Flexible combination of reward information across primates

Moderate confirmation bias enhances decision-making in groups of reinforcement-learning agents

Humans tend to give more weight to information confirming their beliefs than to information that disconfirms them. Nevertheless, this apparent irrationality has been shown to improve individual decision-making under uncertainty. However, little is known about this bias' impact on decision-making in a social context. Here, we investigate the conditions under which confirmation bias is beneficial or detrimental to decision-making under social influence. To do so, we develop a Collective Asymmetric Reinforcement Learning (CARL) model in which artificial agents observe others' actions and rewards, and update this information asymmetrically. We use agent-based simulations to study how confirmation bias affects collective performance on a two-armed bandit task, and how resource scarcity, group size and bias strength modulate this effect. We find that a confirmation bias benefits group learning across a wide range of resource-scarcity conditions. Moreover, we discover that, past a critical bias strength, resource abundance favors the emergence of two different performance regimes, one of which is suboptimal. In addition, we find that this regime bifurcation comes with polarization in small groups of agents. Overall, our results suggest the existence of an optimal, moderate level of confirmation bias for decision-making in a social context.

Influence of surprise on reinforcement learning in younger and older adults

Surprise is a key component of many learning experiences, and yet its precise computational role, and how it changes with age, remain debated. One major challenge is that surprise often occurs jointly with other variables, such as uncertainty and outcome probability. To assess how humans learn from surprising events, and whether aging affects this process, we studied choices while participants learned from bandits with either Gaussian or bi-modal outcome distributions, which decoupled outcome probability, uncertainty, and surprise. A total of 102 participants (51 older, aged 50-73; 51 younger, 19-30 years) chose between three bandits, one of which had a bimodal outcome distribution. Behavioral analyses showed that both age-groups learned the average of the bimodal bandit less well. A trial-by-trial analysis indicated that participants performed choice reversals immediately following large absolute prediction errors, consistent with heightened sensitivity to surprise. This effect was stronger in older adults. Computational models indicated that learning rates in younger as well as older adults were influenced by surprise, rather than uncertainty, but also suggested large interindividual variability in the process underlying learning in our task. Our work bridges between behavioral economics research that has focused on how outcomes with low probability affect choice in older adults, and reinforcement learning work that has investigated age differences in the effects of uncertainty and suggests that older adults overly adapt to surprising events, even when accounting for probability and uncertainty effects.

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.

Dorsal raphe neurons integrate the values of reward amount, delay, and uncertainty in multi-attribute decision-making

The dorsal raphe nucleus (DRN) is implicated in psychiatric disorders that feature impaired sensitivity to reward amount, impulsivity when facing reward delays, and risk-seeking when confronting reward uncertainty. However, it has been unclear whether and how DRN neurons signal reward amount, reward delay, and reward uncertainty during multi-attribute value-based decision-making, where subjects consider these attributes to make a choice. We recorded DRN neurons as monkeys chose between offers whose attributes, namely expected reward amount, reward delay, and reward uncertainty, varied independently. Many DRN neurons signaled offer attributes, and this population tended to integrate the attributes in a manner that reflected monkeys' preferences for amount, delay, and uncertainty. After decision-making, in response to post-decision feedback, these same neurons signaled signed reward prediction errors, suggesting a broader role in tracking value across task epochs and behavioral contexts. Our data illustrate how the DRN participates in value computations, guiding theories about the role of the DRN in decision-making and psychiatric disease.

Heuristics in risky decision-making relate to preferential representation of information

When making choices, individuals differ from one another, as well as from normativity, in how they weigh different types of information. One explanation for this relates to idiosyncratic preferences in what information individuals represent when evaluating choice options. Here, we test this explanation with a simple risky-decision making task, combined with magnetoencephalography (MEG). We examine the relationship between individual differences in behavioral markers of information weighting and neural representation of stimuli pertinent to incorporating that information. We find that the extent to which individuals (N = 19) behaviorally weight probability versus reward information is related to how preferentially they neurally represent stimuli most informative for making probability and reward comparisons. These results are further validated in an additional behavioral experiment (N = 88) that measures stimulus representation as the latency of perceptual detection following priming. Overall, the results suggest that differences in the information individuals consider during choice relate to their risk-taking tendencies.

Risky decisions are influenced by individual attributes as a function of risk preference

It has long been assumed in economic theory that multi-attribute decisions involving several attributes or dimensions - such as probabilities and amounts of money to be earned during risky choices - are resolved by first combining the attributes of each option to form an overall expected value and then comparing the expected values of the alternative options, using a unique evidence accumulation process. A plausible alternative would be performing independent comparisons between the individual attributes and then integrating the results of the comparisons afterwards. Here, we devise a novel method to disambiguate between these types of models, by orthogonally manipulating the expected value of choice options and the relative salience of their attributes. Our results, based on behavioral measures and drift-diffusion models, provide evidence in favor of the framework where information about individual attributes independently impacts deliberation. This suggests that risky decisions are resolved by running in parallel multiple comparisons between the separate attributes - possibly alongside an additional comparison of expected value. This result stands in contrast with the assumption of standard economic theory that choices require a unique comparison of expected values and suggests that at the cognitive level, decision processes might be more distributed than commonly assumed. Beyond our planned analyses, we also discovered that attribute salience affects people of different risk preference type in different ways: risk-averse participants seem to focus more on probability, except when monetary amount is particularly high; risk-neutral/seeking participants, in contrast, seem to focus more on monetary amount, except when probability is particularly low.

The rat frontal orienting field dynamically encodes value for economic decisions under risk

Frontal and parietal cortex are implicated in economic decision-making, but their causal roles are untested. Here we silenced the frontal orienting field (FOF) and posterior parietal cortex (PPC) while rats chose between a cued lottery and a small stable surebet. PPC inactivations produced minimal short-lived effects. FOF inactivations reliably reduced lottery choices. A mixed-agent model of choice indicated that silencing the FOF caused a change in the curvature of the rats' utility function (U = Vρ). Consistent with this finding, single-neuron and population analyses of neural activity confirmed that the FOF encodes the lottery value on each trial. A dynamical model, which accounts for electrophysiological and silencing results, suggests that the FOF represents the current lottery value to compare against the remembered surebet value. These results demonstrate that the FOF is a critical node in the neural circuit for the dynamic representation of action values for choice under risk.

