Abstract Context Representations in Primate Amygdala and Prefrontal Cortex

Reward recalibrates rule representations in human amygdala and hippocampus intracranial recordings

Adaptive behavior requires the ability to shift responding within (intra-dimensional) or between (extra-dimensional) stimulus dimensions when reward contingencies change. Studies of shifting in humans have focused mainly on the prefrontal cortex and/ or have been restricted to indirect measures of neural activity such as fMRI and lesions. Here, we demonstrate the importance of the amygdala and hippocampus by recording local field potentials directly from these regions intracranially in human epilepsy patients. Reward signals were coded in the high frequency gamma activity (HFG; 60-250 Hz) of both regions and synchronised via low frequency (3-5 Hz) phase-locking only after a shift when patients did not already know the rule and it signalled to stop shifting ("Win-Stay"). In contrast, HFG punishment signals were only seen in the amygdala when the rule then changed and it signalled to start shifting ("Lose-Shift"). During decision-making, hippocampal HFG was more inhibited on non-shift relative to shift trials, suggesting a role in preventing interference in rule representation and amygdala HFG was sensitive to stimulus novelty. The findings expand our understanding of human amygdala-hippocampal function and shifting processes, the disruption of which could contribute to shifting deficits in neuropsychiatric disorders.

Social status as a latent variable in the amygdala of observers of social interactions

Successful integration into a hierarchical social group requires knowledge of the status of each individual and of the rules that govern social interactions within the group. In species that lack morphological indicators of status, social status can be inferred by observing the signals exchanged between individuals. We simulated social interactions between macaques by juxtaposing videos of aggressive and appeasing displays, where two individuals appeared in each other's line of sight and their displays were timed to suggest the reciprocation of dominant and subordinate signals. Viewers of these videos successfully inferred the social status of the interacting characters. Dominant individuals attracted more social attention from viewers even when they were not engaged in social displays. Neurons in the viewers' amygdala signaled the status of both the attended (fixated) and the unattended individuals, suggesting that in third-party observers of social interactions, the amygdala jointly signals the status of interacting parties.

Genetically- and spatially-defined basolateral amygdala neurons control food consumption and social interaction

The basolateral amygdala (BLA) contains discrete neuronal circuits that integrate positive or negative emotional information and drive the appropriate innate and learned behaviors. Whether these circuits consist of genetically-identifiable and anatomically segregated neuron types, is poorly understood. Also, our understanding of the response patterns and behavioral spectra of genetically-identifiable BLA neurons is limited. Here, we classified 11 glutamatergic cell clusters in mouse BLA and found that several of them were anatomically segregated in lateral versus basal amygdala, and anterior versus posterior regions of the BLA. Two of these BLA subpopulations innately responded to valence-specific, whereas one responded to mixed - aversive and social - cues. Positive-valence BLA neurons promoted normal feeding, while mixed selectivity neurons promoted fear learning and social interactions. These findings enhance our understanding of cell type diversity and spatial organization of the BLA and the role of distinct BLA populations in representing valence-specific and mixed stimuli.

Neural dynamics underlying minute-timescale persistent behavior in the human brain

The ability to pursue long-term goals relies on a representations of task context that can both be maintained over long periods of time and switched flexibly when goals change. Little is known about the neural substrate for such minute-scale maintenance of task sets. Utilizing recordings in neurosurgical patients, we examined how groups of neurons in the human medial frontal cortex and hippocampus represent task contexts. When cued explicitly, task context was encoded in both brain areas and changed rapidly at task boundaries. Hippocampus exhibited a temporally dynamic code with fast decorrelation over time, preventing cross-temporal generalization. Medial frontal cortex exhibited a static code that decorrelated slowly, allowing generalization across minutes of time. When task context needed to be inferred as a latent variable, hippocampus encoded task context with a static code. These findings reveal two possible regimes for encoding minute-scale task-context representations that were engaged differently based on task demands.

Social Status as a Latent Variable in the Amygdala of Observers of Social Interactions

Successful integration into a hierarchical social group requires knowledge of the status of each individual and of the rules that govern social interactions within the group. In species that lack morphological indicators of status, social status can be inferred by observing the signals exchanged between individuals. We simulated social interactions between macaques by juxtaposing videos of aggressive and appeasing displays where two individuals appeared in each other's line of sight and their displays were timed to suggest the reciprocation of dominant and subordinate signals. Viewers of these videos successfully inferred the social status of the interacting characters. Dominant individuals attracted more social attention from viewers even when they were not engaged in social displays. Neurons in the viewers' amygdala signaled the status of both the attended (fixated) and the unattended individuals suggesting that in third party observers of social interactions, the amygdala signals jointly the status of interacting parties.

Highlights

Monkeys infer the social status of conspecifics from videos of simulated dyadic interactionsDuring fixations neural populations signal the social status of the attended individualsNeurons in the amygdala jointly encode the status of interacting individuals.

In brief

Third party-viewers of pairwise dominant-subordinate interactions infer social status from the observed behaviors. Neurons in the amygdala are tuned to the inferred dominant/subordinate status of both individuals.

Preferences reveal dissociable encoding across prefrontal-limbic circuits

Individual preferences for the flavor of different foods and fluids exert a strong influence on behavior. Most current theories posit that preferences are integrated with other state variables in the orbitofrontal cortex (OFC), which is thought to derive the relative subjective value of available options to guide choice behavior. Here, we report that instead of a single integrated valuation system in the OFC, another complementary one is centered in the ventrolateral prefrontal cortex (vlPFC) in macaques. Specifically, we found that the OFC and vlPFC preferentially represent outcome flavor and outcome probability, respectively, and that preferences are separately integrated into value representations in these areas. In addition, the vlPFC, but not the OFC, represented the probability of receiving the available outcome flavors separately, with the difference between these representations reflecting the degree of preference for each flavor. Thus, both the vlPFC and OFC exhibit dissociable but complementary representations of subjective value, both of which are necessary for decision-making.

Computational role of structure in neural activity and connectivity

One major challenge of neuroscience is identifying structure in seemingly disorganized neural activity. Different types of structure have different computational implications that can help neuroscientists understand the functional role of a particular brain area. Here, we outline a unified approach to characterize structure by inspecting the representational geometry and the modularity properties of the recorded activity and show that a similar approach can also reveal structure in connectivity. We start by setting up a general framework for determining geometry and modularity in activity and connectivity and relating these properties with computations performed by the network. We then use this framework to review the types of structure found in recent studies of model networks performing three classes of computations.

