Abstracts
Christian Doeller, Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig, Germany and Kavli Institute for Systems Neuroscience, NTNU Trondheim, Norway
Cognitive coding in the hippocampal-entorhinal system
Keynote Lecture
The fundamental question in cognitive neuroscience — what are the key neural coding principles underlying higher-level cognition in humans — still remains largely unanswered. In my long-term attempt to tackle this question, I use human memory and the neural population code for space as a model system. I will present evidence from fMRI, MEG and virtual reality experiments for cognitive coding mechanisms in the hippocampal-entorhinal system, such as continuous coding of virtual, mentally simulated, conceptual and visual space, hierarchical scaling of memories and attractor- based integration. Finally, I will provide evidence for an early breakdown of the entorhinal system in a genetic model of Alzheimer’s disease. Identifying neural metrics of cognition could open up the possibilities for new technology to enhance cognition as well as for novel biomarkers for an early detection of neurodegenerative diseases.
Matt Nassar, Brown University
Learning as statistical inference: Neural and computational mechanisms for normative learning
Keynote Lecture
Successful decision making often requires learning from prediction errors, but how much should we learn from any given error? I will examine this question in detail, drawing on an optimal inference model to formalise how we should learn in dynamic environments and a computationally efficient approximation to provide insight into how we could do so by adjusting the rate of learning from moment to moment. I will show behavioural data validating key model predictions in humans, demonstrate a role for the arousal system in setting the learning rate, and dissect the computational roles of neural subsystems upstream of learning rate implementation. I will explore the possibility that learning deficits might emerge from a failure to correctly determine how much should be learned, rather than a failure to represent prediction errors per se, and provide evidence for such an explanation in the case of healthy ageing. Finally, I will re-examine neural architecture of error-driven learning in the context of these results and discuss some future directions emerging from this work.
Weisong Ong, University of Pennsylvania
Cooperation: A social strategy?
Keynote Lecture
We rarely make decisions in a social vacuum; as a life form that has based our success in cooperative activity within communities, many of our actions are chosen with an eye towards vicarious feelings of reward or pain experienced by another. However, there are times in which this desire for empathy is not in complete alignment with self-interest, resulting in the need to rely upon strategic reasoning about the beliefs, desires, and goals of others to make our decisions.
To take a closer look at the neural mechanisms underlying such processes, we use novel paradigms that allow interactive play between two players. We demonstrated that the primates, both human and non-human rely upon recursive reasoning to carry out gameplay. In the non-human primates, we obtained spiking activity in two brain areas, the primate homolog of the temporal-parietal junction (TPJ) which is implicated in mentalising, and the anterior cingulate gyrus, an area connected to empathy and vicarious experience. We found that the neurons in TPJ were able to signal cooperative action, an abstract concept that is independent of realised reward and motor action. This suggests that the capacity to reason strategically is deeply rooted in the social behaviour of primates.
Tomás Ryan, Trinity College Dublin
Information storage in memory engrams
Keynote Lecture
How is memory stored in the brain as information? Over the past 5 years, the ability to label, observe, and manipulate specific neuronal ensembles in an activity-dependent manner has allowed us to identify components of specific memory engrams in the rodent brain. This approach has the potential to revolutionize the study of memory, but our knowledge of memory engrams is still in its infancy. Our work attempts to address fundamental questions at the core of engram formation and function. What kind(s) of plasticity mediates information storage in engram cells? How are engrams maintained? How do engrams represent specific episodes? How are engrams modulated by development? What is the relationship between learned engrams and innate 'ingrams'? In my talk I will discuss significant experimental progress that has been made on some of these issues through published and unpublished studies, and also discuss new approaches that we are developing.