Semi-orthogonal subspaces for value mediate a tradeoff between binding and generalization

When choosing between options, we must associate their values with the action needed to select them. We hypothesize that the brain solves this binding problem through neural population subspaces. To test this hypothesis, we examined neuronal responses in five reward-sensitive regions in macaques performing a risky choice task with sequential offers. Surprisingly, in all areas, the neural population encoded the values of offers presented on the left and right in distinct subspaces. We show that the encoding we observe is sufficient to bind the values of the offers to their respective positions in space while preserving abstract value information, which may be important for rapid learning and generalization to novel contexts. Moreover, after both offers have been presented, all areas encode the value of the first and second offers in orthogonal subspaces. In this case as well, the orthogonalization provides binding. Our binding-by-subspace hypothesis makes two novel predictions borne out by the data. First, behavioral errors should correlate with putative spatial (but not temporal) misbinding in the neural representation. Second, the specific representational geometry that we observe across animals also indicates that behavioral errors should increase when offers have low or high values, compared to when they have medium values, even when controlling for value difference. Together, these results support the idea that the brain makes use of semi-orthogonal subspaces to bind features together.

Widespread coding of navigational variables in prefrontal cortex

To navigate effectively, we must represent information about our location in the environment. Traditional research highlights the role of the hippocampal complex in this process. Spurred by recent research highlighting the widespread cortical encoding of cognitive and motor variables previously thought to have localized function, we hypothesized that navigational variables would be likewise encoded widely, especially in the prefrontal cortex, which is associated with volitional behavior. We recorded neural activity from six prefrontal regions while macaques performed a foraging task in an open enclosure. In all regions, we found strong encoding of allocentric position, allocentric head direction, boundary distance, and linear and angular velocity. These encodings were not accounted for by distance, time to reward, or motor factors. The strength of coding of all variables increased along a ventral-to-dorsal gradient. Together, these results argue that encoding of navigational variables is not localized to the hippocampus and support the hypothesis that navigation is continuous with other forms of flexible cognition in the service of action.

Abstract Value Encoding in Neural Populations But Not Single Neurons

An important open question in neuroeconomics is how the brain represents the value of offers in a way that is both abstract (allowing for comparison) and concrete (preserving the details of the factors that influence value). Here, we examine neuronal responses to risky and safe options in five brain regions that putatively encode value in male macaques. Surprisingly, we find no detectable overlap in the neural codes used for risky and safe options, even when the options have identical subjective values (as revealed by preference) in any of the regions. Indeed, responses are weakly correlated and occupy distinct (semi-orthogonal) encoding subspaces. Notably, however, these subspaces are linked through a linear transform of their constituent encodings, a property that allows for comparison of dissimilar option types. This encoding scheme allows these regions to multiplex decision related processes: they can encode the detailed factors that influence offer value (here, risky and safety) but also directly compare dissimilar offer types. Together these results suggest a neuronal basis for the qualitatively different psychological properties of risky and safe options and highlight the power of population geometry to resolve outstanding problems in neural coding.SIGNIFICANCE STATEMENT To make economic choices, we must have some mechanism for comparing dissimilar offers. We propose that the brain uses distinct neural codes for risky and safe offers, but that these codes are linearly transformable. This encoding scheme has the dual advantage of allowing for comparison across offer types while preserving information about offer type, which in turn allows for flexibility in changing circumstances. We show that responses to risky and safe offers exhibit these predicted properties in five different reward-sensitive regions. Together, these results highlight the power of population coding principles for solving representation problems in economic choice.

Hierarchical organization of rhesus macaque behavior

Primatologists, psychologists and neuroscientists have long hypothesized that primate behavior is highly structured. However, delineating that structure has been impossible due to the difficulties of precision behavioral tracking. Here we analyzed a dataset consisting of continuous measures of the 3D position of two male rhesus macaques (Macaca mulatta) performing three different tasks in a large unrestrained environment over several hours. Using an unsupervised embedding approach on the tracked joints, we identified commonly repeated pose patterns, which we call postures. We found that macaques' behavior is characterized by 49 distinct postures, lasting an average of 0.6 seconds. We found evidence that behavior is hierarchically organized, in that transitions between poses tend to occur within larger modules, which correspond to identifiable actions; these actions are further organized hierarchically. Our behavioral decomposition allows us to identify universal (cross-individual and cross-task) and unique (specific to each individual and task) principles of behavior. These results demonstrate the hierarchical nature of primate behavior, provide a method for the automated ethogramming of primate behavior, and provide important constraints on neural models of pose generation.

Mechanisms of adjustments to different types of uncertainty in the reward environment across mice and monkeys

Despite being unpredictable and uncertain, reward environments often exhibit certain regularities, and animals navigating these environments try to detect and utilize such regularities to adapt their behavior. However, successful learning requires that animals also adjust to uncertainty associated with those regularities. Here, we analyzed choice data from two comparable dynamic foraging tasks in mice and monkeys to investigate mechanisms underlying adjustments to different types of uncertainty. In these tasks, animals selected between two choice options that delivered reward probabilistically, while baseline reward probabilities changed after a variable number (block) of trials without any cues to the animals. To measure adjustments in behavior, we applied multiple metrics based on information theory that quantify consistency in behavior, and fit choice data using reinforcement learning models. We found that in both species, learning and choice were affected by uncertainty about reward outcomes (in terms of determining the better option) and by expectation about when the environment may change. However, these effects were mediated through different mechanisms. First, more uncertainty about the better option resulted in slower learning and forgetting in mice, whereas it had no significant effect in monkeys. Second, expectation of block switches accompanied slower learning, faster forgetting, and increased stochasticity in choice in mice, whereas it only reduced learning rates in monkeys. Overall, while demonstrating the usefulness of metrics based on information theory in examining adaptive behavior, our study provides evidence for multiple types of adjustments in learning and choice behavior according to uncertainty in the reward environment.

