Neural arbitration between social and individual learning systems

Predicting treatment response to ketamine in treatment-resistant depression using auditory mismatch negativity: Study protocol

BACKGROUND

Ketamine has recently attracted considerable attention for its rapid effects on patients with major depressive disorder, including treatment-resistant depression (TRD). Despite ketamine's promising results in treating depression, a significant number of patients do not respond to the treatment, and predicting who will benefit remains a challenge. Although its antidepressant effects are known to be linked to its action as an antagonist of the N-methyl-D-aspartate (NMDA) receptor, the precise mechanisms that determine why some patients respond and others do not are still unclear.

OBJECTIVE

This study aims to understand the computational mechanisms underlying changes in the auditory mismatch negativity (MMN) response following treatment with intravenous ketamine. Moreover, we aim to link the computational mechanisms to their underlying neural causes and use the parameters of the neurocomputational model to make individual treatment predictions.

METHODS

This is a prospective study of 30 patients with TRD who are undergoing intravenous ketamine therapy. Prior to 3 out of 4 ketamine infusions, EEG will be recorded while patients complete the auditory MMN task. Depression, suicidality, and anxiety will be assessed throughout the study and a week after the last ketamine infusion. To translate the effects of ketamine on the MMN to computational mechanisms, we will model changes in the auditory MMN using the hierarchical Gaussian filter, a hierarchical Bayesian model. Furthermore, we will employ a conductance-based neural mass model of the electrophysiological data to link these computational mechanisms to their neural causes.

CONCLUSION

The findings of this study may improve understanding of the mechanisms underlying response and resistance to ketamine treatment in patients with TRD. The parameters obtained from fitting computational models to EEG recordings may facilitate single-patient treatment predictions, which could provide clinically useful prognostic information.

TRIAL REGISTRATION

Clinicaltrials.gov NCT05464264. Registered June 24, 2022.

Heterogeneity in strategy use during arbitration between experiential and observational learning

To navigate our complex social world, it is crucial to deploy multiple learning strategies, such as learning from directly experiencing action outcomes or from observing other people's behavior. Despite the prevalence of experiential and observational learning in humans and other social animals, it remains unclear how people favor one strategy over the other depending on the environment, and how individuals vary in their strategy use. Here, we describe an arbitration mechanism in which the prediction errors associated with each learning strategy influence their weight over behavior. We designed an online behavioral task to test our computational model, and found that while a substantial proportion of participants relied on the proposed arbitration mechanism, there was some meaningful heterogeneity in how people solved this task. Four other groups were identified: those who used a fixed mixture between the two strategies, those who relied on a single strategy and non-learners with irrelevant strategies. Furthermore, groups were found to differ on key behavioral signatures, and on transdiagnostic symptom dimensions, in particular autism traits and anxiety. Together, these results demonstrate how large heterogeneous datasets and computational methods can be leveraged to better characterize individual differences.

Disentangling the contribution of individual and social learning processes in human advice-taking behavior

The study of social learning examines how individuals learn from others by means of observation, imitation, or compliance with advice. However, it still remains largely unknown whether social learning processes have a distinct contribution to behavior, independent from non-social trial-and-error learning that often occurs simultaneously. 153 participants completed a reinforcement learning task, where they were asked to make choices to gain rewards. Advice from an artificial teacher was presented in 60% of the trials, allowing us to compare choice behavior with and without advice. Results showed a strong and reliable tendency to follow advice (test-retest reliability ~0.73). Computational modeling suggested a unique contribution of three distinct learning strategies: (a) individual learning (i.e., learning the value of actions, independent of advice), (b) informed advice-taking (i.e., learning the value of following advice), and (c) non-informed advice-taking (i.e., a constant bias to follow advice regardless of outcome history). Comparing artificial and empirical data provided specific behavioral regression signatures to both informed and non-informed advice taking processes. We discuss the theoretical implications of integrating internal and external information during the learning process.

Hierarchical inference as a source of human biases

The finding that human decision-making is systematically biased continues to have an immense impact on both research and policymaking. Prevailing views ascribe biases to limited computational resources, which require humans to resort to less costly resource-rational heuristics. Here, we propose that many biases in fact arise due to a computationally costly way of coping with uncertainty-namely, hierarchical inference-which by nature incorporates information that can seem irrelevant. We show how, in uncertain situations, Bayesian inference may avail of the environment's hierarchical structure to reduce uncertainty at the cost of introducing bias. We illustrate how this account can explain a range of familiar biases, focusing in detail on the halo effect and on the neglect of base rates. In each case, we show how a hierarchical-inference account takes the characterization of a bias beyond phenomenological description by revealing the computations and assumptions it might reflect. Furthermore, we highlight new predictions entailed by our account concerning factors that could mitigate or exacerbate bias, some of which have already garnered empirical support. We conclude that a hierarchical inference account may inform scientists and policy makers with a richer understanding of the adaptive and maladaptive aspects of human decision-making.

Teachers recruit mentalizing regions to represent learners' beliefs

Teaching enables humans to impart vast stores of culturally specific knowledge and skills. However, little is known about the neural computations that guide teachers' decisions about what information to communicate. Participants (N = 28) played the role of teachers while being scanned using fMRI; their task was to select examples that would teach learners how to answer abstract multiple-choice questions. Participants' examples were best described by a model that selects evidence that maximizes the learner's belief in the correct answer. Consistent with this idea, participants' predictions about how well learners would do closely tracked the performance of an independent sample of learners (N = 140) who were tested on the examples they had provided. In addition, regions that play specialized roles in processing social information, namely the bilateral temporoparietal junction and middle and dorsal medial prefrontal cortex, tracked learners' posterior belief in the correct answer. Our results shed light on the computational and neural architectures that support our extraordinary abilities as teachers.