Detection of latent brain states from baseline neural activity in the amygdala

The amygdala responds to a large variety of socially and emotionally salient environmental and interoceptive stimuli. The context in which these stimuli occur determines their social and emotional significance. In canonical neurophysiological studies, the fast-paced succession of stimuli and events induce phasic changes in neural activity. During inter-trial intervals neural activity is expected to return to a stable and relatively featureless baseline. Context, such as the presence of a social partner, or the similarity of trials in a blocked design, induces brain states that can transcend the fast-paced succession of stimuli and can be recovered from the baseline firing rate of neurons. Indeed, the baseline firing rates of neurons in the amygdala change between blocks of trials of gentle grooming touch, delivered by a trusted social partner, and non-social airflow stimuli, delivered by a computer-controlled air valve. In this experimental paradigm, the presence of the groomer alone was sufficient to induce small but significant changes in baseline firing rates. Here, we examine local field potentials (LFP) recorded during these baseline periods to determine whether context was encoded by network dynamics that emerge in the local field potentials from the activity of large ensembles of neurons. We found that machine learning techniques can reliably decode social vs. non-social context from spectrograms of baseline local field potentials. Notably, decoding accuracy improved significantly with access to broad-band information. No significant differences were detected between the nuclei of the amygdala that receive direct or indirect inputs from areas of the prefrontal cortex known to coordinate flexible, context-dependent behaviors. The lack of nuclear specificity suggests that context-related synaptic inputs arise from a shared source, possibly interoceptive inputs that signal the sympathetic- vs. parasympathetic-dominated states characterizing non-social and social blocks, respectively.

The representational geometry of emotional states in basolateral amygdala

Sensory stimuli associated with aversive outcomes cause multiple behavioral responses related to an animal's evolving emotional state, but neural mechanisms underlying these processes remain unclear. Here aversive stimuli were presented to mice, eliciting two responses reflecting fear and flight to safety: tremble and ingress into a virtual burrow. Inactivation of basolateral amygdala (BLA) eliminated differential responses to aversive and neutral stimuli without eliminating responses themselves, suggesting BLA signals valence, not motor commands. However, two-photon imaging revealed that neurons typically exhibited mixed selectivity for stimulus identity, valence, tremble and/or ingress. Despite heterogeneous selectivity, BLA representational geometry was lower-dimensional when encoding valence, tremble and safety, enabling generalization of emotions across conditions. Further, tremble and valence coding directions were orthogonal, allowing linear readouts to specialize. Thus BLA representational geometry confers two computational properties that identify specialized neural circuits encoding variables describing emotional states: generalization across conditions, and readouts lacking interference from other readouts.

Spontaneously emerging internal models of visual sequences combine abstract and event-specific information in the prefrontal cortex

When exposed to sensory sequences, do macaque monkeys spontaneously form abstract internal models that generalize to novel experiences? Here, we show that neuronal populations in macaque ventrolateral prefrontal cortex jointly encode visual sequences by separate codes for the specific pictures presented and for their abstract sequential structure. We recorded prefrontal neurons while macaque monkeys passively viewed visual sequences and sequence mismatches in the local-global paradigm. Even without any overt task or response requirements, prefrontal populations spontaneously form representations of sequence structure, serial order, and image identity within distinct but superimposed neuronal subspaces. Representations of sequence structure rapidly update following single exposure to a mismatch sequence, while distinct populations represent mismatches for sequences of different complexity. Finally, those representations generalize across sequences following the same repetition structure but comprising different images. These results suggest that prefrontal populations spontaneously encode rich internal models of visual sequences reflecting both content-specific and abstract information.

Social isolation recruits amygdala-cortical circuitry to escalate alcohol drinking

How do social factors impact the brain and contribute to increased alcohol drinking? We found that social rank predicts alcohol drinking, where subordinates drink more than dominants. Furthermore, social isolation escalates alcohol drinking, particularly impacting subordinates who display a greater increase in alcohol drinking compared to dominants. Using cellular resolution calcium imaging, we show that the basolateral amygdala-medial prefrontal cortex (BLA-mPFC) circuit predicts alcohol drinking in a rank-dependent manner, unlike non-specific BLA activity. The BLA-mPFC circuit becomes hyperexcitable during social isolation, detecting social isolation states. Mimicking the observed increases in BLA-mPFC activity using optogenetics was sufficient to increase alcohol drinking, suggesting the BLA-mPFC circuit may be a neural substrate for the negative impact of social isolation. To test the hypothesis that the BLA-mPFC circuit conveys a signal induced by social isolation to motivate alcohol consumption, we first determined if this circuit detects social information. Leveraging optogenetics in combination with calcium imaging and computer vision pose tracking, we found that BLA-mPFC circuitry governs social behavior and neural representation of social contact. We further show that BLA-mPFC stimulation mimics social isolation-induced mPFC encoding of sucrose and alcohol, and inhibition of the BLA-mPFC circuit decreases alcohol drinking following social isolation. Collectively, these data suggest the amygdala-cortical circuit mirrors a neural encoding state similar to social isolation and underlies social isolation-associated alcohol drinking.

Differential functional organization of amygdala-medial prefrontal cortex networks in macaque and human

Over the course of evolution, the amygdala (AMG) and medial frontal cortex (mPFC) network, involved in behavioral adaptation, underwent structural changes in the old-world monkey and human lineages. Yet, whether and how the functional organization of this network differs remains poorly understood. Using resting-state functional magnetic resonance imagery, we show that the functional connectivity (FC) between AMG nuclei and mPFC regions differs between humans and awake macaques. In humans, the AMG-mPFC FC displays U-shaped pattern along the corpus callosum: a positive FC with the ventromedial prefrontal (vmPFC) and anterior cingulate cortex (ACC), a negative FC with the anterior mid-cingulate cortex (MCC), and a positive FC with the posterior MCC. Conversely, in macaques, the negative FC shifted more ventrally at the junction between the vmPFC and the ACC. The functional organization divergence of AMG-mPFC network between humans and macaques might help understanding behavioral adaptation abilities differences in their respective socio-ecological niches.

Neuronal Population Encoding of Identity in Primate Prefrontal Cortex

The ventrolateral prefrontal cortex (VLPFC) shows robust activation during the perception of faces and voices. However, little is known about what categorical features of social stimuli drive neural activity in this region. Since perception of identity and expression are critical social functions, we examined whether neural responses to naturalistic stimuli were driven by these two categorical features in the prefrontal cortex. We recorded single neurons in the VLPFC, while two male rhesus macaques (Macaca mulatta) viewed short audiovisual videos of unfamiliar conspecifics making expressions of aggressive, affiliative, and neutral valence. Of the 285 neurons responsive to the audiovisual stimuli, 111 neurons had a main effect (two-way ANOVA) of identity, expression, or their interaction in their stimulus-related firing rates; however, decoding of expression and identity using single-unit firing rates rendered poor accuracy. Interestingly, when decoding from pseudo-populations of recorded neurons, the accuracy for both expression and identity increased with population size, suggesting that the population transmitted information relevant to both variables. Principal components analysis of mean population activity across time revealed that population responses to the same identity followed similar trajectories in the response space, facilitating segregation from other identities. Our results suggest that identity is a critical feature of social stimuli that dictates the structure of population activity in the VLPFC, during the perception of vocalizations and their corresponding facial expressions. These findings enhance our understanding of the role of the VLPFC in social behavior.