Andreas Brandmaier, Max Planck Institute for Human Development
and Rogier Kievit, University of Cambridge
Teaching Lecture
Structural equation models for cognitive neuroscience: A whirlwind tour of principles and applications Structural equation models (SEM) are multivariate models that express relationships among observed and latent variables. They subsume and generalise a variety of common analysis techniques such as regression, t test, analysis of variance, latent factor models, growth curves, or autoregressive models. Explicit measurement models allow us to move from a noisy observed level to a construct level accommodating variables of greater reliability and validity. We illustrate a variety of modern SEM applications showcasing the benefits of SEM using both cross-sectional and longitudinal models of brain and cognition, including multiple-cause multiple-indicator models, latent growth curve models, and latent change score models. We conclude with a novel machine- learning-inspired estimation technique for estimating regularised SEM that promises more robust results for a particular common challenge in cognitive neuroscience, namely when estimating large models with relatively few observations.
Stephen Fleming, University College London
Why metacognition matters for (computational) psychiatry
Teaching Lecture
Metacognition refers to the ability to introspect about and control other cognitive processes. In my talk I will review recent progress in our understanding of how metacognition works, and ask questions such as: How can we measure metacognition in the lab? What are the leading theories of how we introspect? Is metacognition uniquely human, and when did it appear in evolution and development? I will argue that metacognition research holds particular promise for computational psychiatry, as it seeks to develop models of how we think and feel about our behaviour. Such models may thus provide a powerful framework for bridging the gap between behaviour and subjective experience. In a recent web-based study we found initial evidence in support of this view, showing that metacognitive aspects of perceptual decision making track variability in psychiatric symptoms in the general population.
Hauke Heekeren, Freie Universität Berlin
Social learning and decision making across development
Teaching Lecture
Social learning plays a crucial role in human development, everyday life, and society. Especially when decisions are difficult, people rely on advice or recommendations regarding a decision or course of action. In my lecture, I will review recent work that we did on social learning and decision making in different age groups using a combination of computational modelling and functional neuroimaging.
Quentin Huys, University College London
Fitting RL models to data
Teaching Lecture
Reinforcement learning is one of the key techniques in the analysis of choices. In this tutorial, I will first give a crash course in reinforcement learning, introducing value functions and various ways of solving them. In the second part, I will discuss how to fit these models to experimental data and provide a worked example. Finally, I will briefly describe a few biological and psychiatric applications.
Nicolas Schuck, Max Planck Institute for Human Development
Reinforcement learning revisited
Teaching Lecture
The framework of reinforcement learning has helped to understand neural and computational issues underlying value-based decision making. While the bulk of evidence in the past was concerned with the representation of values, rewards and prediction errors, a number of novel findings have brought about significant reinterpretations and advancements. These findings highlight that neural value and prediction error signals carry information about the ongoing task and its structure, their computation takes uncertainty in the underlying task structure into account, and that the brain may employ the prefrontal cortex to represent the task structure itself. In addition, several lines of research have suggested that reactivation of episodic experiences may play a larger role in value- based updating and planning.
Douglas Garrett
The Lifespan Neural Dynamics Group (LNDG)
Max Planck UCL Centre Group Presentation
Our work seeks to understand why human brain activity fluctuates so markedly from moment to moment. Typically, researchers conceive of signal variability as neural “noise,” a nuisance factor that presumably interferes with the efficiency of neural processes. However, increasing evidence supports that moment-to-moment signal variability is needed for neural communication, flexibility, and adaptability. We have been at the forefront of an emerging line of neuroimaging research that has focused on the manner in which brain signal variability reflects broad-scale cognition and ageing, in particular showing that older, poorer performers exhibit less signal variability across moments. Broadly, we examine brain signal variability and dynamics in relation to six core research foci: lifespan development, cognition, neuromodulation, structural/functional connectivity, transcranial stimulation, and modelling/methods. This talk will focus on two lines of LNDG research. First, dopamine (DA) continues to emerge as a candidate mechanism underlying such individual differences in neural dynamics and cognition, with findings from DA agonism, antagonism, and PET binding potential supporting this brain–behaviour relationship. Further, under the working hypothesis that variability of neural activity is a prerequisite for cognitive adaptability, we also study the extent to which humans flexibly regulate neural variability (entropy) within-person to maximise reward during decision making. In recent work, we encouraged participants to use either a “liberal” (by penalising misses) or “conservative” (by penalising false alarms) response criterion for reporting perceptual targets. We found that greater upregulation of entropy in frontal electrodes during the liberal condition (liberal minus conservative) solely predicted more liberal shifts in signal-detection-theoretic criterion (bias). Thus, we provide first evidence that those who are able to more flexibly modulate neural variability also more flexibly adjust their behavioural responses to maximise reward.
