Negative Reward Prediction Error Research Articles

The effect of reward prediction errors on subjective affect depends on outcome valence and decision context.

The valence of an individual's emotional response to an event is often thought to depend on their prior expectations for the event: better-than-expected outcomes produce positive affect and worse-than-expected outcomes produce negative affect. In recent years, this hypothesis has been instantiated within influential computational models of subjective affect that assume the valence of affect is driven by reward prediction errors. However, there remain a number of open questions regarding this association. In this project, we investigated the moderating effects of outcome valence and decision context (Experiment 1: free vs. forced choices; Experiment 2: trials with vs. trials without counterfactual feedback) on the effects of reward prediction errors on subjective affect. We conducted two online experiments (N = 300 in total) of general-population samples recruited via Prolific to complete a risky decision-making task with an embedded high-resolution sampling of subjective affect. Hierarchical Bayesian computational modeling revealed that both outcome amount and reward prediction errors influenced subjective affect, but that the effects of reward prediction errors were moderated by both outcome valence and decision context. Specifically, we found evidence that only negative reward prediction errors (worse-than-expected outcomes) influenced subjective affect, with no significant effect of positive reward prediction errors (better-than-expected outcomes). Moreover, these effects were only apparent on trials in which participants made a choice freely (but not on forced-choice trials) and when counterfactual feedback was absent (but not when counterfactual feedback was present). These results deepen our understanding of the effects of reward prediction errors on subjective affect. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Dynamics Learning Rate Bias in Pigeons: Insights from Reinforcement Learning and Neural Correlates

Research in reinforcement learning indicates that animals respond differently to positive and negative reward prediction errors, which can be calculated by assuming learning rate bias. Many studies have shown that humans and other animals have learning rate bias during learning, but it is unclear whether and how the bias changes throughout the entire learning process. Here, we recorded the behavior data and the local field potentials (LFPs) in the striatum of five pigeons performing a probabilistic learning task. Reinforcement learning models with and without learning rate biases were used to dynamically fit the pigeons’ choice behavior and estimate the option values. Furthemore, the correlation between the striatal LFPs power and the model-estimated option values was explored. We found that the pigeons’ learning rate bias shifted from negative to positive during the learning process, and the striatal Gamma (31 to 80 Hz) power correlated with the option values modulated by dynamic learning rate bias. In conclusion, our results support the hypothesis that pigeons employ a dynamic learning strategy in the learning process from both behavioral and neural aspects, providing valuable insights into reinforcement learning mechanisms of non-human animals.

Asymmetric coding of reward prediction errors in human insula and dorsomedial prefrontal cortex

The signed value and unsigned salience of reward prediction errors (RPEs) are critical to understanding reinforcement learning (RL) and cognitive control. Dorsomedial prefrontal cortex (dMPFC) and insula (INS) are key regions for integrating reward and surprise information, but conflicting evidence for both signed and unsigned activity has led to multiple proposals for the nature of RPE representations in these brain areas. Recently developed RL models allow neurons to respond differently to positive and negative RPEs. Here, we use intracranially recorded high frequency activity (HFA) to test whether this flexible asymmetric coding strategy captures RPE coding diversity in human INS and dMPFC. At the region level, we found a bias towards positive RPEs in both areas which paralleled behavioral adaptation. At the local level, we found spatially interleaved neural populations responding to unsigned RPE salience and valence-specific positive and negative RPEs. Furthermore, directional connectivity estimates revealed a leading role of INS in communicating positive and unsigned RPEs to dMPFC. These findings support asymmetric coding across distinct but intermingled neural populations as a core principle of RPE processing and inform theories of the role of dMPFC and INS in RL and cognitive control.