Stimulating human prefrontal cortex increases reward learning

Work in computational psychiatry suggests that mood disorders may stem from aberrant reinforcement learning processes. Specifically, it has been proposed that depressed individuals believe that negative events are more informative than positive events, resulting in higher learning rates from negative outcomes (Pulcu and Browning, 2019). In this proof-of-concept study, we investigated whether transcranial direct current stimulation (tDCS) applied to dorsolateral prefrontal cortex, as commonly used in depression treatment trials, might change learning rates for affective outcomes. Healthy adults completed an established reinforcement learning task (Pulcu and Browning, 2017) in which the information content of reward and loss outcomes was manipulated by varying the volatility of stimulus-outcome associations. Learning rates on the tasks were quantified using computational models. Stimulation over dorsolateral prefrontal cortex (DLPFC) but not motor cortex (M1) increased learning rates specifically for reward outcomes. The effects of prefrontal tDCS were cognitive state-dependent: tDCS applied during task performance increased learning rates for wins; tDCS applied before task performance decreased both win and loss learning rates. A replication study confirmed the key finding that tDCS to DLPFC during task performance increased learning rates specifically for rewards. Taken together, these findings demonstrate the potential of tDCS for modulating computational parameters of reinforcement learning that are relevant to mood disorders.

Experience-driven recalibration of learning from surprising events

Different environments favor different patterns of adaptive learning. A surprising event that in one context would accelerate belief updating might, in another context, be downweighted as a meaningless outlier. Here, we investigated whether people would spontaneously regulate the influence of surprise on learning in response to event-by-event experiential feedback. Across two experiments, we examined whether participants performing a perceptual judgment task under spatial uncertainty (n = 29, n = 63) adapted their patterns of predictive gaze according to the informativeness or uninformativeness of surprising events in their current environment. Uninstructed predictive eye movements exhibited a form of metalearning in which surprise came to modulate event-by-event learning rates in opposite directions across contexts. Participants later appropriately readjusted their patterns of adaptive learning when the statistics of the environment underwent an unsignaled reversal. Although significant adjustments occurred in both directions, performance was consistently superior in environments in which surprising events reflected meaningful change, potentially reflecting a bias towards interpreting surprise as informative and/or difficulty ignoring salient outliers. Our results provide evidence for spontaneous, context-appropriate recalibration of the role of surprise in adaptive learning.

Systematic comparison of risky choices in humans and monkeys

The past decades have seen tremendous progress in fundamental studies on economic choice in humans. However, elucidation of the underlying neuronal processes requires invasive neurophysiological studies that are met with difficulties in humans. Monkeys as evolutionary closest relatives offer a solution. The animals display sophisticated and well-controllable behavior that allows to implement key constructs of proven economic choice theories. However, the similarity of economic choice between the two species has never been systematically investigated. We investigated compliance with the independence axiom (IA) of expected utility theory as one of the most demanding choice tests and compared IA violations between humans and monkeys. Using generalized linear modeling and cumulative prospect theory (CPT), we found that humans and monkeys made comparable risky choices, although their subjective values (utilities) differed. These results suggest similar fundamental choice mechanism across these primate species and encourage to study their underlying neurophysiological mechanisms.

Complex economic decisions from simple neurocognitive processes: the role of interactive attention

Neurocognitive theories of value-based choice propose that people additively accumulate choice attributes when making decisions. These theories cannot explain the emergence of complex multiplicative preferences such as those assumed by prospect theory and other economic models. We investigate an interactive attention mechanism, according to which attention to attributes (like payoffs) depends on other attributes (like probabilities) attended to previously. We formalize this mechanism using a Markov attention model combined with an accumulator decision process, and test our model on eye-tracking and mouse-tracking data in risky choice. Our tests show that interactive attention is necessary to make good choices, that most participants display interactive attention and that allowing for interactive attention in accumulation-based decision models improves their predictions. By equipping established decision models with sophisticated attentional dynamics, we extend these models to describe complex economic choice, and in the process, we unify two prominent theoretical approaches to studying value-based decision making.

Neural mechanisms underlying the hierarchical construction of perceived aesthetic value

Little is known about how the brain computes the perceived aesthetic value of complex stimuli such as visual art. Here, we used computational methods in combination with functional neuroimaging to provide evidence that the aesthetic value of a visual stimulus is computed in a hierarchical manner via a weighted integration over both low and high level stimulus features contained in early and late visual cortex, extending into parietal and lateral prefrontal cortices. Feature representations in parietal and lateral prefrontal cortex may in turn be utilized to produce an overall aesthetic value in the medial prefrontal cortex. Such brain-wide computations are not only consistent with a feature-based mechanism for value construction, but also resemble computations performed by a deep convolutional neural network. Our findings thus shed light on the existence of a general neurocomputational mechanism for rapidly and flexibly producing value judgements across an array of complex novel stimuli and situations.

The PRO model accounts for the anterior cingulate cortex role in risky decision-making and monitoring

The anterior cingulate cortex (ACC) has been implicated in a number of functions, including performance monitoring and decision-making involving effort. The prediction of responses and outcomes (PRO) model has provided a unified account of much human and monkey ACC data involving anatomy, neurophysiology, EEG, fMRI, and behavior. We explored the computational nature of ACC with the PRO model, extending it to account specifically for both human and macaque monkey decision-making under risk, including both behavioral and neural data. We show that the PRO model can account for a number of additional effects related to outcome prediction, decision-making under risk, gambling behavior. In particular, we show that the ACC represents the variance of uncertain outcomes, suggesting a link between ACC function and mean-variance theories of decision making. The PRO model provides a unified account of a large set of data regarding the ACC.

How does decision-making change during challenging times?

Prospect Theory, proposed and developed by Kahneman and Tversky, demonstrated that people do not make rational decisions based on expected utility, but are instead biased by specific cognitive tendencies leading to neglect, under- or over- consider information, depending on the context of presentation. In this vein, the present paper focuses on whether and how individual decision-making attitudes are prone to change in the presence of globally challenging events. We ran three partial replications of the Kahneman and Tversky (1979) paper, focusing on a set of eight prospects, after a terror attack (Paris, November 2015, 134 subjects) and during the Covid-19 pandemic, both during the first lockdown in Italy (Spring 2020, 176 subjects) and after the first reopening (140 subjects). The results confirm patterns of choice characterizing uncertain times, as shown by previous literature. In particular, we note significant increase of risk aversion, both in the gain and in the loss domains, that consistently emerged in the three replications. Given the nature of our sample, and the heterogeneity between the three periods investigated, we suggest that the phenomenon we present can be explained stress-related effects on decision making rather than by other economic effects, such as the income effect.