Tobias Hauser
Developmental Computational Psychiatry Group
Max Planck UCL Centre Group Presentation
Many psychiatric disorders arise during adolescence, a period when brain and cognition undergo substantial developmental changes. However, it is unclear how aberrant developmental trajectories lead to the emergence of psychiatric disorders. The newly established developmental computational psychiatry group addresses these questions by studying the development of neurocognitive functions and how they go awry in psychiatric patients. During this talk I will illustrate how these methods can provide novel insight into the pathological mechanisms underlying psychiatric disorders, such as obsessive-compulsive disorder.
Quentin Huys
Clinical Translation Group
Max Planck UCL Centre Group Presentation
Quentin's group is a new addition to the Max Planck UCL Centre, having begun in August 2018. He and his group will continue to focus on using computational techniques to advance clinical problems. It will rely on both data-driven and theory-driven approaches and the embedding of experiments within clinical practice. In the past, he has focused on the development of methods for the analysis of behaviour and decision making, and the use of computational neuroscience for the prediction of clinically relevant outcomes such as relapses after alcohol detoxification or after discontinuing antidepressants.
Robb Rutledge
Decision Making and Emotion Group
Max Planck UCL Centre Group Presentation
The subjective well-being or happiness of individuals is an important metric for societies, but we know little about how the cumulative influence of daily life events are aggregated into subjective feelings. Using computational modelling, we consistently find that momentary happiness in decision tasks is explained not by task earnings, but by the combined influence of past rewards and expectations. The robustness of this account is evident in large-scale smartphone-based replications. We have applied this computational approach in mood disorders including major depression and bipolar disorder. In ongoing research, we quantify how mood is determined in relation to a wide variety of decision variables (e.g., effort, uncertainty, control). We use smartphones to longitudinally assess patients to try to predict symptoms over weeks or months based on our computational moods for mood dynamics and decision processes.
Manuel Arnold
Identifying parameter heterogeneity in autoregressive cross-lagged panel models
IMPRS COMP2PSYCH Fellows' Presentation
Structural equation models (SEM) are widely applied in the behavioural and social sciences to analyse the relationship between observed and latent variables. A standard assumption underlying many SEMs is that parameter values are equal for all observations in the sample. Thus, researchers need to determine whether relevant heterogeneity exists in their samples or they run the risk of reporting meaningless parameter estimates and inaccurate standard errors. During the last decades, several SEM extensions have been developed to identify and account for heterogeneity. One of those approaches is “individual parameter contribution” (IPC) regression, proposed by Oberski (2013). IPC regression is conducted in three steps. First, a theory-guided SEM is fitted. Second, the contributions of every individual to the model parameters are calculated based on the first-step model. Third, heterogeneity in the model parameters is explained by regressing the contributions on grouping variables or individual characteristics. This talk aims to illustrate how IPC regression can be used as a data-driven procedure to detect and provide estimates of individual or group differences in contemporary longitudinal structural equation models, focusing on autoregressive panel models in discrete and continuous time. It will be shown that substantial heterogeneity in the autoregressive parameters of a panel model may bias the IPC regression estimates. A novel bias correction procedure based on Fisher scoring will be presented.