Early selection of task-relevant features through population gating

Brains can gracefully weed out irrelevant stimuli to guide behavior. This feat is believed to rely on a progressive selection of task-relevant stimuli across the cortical hierarchy, but the specific across-area interactions enabling stimulus selection are still unclear. Here, we propose that population gating, occurring within primary auditory cortex (A1) but controlled by top-down inputs from prelimbic region of medial prefrontal cortex (mPFC), can support across-area stimulus selection. Examining single-unit activity recorded while rats performed an auditory context-dependent task, we found that A1 encoded relevant and irrelevant stimuli along a common dimension of its neural space. Yet, the relevant stimulus encoding was enhanced along an extra dimension. In turn, mPFC encoded only the stimulus relevant to the ongoing context. To identify candidate mechanisms for stimulus selection within A1, we reverse-engineered low-rank RNNs trained on a similar task. Our analyses predicted that two context-modulated neural populations gated their preferred stimulus in opposite contexts, which we confirmed in further analyses of A1. Finally, we show in a two-region RNN how population gating within A1 could be controlled by top-down inputs from PFC, enabling flexible across-area communication despite fixed inter-areal connectivity.

Multiple routes of communication within the amygdala-mPFC network: A comparative approach in humans and macaques

The network formed by the amygdala (AMG) and the medial Prefrontal Cortex (mPFC), at the interface between our internal and external environment, has been shown to support some important aspects of behavioral adaptation. Whether and how the anatomo-functional organization of this network evolved across primates remains unclear. Here, we compared AMG nuclei morphological characteristics and their functional connectivity with the mPFC in humans and macaques to identify potential homologies and differences between these species. Based on selected studies, we highlight two subsystems within the AMG-mPFC circuits, likely involved in distinct temporal dynamics of integration during behavioral adaptation. We also show that whereas the mPFC displays a large expansion but a preserved intrinsic anatomo-functional organization, the AMG displays a volume reduction and morphological changes related to specific nuclei. We discuss potential commonalities and differences in the dialogue between AMG nuclei and mPFC in humans and macaques based on available data.

The Computational and Neural Bases of Context-Dependent Learning

Flexible behavior requires the creation, updating, and expression of memories to depend on context. While the neural underpinnings of each of these processes have been intensively studied, recent advances in computational modeling revealed a key challenge in context-dependent learning that had been largely ignored previously: Under naturalistic conditions, context is typically uncertain, necessitating contextual inference. We review a theoretical approach to formalizing context-dependent learning in the face of contextual uncertainty and the core computations it requires. We show how this approach begins to organize a large body of disparate experimental observations, from multiple levels of brain organization (including circuits, systems, and behavior) and multiple brain regions (most prominently the prefrontal cortex, the hippocampus, and motor cortices), into a coherent framework. We argue that contextual inference may also be key to understanding continual learning in the brain. This theory-driven perspective places contextual inference as a core component of learning.

Multiplexing working memory and time in the trajectories of neural networks

Working memory (WM) and timing are generally considered distinct cognitive functions, but similar neural signatures have been implicated in both. To explore the hypothesis that WM and timing may rely on shared neural mechanisms, we used psychophysical tasks that contained either task-irrelevant timing or WM components. In both cases, the task-irrelevant component influenced performance. We then developed recurrent neural network (RNN) simulations that revealed that cue-specific neural sequences, which multiplexed WM and time, emerged as the dominant regime that captured the behavioural findings. During training, RNN dynamics transitioned from low-dimensional ramps to high-dimensional neural sequences, and depending on task requirements, steady-state or ramping activity was also observed. Analysis of RNN structure revealed that neural sequences relied primarily on inhibitory connections, and could survive the deletion of all excitatory-to-excitatory connections. Our results indicate that in some instances WM is encoded in time-varying neural activity because of the importance of predicting when WM will be used.

Preferences reveal separable valuation systems in prefrontal-limbic circuits

Individual preferences for the flavor of different foods and fluids exert a strong influence on behavior. Most current theories posit that preferences are integrated with other state variables in orbitofrontal cortex (OFC), which is thought to derive the relative subjective value of available options to drive choice behavior. Here we report that instead of a single integrated valuation system in OFC, another separate one is centered in ventrolateral prefrontal cortex (vlPFC) in macaque monkeys. Specifically, we found that OFC and vlPFC preferentially represent outcome flavor and outcome probability, respectively, and that preferences are separately integrated into these two aspects of subjective valuation. In addition, vlPFC, but not OFC, represented the outcome probability for the two options separately, with the difference between these representations reflecting the degree of preference. Thus, there are at least two separable valuation systems that work in concert to guide choices and that both are biased by preferences.

Glutamatergic and GABAergic neurons in pontine central gray mediate opposing valence-specific behaviors through a global network

Extracting the valence of environmental cues is critical for animals' survival. How valence in sensory signals is encoded and transformed to produce distinct behavioral responses remains not well understood. Here, we report that the mouse pontine central gray (PCG) contributes to encoding both negative and positive valences. PCG glutamatergic neurons were activated selectively by aversive, but not reward, stimuli, whereas its GABAergic neurons were preferentially activated by reward signals. The optogenetic activation of these two populations resulted in avoidance and preference behavior, respectively, and was sufficient to induce conditioned place aversion/preference. Suppression of them reduced sensory-induced aversive and appetitive behaviors, respectively. These two functionally opponent populations, receiving a broad range of inputs from overlapping yet distinct sources, broadcast valence-specific information to a distributed brain network with distinguishable downstream effectors. Thus, PCG serves as a critical hub to process positive and negative valences of incoming sensory signals and drive valence-specific behaviors with distinct circuits.

Subgenual and Hippocampal Pathways in Amygdala Are Set to Balance Affect and Context Processing