The Promise of Behavioral Tracking Systems for Advancing Primate Animal Welfare

Recent years have witnessed major advances in the ability of computerized systems to track the positions of animals as they move through large and unconstrained environments. These systems have so far been a great boon in the fields of primatology, psychology, neuroscience, and biomedicine. Here, we discuss the promise of these technologies for animal welfare. Their potential benefits include identifying and reducing pain, suffering, and distress in captive populations, improving laboratory animal welfare within the context of the three Rs of animal research (reduction, refinement, and replacement), and applying our understanding of animal behavior to increase the "natural" behaviors in captive and wild populations facing human impact challenges. We note that these benefits are often incidental to the designed purpose of these tracking systems, a reflection of the fact that animal welfare is not inimical to research progress, but instead, that the aligned interests between basic research and welfare hold great promise for improvements to animal well-being.

A structural and functional subdivision in central orbitofrontal cortex

Economic choice requires many cognitive subprocesses, including stimulus detection, valuation, motor output, and outcome monitoring; many of these subprocesses are associated with the central orbitofrontal cortex (cOFC). Prior work has largely assumed that the cOFC is a single region with a single function. Here, we challenge that unified view with convergent anatomical and physiological results from rhesus macaques. Anatomically, we show that the cOFC can be subdivided according to its much stronger (medial) or weaker (lateral) bidirectional anatomical connectivity with the posterior cingulate cortex (PCC). We call these subregions cOFCm and cOFCl, respectively. These two subregions have notable functional differences. Specifically, cOFCm shows enhanced functional connectivity with PCC, as indicated by both spike-field coherence and mutual information. The cOFCm-PCC circuit, but not the cOFCl-PCC circuit, shows signatures of relaying choice signals from a non-spatial comparison framework to a spatially framed organization and shows a putative bidirectional mutually excitatory pattern.

The computational roots of positivity and confirmation biases in reinforcement learning

Humans do not integrate new information objectively: outcomes carrying a positive affective value and evidence confirming one's own prior belief are overweighed. Until recently, theoretical and empirical accounts of the positivity and confirmation biases assumed them to be specific to 'high-level' belief updates. We present evidence against this account. Learning rates in reinforcement learning (RL) tasks, estimated across different contexts and species, generally present the same characteristic asymmetry, suggesting that belief and value updating processes share key computational principles and distortions. This bias generates over-optimistic expectations about the probability of making the right choices and, consequently, generates over-optimistic reward expectations. We discuss the normative and neurobiological roots of these RL biases and their position within the greater picture of behavioral decision-making theories.

Serotonin neurons modulate learning rate through uncertainty

Regulating how fast to learn is critical for flexible behavior. Learning about the consequences of actions should be slow in stable environments, but accelerate when that environment changes. Recognizing stability and detecting change are difficult in environments with noisy relationships between actions and outcomes. Under these conditions, theories propose that uncertainty can be used to modulate learning rates ("meta-learning"). We show that mice behaving in a dynamic foraging task exhibit choice behavior that varied as a function of two forms of uncertainty estimated from a meta-learning model. The activity of dorsal raphe serotonin neurons tracked both types of uncertainty in the foraging task as well as in a dynamic Pavlovian task. Reversible inhibition of serotonin neurons in the foraging task reproduced changes in learning predicted by a simulated lesion of meta-learning in the model. We thus provide a quantitative link between serotonin neuron activity, learning, and decision making.

A Normative Account of Confirmation Bias During Reinforcement Learning

Reinforcement learning involves updating estimates of the value of states and actions on the basis of experience. Previous work has shown that in humans, reinforcement learning exhibits a confirmatory bias: when the value of a chosen option is being updated, estimates are revised more radically following positive than negative reward prediction errors, but the converse is observed when updating the unchosen option value estimate. Here, we simulate performance on a multi-arm bandit task to examine the consequences of a confirmatory bias for reward harvesting. We report a paradoxical finding: that confirmatory biases allow the agent to maximize reward relative to an unbiased updating rule. This principle holds over a wide range of experimental settings and is most influential when decisions are corrupted by noise. We show that this occurs because on average, confirmatory biases lead to overestimating the value of more valuable bandits and underestimating the value of less valuable bandits, rendering decisions overall more robust in the face of noise. Our results show how apparently suboptimal learning rules can in fact be reward maximizing if decisions are made with finite computational precision.

Computational models of adaptive behavior and prefrontal cortex

The real world is uncertain, and while ever changing, it constantly presents itself in terms of new sets of behavioral options. To attain the flexibility required to tackle these challenges successfully, most mammalian brains are equipped with certain computational abilities that rely on the prefrontal cortex (PFC). By examining learning in terms of internal models associating stimuli, actions, and outcomes, we argue here that adaptive behavior relies on specific interactions between multiple systems including: (1) selective models learning stimulus-action associations through rewards; (2) predictive models learning stimulus- and/or action-outcome associations through statistical inferences anticipating behavioral outcomes; and (3) contextual models learning external cues associated with latent states of the environment. Critically, the PFC combines these internal models by forming task sets to drive behavior and, moreover, constantly evaluates the reliability of actor task sets in predicting external contingencies to switch between task sets or create new ones. We review different models of adaptive behavior to demonstrate how their components map onto this unifying framework and specific PFC regions. Finally, we discuss how our framework may help to better understand the neural computations and the cognitive architecture of PFC regions guiding adaptive behavior.

Inducing Affective Learning Biases with Cognitive Training and Prefrontal tDCS: A Proof-of-Concept Study

Background

Cognitive models of mood disorders emphasize a causal role of negative affective biases in depression. Computational work suggests that these biases may stem from a belief that negative events have a higher information content than positive events, resulting in preferential processing of and learning from negative outcomes. Learning biases therefore represent a promising target for therapeutic interventions. In this proof-of-concept study in healthy volunteers, we assessed the malleability of biased reinforcement learning using a novel cognitive training paradigm and concurrent transcranial direct current stimulation (tDCS).

