preliminary, as further experiments are needed to compare performance within and outside of a deceptive context to clarify the effects of deceptive intent.

Neural correlates of transitive inference: An SDM meta-analysis on 32 fMRI studies

Transitive inference (TI) is a critical capacity involving the integration of relevant information into prior knowledge structure for drawing novel inferences on unobserved relationships. To date, the neural correlates of TI remain unclear due to the small sample size and heterogeneity of various experimental tasks from individual studies. Here, the meta-analysis on 32 fMRI studies was performed to detect brain activation patterns of TI and its three paradigms (spatial inference, hierarchical inference, and associative inference). We found the hippocampus, prefrontal cortex (PFC), putamen, posterior parietal cortex (PPC), retrosplenial cortex (RSC), supplementary motor area (SMA), precentral gyrus (PreCG), and median cingulate cortex (MCC) were engaged in TI. Specifically, the RSC was implicated in the associative inference, whereas PPC, SMA, PreCG, and MCC were implicated in the hierarchical inference. In addition, the hierarchical inference and associative inference both evoked activation in the hippocampus, medial PFC, and PCC. Although the meta-analysis on spatial inference did not generate a reliable result due to insufficient amount of investigations, the present work still offers a new insight for better understanding the neural basis underlying TI.

Adaptive cognitive maps for curved surfaces in the 3D world

Terrains in a 3D world can be undulating. Yet, most prior research has exclusively investigated spatial representations on a flat surface, leaving a 2D cognitive map as the dominant model in the field. Here, we investigated whether humans represent a curved surface by building a dimension-reduced flattened 2D map or a full 3D map. Participants learned the location of objects positioned on a flat and curved surface in a virtual environment by driving on the concave side of the surface (Experiment 1), driving and looking vertically (Experiment 2), or flying (Experiment 3). Subsequently, they were asked to retrieve either the path distance or the 3D Euclidean distance between the objects. Path distance estimation was good overall, but we found a significant underestimation bias for the path distance on the curve, suggesting an influence of potential 3D shortcuts, even though participants were only driving on the surface. Euclidean distance estimation was better when participants were exposed more to the global 3D structure of the environment by looking and flying. These results suggest that the representation of the 2D manifold, embedded in a 3D world, is neither purely 2D nor 3D. Rather, it is flexible and dependent on the behavioral experience and demand.

The Structure of Systematicity in the Brain

A hallmark of human intelligence is the ability to adapt to new situations, by applying learned rules to new content (systematicity) and thereby enabling an open-ended number of inferences and actions (generativity). Here, we propose that the human brain accomplishes these feats through pathways in the parietal cortex that encode the abstract structure of space, events, and tasks, and pathways in the temporal cortex that encode information about specific people, places, and things (content). Recent neural network models show how the separation of structure and content might emerge through a combination of architectural biases and learning, and these networks show dramatic improvements in the ability to capture systematic, generative behavior. We close by considering how the hippocampal formation may form integrative memories that enable rapid learning of new structure and content representations.

The neural representation of absolute direction during mental navigation in conceptual spaces

When humans mentally “navigate” bidimensional uniform conceptual spaces, they recruit the same grid-like and distance codes typically evoked when exploring the physical environment. Here, using fMRI, we show evidence that conceptual navigation also elicits another kind of spatial code: that of absolute direction. This code is mostly localized in the medial parietal cortex, where its strength predicts participants’ comparative semantic judgments. It may provide a complementary mechanism for conceptual navigation outside the hippocampal formation.

Viganò et al. use fMRI in healthy human participants to show that conceptual navigation elicits a spatial code for absolute direction in the medial parietal cortex. Their findings are suggestive of a complementary mechanism for conceptual navigation outside the hippocampal formation.