The amygdala, hippocampus, and subgenual cortex area 25 (A25) are engaged in complex cognitive-emotional processes. Yet pathway interactions from hippocampus and A25 with postsynaptic sites in amygdala remain largely unknown. In rhesus monkeys of both sexes, we studied with neural tracers how pathways from A25 and hippocampus interface with excitatory and inhibitory microcircuits in amygdala at multiple scales. We found that both hippocampus and A25 innervate distinct as well as overlapping sites of the basolateral (BL) amygdalar nucleus. Unique hippocampal pathways heavily innervated the intrinsic paralaminar basolateral nucleus, which is associated with plasticity. In contrast, orbital A25 preferentially innervated another intrinsic network, the intercalated masses, an inhibitory reticulum that gates amygdalar autonomic output and inhibits fear-related behaviors. Finally, using high-resolution confocal and electron microscopy (EM), we found that among inhibitory postsynaptic targets in BL, both hippocampal and A25 pathways preferentially formed synapses with calretinin (CR) neurons, which are known for disinhibition and may enhance excitatory drive in the amygdala. Among other inhibitory postsynaptic sites, A25 pathways innervated the powerful parvalbumin (PV) neurons which may flexibly regulate the gain of neuronal assemblies in the BL that affect the internal state. In contrast, hippocampal pathways innervated calbindin (CB) inhibitory neurons, which modulate specific excitatory inputs for processing context and learning correct associations. Common and unique patterns of innervation in amygdala by hippocampus and A25 have implications for how complex cognitive and emotional processes may be selectively disrupted in psychiatric disorders.SIGNIFICANCE STATEMENT The hippocampus, subgenual A25, and amygdala are associated with learning, memory, and emotions. We found that A25 is poised to affect diverse amygdalar processes, from emotional expression to fear learning by innervating the basal complex and the intrinsic intercalated masses. Hippocampal pathways uniquely interacted with another intrinsic amygdalar nucleus which is associated with plasticity, suggesting flexible processing of signals in context for learning. In the basolateral (BL) amygdala, which has a role in fear learning, both hippocampal and A25 interacted preferentially with disinhibitory neurons, suggesting a boost in excitation. The two pathways diverged in innervating other classes of inhibitory neurons, suggesting circuit specificities that could become perturbed in psychiatric diseases.

Environmental context-dependent activation of dopamine neurons via putative amygdala-nigra pathway in macaques

Seeking out good and avoiding bad objects is critical for survival. In practice, objects are rarely good every time or everywhere, but only at the right time or place. Whereas the basal ganglia (BG) are known to mediate goal-directed behavior, for example, saccades to rewarding objects, it remains unclear how such simple behaviors are rendered contingent on higher-order factors, including environmental context. Here we show that amygdala neurons are sensitive to environments and may regulate putative dopamine (DA) neurons via an inhibitory projection to the substantia nigra (SN). In male macaques, we combined optogenetics with multi-channel recording to demonstrate that rewarding environments induce tonic firing changes in DA neurons as well as phasic responses to rewarding events. These responses may be mediated by disinhibition via a GABAergic projection onto DA neurons, which in turn is suppressed by an inhibitory projection from the amygdala. Thus, the amygdala may provide an additional source of learning to BG circuits, namely contingencies imposed by the environment.

Basolateral amygdala and orbitofrontal cortex, but not dorsal hippocampus, are necessary for the control of reward-seeking by occasion setters

Reward-seeking in the world is driven by cues that can have ambiguous predictive and motivational value. To produce adaptive, flexible reward-seeking, it is necessary to exploit occasion setters, other distinct features in the environment, to resolve the ambiguity of Pavlovian reward-paired cues. Despite this, very little research has investigated the neurobiological underpinnings of occasion setting, and as a result little is known about which brain regions are critical for occasion setting. To address this, we exploited a recently developed task that was amenable to neurobiological inquiry where a conditioned stimulus is only predictive of reward delivery if preceded in time by the non-overlapping presentation of a separate cue-an occasion setter. This task required male rats to maintain and link cue-triggered expectations across time to produce adaptive reward-seeking. We interrogated the contributions of the basolateral amygdala and orbitofrontal cortex to occasion setting as these regions are thought to be critical for the computation and exploitation of state value, respectively. Reversible inactivation of either structure prior to the occasion-setting task resulted in a profound inability of rats to use the occasion setter to guide reward-seeking. In contrast, inactivation of the dorsal hippocampus, a region fundamental for context-specific responding was without effect nor did inactivation of the basolateral amygdala or orbitofrontal cortex in a standard Pavlovian conditioning preparation affect conditioned responding. We conclude that neural activity within the orbitofrontal cortex and basolateral amygdala circuit is necessary to update and resolve ambiguity in the environment to promote cue-driven reward-seeking.

Parametric control of flexible timing through low-dimensional neural manifolds

Biological brains possess an unparalleled ability to adapt behavioral responses to changing stimuli and environments. How neural processes enable this capacity is a fundamental open question. Previous works have identified two candidate mechanisms: a low-dimensional organization of neural activity and a modulation by contextual inputs. We hypothesized that combining the two might facilitate generalization and adaptation in complex tasks. We tested this hypothesis in flexible timing tasks where dynamics play a key role. Examining trained recurrent neural networks, we found that confining the dynamics to a low-dimensional subspace allowed tonic inputs to parametrically control the overall input-output transform, enabling generalization to novel inputs and adaptation to changing conditions. Reverse-engineering and theoretical analyses demonstrated that this parametric control relies on a mechanism where tonic inputs modulate the dynamics along non-linear manifolds while preserving their geometry. Comparisons with data from behaving monkeys confirmed the behavioral and neural signatures of this mechanism.

A context-dependent switch from sensing to feeling in the primate amygdala

The skin transmits affective signals that integrate into our social vocabulary. As the socio-affective aspects of touch are likely processed in the amygdala, we compare neural responses to social grooming and gentle airflow recorded from the amygdala and the primary somatosensory cortex of non-human primates. Neurons in the somatosensory cortex respond to both types of tactile stimuli. In the amygdala, however, neurons do not respond to individual grooming sweeps even though grooming elicits autonomic states indicative of positive affect. Instead, many show changes in baseline firing rates that persist throughout the grooming bout. Such baseline fluctuations are attributed to social context because the presence of the groomer alone can account for the observed changes in baseline activity. It appears, therefore, that during grooming, the amygdala stops responding to external inputs on a short timescale but remains responsive to social context (or the associated affective states) on longer time scales.

Recurrent networks endowed with structural priors explain suboptimal animal behavior

The strategies found by animals facing a new task are determined both by individual experience and by structural priors evolved to leverage the statistics of natural environments. Rats quickly learn to capitalize on the trial sequence correlations of two-alternative forced choice (2AFC) tasks after correct trials but consistently deviate from optimal behavior after error trials. To understand this outcome-dependent gating, we first show that recurrent neural networks (RNNs) trained in the same 2AFC task outperform rats as they can readily learn to use across-trial information both after correct and error trials. We hypothesize that, although RNNs can optimize their behavior in the 2AFC task without any a priori restrictions, rats' strategy is constrained by a structural prior adapted to a natural environment in which rewarded and non-rewarded actions provide largely asymmetric information. When pre-training RNNs in a more ecological task with more than two possible choices, networks develop a strategy by which they gate off the across-trial evidence after errors, mimicking rats' behavior. Population analyses show that the pre-trained networks form an accurate representation of the sequence statistics independently of the outcome in the previous trial. After error trials, gating is implemented by a change in the network dynamics that temporarily decouple the categorization of the stimulus from the across-trial accumulated evidence. Our results suggest that the rats' suboptimal behavior reflects the influence of a structural prior that reacts to errors by isolating the network decision dynamics from the context, ultimately constraining the performance in a 2AFC laboratory task.