Brain connectivity modulation by Bayesian surprise in relation to control demand drives cognitive flexibility via control engagement

Human control is characterized by its flexibility and adaptability in response to the conditional probability in the environment. Previous studies have revealed that efficient conflict control could be attained by predicting and adapting to the changing control demand. However, it is unclear whether cognitive flexibility could also be gained by predicting and adapting to the changing control demand. The present study aimed to explore this issue by combining the model-based analyses of behavioral and neuroimaging data with a probabilistic cued task switching paradigm. We demonstrated that the Bayesian surprise (i.e. unsigned precision-weighted prediction error [PE]) negatively modulated the connections among stimulus processing brain regions and control regions/networks. The effect of Bayesian surprise modulation on these connections guided control engagement as reflected by the control PE effect on behavior, which in turn facilitated cognitive flexibility. These results bridge a gap in the literature by illustrating the neural and behavioral effect of control demand prediction (or PE) on cognitive flexibility and offer novel insights into the source of switch cost and the mechanism of cognitive flexibility.

Emotion dynamics as hierarchical Bayesian inference in time

What fundamental property of our environment would be most valuable and optimal in characterizing the emotional dynamics we experience in daily life? Empirical work has shown that an accurate estimation of uncertainty is necessary for our optimal perception, learning, and decision-making. However, the role of this uncertainty in governing our affective dynamics remains unexplored. Using Bayesian encoding, decoding and computational modeling, on a large-scale neuroimaging and behavioral data on a passive movie-watching task, we showed that emotions naturally arise due to ongoing uncertainty estimations about future outcomes in a hierarchical neural architecture. Several prefrontal subregions hierarchically encoded a lower-dimensional signal that highly correlated with the evolving uncertainty. Crucially, the lateral orbitofrontal cortex (lOFC) tracked the temporal fluctuations of this uncertainty and was predictive of the participants' predisposition to anxiety. Furthermore, we observed a distinct functional double-dissociation within OFC with increased connectivity between medial OFC and DMN, while with that of lOFC and FPN in response to the evolving affect. Finally, we uncovered a temporally predictive code updating an individual's beliefs spontaneously with fluctuating outcome uncertainty in the lOFC. A biologically relevant and computationally crucial parameter in the theories of brain function, we propose uncertainty to be central to the definition of complex emotions.

The computational relationship between reinforcement learning, social inference, and paranoia

Theoretical accounts suggest heightened uncertainty about the state of the world underpin aberrant belief updates, which in turn increase the risk of developing a persecutory delusion. However, this raises the question as to how an agent's uncertainty may relate to the precise phenomenology of paranoia, as opposed to other qualitatively different forms of belief. We tested whether the same population (n = 693) responded similarly to non-social and social contingency changes in a probabilistic reversal learning task and a modified repeated reversal Dictator game, and the impact of paranoia on both. We fitted computational models that included closely related parameters that quantified the rigidity across contingency reversals and the uncertainty about the environment/partner. Consistent with prior work we show that paranoia was associated with uncertainty around a partner's behavioural policy and rigidity in harmful intent attributions in the social task. In the non-social task we found that pre-existing paranoia was associated with larger decision temperatures and commitment to suboptimal cards. We show relationships between decision temperature in the non-social task and priors over harmful intent attributions and uncertainty over beliefs about partners in the social task. Our results converge across both classes of model, suggesting paranoia is associated with a general uncertainty over the state of the world (and agents within it) that takes longer to resolve, although we demonstrate that this uncertainty is expressed asymmetrically in social contexts. Our model and data allow the representation of sociocognitive mechanisms that explain persecutory delusions and provide testable, phenomenologically relevant predictions for causal experiments.

The devil is in the task structure

Conflicting evidence about how the brain processes social and individual learning stems from which type of information is presented as the primary source of knowledge during experiments.

The prefrontal cortex and (uniquely) human cooperation: a comparative perspective

Humans have an exceptional ability to cooperate relative to many other species. We review the neural mechanisms supporting human cooperation, focusing on the prefrontal cortex. One key feature of human social life is the prevalence of cooperative norms that guide social behavior and prescribe punishment for noncompliance. Taking a comparative approach, we consider shared and unique aspects of cooperative behaviors in humans relative to nonhuman primates, as well as divergences in brain structure that might support uniquely human aspects of cooperation. We highlight a medial prefrontal network common to nonhuman primates and humans supporting a foundational process in cooperative decision-making: valuing outcomes for oneself and others. This medial prefrontal network interacts with lateral prefrontal areas that are thought to represent cooperative norms and modulate value representations to guide behavior appropriate to the local social context. Finally, we propose that more recently evolved anterior regions of prefrontal cortex play a role in arbitrating between cooperative norms across social contexts, and suggest how future research might fruitfully examine the neural basis of norm arbitration.

Neural arbitration between social and individual learning systems

Decision making requires integrating knowledge gathered from personal experiences with advice from others. The neural underpinnings of the process of arbitrating between information sources has not been fully elucidated. In this study, we formalized arbitration as the relative precision of predictions, afforded by each learning system, using hierarchical Bayesian modeling. In a probabilistic learning task, participants predicted the outcome of a lottery using recommendations from a more informed advisor and/or self-sampled outcomes. Decision confidence, as measured by the number of points participants wagered on their predictions, varied with our definition of arbitration as a ratio of precisions. Functional neuroimaging demonstrated that arbitration signals were independent of decision confidence and involved modality-specific brain regions. Arbitrating in favor of self-gathered information activated the dorsolateral prefrontal cortex and the midbrain, whereas arbitrating in favor of social information engaged the ventromedial prefrontal cortex and the amygdala. These findings indicate that relative precision captures arbitration between social and individual learning systems at both behavioral and neural levels.