Remembering unexpected beauty: Contributions of the ventral striatum to the processing of reward prediction errors regarding the facial attractiveness in face memory

The COVID-19 pandemic has led people to predict facial attractiveness from partially covered faces. Differences in the predicted and observed facial attractiveness (i.e., masked and unmasked faces, respectively) are defined as reward prediction error (RPE) in a social context. Cognitive neuroscience studies have elucidated the neural mechanisms underlying RPE-induced memory improvements in terms of monetary rewards. However, little is known about the mechanisms underlying RPE-induced memory modulation in terms of social rewards. To elucidate this, the present functional magnetic resonance imaging (fMRI) study investigated activity and functional connectivity during face encoding. In encoding trials, participants rated the predicted attractiveness of faces covered except for around the eyes (prediction phase) and then rated the observed attractiveness of these faces without any cover (outcome phase). The difference in ratings between these phases was defined as RPE in facial attractiveness, and RPE was categorized into positive RPE (increased RPE from the prediction to outcome phases), negative RPE (decreased RPE from the prediction to outcome phases), and non-RPE (no difference in RPE between the prediction and outcome phases). During retrieval, participants were presented with individual faces that had been seen and unseen in the encoding trials, and were required to judge whether or not each face had been seen in the encoding trials. Univariate activity in the ventral striatum (VS) exhibited a linear increase with increased RPE in facial attractiveness. In the multivariate pattern analysis (MVPA), activity patterns in the VS and surrounding areas (extended VS) significantly discriminated between positive/negative RPE and non-RPE. In the functional connectivity analysis, significant functional connectivity between the extended VS and the hippocampus was observed most frequently in positive RPE. Memory improvements by face-based RPE could be involved in functional networks between the extended VS (representing RPE) and the hippocampus, and the interaction could be modulated by RPE values in a social context.

Involvement of cortical input to the rostromedial tegmental nucleus in aversion to foot shock.

The rostromedial tegmental nucleus (RMTg) encodes negative reward prediction error (RPE) and plays an important role in guiding behavioral responding to aversive stimuli. Previous research has focused on regulation of RMTg activity by the lateral habenula despite studies revealing RMTg afferents from other regions including the frontal cortex. The current study provides a detailed anatomical and functional analysis of cortical input to the RMTg of male rats. Retrograde tracing uncovered dense cortical input to the RMTg spanning the medial prefrontal cortex, the orbitofrontal cortex and anterior insular cortex. Afferents were most dense in the dorsomedial subregion of the PFC (dmPFC), an area that is also implicated in both RPE signaling and aversive responding. RMTg-projecting dmPFC neurons originate in layer V, are glutamatergic, and collateralize to select brain regions. In-situ mRNA hybridization revealed that neurons in this circuit are predominantly D1 receptor-expressing with a high degree of D2 receptor colocalization. Consistent with cFos induction in this neural circuit during exposure to foot shock and shock-predictive cues, optogenetic stimulation of dmPFC terminals in the RMTg drove avoidance. Lastly, acute slice electrophysiology and morphological studies revealed that exposure to repeated foot shock resulted in significant physiological and structural changes consistent with a loss of top-down modulation of RMTg-mediated signaling. Altogether, these data reveal the presence of a prominent cortico-subcortical projection involved in adaptive behavioral responding to aversive stimuli such as foot shock and provide a foundation for future work aimed at exploring alterations in circuit function in diseases characterized by deficits in cognitive control over reward and aversion.

Prediction errors arising from switches between major and minor modes in music: An fMRI study

The major and minor modes in Western music have positive and negative connotations, respectively. The present fMRI study examined listeners’ neural responses to switches between major and minor modes. We manipulated the final chords of J. S. Bach’s keyboard pieces so that each major-mode passage ended with either the major (Major-Major) or minor (Major-Minor) tonic chord, and each minor-mode passage ended with either the minor (Minor-Minor) or major (Minor-Major) tonic chord. If the final major and minor chords have positive and negative reward values respectively, the Major-Minor and Minor-Major stimuli would cause negative and positive reward prediction errors (RPEs) respectively in a listener’s brain. We found that activity in a frontoparietal network was significantly higher for Major-Minor than for Major-Major. Based on previous research, these results support the idea that a major-to-minor switch causes negative RPE. The contrast of Minor-Major minus Minor-Minor yielded activation in the ventral insula and visual cortex, speaking against the idea that a minor-to-major switch causes positive RPE. We discuss our results in relation to executive functions and the emotional connotations of major versus minor modes.