Goals, usefulness and abstraction in value-based choice

Colombian drug lord Pablo Escobar, while on the run, purportedly burned two million dollars in banknotes to keep his daughter warm. A stark reminder that, in life, circumstances and goals can quickly change, forcing us to reassess and modify our values on-the-fly. Studies in decision-making and neuroeconomics have often implicitly equated value to reward, emphasising the hedonic and automatic aspect of the value computation, while overlooking its functional (concept-like) nature. Here we outline the computational and biological principles that enable the brain to compute the usefulness of an option or action by creating abstractions that flexibly adapt to changing goals. We present different algorithmic architectures, comparing ideas from artificial intelligence (AI) and cognitive neuroscience with psychological theories and, when possible, drawing parallels.

Importance of prefrontal meta control in human-like reinforcement learning

Recent investigation on reinforcement learning (RL) has demonstrated considerable flexibility in dealing with various problems. However, such models often experience difficulty learning seemingly easy tasks for humans. To reconcile the discrepancy, our paper is focused on the computational benefits of the brain's RL. We examine the brain's ability to combine complementary learning strategies to resolve the trade-off between prediction performance, computational costs, and time constraints. The complex need for task performance created by a volatile and/or multi-agent environment motivates the brain to continually explore an ideal combination of multiple strategies, called meta-control. Understanding these functions would allow us to build human-aligned RL models.

Calcium activity is a degraded estimate of spikes

Recording action potentials extracellularly during behavior has led to fundamental discoveries regarding neural function-hippocampal neurons respond to locations in space,1 motor cortex neurons encode movement direction,2 and dopamine neurons signal reward prediction errors3-observations undergirding current theories of cognition,4 movement,5 and learning.6 Recently it has become possible to measure calcium flux, an internal cellular signal related to spiking. The ability to image calcium flux in anatomically7,8 or genetically9 identified neurons can extend our knowledge of neural circuit function by allowing activity to be monitored in specific cell types or projections, or in the same neurons across many days. However, while initial studies were grounded in prior unit recording work, it has become fashionable to assume that calcium is identical to spiking, even though the spike-to-fluorescence transformation is nonlinear, noisy, and unpredictable under real-world conditions.10 It remains an open question whether calcium provides a high-fidelity representation of single-unit activity in awake, behaving subjects. Here, we have addressed this question by recording both signals in the lateral orbitofrontal cortex (OFC) of rats during olfactory discrimination learning. Activity in the OFC during olfactory learning has been well-studied in humans,11,12,13,14 nonhuman primates,15,16 and rats,17,18,19,20,21 where it has been shown to signal information about both the sensory properties of odor cues and the rewards they predict. Our single-unit results replicated prior findings, whereas the calcium signal provided only a degraded estimate of the information available in the single-unit spiking, reflecting primarily reward value.

The implications of categorical and category-free mixed selectivity on representational geometries

The firing rates of individual neurons displaying mixed selectivity are modulated by multiple task variables. When mixed selectivity is nonlinear, it confers an advantage by generating a high-dimensional neural representation that can be flexibly decoded by linear classifiers. Although the advantages of this coding scheme are well accepted, the means of designing an experiment and analyzing the data to test for and characterize mixed selectivity remain unclear. With the growing number of large datasets collected during complex tasks, the mixed selectivity is increasingly observed and is challenging to interpret correctly. We review recent approaches for analyzing and interpreting neural datasets and clarify the theoretical implications of mixed selectivity in the variety of forms that have been reported in the literature. We also aim to provide a practical guide for determining whether a neural population has linear or nonlinear mixed selectivity and whether this mixing leads to a categorical or category-free representation.

Self-Enhancement and the Medial Prefrontal Cortex: The Convergence of Clinical and Experimental Findings

Self-enhancement (SE) is often overlooked as a fundamental cognitive ability mediated via the Prefrontal Cortex (PFC). Here, we present research that establishes the relationship between the PFC, SE, and the potential evolved beneficial mechanisms. Specifically, we believe there is now enough evidence to speculate that SE exists to provide significant benefits and should be considered a normal aspect of the self. Whatever the metabolic or social cost, the upside of SE is great enough that it is a core and fundamental psychological construct. Furthermore, though entirely theoretical, we suggest that a critical reason the PFC has evolved so significantly in Homo sapiens is to, in part, sustain SE. We, therefore, elaborate on its proximate and ultimate mechanisms.

Neural signals implicated in the processing of appetitive and aversive events in social and non-social contexts

In 2014, we participated in a special issue of Frontiers examining the neural processing of appetitive and aversive events. Specifically, we reviewed brain areas that contribute to the encoding of prediction errors and value versus salience, attention and motivation. Further, we described how we disambiguated these cognitive processes and their neural substrates by using paradigms that incorporate both appetitive and aversive stimuli. We described a circuit in which the orbitofrontal cortex (OFC) signals expected value and the basolateral amygdala (BLA) encodes the salience and valence of both appetitive and aversive events. This information is integrated by the nucleus accumbens (NAc) and dopaminergic (DA) signaling in order to generate prediction and prediction error signals, which guide decision-making and learning via the dorsal striatum (DS). Lastly, the anterior cingulate cortex (ACC) is monitoring actions and outcomes, and signals the need to engage attentional control in order to optimize behavioral output. Here, we expand upon this framework, and review our recent work in which within-task manipulations of both appetitive and aversive stimuli allow us to uncover the neural processes that contribute to the detection of outcomes delivered to a conspecific and behaviors in social contexts. Specifically, we discuss the involvement of single-unit firing in the ACC and DA signals in the NAc during the processing of appetitive and aversive events in both social and non-social contexts.

Cognitive experience alters cortical involvement in goal-directed navigation

Neural activity in the mammalian cortex has been studied extensively during decision tasks, and recent work aims to identify under what conditions cortex is actually necessary for these tasks. We discovered that mice with distinct cognitive experiences, beyond sensory and motor learning, use different cortical areas and neural activity patterns to solve the same navigation decision task, revealing past learning as a critical determinant of whether cortex is necessary for goal-directed navigation. We used optogenetics and calcium imaging to study the necessity and neural activity of multiple cortical areas in mice with different training histories. Posterior parietal cortex and retrosplenial cortex were mostly dispensable for accurate performance of a simple navigation task. In contrast, these areas were essential for the same simple task when mice were previously trained on complex tasks with delay periods or association switches. Multiarea calcium imaging showed that, in mice with complex-task experience, single-neuron activity had higher selectivity and neuron–neuron correlations were weaker, leading to codes with higher task information. Therefore, past experience is a key factor in determining whether cortical areas have a causal role in goal-directed navigation.