Rachel Bedder
A computational model of mood and future prospects
IMPRS COMP2PSYCH Fellows' Presentation
Our mood is affected not only by what is currently happening, but also by what we believe is likely to happen in the future. However, the cognitive mechanisms by which future prospects impact current mood are unknown. Understanding these mechanisms is particularly important in clinical conditions characterised by rumination about possible negative events. I will present the results of a probabilistic reward paradigm (N = 42) and a replication (N = 38) with frequent experience sampling on mood. Participants made decisions between safer and riskier options in Gain blocks, where they can win points and in Loss blocks where they can only lose points. Participants were shown whether the next two blocks would consist of Gain or Loss trials to assess the effect that future prospects have on mood and risk taking. The two experiments show increased mood when future prospects are positive and decreased mood when future prospects are negative. I introduce a new computational model to account for the impact of future prospects on mood fluctuations. This task is being adapted for use in a new smartphone application ‘The Happiness Project’, which is optimised for collecting longitudinal data in clinical trials. I will also present results (N = 44,158) from the smartphone application ‘The Great Brain Experiment’ showing that the propensity to engage in loss-related risk taking increases throughout the day.
Rasmus Bruckner
Computational mechanisms of human state-action-reward contingency learning under perceptual uncertainty
IMPRS COMP2PSYCH Fellows' Presentation
To successfully interact with an everchanging world imbued with uncertainties, humans often have to learn probabilistic state-action-reward contingencies. Reinforcement learning algorithms have been able to provide a mechanistic picture of the neurocomputational principles that govern such learning and decision processes. However, standard reinforcement learning algorithms assume that the environmental state is fully observable. Humans, on the other hand, often have to learn the expected reward of choice options under considerable perceptual uncertainty. In this project, we investigate the computational principles that govern probabilistic state-action-reward learning under perceptual uncertainty. To this end, we designed an integrated perceptual and economic decision making learning task and acquired behavioural data from 52 human participants. To interpret the participants’ choice data, we developed a set of artificial agents, which describe a range of cognitive-computational strategies. These strategies range from Bayes-optimal exploitative decision making that takes perceptual uncertainty parametrically into account to fully random choice policies. Our behavioural modelling initiative favoured an agent model that suggests that humans integrate their subjective perceptual uncertainty when learning probabilistic state- action-reward contingencies. They tend, however, to underestimate the degree they should do so from a normative Bayes-optimal perspective.
Benjamin Chew
Endogenous fluctuations in the dopaminergic midbrain drive choice variability
IMPRS COMP2PSYCH Fellows' Presentation
Humans are often inconsistent in their choices, even when facing identical offers. Economic theory struggles to account for these inconsistencies. A potential physiological driver may be endogenous fluctuations in brain areas supporting economic decision making. Here, using real-time functional magnetic resonance imaging we asked whether ongoing fluctuations in activity that encompasses dopaminergic midbrain influence economic decisions. We show that endogenous activity is predictive of choice with subjects having a greater appetite for a risky choice when dopaminergic midbrain activity is low. Computational modelling reveals this effect involves an increased tendency to take risks irrespective of option value. We conclude that endogenous fluctuations in dopaminergic midbrain activity are a critical driver of choice variability, identifying a physiological basis for irrational choice.