Methods

In two studies, young healthy adults completed two sessions of negative (n = 20) or positive (n = 20) training designed to selectively increase learning from loss or win outcomes, respectively. During training active or sham tDCS was applied bilaterally to dorsolateral prefrontal cortex. Analyses tested for changes both in learning rates and win- and loss-driven behaviour. Potential positive/negative emotional transfer of win/loss learning was assessed by a facial emotion recognition task and mood questionnaires.

Results

Negative and positive training increased learning rates for losses and wins, respectively. With negative training, there was also a trend for win (but not loss) learning rates to decrease over successive task blocks. After negative training, there was evidence for near transfer in the form of an increase in loss-driven choices when participants performed a similar (untrained) task. There was no change in far transfer measures of emotional face processing or mood. tDCS had no effect on any aspect of behaviour.

Discussion and Conclusions

Negative training induced a mild negative bias in healthy adults as reflected in loss-driven choice behaviour. Prefrontal tDCS had no effect. Further research is needed to assess if this training procedure can be adapted to enhance learning from positive outcomes and whether effects translate to affective disorders.

Electronic supplementary material

The online version of this article (10.1007/s10608-020-10146-9) contains supplementary material, which is available to authorized users.

Timescales of Cognition in the Brain

We live in a world that changes on many timescales. To learn and make decisions appropriately, the human brain has evolved to integrate various types of information, such as sensory evidence and reward feedback, on multiple timescales. This is reflected in cortical hierarchies of timescales consisting of heterogeneous neuronal activities and expression of genes related to neurotransmitters critical for learning. We review the recent findings on how timescales of sensory and reward integration are affected by the temporal properties of sensory and reward signals in the environment. Despite existing evidence linking behavioral and neuronal timescales, future studies must examine how neural computations at multiple timescales are adjusted and combined to influence behavior flexibly.

Additively Combining Utilities and Beliefs: Research Gaps and Algorithmic Developments

Value-based decision making in complex environments, such as those with uncertain and volatile mapping of reward probabilities onto options, may engender computational strategies that are not necessarily optimal in terms of normative frameworks but may ensure effective learning and behavioral flexibility in conditions of limited neural computational resources. In this article, we review a suboptimal strategy - additively combining reward magnitude and reward probability attributes of options for value-based decision making. In addition, we present computational intricacies of a recently developed model (named MIX model) representing an algorithmic implementation of the additive strategy in sequential decision-making with two options. We also discuss its opportunities; and conceptual, inferential, and generalization issues. Furthermore, we suggest future studies that will reveal the potential and serve the further development of the MIX model as a general model of value-based choice making.

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.

Choice-relevant information transformation along a ventrodorsal axis in the medial prefrontal cortex

Choice-relevant brain regions in prefrontal cortex may progressively transform information about options into choices. Here, we examine responses of neurons in four regions of the medial prefrontal cortex as macaques performed two-option risky choices. All four regions encode economic variables in similar proportions and show similar putative signatures of key choice-related computations. We provide evidence to support a gradient of function that proceeds from areas 14 to 25 to 32 to 24. Specifically, we show that decodability of twelve distinct task variables increases along that path, consistent with the idea that regions that are higher in the anatomical hierarchy make choice-relevant variables more separable. We also show progressively longer intrinsic timescales in the same series. Together these results highlight the importance of the medial wall in choice, endorse a specific gradient-based organization, and argue against a modular functional neuroanatomy of choice.

Partial integration of the components of value in anterior cingulate cortex

Evaluation often involves integrating multiple determinants of value, such as the different possible outcomes in risky choice. A brain region can be placed either before or after a presumed evaluation stage by measuring how responses of its neurons depend on multiple determinants of value. A brain region could also, in principle, show partial integration, which would indicate that it occupies a middle position between (preevaluative) nonintegration and (postevaluative) full integration. Existing mathematical techniques cannot distinguish full from partial integration and therefore risk misidentifying regional function. Here we use a new Bayesian regression-based approach to analyze responses of neurons in dorsal anterior cingulate cortex (dACC) to risky offers. We find that dACC neurons only partially integrate across outcome dimensions, indicating that dACC cannot be assigned to either a pre- or postevaluative position. Neurons in dACC also show putative signatures of value comparison, thereby demonstrating that comparison does not require complete evaluation before proceeding. (PsycInfo Database Record (c) 2020 APA, all rights reserved).

Nonhuman Primates Satisfy Utility Maximization in Compliance with the Continuity Axiom of Expected Utility Theory

Expected Utility Theory (EUT), the first axiomatic theory of risky choice, describes choices as a utility maximization process: decision makers assign a subjective value (utility) to each choice option and choose the one with the highest utility. The continuity axiom, central to Expected Utility Theory and its modifications, is a necessary and sufficient condition for the definition of numerical utilities. The axiom requires decision makers to be indifferent between a gamble and a specific probabilistic combination of a more preferred and a less preferred gamble. While previous studies demonstrated that monkeys choose according to combinations of objective reward magnitude and probability, a concept-driven experimental approach for assessing the axiomatically defined conditions for maximizing utility by animals is missing. We experimentally tested the continuity axiom for a broad class of gamble types in 4 male rhesus macaque monkeys, showing that their choice behavior complied with the existence of a numerical utility measure as defined by the economic theory. We used the numerical quantity specified in the continuity axiom to characterize subjective preferences in a magnitude-probability space. This mapping highlighted a trade-off relation between reward magnitudes and probabilities, compatible with the existence of a utility function underlying subjective value computation. These results support the existence of a numerical utility function able to describe choices, allowing for the investigation of the neuronal substrates responsible for coding such rigorously defined quantity.SIGNIFICANCE STATEMENT A common assumption of several economic choice theories is that decisions result from the comparison of subjectively assigned values (utilities). This study demonstrated the compliance of monkey behavior with the continuity axiom of Expected Utility Theory, implying a subjective magnitude-probability trade-off relation, which supports the existence of numerical utility directly linked to the theoretical economic framework. We determined a numerical utility measure able to describe choices, which can serve as a correlate for the neuronal activity in the quest for brain structures and mechanisms guiding decisions.