Context coding in the mouse nucleus accumbens modulates motivationally relevant information

Neural activity in the nucleus accumbens (NAc) is thought to track fundamentally value-centric quantities linked to reward and effort. However, the NAc also contributes to flexible behavior in ways that are difficult to explain based on value signals alone, raising the question of if and how nonvalue signals are encoded in NAc. We recorded NAc neural ensembles while head-fixed mice performed an odor-based biconditional discrimination task where an initial discrete cue modulated the behavioral significance of a subsequently presented reward-predictive cue. We extracted single-unit and population-level correlates related to the cues and found value-independent coding for the initial, context-setting cue. This context signal occupied a population-level coding space orthogonal to outcome-related representations and was predictive of subsequent behaviorally relevant responses to the reward-predictive cues. Together, these findings support a gating model for how the NAc contributes to behavioral flexibility and provide a novel population-level perspective from which to view NAc computations.

Orthogonal representations for robust context-dependent task performance in brains and neural networks

How do neural populations code for multiple, potentially conflicting tasks? Here we used computational simulations involving neural networks to define "lazy" and "rich" coding solutions to this context-dependent decision-making problem, which trade off learning speed for robustness. During lazy learning the input dimensionality is expanded by random projections to the network hidden layer, whereas in rich learning hidden units acquire structured representations that privilege relevant over irrelevant features. For context-dependent decision-making, one rich solution is to project task representations onto low-dimensional and orthogonal manifolds. Using behavioral testing and neuroimaging in humans and analysis of neural signals from macaque prefrontal cortex, we report evidence for neural coding patterns in biological brains whose dimensionality and neural geometry are consistent with the rich learning regime.

Neurophysiological Evidence for Cognitive Map Formation during Sequence Learning

Humans deftly parse statistics from sequences. Some theories posit that humans learn these statistics by forming cognitive maps, or underlying representations of the latent space which links items in the sequence. Here, an item in the sequence is a node, and the probability of transitioning between two items is an edge. Sequences can then be generated from walks through the latent space, with different spaces giving rise to different sequence statistics. Individual or group differences in sequence learning can be modeled by changing the time scale over which estimates of transition probabilities are built, or in other words, by changing the amount of temporal discounting. Latent space models with temporal discounting bear a resemblance to models of navigation through Euclidean spaces. However, few explicit links have been made between predictions from Euclidean spatial navigation and neural activity during human sequence learning. Here, we use a combination of behavioral modeling and intracranial encephalography (iEEG) recordings to investigate how neural activity might support the formation of space-like cognitive maps through temporal discounting during sequence learning. Specifically, we acquire human reaction times from a sequential reaction time task, to which we fit a model that formulates the amount of temporal discounting as a single free parameter. From the parameter, we calculate each individual's estimate of the latent space. We find that neural activity reflects these estimates mostly in the temporal lobe, including areas involved in spatial navigation. Similar to spatial navigation, we find that low-dimensional representations of neural activity allow for easy separation of important features, such as modules, in the latent space. Lastly, we take advantage of the high temporal resolution of iEEG data to determine the time scale on which latent spaces are learned. We find that learning typically happens within the first 500 trials, and is modulated by the underlying latent space and the amount of temporal discounting characteristic of each participant. Ultimately, this work provides important links between behavioral models of sequence learning and neural activity during the same behavior, and contextualizes these results within a broader framework of domain general cognitive maps.

Cortical Granularity Shapes the Organization of Afferent Paths to the Amygdala and Its Striatal Targets in Nonhuman Primate

The prefrontal cortex (PFC) and insula, amygdala, and striatum form interconnected networks that drive motivated behaviors. We previously found a connectional trend in which granularity of the ventromedial and orbital PFC/insula predicted connections to the amygdala, and also the breadth of amygdalo-striatal efferents, including projections beyond the "classic" ventral striatum. To further interrogate connectional relationships among the cortex, amygdala, and striatum, and to further define the "limbic" (amygdala-recipient) striatum, we conducted tract tracing studies in two cohorts of macaques (male n = 14, female n = 1). We focused on the cortico-amygdalo-striatal (indirect) and cortico-"limbic" striatal (direct) paths originating in the entire PFC and insula. Larger datasets and a quantitative approach revealed "cortical rules" in which cortical granularity predicts the complexity and location of projections to both the basal nucleus of the amygdala and striatum. Remarkably, projections from "cortical-like" basal nucleus to the striatum followed similar patterns. In both "direct" and "indirect" paths to the "limbic" striatum, agranular cortices formed a "foundational," broad projection, and were joined by inputs from progressively more differentiated cortices. In amygdalo-striatal paths, the ventral basal nucleus was the "foundational" input, with progressively more dorsal basal nucleus regions gradually adding inputs as the "limbic" striatum extended caudally. Together, the "indirect" and "direct" paths followed consistent principles in which cortical granularity dictated the strength and complexity of projections at their targets. Cluster analyses independently confirmed these connectional trends, and also highlighted connectional features that predicted termination in specific subregions of the basal nucleus and "limbic" striatum.SIGNIFICANCE STATEMENT The "limbic" system broadly refers to brain circuits that coordinate emotional responses. Here, we investigate circuits of the amygdala, which are involved in coding the emotional value of external cues, and their influence on the striatum. Regions of prefrontal cortex (PFC) and insula form gradients of overlapping inputs to the amygdala's basal nucleus, which feed forward to the striatum. Direct cortical inputs to these "amygdala-recipient" striatal areas are surprisingly organized according to similar principles but subtly shift from the "classic" ventral striatum to the caudal ventral striatum. Together, these distinct subsystems, cortico-amygdalo-striatal circuits and direct cortico-striatal circuits, provide substantial opportunity for different levels of internal, sensory, and external experiences to be integrated within the striatum, a major motor-behavioral interface.

Emotional Context Sculpts Action Goal Representations in the Lateral Frontal Pole

Emotional states provide an ever-present source of contextual information that should inform behavioral goals. Despite the ubiquity of emotional signals in our environment, the neural mechanisms underlying their influence on goal-directed action remains unclear. Prior work suggests that the lateral frontal pole (FPl) is uniquely positioned to integrate affective information into cognitive control representations. We used pattern similarity analysis to examine the content of representations in FPl and interconnected mid-lateral prefrontal and amygdala circuitry. Healthy participants (n = 37; n = 21 females) were scanned while undergoing an event-related Affective Go/No-Go task, which requires goal-oriented action selection during emotional processing. We found that FPl contained conjunctive emotion-action goal representations that were related to successful cognitive control during emotional processing. These representations differed from conjunctive emotion-action goal representations found in the basolateral amygdala. While robust action goal representations were present in mid-lateral prefrontal cortex, they were not modulated by emotional valence. Finally, converging results from functional connectivity and multivoxel pattern analyses indicated that FPl emotional valence signals likely originated from interconnected subgenual anterior cingulate cortex (ACC) (BA25), which was in turn functionally coupled with the amygdala. Thus, our results identify a key pathway by which internal emotional states influence goal-directed behavior.SIGNIFICANCE STATEMENT Optimal functioning in everyday life requires behavioral regulation that flexibly adapts to dynamically changing emotional states. However, precisely how emotional states influence goal-directed action remains unclear. Unveiling the neural architecture that supports emotion-goal integration is critical for our understanding of disorders such as psychopathy, which is characterized by deficits in incorporating emotional cues into goals, as well as mood and anxiety disorders, which are characterized by impaired goal-based emotion regulation. Our study identifies a key circuit through which emotional states influence goal-directed behavior. This circuitry comprised the lateral frontal pole (FPl), which represented integrated emotion-goal information, as well as interconnected amygdala and subgenual ACC, which conveyed emotional signals to FPl.