Individuals with problem gambling and obsessive-compulsive disorder learn through distinct reinforcement mechanisms.

Obsessive-compulsive disorder (OCD) and pathological gambling (PG) are accompanied by deficits in behavioural flexibility. In reinforcement learning, this inflexibility can reflect asymmetric learning from outcomes above and below expectations. In alternative frameworks, it reflects perseveration independent of learning. Here, we examine evidence for asymmetric reward-learning in OCD and PG by leveraging model-based functional magnetic resonance imaging (fMRI). Compared with healthy controls (HC), OCD patients exhibited a lower learning rate for worse-than-expected outcomes, which was associated with the attenuated encoding of negative reward prediction errors in the dorsomedial prefrontal cortex and the dorsal striatum. PG patients showed higher and lower learning rates for better- and worse-than-expected outcomes, respectively, accompanied by higher encoding of positive reward prediction errors in the anterior insula than HC. Perseveration did not differ considerably between the patient groups and HC. These findings elucidate the neural computations of reward-learning that are altered in OCD and PG, providing a potential account of behavioural inflexibility in those mental disorders.

A computational neuroimaging study of reinforcement learning and goal-directed exploration in schizophrenia spectrum disorders.

Prior evidence indicates that negative symptom severity and cognitive deficits, in people with schizophrenia (PSZ), relate to measures of reward-seeking and loss-avoidance behavior (implicating the ventral striatum/VS), as well as uncertainty-driven exploration (reliant on rostrolateral prefrontal cortex/rlPFC). While neural correlates of reward-seeking and loss-avoidance have been examined in PSZ, neural correlates of uncertainty-driven exploration have not. Understanding neural correlates of uncertainty-driven exploration is an important next step that could reveal insights to how this mechanism of cognitive and negative symptoms manifest at a neural level. We acquired fMRI data from 29 PSZ and 36 controls performing the Temporal Utility Integration decision-making task. Computational analyses estimated parameters corresponding to learning rates for both positive and negative reward prediction errors (RPEs) and the degree to which participates relied on representations of relative uncertainty. Trial-wise estimates of expected value, certainty, and RPEs were generated to model fMRI data. Behaviorally, PSZ demonstrated reduced reward-seeking behavior compared to controls, and negative symptoms were positively correlated with loss-avoidance behavior. This finding of a bias toward loss avoidance learning in PSZ is consistent with previous work. Surprisingly, neither behavioral measures of exploration nor neural correlates of uncertainty in the rlPFC differed significantly between groups. However, we showed that trial-wise estimates of relative uncertainty in the rlPFC distinguished participants who engaged in exploratory behavior from those who did not. rlPFC activation was positively associated with intellectual function. These results further elucidate the nature of reinforcement learning and decision-making in PSZ and healthy volunteers.

Neural mechanism underlying depressive-like state associated with social status loss

Downward social mobility is a well-known mental risk factor for depression, but its neural mechanism remains elusive. Here, by forcing mice to lose against their subordinates in a non-violent social contest, we lower their social ranks stably and induce depressive-like behaviors. These rank-decline-associated depressive-like behaviors can be reversed by regaining social status. Invivo fiber photometry and single-unit electrophysiological recording show that forced loss, but not natural loss, generates negative reward prediction error (RPE). Through the lateral hypothalamus, the RPE strongly activates the brain's anti-reward center, the lateral habenula (LHb). LHb activation inhibits the medial prefrontal cortex (mPFC) that controls social competitiveness and reinforces retreats in contests. These results reveal the core neural mechanisms mutually promoting social status loss and depressive behaviors. The intertwined neuronal signaling controlling mPFC and LHb activities provides a mechanistic foundation for the crosstalk between social mobility and psychological disorder, unveiling a promising target for intervention.

Opponent Learning with Different Representations in the Cortico-Basal Ganglia Circuits.