The description-experience gap: a challenge for the neuroeconomics of decision-making under uncertainty

The experimental investigation of decision-making in humans relies on two distinct types of paradigms, involving either description- or experience-based choices. In description-based paradigms, decision variables (i.e. payoffs and probabilities) are explicitly communicated by means of symbols. In experience-based paradigms decision variables are learnt from trial-by-trial feedback. In the decision-making literature, 'description-experience gap' refers to the fact that different biases are observed in the two experimental paradigms. Remarkably, well-documented biases of description-based choices, such as under-weighting of rare events and loss aversion, do not apply to experience-based decisions. Here, we argue that the description-experience gap represents a major challenge, not only to current decision theories, but also to the neuroeconomics research framework, which relies heavily on the translation of neurophysiological findings between human and non-human primate research. In fact, most non-human primate neurophysiological research relies on behavioural designs that share features of both description- and experience-based choices. As a consequence, it is unclear whether the neural mechanisms built from non-human primate electrophysiology should be linked to description-based or experience-based decision-making processes. The picture is further complicated by additional methodological gaps between human and non-human primate neuroscience research. After analysing these methodological challenges, we conclude proposing new lines of research to address them. This article is part of the theme issue 'Existence and prevalence of economic behaviours among non-human primates'.

Dissociable roles of cortical excitation-inhibition balance during patch-leaving versus value-guided decisions

In a dynamic world, it is essential to decide when to leave an exploited resource. Such patch-leaving decisions involve balancing the cost of moving against the gain expected from the alternative patch. This contrasts with value-guided decisions that typically involve maximizing reward by selecting the current best option. Patterns of neuronal activity pertaining to patch-leaving decisions have been reported in dorsal anterior cingulate cortex (dACC), whereas competition via mutual inhibition in ventromedial prefrontal cortex (vmPFC) is thought to underlie value-guided choice. Here, we show that the balance between cortical excitation and inhibition (E/I balance), measured by the ratio of GABA and glutamate concentrations, plays a dissociable role for the two kinds of decisions. Patch-leaving decision behaviour relates to E/I balance in dACC. In contrast, value-guided decision-making relates to E/I balance in vmPFC. These results support mechanistic accounts of value-guided choice and provide evidence for a role of dACC E/I balance in patch-leaving decisions.

Non-human primates use combined rules when deciding under ambiguity

Decision outcomes in unpredictable environments may not have exact known probabilities. Yet the predictability level of outcomes matters in decisions, and animals, including humans, generally avoid ambiguous options. Managing ambiguity may be more challenging and requires stronger cognitive skills than decision-making under risk, where decisions involve known probabilities. Here we compare decision-making in capuchins, macaques, orangutans, gorillas, chimpanzees and bonobos in risky and ambiguous contexts. Subjects were shown lotteries (a tray of potential rewards, some large, some small) and could gamble a medium-sized food item to obtain one of the displayed rewards. The odds of winning and losing varied and were accessible in the risky context (all rewards were visible) or partially available in the ambiguous context (some rewards were covered). In the latter case, the level of information varied from fully ambiguous (individuals could not guess what was under the covers) to predictable (individuals could guess). None of the species avoided gambling in ambiguous lotteries and gambling rates were high if at least two large rewards were visible. Capuchins and bonobos ignored the covered items and gorillas and macaques took the presence of potential rewards into account, but only chimpanzees and orangutans could consistently build correct expectations about the size of the covered rewards. Chimpanzees and orangutans combined decision rules according to the number of large visible rewards and the level of predictability, a process resembling conditional probabilities assessment in humans. Despite a low sample size, this is the first evidence in non-human primates that a combination of several rules can underlie choices made in an unpredictable environment. Our finding that non-human primates can deal with the uncertainty of an outcome when exchanging one food item for another is a key element to the understanding of the evolutionary origins of economic behaviour. This article is part of the theme issue 'Existence and prevalence of economic behaviours among non-human primates'.

Cognitiva Speciebus: Towards a Linnaean Approach to Cognition

Research points to the limitations of approaches to decision-making, that rest on general 'Newtonian principles' derived from unitary a priori conceptions of rationality. To understand how the mind exploits environments, we instead propose a process of more open-ended discovery and systematization in the mold of Linnaeus's famous taxonomy of plants.

Momentary subjective well-being depends on learning and not reward

Subjective well-being or happiness is often associated with wealth. Recent studies suggest that momentary happiness is associated with reward prediction error, the difference between experienced and predicted reward, a key component of adaptive behaviour. We tested subjects in a reinforcement learning task in which reward size and probability were uncorrelated, allowing us to dissociate between the contributions of reward and learning to happiness. Using computational modelling, we found convergent evidence across stable and volatile learning tasks that happiness, like behaviour, is sensitive to learning-relevant variables (i.e. probability prediction error). Unlike behaviour, happiness is not sensitive to learning-irrelevant variables (i.e. reward prediction error). Increasing volatility reduces how many past trials influence behaviour but not happiness. Finally, depressive symptoms reduce happiness more in volatile than stable environments. Our results suggest that how we learn about our world may be more important for how we feel than the rewards we actually receive.

Many people believe they would be happier if only they had more money. And events such as winning the lottery or receiving a large pay rise do make people happy, at least temporarily. But recent studies suggest that the main factor driving happiness on such occasions is not the size of the reward received. Instead, it is how well that reward matches up with expectations. Receiving a 10% pay rise when you were expecting 1% will make you feel happier than receiving 10% when you had been expecting 20%.

This difference between an expected and an actual reward is referred to as a reward prediction error. Reward prediction errors have a key role in learning. They motivate people to repeat behaviours that led to unexpectedly large rewards. But they also enable people to update their beliefs about the world, which is rewarding in itself. Could it be that reward prediction errors are associated with happiness mainly because they help us understand the world a little better than before?

To test this idea, Blain and Rutledge designed a task in which the likelihood of receiving a reward was unrelated to the size of the reward. This study design makes it possible to separate out the contributions of learning versus reward to moment-by-moment happiness.