Distinct Neural Representations of Content and Ordinal Structure in Auditory Sequence Memory

Two forms of information, frequency (content) and ordinal position (structure), have to be stored when retaining a sequence of auditory tones in working memory (WM). However, the neural representations and coding characteristics of content and structure, particularly during WM maintenance, remain elusive. Here, in two EEG studies in human participants (both sexes), by transiently perturbing the "activity-silent" WM retention state and decoding the reactivated WM information, we demonstrate that content and structure are stored in a dissociative manner with distinct characteristics throughout WM process. First, each tone in the sequence is associated with two codes in parallel, characterizing its frequency and ordinal position, respectively. Second, during retention, a structural retrocue successfully reactivates structure but not content, whereas a following white noise triggers content but not structure. Third, structure representation remains stable, whereas content code undergoes a dynamic transformation through memory progress. Finally, the noise-triggered content reactivations during retention correlate with subsequent WM behavior. Overall, our results support distinct content and structure representations in auditory WM and provide an efficient approach to access the silently stored WM information in the human brain. The dissociation of content and structure could facilitate efficient memory formation via generalizing stable structure to new auditory contents.SIGNIFICANCE STATEMENT In memory experiences, contents do not exist independently but are linked with each other via ordinal structure. For instance, recalling a piece of favorite music relies on correct ordering (sequence structure) of musical tones (content). How are the structure and content for an auditory temporally structured experience maintained in working memory? Here, by using impulse-response approach and time-resolved representational dissimilarity analysis on human EEG recordings in an auditory working memory task, we reveal that content and structure are stored in a dissociated way, which would facilitate efficient and rapid memory formation through generalizing stable structure knowledge to new auditory inputs.

Clone-structured graph representations enable flexible learning and vicarious evaluation of cognitive maps

Cognitive maps are mental representations of spatial and conceptual relationships in an environment, and are critical for flexible behavior. To form these abstract maps, the hippocampus has to learn to separate or merge aliased observations appropriately in different contexts in a manner that enables generalization and efficient planning. Here we propose a specific higher-order graph structure, clone-structured cognitive graph (CSCG), which forms clones of an observation for different contexts as a representation that addresses these problems. CSCGs can be learned efficiently using a probabilistic sequence model that is inherently robust to uncertainty. We show that CSCGs can explain a variety of cognitive map phenomena such as discovering spatial relations from aliased sensations, transitive inference between disjoint episodes, and formation of transferable schemas. Learning different clones for different contexts explains the emergence of splitter cells observed in maze navigation and event-specific responses in lap-running experiments. Moreover, learning and inference dynamics of CSCGs offer a coherent explanation for disparate place cell remapping phenomena. By lifting aliased observations into a hidden space, CSCGs reveal latent modularity useful for hierarchical abstraction and planning. Altogether, CSCG provides a simple unifying framework for understanding hippocampal function, and could be a pathway for forming relational abstractions in artificial intelligence.

Neural state space alignment for magnitude generalization in humans and recurrent networks

A prerequisite for intelligent behavior is to understand how stimuli are related and to generalize this knowledge across contexts. Generalization can be challenging when relational patterns are shared across contexts but exist on different physical scales. Here, we studied neural representations in humans and recurrent neural networks performing a magnitude comparison task, for which it was advantageous to generalize concepts of "more" or "less" between contexts. Using multivariate analysis of human brain signals and of neural network hidden unit activity, we observed that both systems developed parallel neural "number lines" for each context. In both model systems, these number state spaces were aligned in a way that explicitly facilitated generalization of relational concepts (more and less). These findings suggest a previously overlooked role for neural normalization in supporting transfer of a simple form of abstract relational knowledge (magnitude) in humans and machine learning systems.

Using deep reinforcement learning to reveal how the brain encodes abstract state-space representations in high-dimensional environments

Humans possess an exceptional aptitude to efficiently make decisions from high-dimensional sensory observations. However, it is unknown how the brain compactly represents the current state of the environment to guide this process. The deep Q-network (DQN) achieves this by capturing highly nonlinear mappings from multivariate inputs to the values of potential actions. We deployed DQN as a model of brain activity and behavior in participants playing three Atari video games during fMRI. Hidden layers of DQN exhibited a striking resemblance to voxel activity in a distributed sensorimotor network, extending throughout the dorsal visual pathway into posterior parietal cortex. Neural state-space representations emerged from nonlinear transformations of the pixel space bridging perception to action and reward. These transformations reshape axes to reflect relevant high-level features and strip away information about task-irrelevant sensory features. Our findings shed light on the neural encoding of task representations for decision-making in real-world situations.