The direct and indirect pathways of the basal ganglia (BG) have been suggested to learn mainly from positive and negative feedbacks, respectively. Since these pathways unevenly receive inputs from different cortical neuron types and/or regions, they may preferentially use different state/action representations. We explored whether such a combined use of different representations, coupled with different learning rates from positive and negative reward prediction errors (RPEs), has computational benefits. We modeled animal as an agent equipped with two learning systems, each of which adopted individual representation (IR) or successor representation (SR) of states. With varying the combination of IR or SR and also the learning rates from positive and negative RPEs in each system, we examined how the agent performed in a dynamic reward navigation task. We found that combination of SR-based system learning mainly from positive RPEs and IR-based system learning mainly from negative RPEs could achieve a good performance in the task, as compared with other combinations. In such a combination of appetitive SR-based and aversive IR-based systems, both systems show activities of comparable magnitudes with opposite signs, consistent with the suggested profiles of the two BG pathways. Moreover, the architecture of such a combination provides a novel coherent explanation for the functional significance and underlying mechanism of diverse findings about the cortico-BG circuits. These results suggest that particularly combining different representations with appetitive and aversive learning could be an effective learning strategy in certain dynamic environments, and it might actually be implemented in the cortico-BG circuits.

Effort Reinforces Learning.

Humans routinely learn the value of actions by updating their expectations based on past outcomes - a process driven by reward prediction errors (RPEs). Importantly, however, implementing a course of action also requires the investment of effort. Recent work has revealed a close link between the neural signals involved in effort exertion and those underpinning reward-based learning, but the behavioral relationship between these two functions remains unclear. Across two experiments, we tested healthy male and female human participants (N = 140) on a reinforcement learning task in which they registered their responses by applying physical force to a pair of hand-held dynamometers. We examined the effect of effort on learning by systematically manipulating the amount of force required to register a response during the task. Our key finding, replicated across both experiments, was that greater effort increased learning rates following positive outcomes and decreased them following negative outcomes, which corresponded to a differential effect of effort in boosting positive RPEs and blunting negative RPEs. Interestingly, this effect was most pronounced in individuals who were more averse to effort in the first place, raising the possibility that the investment of effort may have an adaptive effect on learning in those less motivated to exert it. By integrating principles of reinforcement learning with neuroeconomic approaches to value-based decision-making, we show that the very act of investing effort modulates one's capacity to learn, and demonstrate how these functions may operate within a common computational framework.SIGNIFICANCE STATEMENT Recent work suggests that learning and effort may share common neurophysiological substrates. This raises the possibility that the very act of investing effort influences learning. Here, we tested whether effort modulates teaching signals in a reinforcement learning paradigm. Our results showed that effort resulted in more efficient learning from positive outcomes and less efficient learning from negative outcomes. Interestingly, this effect varied across individuals, and was more pronounced in those who were more averse to investing effort in the first place. These data highlight the importance of motivational factors in a common framework of reward-based learning, which integrates the computational principles of reinforcement learning with those of value-based decision-making.

Choice-selective sequences dominate in cortical relative to thalamic inputs to NAc to support reinforcement learning.

SUMMARYHow are actions linked with subsequent outcomes to guide choices? The nucleus accumbens, which is implicated in this process, receives glutamatergic inputs from the prelimbic cortex and midline regions of the thalamus. However, little is known about whether and how representations differ across these input pathways. By comparing these inputs during a reinforcement learning task in mice, we discovered that prelimbic cortical inputs preferentially represent actions and choices, whereas midline thalamic inputs preferentially represent cues. Choice-selective activity in the prelimbic cortical inputs is organized in sequences that persist beyond the outcome. Through computational modeling, we demonstrate that these sequences can support the neural implementation of reinforcement-learning algorithms, in both a circuit model based on synaptic plasticity and one based on neural dynamics. Finally, we test and confirm a prediction of our circuit models by direct manipulation of nucleus accumbens input neurons.

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.