In the task, volunteers had to decide which of two cars would win a race. In the ‘stable’ condition, one of the cars always had an 80% chance of winning. In the ‘volatile’ condition, one car had an 80% chance of winning for the first 20 trials. The other car then had an 80% chance of winning for the next 20 trials. The volunteers were not told these probabilities in advance, but had to work them out by playing the game. However, on every trial, the volunteers were shown the reward they would receive if they chose either of the cars and that car went on to win. The size of the rewards varied at random and was unrelated to the likelihood of a car winning.

Every few trials, the volunteers were asked to indicate their current level of happiness on a scale. The results showed that volunteers were happier after winning than after losing. On average they were also happier in the stable condition than in the volatile condition. This was especially true for volunteers with pre-existing symptoms of depression. Moreover, happiness after wins did not depend on how large the reward they got was, but instead simply on how surprised they were to win.

These results suggest that how we learn about the world around us can be more important for how we feel than rewards we receive directly. Measuring happiness in various types of environment could help us understand factors affecting mental health. The current results suggest, for example, that uncertain environments may be especially unpleasant for people with depression. Further research is needed to understand why this might be the case. In the real world, rewards are often uncertain and infrequent, but learning may nevertheless have the potential to boost happiness.

How Outcome Uncertainty Mediates Attention, Learning, and Decision-Making

Animals and humans evolved sophisticated nervous systems that endowed them with the ability to form internal-models or beliefs and make predictions about the future to survive and flourish in a world in which future outcomes are often uncertain. Crucial to this capacity is the ability to adjust behavioral and learning policies in response to the level of uncertainty. Until recently, the neuronal mechanisms that could underlie such uncertainty-guided control have been largely unknown. In this review, I discuss newly discovered neuronal circuits in primates that represent uncertainty about future rewards and propose how they guide information-seeking, attention, decision-making, and learning to help us survive in an uncertain world. Lastly, I discuss the possible relevance of these findings to learning in artificial systems.

Learning arbitrary stimulus-reward associations for naturalistic stimuli involves transition from learning about features to learning about objects

Most cognitive processes are studied using abstract or synthetic stimuli with specific features to fully control what is presented to subjects. However, recent studies have revealed enhancements of cognitive capacities (such as working memory) when processing naturalistic versus abstract stimuli. Using abstract stimuli constructed from distinct visual features (e.g., color and shape), we have recently shown that human subjects can learn multidimensional stimulus-reward associations via initially estimating reward value of individual features (feature-based learning) before gradually switching to learning about reward value of individual stimuli (object-based learning). Here, we examined whether similar strategies are adopted during learning about naturalistic stimuli that are clearly perceived as objects (instead of a combination of features) and contain both task-relevant and irrelevant features. We found that similar to learning about abstract stimuli, subjects initially adopted feature-based learning more strongly before transitioning to object-based learning. However, there were three key differences between learning about naturalistic and abstract stimuli. First, compared with abstract stimuli, the initial learning strategy was less feature-based for naturalistic stimuli. Second, subjects transitioned to object-based learning faster for naturalistic stimuli. Third, unexpectedly, subjects were more likely to adopt feature-based learning for naturalistic stimuli, both at the steady state and overall. These results suggest that despite the stronger tendency to perceive naturalistic stimuli as objects, which leads to greater likelihood of using object-based learning as the initial strategy and a faster transition to object-based learning, the influence of individual features on learning is stronger for these stimuli such that ultimately the object-based strategy is adopted less. Overall, our findings suggest that feature-based learning is a general initial strategy for learning about reward value of all types of multi-dimensional stimuli.

Multiple timescales of neural dynamics and integration of task-relevant signals across cortex

A long-lasting challenge in neuroscience has been to find a set of principles that could be used to organize the brain into distinct areas with specific functions. Recent studies have proposed the orderly progression in the time constants of neural dynamics as an organizational principle of cortical computations. However, relationships between these timescales and their dependence on response properties of individual neurons are unknown, making it impossible to determine how mechanisms underlying such a computational principle are related to other aspects of neural processing. Here, we developed a comprehensive method to simultaneously estimate multiple timescales in neuronal dynamics and integration of task-relevant signals along with selectivity to those signals. By applying our method to neural and behavioral data during a dynamic decision-making task, we found that most neurons exhibited multiple timescales in their response, which consistently increased from parietal to prefrontal and cingulate cortex. While predicting rates of behavioral adjustments, these timescales were not correlated across individual neurons in any cortical area, resulting in independent parallel hierarchies of timescales. Additionally, none of these timescales depended on selectivity to task-relevant signals. Our results not only suggest the existence of multiple canonical mechanisms for increasing timescales of neural dynamics across cortex but also point to additional mechanisms that allow decorrelation of these timescales to enable more flexibility.

A brain network supporting social influences in human decision-making

Humans learn from their own trial-and-error experience and observing others. However, it remains unknown how brain circuits compute expected values when direct learning and social learning coexist in uncertain environments. Using a multiplayer reward learning paradigm with 185 participants (39 being scanned) in real time, we observed that individuals succumbed to the group when confronted with dissenting information but observing confirming information increased their confidence. Leveraging computational modeling and functional magnetic resonance imaging, we tracked direct valuation through experience and vicarious valuation through observation and their dissociable, but interacting neural representations in the ventromedial prefrontal cortex and the anterior cingulate cortex, respectively. Their functional coupling with the right temporoparietal junction representing instantaneous social information instantiated a hitherto uncharacterized social prediction error, rather than a reward prediction error, in the putamen. These findings suggest that an integrated network involving the brain's reward hub and social hub supports social influence in human decision-making.

Consistent patterns of distractor effects during decision making

The value of a third potential option or distractor can alter the way in which decisions are made between two other options. Two hypotheses have received empirical support: that a high value distractor improves the accuracy with which decisions between two other options are made and that it impairs accuracy. Recently, however, it has been argued that neither observation is replicable. Inspired by neuroimaging data showing that high value distractors have different impacts on prefrontal and parietal regions, we designed a dual route decision-making model that mimics the neural signals of these regions. Here we show in the dual route model and empirical data that both enhancement and impairment effects are robust phenomena but predominate in different parts of the decision space defined by the options' and the distractor's values. However, beyond these constraints, both effects co-exist under similar conditions. Moreover, both effects are robust and observable in six experiments.