A Novel Framework for Understanding the Pattern Identification of Traditional Asian Medicine From the Machine Learning Perspective

Pattern identification (PI), a unique diagnostic system of traditional Asian medicine, is the process of inferring the pathological nature or location of lesions based on observed symptoms. Despite its critical role in theory and practice, the information processing principles underlying PI systems are generally unclear. We present a novel framework for comprehending the PI system from a machine learning perspective. After a brief introduction to the dimensionality of the data, we propose that the PI system can be modeled as a dimensionality reduction process and discuss analytical issues that can be addressed using our framework. Our framework promotes a new approach in understanding the underlying mechanisms of the PI process with strong mathematical tools, thereby enriching the explanatory theories of traditional Asian medicine.

Anterior cingulate neurons signal neutral cue pairings during sensory preconditioning

Of all frontocortical subregions, the anterior cingulate cortex (ACC) has perhaps the most overlapping theories of function.1-3 Recording studies in rats, humans, and other primates have reported diverse neural responses that support many theories,4-12 yet nearly all these studies have in common tasks in which one event reliably predicts another. This leaves open the possibility that ACC represents associative pairing of events, independent of their overt biological significance. Sensory preconditioning13 provides an opportunity to test this. In the first phase, preconditioning, value-neutral sensory stimuli are paired (A→B). To test whether this was learned, subjects are given standard conditioning during which one of the previously neutral sensory cues is paired with a biologically meaningful outcome (B→outcome). During the final probe test, the neutral cue which was never paired with a biologically meaningful outcome is presented alone (A→) and will elicit a conditional response, suggesting that subjects had learned the associative structure during preconditioning and use that knowledge to infer presentation of the biologically relevant outcome (A→B→outcome). Inference-based responding demonstrates a fundamental property of model-based reasoning14,15 and requires learning of the associations between neutral stimuli before rewards are introduced.16-19 ACC neurons developed firing patterns that reflected the learning of sensory associations during preconditioning, even though no rewards were present. The strength of these correlates predicted rats' ability to later mobilize and use that associative information during the probe test. These results demonstrate that clear biological significance is not necessary to produce correlates of learning in ACC.

Foraging with the frontal cortex: A cross-species evaluation of reward-guided behavior

Efficient foraging is essential to survival and depends on frontal cortex in mammals. Because of its role in psychiatric disorders, frontal cortex and its contributions to reward procurement have been studied extensively in both rodents and non-human primates. How frontal cortex of these animal models compares is a source of intense debate. Here we argue that translating findings from rodents to non-human primates requires an appreciation of both the niche in which each animal forages as well as the similarities in frontal cortex anatomy and function. Consequently, we highlight similarities and differences in behavior and anatomy, before focusing on points of convergence in how parts of frontal cortex contribute to distinct aspects of foraging in rats and macaques, more specifically. In doing so, our aim is to emphasize where translation of frontal cortex function between species is clearer, where there is divergence, and where future work should focus. We finish by highlighting aspects of foraging for which have received less attention but we believe are critical to uncovering how frontal cortex promotes survival in each species.

Revealing nonlinear neural decoding by analyzing choices

Sensory data about most natural task-relevant variables are entangled with task-irrelevant nuisance variables. The neurons that encode these relevant signals typically constitute a nonlinear population code. Here we present a theoretical framework for quantifying how the brain uses or decodes its nonlinear information. Our theory obeys fundamental mathematical limitations on information content inherited from the sensory periphery, describing redundant codes when there are many more cortical neurons than primary sensory neurons. The theory predicts that if the brain uses its nonlinear population codes optimally, then more informative patterns should be more correlated with choices. More specifically, the theory predicts a simple, easily computed quantitative relationship between fluctuating neural activity and behavioral choices that reveals the decoding efficiency. This relationship holds for optimal feedforward networks of modest complexity, when experiments are performed under natural nuisance variation. We analyze recordings from primary visual cortex of monkeys discriminating the distribution from which oriented stimuli were drawn, and find these data are consistent with the hypothesis of near-optimal nonlinear decoding.

The memory trace of a stressful episode

Stress influences episodic memory formation via noradrenaline and glucocorticoid effects on amygdala and hippocampus. A common finding is the improvement of memory for central aspects of a stressful episode. This is putatively related to changes in the neural representations of specific experiences, i.e., their memory traces. Here we show that the memory improvement for objects that were encountered in a stressful episode relates to differences in the neural representations of these objects in the amygdala. Using functional magnetic resonance imaging, we found that stress specifically altered the representations of central objects: compared to control objects, they became more similar to one another and more distinct from objects that were not part of this episode. Furthermore, higher similarity of central objects to the main stressor-the faces of the stress-inducing committee members-predicted better memory. This suggests that the central objects were closely integrated into a stressor-centered memory representation. Our findings provide mechanistic insights into how stress shapes the memory trace and have profound implications for neurocognitive models of stressful and emotional memory.

Optogenetic study of central medial and paraventricular thalamic projections to the basolateral amygdala

The central medial (CMT) and paraventricular (PVT) thalamic nuclei project strongly to the basolateral amygdala (BL). Similarities between the responsiveness of CMT, PVT, and BL neurons suggest that these nuclei strongly influence BL activity. Supporting this possibility, an electron microscopic study reported that, in contrast with other extrinsic afferents, CMT and PVT axon terminals form very few synapses with BL interneurons. However, since limited sampling is a concern in electron microscopic studies, the present investigation was undertaken to compare the impact of CMT and PVT thalamic inputs on principal and local-circuit BL neurons with optogenetic methods and whole cell recordings in vitro. Optogenetic stimulation of CMT and PVT axons elicited glutamatergic excitatory postsynaptic potentials (EPSPs) or excitatory postsynaptic currents (EPSCs) in principal cells and interneurons, but they generally had a longer latency in interneurons. Moreover, after blockade of polysynaptic interactions with tetrodotoxin (TTX), a lower proportion of interneurons (50%) than principal cells (90%) remained responsive to CMT and PVT inputs. Although the presence of TTX-resistant responses in some interneurons indicates that CMT and PVT inputs directly contact some local-circuit cells, their lower incidence and amplitude after TTX suggest that CMT and PVT inputs form fewer synapses with them than with principal BL cells. Together, these results indicate that CMT and PVT inputs mainly contact principal BL neurons such that when CMT or PVT neurons fire, limited feedforward inhibition counters their excitatory influence over principal BL cells. However, CMT and PVT axons can also recruit interneurons indirectly, via the activation of principal cells, thereby generating feedback inhibition.NEW & NOTEWORTHY Midline thalamic (MTh) nuclei contribute major projections to the basolateral amygdala (BL). Similarities between the responsiveness of MTh and BL neurons suggest that MTh neurons exert a significant influence over BL activity. Using optogenetic techniques, we show that MTh inputs mainly contact principal BL neurons such that when MTh neurons fire, little feedforward inhibition counters their excitatory influence over principal cells. Thus, MTh inputs may be major determinants of BL activity.