Differential effects of positive and negative reward prediction error on saccade response time adaptation in a reversal learning visual stop-signal task

Differential effects of positive and negative reward prediction error on saccade response time adaptation in a reversal learning visual stop-signal task

An association between prediction errors and risk-seeking: Theory and behavioral evidence.

Reward prediction errors (RPEs) and risk preferences have two things in common: both can shape decision making behavior, and both are commonly associated with dopamine. RPEs drive value learning and are thought to be represented in the phasic release of striatal dopamine. Risk preferences bias choices towards or away from uncertainty; they can be manipulated with drugs that target the dopaminergic system. Based on the common neural substrate, we hypothesize that RPEs and risk preferences are linked on the level of behavior as well. Here, we develop this hypothesis theoretically and test it empirically. First, we apply a recent theory of learning in the basal ganglia to predict how RPEs influence risk preferences. We find that positive RPEs should cause increased risk-seeking, while negative RPEs should cause risk-aversion. We then test our behavioral predictions using a novel bandit task in which value and risk vary independently across options. Critically, conditions are included where options vary in risk but are matched for value. We find that our prediction was correct: participants become more risk-seeking if choices are preceded by positive RPEs, and more risk-averse if choices are preceded by negative RPEs. These findings cannot be explained by other known effects, such as nonlinear utility curves or dynamic learning rates.

Monitoring and proactive control of visual search speed-accuracy tradeoff by supplementary eye field

Neurophysiological mechanisms of speed-accuracy tradeoff (SAT) have only recently been investigated. Previous studies with macaque monkeys showed that SAT of inefficient visual search was accomplished by modulation of salience map evidence representations in frontal eye field (FEF) and superior colliculus (SC). Saccade initiation occurred when movement activity was either reduced or equivalent in Accurate-cued relative to Fast-cued trials. Targeting errors occurred when salience neurons in the FEF and the SC treated distractors as targets. Here, we report new observations about SAT performance and neurophysiological results from the supplementary eye field (SEF), located in medial frontal cortex, which contributes to executive control of gaze behavior but not to saccade target selection. In two monkeys, we found natural preference for quick responding, with more choice inaccuracies. On trials with choice errors, the monkeys often executed a post-primary saccade to the foregone target, which was generated later in the Accurate than the Fast condition. SEF neurons signaled choice errors, and the magnitude of this signal predicted whether the post-primary saccade was corrective in nature. In the Accurate condition, when slower responding was required, the rate of choice errors decreased, but the rate of premature timing errors increased. After timing errors, SEF neurons signaled negative reward prediction error at the time of expected reward. We assessed the distribution of signaling of choice errors and reward prediction errors across the sample of SEF neurons. Some neurons signaled both types of error, whereas others signaled either choice error or reward prediction error. We also found evidence for proactive control of response time. Baseline activity was a strong predictor of response time in both Fast and Accurate conditions. Taken together, these results indicate that SEF may be a source of the modulation observed in FEF and SC and inform new models of distributed decision making.

Sensitivity to negative and positive feedback as a stable and enduring behavioural trait in rats

RationaleAccording to psychological theories, cognitive distortions play a pivotal role in the aetiology and recurrence of mood disorders. Although clinical evidence for the coexistence of depression and altered sensitivity to performance feedback is relatively coherent, we still do not know whether increased or decreased sensitivity to positive or negative feedback is associated with ‘pro-depressive’ profile in healthy subjects.ObjectiveOur research has been designed to answer this question, and here, we present the first steps in that direction.MethodsUsing a rat version of the probabilistic reversal-learning (PRL) paradigm, we evaluated how sensitivity to negative and positive feedback influences other cognitive processes associated with mood disorders, such as motivation in the progressive ratio schedule of reinforcement (PRSR) paradigm, hedonic status in the sucrose preference (SP) test, locomotor and exploratory activity in the open field (OF) test, and anxiety in the light/dark box (LDB) test.ResultsThe results of our study demonstrated for the first time that in rodents, sensitivity to negative and positive feedback could be considered a stable and enduring behavioural trait. Importantly, we also showed that these traits are independent of each other and that trait sensitivity to positive feedback is associated with cognitive flexibility in the PRL test. The computational modelling results also revealed that in animals classified as sensitive to positive feedback, the α learning rates for both positive and negative reward prediction errors were higher than those in animals classified as insensitive. We observed no statistically significant interactions between sensitivity to negative or positive feedback and the parameters measured in the PRSR, SP, OF or LDB tests.ConclusionsFurther studies using animal models of depression based on chronic stress should reveal whether sensitivity to feedback is a latent trait that when interacts with stressful life events, could produce correlates of depressive symptoms in rats.