Boys' advantage on the fractions number line is mediated by visuospatial attention: Evidence for a parietal-spatial contribution to number line learning

The study tested the hypotheses that boys will have an advantage learning the fractions number line and this advantage will be mediated by spatial abilities. Fractions number line and, as a contrast, fractions arithmetic performance were assessed for 342 adolescents, as was their intelligence, working memory, and various spatial abilities. Boys showed smaller placement errors on the fractions number line (d = -0.22) and correctly solved more fractions arithmetic problems (d = 0.23) than girls. Working memory and intelligence predicted performance on both fractions measures, and a measure of visuospatial attention uniquely predicted number line performance and fully mediated the sex difference. Visuospatial working memory uniquely predicted fractions arithmetic performance and fully mediated the sex difference. The results help to clarify the nuanced relations between spatial abilities and formal mathematics learning and the sex differences that often emerge in mathematical domains that have a visuospatial component.

Integrating when and what information in the left parietal lobe allows language rule generalization

A crucial aspect when learning a language is discovering the rules that govern how words are combined in order to convey meanings. Because rules are characterized by sequential co-occurrences between elements (e.g., “These cupcakes are unbelievable”), tracking the statistical relationships between these elements is fundamental. However, purely bottom-up statistical learning alone cannot fully account for the ability to create abstract rule representations that can be generalized, a paramount requirement of linguistic rules. Here, we provide evidence that, after the statistical relations between words have been extracted, the engagement of goal-directed attention is key to enable rule generalization. Incidental learning performance during a rule-learning task on an artificial language revealed a progressive shift from statistical learning to goal-directed attention. In addition, and consistent with the recruitment of attention, functional MRI (fMRI) analyses of late learning stages showed left parietal activity within a broad bilateral dorsal frontoparietal network. Critically, repetitive transcranial magnetic stimulation (rTMS) on participants’ peak of activation within the left parietal cortex impaired their ability to generalize learned rules to a structurally analogous new language. No stimulation or rTMS on a nonrelevant brain region did not have the same interfering effect on generalization. Performance on an additional attentional task showed that this rTMS on the parietal site hindered participants’ ability to integrate “what” (stimulus identity) and “when” (stimulus timing) information about an expected target. The present findings suggest that learning rules from speech is a two-stage process: following statistical learning, goal-directed attention—involving left parietal regions—integrates “what” and “when” stimulus information to facilitate rapid rule generalization.

This study uses repetitive transcranial stimulation to show that learning language rules from speech is a two-stage process; following statistical learning, goal-directed attention (involving left parietal regions) integrates "what" and "when" stimulus information to facilitate rapid rule generalization.

The growth and form of knowledge networks by kinesthetic curiosity

Throughout life, we might seek a calling, companions, skills, entertainment, truth, self-knowledge, beauty, and edification. The practice of curiosity can be viewed as an extended and open-ended search for valuable information with hidden identity and location in a complex space of interconnected information. Despite its importance, curiosity has been challenging to computationally model because the practice of curiosity often flourishes without specific goals, external reward, or immediate feedback. Here, we show how network science, statistical physics, and philosophy can be integrated into an approach that coheres with and expands the psychological taxonomies of specific-diversive and perceptual-epistemic curiosity. Using this interdisciplinary approach, we distill functional modes of curious information seeking as searching movements in information space. The kinesthetic model of curiosity offers a vibrant counterpart to the deliberative predictions of model-based reinforcement learning. In doing so, this model unearths new computational opportunities for identifying what makes curiosity curious.

Knowledge Across Reference Frames: Cognitive Maps and Image Spaces

In human and non-human animals, conceptual knowledge is partially organized according to low-dimensional geometries that rely on brain structures and computations involved in spatial representations. Recently, two separate lines of research have investigated cognitive maps, that are associated with the hippocampal formation and are similar to world-centered representations of the environment, and image spaces, that are associated with the parietal cortex and are similar to self-centered spatial relationships. We review evidence supporting cognitive maps and image spaces, and we propose a hippocampal-parietal network that can account for the organization and retrieval of knowledge across multiple reference frames. We also suggest that cognitive maps and image spaces may be two manifestations of a more general propensity of the mind to create low-dimensional internal models.