Simon Ciranka
Navigating uncertainty across adolescence
IMPRS COMP2PSYCH Fellows' Presentation
Adopting efficient strategies to approach an environment of uncertainty is a developmental challenge. When unsure how to decide by means of observing others, we learn about the world while circumventing costly trial and error learning. We investigate developmental differences in how people use social information while dealing with different types of uncertainty. We focus on adolescence because adolescence is a life period associated with risky decision policies under uncertainty. However, laboratory assessments of adolescent decision making and social influence of the latter often focus on one aspect of uncertainty: risk. Real-life scenarios are usually more complex, thus laboratory results often are only weakly associated with real-life decision making. Real life’s uncertainty is usually Knightian. Hence, the decision maker does not know how likely an outcome is, and even if she reduced her uncertainty via learning, outcomes would still occur probabilistically. If feelings of uncertainty play a role in social influence, employing experimental paradigms that reflect uncertainty is important to identify correlates of social influence outside the laboratory. We compare social impact in pre-, mid- and post adolescent participants making decisions under risk and uncertainty in a gambling and a Bayesian learning paradigm, respectively. In both conditions, participants are repeatedly asked to decide between a safe or a risky option. Safe options provide a definite but small bonus to their final payment, risky options a larger but probabilistic one. After completing one solitary condition in each task, a social condition is introduced. Here decisions of another unknown subject–matched to solitary choices are shown to the participants before they make their own choice. During the first task participants will perform social and non-social versions of a gamble paradigm with known probabilities. In a second task, participants need to infer the outcome probabilities of their decision problems by observing differently sized samples from the outcome distribution and then base their choice upon their subjective estimate of outcome distributions. To comprehend the role of others in adolescent maladaptive behaviours such as substance abuse, it is important to understand the developmental idiosyncrasies of peer influence in the computations that determine their choices. To do this, we utilise hierarchical Bayesian models that are able to test competing hypothesis about social influence on adolescent risk taking.
Magda Dubois
Apathy as aberrant maturational separation of dopamine pathways?
IMPRS COMP2PSYCH Fellows' Presentation
Apathy is characterised by a lack of motivation and the inability to engage in goal-directed actions. Apathy is an important feature in several neurological and psychiatric disorders. Many apathy- related disorders, such as depression or schizophrenia, emerge during adolescence. Apathy is thought to be due to an imbalanced trade-off between reward and effort and recent studies have shown that when subjects learn about effort and reward associated with a stimulus, distinct regions of the dopamine network (dorsomedial prefrontal cortex and ventral striatum, respectively) are involved. In subjects with increased apathy, effort and reward prediction errors overlap in the dorsomedial prefrontal cortex to a greater degree which suggests that the mesolimbic and the mesocortical pathways might be less well differentiated. Importantly, maturation of mesocortical projections takes place only during adolescence when they extend from ventral striatum to prefrontal target areas. This suggests that a spatial separation of effort and reward prediction errors might not be present before adulthood and that apathy might arise if this mechanism goes awry during adolescence, leading to an imbalance between striatal and prefrontal dopamine. During my presentation, I will present how I will tackle these questions in my PhD.
Samuel Ereira
Learning the distinction between self and other
IMPRS COMP2PSYCH Fellows' Presentations
We recently found that agent-specific learning signals can be measured from neural activity. Furthermore, the degree of agent-specificity in these signals is a predictor of interindividual differences in self–other distinction abilities and subclinical traits. Our new project aims to explain where this variability in self–other distinction emerges from. Why are some signals more agent- specific than others? We used a training paradigm to test the hypothesis that the extent of self- other distinction is learned via inferred statistical dependencies between self-attributed signals and other-attributed signals. I will present behavioural data that shows evidence that self–other distinction can be learned in this way, as well as some preliminary fMRI data.
Alexandra Hopkins
Biases in learning and decision making in anxiety
IMPRS COMP2PSYCH Fellows' Presentations
Anxiety disorders are the most common mental health condition, with lifetime prevalence as high as 33.7%. Despite being stratified according to the focus of the anxiety, patients across the diagnostic spectrum tend to display prominent biases in learning and decision making, which are often held even in the face of strong counterevidence. My PhD project looks to investigate how these biases contribute to the aetiology and maintenance of anxiety.
I will firstly present work using a computational model that aimed to delineate the mechanisms underlying biased learning about the self in people who display high fear of negative evaluation, a symptom central to social anxiety disorder. Our findings indicate that higher anxiety is associated with enhanced learning about negative information specifically and that positive information processing is intact. Importantly, this negative bias only happens with self-referential information. This is important for informing therapeutical interventions.