The roles of habenula and related neural circuits in neuropsychiatric diseases

The habenula is a small and bilateral nucleus above dorsal thalamus, which contains several different types of neurons. The habenula has extensive connections with the forebrain, septum and monoaminergic nuclei in the midbrain and brainstem. Habenula is known as an 'anti-reward' nucleus, which can be activated by aversive stimulus and negative reward prediction errors. Accumulating researchs have implicated that the habenula is involved in several behaviors crucial to survival. Meanwhile, the roles of the habenula in neuropsychiatric diseases have received increasing attention. This review summaries the studies regarding the roles of habenula and the related circuits in neuropathic pain, depression, drug addiction and schizophrenia, and discusses the possibility to use the habenula as a treatment target.

Neural Signatures of Prediction Errors in a Decision-Making Task Are Modulated by Action Execution Failures

Decisions must be implemented through actions, and actions are prone to error. As such, when an expected outcome is not obtained, an individual should be sensitive to not only whether the choice itself was suboptimal but also whether the action required to indicate that choice was executed successfully. The intelligent assignment of credit to action execution versus action selection has clear ecological utility for the learner. To explore this, we used a modified version of a classic reinforcement learning task in which feedback indicated whether negative prediction errors were, or were not, associated with execution errors. Using fMRI, we asked if prediction error computations in the human striatum, a key substrate in reinforcement learning and decision making, are modulated when a failure in action execution results in the negative outcome. Participants were more tolerant of non-rewarded outcomes when these resulted from execution errors versus when execution was successful, but reward was withheld. Consistent with this behavior, a model-driven analysis of neural activity revealed an attenuation of the signal associated with negative reward prediction errors in the striatum following execution failures. These results converge with other lines of evidence suggesting that prediction errors in the mesostriatal dopamine system integrate high-level information during the evaluation of instantaneous reward outcomes.

Decreases in Cued Reward Seeking After Reward-Paired Inhibition of Mesolimbic Dopamine

Reward-paired optogenetic manipulation of dopamine neurons can increase or decrease behavioral responding to antecedent cues when subjects have the opportunity for new learning, in accordance with a dopamine-mediated error learning signal. Here we examined the impact of reward-paired dopamine neuron inhibition on behavioral responding to reward-predictive cues after subjects had learned. We trained male TH-IRES-Cre mice to lever press for food reward in a progressive ratio procedure, a 2-cue choice procedure, or when continuously reinforced; in all procedures, completion of the response requirement was signaled by an auditory cue presented prior to food delivery. After training, mice underwent successive sessions in which optogenetic inhibition of dopamine neurons was triggered during food receipt. Rather than mimic brief inhibitions associated with negative reward prediction errors, we applied inhibition throughout the ingestion period on each trial. We found in all procedures that optogenetic inhibition of dopamine neurons during reward receipt decreased behavioral responding to the preceding reward-predictive cue over days, a behavioral change observed during time periods without optogenetic neuronal inhibition. Extinction-like behavioral responding was selective for learned associations: it was observed in the 2-cue choice procedure in which each subject was trained on two associations and inhibition was paired with reward for only one of the associations. Thus, inhibition during reward receipt can decrease responding to reward-predictive cues, sharing some features of behavioral extinction. These findings suggest changes in mesolimbic dopaminergic transmission at the time of experienced reward impacts subsequent responding to cues in well-trained subjects as predicted for a learning signal.