Mechanisms of competitive selection: A canonical neural circuit framework

Competitive selection, the transformation of multiple competing sensory inputs and internal states into a unitary choice, is a fundamental component of animal behavior. Selection behaviors have been studied under several intersecting umbrellas including decision-making, action selection, perceptual categorization, and attentional selection. Neural correlates of these behaviors and computational models have been investigated extensively. However, specific, identifiable neural circuit mechanisms underlying the implementation of selection remain elusive. Here, we employ a first principles approach to map competitive selection explicitly onto neural circuit elements. We decompose selection into six computational primitives, identify demands that their execution places on neural circuit design, and propose a canonical neural circuit framework. The resulting framework has several links to neural literature, indicating its biological feasibility, and has several common elements with prominent computational models, suggesting its generality. We propose that this framework can help catalyze experimental discovery of the neural circuit underpinnings of competitive selection.

From high- to one-dimensional dynamics of decision making: testing simplifications in attractor models

Computational models introduce simplifications that need to be understood and validated. For attractor models of decision making, the main simplification is the high-level representation of different sub-processes of the complex decision system in one dynamic description of the overall process dynamics. This simplification implies that the overall process dynamics of the decision system are independent from specific values handled in different sub-processes. Here, we test the validity of this simplification empirically by investigating choice perseveration in a nonverbal, value-based decision task. Specifically, we tested whether choice perseveration occurred irrespectively of the attribute dimension as suggested by a simulation of the computational model. We find evidence supporting the validity of the simplification. We conclude that the simplification might capture mechanistic aspects of decision-making processes, and that the summation of the overall process dynamics of decision systems into one single variable is a valid approach in computational modeling. Supplement materials such as empirical data, analysis scripts, and the computational model are publicly available at the Open Science Framework (osf.io/7fb5q).

Activity in orbitofrontal neuronal ensembles reflects inhibitory control

Stopping, or inhibition, is a form of self-control that is a core element of flexible and adaptive behavior. Its neural origins remain unclear. Some views hold that inhibition decisions reflect the aggregation of widespread and diverse pieces of information, including information arising in ostensible core reward regions (i.e., outside the canonical executive system). We recorded activity of single neurons in the orbitofrontal cortex (OFC) of macaques, a region associated with economic decisions, and whose role in inhibition is debated. Subjects performed a classic inhibition task known as the stop signal task. Ensemble decoding analyses reveal a clear firing rate pattern that distinguishes successful from failed inhibition and that begins after the stop signal and before the stop signal reaction time (SSRT). We also found a different and orthogonal ensemble pattern that distinguishes successful from failed stopping before the beginning of the trial. These signals were distinct from, and orthogonal to, value encoding, which was also observed in these neurons. The timing of the early and late signals was, respectively, consistent with the idea that neuronal activity in OFC encodes inhibition both proactively and reactively.

Human Decision-Making beyond the Rational Decision Theory

Two recent studies (Farashahi et al. and Rouault et al.) provide compelling evidence refuting the Subjective Expected Utility (SEU) hypothesis as a ground model describing human decision-making. Together, these studies pave the way towards a new model that subsumes the notion of decision-making and adaptive behavior into a single account.

Macaques are risk-averse in a freely moving foraging task

Rhesus macaques (Macaca mulatta) appear to be robustly risk-seeking in computerized gambling tasks typically used for electrophysiology. This behavior distinguishes them from many other animals, which are risk-averse, albeit measured in more naturalistic contexts. We wondered whether macaques' risk preferences reflect their evolutionary history or derive from the less naturalistic elements of task design associated with the demands of physiological recording. We assessed macaques' risk attitudes in a task that is somewhat more naturalistic than many that have previously been used: subjects foraged at four feeding stations in a large enclosure. Patches (i.e., stations), provided either stochastically or non-stochastically depleting rewards. Subjects' patch residence times were longer at safe than at risky stations, indicating a preference for safe options. This preference was not attributable to a win-stay-lose-shift heuristic and reversed as the environmental richness increased. These findings highlight the lability of risk attitudes in macaques and support the hypothesis that the ecological validity of a task can influence the expression of risk preference.

Combinations of low-level and high-level neural processes account for distinct patterns of context-dependent choice

Context effects have been explained by either low-level neural adjustments or high-level cognitive processes but not their combination. It is currently unclear how these processes interact to shape individuals’ responses to context. Here, we used a large cohort of human subjects in experiments involving choice between two or three gambles in order to study the dependence of context effects on neural adaptation and individuals’ risk attitudes. Our experiments did not provide any evidence that neural adaptation on long timescales (~100 trials) contributes to context effects. Using post-hoc analyses we identified two groups of subjects with distinct patterns of responses to decoys, both of which depended on individuals’ risk aversion. Subjects in the first group exhibited strong, consistent decoy effects and became more risk averse due to decoy presentation. In contrast, subjects in the second group did not show consistent decoy effects and became more risk seeking. The degree of change in risk aversion due to decoy presentation was positively correlated with the original degrees of risk aversion. To explain these results and reveal underlying neural mechanisms, we developed new models incorporating both low- and high-level processes and used these models to fit individuals’ choice behavior. We found that observed distinct patterns of decoy effects can be explained by a combination of adjustments in neural representations and competitive weighting of reward attributes, both of which depend on risk aversion but in opposite directions. Altogether, our results demonstrate how a combination of low- and high-level processes shapes choice behavior in more naturalistic settings, modulates overall risk preference, and explains distinct behavioral phenotypes.

A large body of experimental work has illustrated that the introduction of a new, and often irrelevant, option can influence preference among the existing options, a phenomenon referred to as context or decoy effects. For example, introducing a new option that is worse than one of the two existing options in all its attributes but better than the alternative option in some attributes (and thus should not ever be selected) can increase the preference for the former option. Context effects have been explained by high-level cognitive processes—such as comparisons and competitions between attributes—or low-level adjustments of neural representations. However, it is unclear how these processes interact to shape individuals’ responses to context. Here, we show that both high-level cognitive processes and low-level neural adjustments shift risk preference during choice between multiple risky options but in opposite directions. Moreover, we demonstrate that combinations of these processes can account for distinct patterns of context effects in human subjects.