In addition to these self-biases, which may act in a way that prevents appropriate social learning, anxious individuals display a tendency to incorrectly infer causal statistics in the environment, displaying an enhanced expectation for negative events to occur. An example of this is in a near- miss experience, where an anxiety reaction occurs despite the feared outcome never manifesting. I will present findings of a novel task, designed to elucidate how near-miss inference can interfere with appropriate learning about the environment. I will present computational modelling work that suggests this bias is specific to the threat domain and is influenced by biased estimations of uncertainty.
Julian Kosciessa
Cross-modal signatures of dynamic attentional allocation across the adult lifespan
IMPRS COMP2PSYCH Fellows' Presentation
Dynamic attentional allocation is fundamental for purposefully engaging with complex environments. While the neural mechanisms underlying selective attention have received much emphasis in the extant literature, how brain dynamics support flexible switching between attentional targets and how these dynamics differ by age has remained largely elusive. To address these questions, we conducted a parallel EEG-fMRI experiment probing dynamic neural regimes that support attentional switching between parametrically varying amounts of targets across younger and older adults. In particular, we consider the role of low-frequency rhythms as a potential mechanism for the dynamic coordination of macroscopic brain networks. We further investigate the extent to which thalamic BOLD dynamics jointly reflect such coordination.
Yunzhe Liu
Structural replay: Generalisation through factorised representation
IMPRS COMP2PSYCH Fellows' Presentation
Understanding the relationship among entities is crucial for complex behaviour. Abstracting such relational structure from past sensory experiences enables an agent to generalise to a new environment that shares a similar structure. In space, the structure of environment was thought to be maintained during rest by fast played-out sequences of past experience (i.e., 'replay'). However, it is unclear whether replay represents the underlying structure or simply the order of experience, especially in a general setting, unlike space. I am planning to tackle this question through both experimentally and theoretically.
In the experimental work, we trained human subjects on a non-spatial relational structure that was shared across different sets of objects, then presented them with a novel set of objects in an order different from the order implied by the structure. Across two studies we showed that human replay followed the underlying structure rather than the order of visual experience. Such structural replay transitioned to reverse direction following receipt of a reward, conforming to known characteristics of experienced replay in rodents. Furthermore, we demonstrated that a structural representation can be factorised from sensory experiences and guide novel objects to replay in structure-confined order. This abstract relational structure was encoded explicitly prior to the experience of the objects. Together, the two studies demonstrated the existence and function of structural replay.
In the theoretical work, I aim to understand what kinds of replay would be useful to make artificial deep neural networks learn more efficiently. Recent developments of a "hippocampal-entorhinal" deep net provided a promising direction. I will propose a tentative deep neural network architecture that employs reinforcement learning, hippocampal-entorhinal representation, and predictive coding, and has the potential capability of generalisation.
Matthew Nour
Dopaminergic signals for belief updates but not surprise in the midbrain and ventral striatum
IMPRS COMP2PSYCH Fellows' Presentation
Distinguishing between meaningful and meaningless sensory information is fundamental to forming accurate representations of the world. Dopamine is implicated in encoding belief updates about internal models of the environment, as well as processing the meaningful information content of observations. However, direct evidence for dopamine’s role in human belief updating is lacking. We addressed this question in volunteer healthy participants who performed a model-based fMRI task, and in whom we also acquired PET imaging measures of static and dynamic dopamine. We show that belief updates about task structure, but not sensory unexpectedness, are encoded in midbrain and ventral striatum activity. Using PET, we show this neural encoding is inversely related to dopamine-2/3 receptor availability in the midbrain and dexamphetamine-induced dopamine release capacity in the striatum. Trial-by-trial analysis of task performance indicated that subclinical paranoid ideation is inversely related to behavioural sensitivity to the meaningful content of observations. The findings provide the first direct evidence implicating dopamine in model-based belief updating in humans, and have implications for understating the pathophysiology of psychotic disorders where dopamine function is disrupted.
Liliana Polyanska
Fear and perceptual uncertainty
IMPRS COMP2PSYCH Fellows' Presentation
My project will investigate the relationship between fear and perceptual uncertainty. Their interplay and resulting influence on such constructs as decision making, fear generalisation, and response to an uncertain threat will be examined from a neurocognitive and computational perspective.