The hierarchical construction of value

Here we argue that the assignment of subjective value to potential outcomes at the time of decision-making is an active process, in which individual features of a potential outcome of varying degrees of abstraction are represented hierarchically and integrated in a weighted fashion to produce an overall value judgment. We implicate the lateral orbital and medial prefrontal cortex in this function, situating these areas more broadly within a hierarchical integration process that takes place throughout the cortex for the ultimate purpose of valuing options to guide decisions.

Over-representation of fundamental decision variables in the prefrontal cortex underlies decision bias

The brain is organized into anatomically distinct structures consisting of a variety of projection neurons. While such evolutionarily conserved neural circuit organization underlies the innate ability of animals to swiftly adapt to environments, they can cause biased cognition and behavior. Although recent studies have begun to address the causal importance of projection-neuron types as distinct computational units, it remains unclear how projection types are functionally organized in encoding variables during cognitive tasks. This review focuses on the neural computation of decision making in the prefrontal cortex and discusses what decision variables are encoded by single neurons, neuronal populations, and projection type, alongside how specific projection types constrain decision making. We focus particularly on "over-representations" of distinct decision variables in the prefrontal cortex that reflect the biological and subjective significance of the variables for the decision makers. We suggest that task-specific over-representation in the prefrontal cortex involves the refinement of the given decision making, while generalized over-representation of fundamental decision variables is associated with suboptimal decision biases, including pathological ones such as those in patients with psychiatric disorders. Such over-representation of the fundamental decision variables in the prefrontal cortex appear to be tightly constrained by afferent and efferent connections that can be optogenetically intervened on. These ideas may provide critical insights into potential therapeutic targets for psychiatric disorders, including addiction and depression.

Functions of primate amygdala neurons in economic decisions and social decision simulation

Long implicated in aversive processing, the amygdala is now recognized as a key component of the brain systems that process rewards. Beyond reward valuation, recent findings from single-neuron recordings in monkeys indicate that primate amygdala neurons also play an important role in decision-making. The reward value signals encoded by amygdala neurons constitute suitable inputs to economic decision processes by being sensitive to reward contingency, relative reward quantity and temporal reward structure. During reward-based decisions, individual amygdala neurons encode both the value inputs and corresponding choice outputs of economic decision processes. The presence of such value-to-choice transitions in single amygdala neurons, together with other well-defined signatures of decision computation, indicate that a decision mechanism may be implemented locally within the primate amygdala. During social observation, specific amygdala neurons spontaneously encode these decision signatures to predict the choices of social partners, suggesting neural simulation of the partner's decision-making. The activity of these 'simulation neurons' could arise naturally from convergence between value neurons and social, self-other discriminating neurons. These findings identify single-neuron building blocks and computational architectures for decision-making and social behavior in the primate amygdala. An emerging understanding of the decision function of primate amygdala neurons can help identify potential vulnerabilities for amygdala dysfunction in human conditions afflicting social cognition and mental health.

Dynamical prefrontal population coding during defensive behaviours

Coping with threatening situations requires both identifying stimuli that predict danger and selecting adaptive behavioural responses to survive¹. The dorsomedial prefrontal cortex (dmPFC) is a critical structure that is involved in the regulation of threat-related behaviour^2-4. However, it is unclear how threat-predicting stimuli and defensive behaviours are associated within prefrontal networks to successfully drive adaptive responses. Here we used a combination of extracellular recordings, neuronal decoding approaches, pharmacological and optogenetic manipulations to show that, in mice, threat representations and the initiation of avoidance behaviour are dynamically encoded in the overall population activity of dmPFC neurons. Our data indicate that although dmPFC population activity at stimulus onset encodes sustained threat representations driven by the amygdala, it does not predict action outcome. By contrast, transient dmPFC population activity before the initiation of action reliably predicts avoided from non-avoided trials. Accordingly, optogenetic inhibition of prefrontal activity constrained the selection of adaptive defensive responses in a time-dependent manner. These results reveal that the adaptive selection of defensive responses relies on a dynamic process of information linking threats with defensive actions, unfolding within prefrontal networks.

Neural representation of abstract task structure during generalization

Cognitive models in psychology and neuroscience widely assume that the human brain maintains an abstract representation of tasks. This assumption is fundamental to theories explaining how we learn quickly, think creatively, and act flexibly. However, neural evidence for a verifiably generative abstract task representation has been lacking. Here, we report an experimental paradigm that requires forming such a representation to act adaptively in novel conditions without feedback. Using functional magnetic resonance imaging, we observed that abstract task structure was represented within left mid-lateral prefrontal cortex, bilateral precuneus, and inferior parietal cortex. These results provide support for the neural instantiation of the long-supposed abstract task representation in a setting where we can verify its influence. Such a representation can afford massive expansions of behavioral flexibility without additional experience, a vital characteristic of human cognition.

Orbitofrontal State Representations Are Related to Choice Adaptations and Reward Predictions

Animals can categorize the environment into "states," defined by unique sets of available action-outcome contingencies in different contexts. Doing so helps them choose appropriate actions and make accurate outcome predictions when in each given state. State maps have been hypothesized to be held in the orbitofrontal cortex (OFC), an area implicated in decision-making and encoding information about outcome predictions. Here we recorded neural activity in OFC in 6 male rats to test state representations. Rats were trained on an odor-guided choice task consisting of five trial blocks containing distinct sets of action-outcome contingencies, constituting states, with unsignaled transitions between them. OFC neural ensembles were analyzed using decoding algorithms. Results indicate that the vast majority of OFC neurons contributed to representations of the current state at any point in time, independent of odor cues and reward delivery, even at the level of individual neurons. Across state transitions, these representations gradually integrated evidence for the new state; the rate at which this integration happened in the prechoice part of the trial was related to how quickly the rats' choices adapted to the new state. Finally, OFC representations of outcome predictions, often thought to be the primary function of OFC, were dependent on the accuracy of OFC state representations.SIGNIFICANCE STATEMENT A prominent hypothesis proposes that orbitofrontal cortex (OFC) tracks current location in a "cognitive map" of state space. Here we tested this idea in detail by analyzing neural activity recorded in OFC of rats performing a task consisting of a series of states, each defined by a set of available action-outcome contingencies. Results show that most OFC neurons contribute to state representations and that these representations are related to the rats' decision-making and OFC reward predictions. These findings suggest new interpretations of emotional dysregulation in pathologies, such as addiction, which have long been known to be related to OFC dysfunction.