Max Rollwage
The influence of subjective confidence on post-decision evidence integration and changes of mind
IMPRS COMP2PSYCH Fellows' Presentation
Recent research suggests that confidence in a decision influences subsequent information processing and changes of mind. However, task performance and subjective confidence are tightly correlated. As a consequence, isolating and examining the unique impact of confidence on subsequent cognitive processes and behaviour is difficult. A new psychophysical manipulation (i.e., positive evidence) enabled us to experimentally manipulate subjective confidence while keeping performance. Through this dissociation we were able to investigate the specific influence of subjective confidence on changes of mind and post-decision information processing. In two behavioural experiments we showed that subjective confidence in an initial decision predicted later changes of mind. When people were highly confident (independent of actual performance) in their first decision, they were less likely to later change their minds. Drift-diffusion modelling revealed that this effect was driven by participants selectively accumulating evidence for their chosen option when being highly confident (i.e., confidence leading to a boost in confirmation bias). In a MEG study we used decoding techniques to further dissociate neural activity reflecting objective evidence in favour of a decision from neural signatures of subjective confidence, again confirming that changes of mind were specifically predicted by neural signatures of subjective confidence.
Yuki Shimura
The impact of preferences on emotional responses to altruism
IMPRS COMP2PSYCH Fellows' Presentation
Research in the field of psychology suggests that a possible motive for generosity is the increased happiness with which it is associated. However, it is unknown whether being altruistic increases happiness in everyone or only in people who are prosocial and happier when they take prosocial actions. I hypothesise that it is not being altruistic that increases happiness, but rather taking actions consistent with one’s values. My goal is to understand whether happiness from altruistic behaviour reflects values that are different from person to person, or if everyone is happier as a result of taking altruistic actions. A study by Rutledge et al. (2016) extracted individual dispositions in the social domain that impact momentary subjective wellbeing captured in a computational model as guilt and envy parameters. The study found that people with stronger guilt parameters were more generous in the dictator game than their counterparts with stronger envy parameters. Adapting the social decision-making task used in that study, we will use envy and guilt parameters as proxies for people’s values and test whether people who have a high envy perimeter will not be happier when forced to be altruistic (e.g., giving money to another person or to a charity) even when the actions do not cost them anything. We predict that people with a high guilt parameter will have greater happiness following altruistic actions, even if those actions are forced. If this prediction is accurate, we will test whether positive feedback from other people or charities at the midpoint of the experiment could shift values, leading to a change in behaviour and in happiness.
Lennart Wittkuhn
Towards tracking fast neural replay events in humans using fMRI
IMPRS COMP2PSYCH Fellows' Presentation
One of the most consequential findings in memory research of the past decades has been the discovery of fast sequential reactivation of memory representations within the hippocampus. This phenomenon, called replay, is thought to be a key mechanism for neural storing and computing relations between previously encountered locations and objects and has been linked to memory consolidation as well as planning and decision-making processes. Despite the potentially far- reaching role of this process in higher cognitive functions, the vast majority of studies investigating hippocampal replay have focused on rodents in spatial paradigms. This reliance on animals is mainly caused by the difficulty to investigate the sequential nature of fast-paced replay events (lasting on the order of 100 ms in humans) using temporally low resolved fMRI.
In my PhD project, I intend to overcome this obstacle and investigate hippocampal replay in humans using fMRI. Based on a recent fMRI study in which evidence for replay was found in sequential statistics of pattern activations in the human hippocampus during rest (Schuck & Niv, in prep), I aim to verify and extend the possibility to identify fast sequences of neural events. In my current study, I investigate the statistical properties of BOLD activation patterns following the display of fast sequences of visual objects to develop tools that allow the detection of fast-paced neural events with fMRI in general. In my talk, I will present preliminary results and discuss the implications of my findings for further analyses and subsequent research projects.