Biology subjects

Pulcu, E.

Publications and source records attributed to Pulcu, E..

10 recordsLinked to original sources

bioRxiv2026

Valence-specific representation of uncertainty in the anterior cingulate is impaired in those with affective symptoms

The ability to seek reward and avoid punishment is a fundamental survival instinct. In natural environments, however, the statistics of rewards and punishments can change independently of one another. In addition, individuals experiencing anxiety and depression may be selectively biased to process rewards and punishments differently. Here, we examine how humans adapt their behavior in such dynamic environments, using a task where cues are associated with independently changing probabilities of rewards (wins) and punishments (losses). We demonstrate that participants dynamically adjust their learning rates based on the relative volatility of each valence. Neuroimaging reveals that this behavioral flexibility is supported by valence-specific segregation of volatility signals within distinct subregions of the anterior cingulate cortex (ACC: perigenual and dorsal). Furthermore, individuals with higher levels of anxiety and depression exhibit a relative reduction in loss-volatility learning adaptation, accompanied by less distinct neural encoding of win and loss volatility in both ACC subregions. Together, our findings indicate that flexible adaptation to separate reward and punishment contingencies relies on distinct, valence-specific tracking of volatility within the prefrontal cortex. These findings suggest a potential computational and neuroanatomical framework for understanding maladaptive learning in affective disorders.

neuroscience↗

bioRxiv2026

Ketamine attenuates habenula activity in response to aversive outcomes during Pavlovian learning

Ketamine is an NMDA receptor antagonist with rapid-antidepressant properties when administered at a sub-anesthetic dose. Preclinical models indicate that a direct injection of ketamine into lateral habenula (Hb), a small midbrain structure with an evolutionarily preserved role in aversive learning across mammals, can rapidly relieve depression-like behavior. However, there is limited evidence to explain how ketamine acts on the function of the human habenula. In a translational computational neuroscience study, 70 healthy adult volunteers were randomised in a 1:1 ratio to receive ketamine or placebo (NaCl 0.9%). We used an aversive Pavlovian conditioning paradigm combined with 7-Tesla functional neuroimaging to show that ketamine attenuates habenula response during aversive stimuli expectations and outcomes 24 hours post-infusion. We further present preliminary evidence suggesting that when aversive learning occurs after ketamine infusion, reduced habenula activity during the learning process may lead to downstream effects that diminish the aversive impact of negative affective memories. These findings provide translational support for preclinical models of ketamines mechanisms in humans.

neuroscience↗

bioRxiv2023

Humans adapt rationally to approximate estimates of uncertainty

Efficient learning requires estimation of, and adaptation to, different forms of uncertainty. If uncertainty is caused by randomness in outcomes (noise), observed events should have less influence on beliefs, whereas if uncertainty is caused by a change in the process being estimated (volatility) the influence of events should increase. Previous work has demonstrated that humans respond appropriately to changes in volatility, but there is less evidence of a rational response to noise. Here, we test adaptation to variable levels of volatility and noise in human participants, using choice behaviour and pupillometry as a measure of the central arousal system. We find that participants adapt as expected to changes in volatility, but not to changes in noise. Using a Bayesian observer model, we demonstrate that participants are, in fact, adapting to estimated noise, but that their estimates are imprecise, leading them to misattribute it as volatility and thus to respond inappropriately.

neuroscience↗

bioRxiv2023

Individualistic attitudes in iterative Prisoners Dilemma undermine evolutionary fitness and may drive cooperative human players to extinction

Humans perform inarguably the richest plethora of prosocial behaviors in the animal kingdom. Prosocial behaviors such as interpersonal cooperation and altruistic punishment are important for understanding how humans navigate their social environment. The success and failure of strategies human players device have implications for determining their long-term socio-economic/evolutionary fitness. Following the footsteps of Press and Dysons (2012), I implemented their evolutionary game theoretic models developed for the Iterated Prisoners Dilemma (a behavioral economic probe of interpersonal cooperation) and analysed human choice behavior in the Ultimatum Game (a behavioral economic probe of altruistic punishment) involving 50 human participants versus computerized opponents with prosocial or individualistic social value orientation. Although the results indicate that it is more likely to break cycles of mutual defection in ecosystems in which humans interact with individualistic opponents, analysis of social-economic fitness at the Markov stationary states suggested that this comes at an evolutionary cost. Overall, human players acted in a significantly more cooperative manner than their opponents, but they failed to overcome extortion from individualistic agents, risking "extinction" in 70% of the cases. These findings demonstrate human players might be short-sighted and social interactive decision strategies they device while adjusting to different types of opponents may not be optimal in the long run.

evolutionary biology↗

bioRxiv2021

Value-based decision-making between affective memories

Affective biases can influence how past events are recalled from memory. However, the mechanisms underlying how discrete affective events shape memory formation and subsequent recall are not well understood. Further understanding this is important given the central role of negative biases in affective memory recall in depression and antidepressant drug action. In order to capture cognitive processes associated with affective memory formation and recall, we studied value-based decision-making between affective memories in two within-subject experiments (n=45 and n=74). Our findings suggest that discrete affective events, created by large magnitude Wheel of Fortune (WoF) outcomes, influence affective memory formation processes during reinforcement-learning (RL). After 24 hours, we show that healthy volunteers display stable preferences during value-based recall of affective memories in a binary decision-making task. Computational modelling of these preferences demonstrated a positive bias during value-based recall, induced by previously winning in the WoF. We further showed that value-based decision-making between affective memories engages the pupil-linked central arousal systems, leading to pupil constriction prior to, and differential pupil dilation after the decision onset depending on the valence of the chosen options. Taken together, we demonstrate that mechanisms underlying human affective memory systems can be described by RL and probability weighting models. This approach could be used as a translational assay to study the effects of novel antidepressants.

neuroscience↗

bioRxiv2019

Pupillary correlates of computations underlying human responder behaviour in the Ultimatum Game

Negotiating with others about how finite resources should be distributed is an important aspect of human social life. However, little is known about mechanisms underlying human social-interactive decision-making. Here, we report results from a novel iterative Ultimatum Game (UG) task, in which the proposers facial emotions and offer amounts were sampled probabilistically based on the participants decisions, creating a gradually evolving social-interactive decision-making environment. Our model-free results confirm the prediction that both the proposers facial emotions and the offer amount influence human choice behaviour. These main effects demonstrate that biases in facial emotion recognition also contribute to violations of the Rational Actor model (i.e. all offers should be accepted). Model-based analyses extend these findings, indicating that participants decisions are guided by an aversion to inequality in the UG. We highlight that the proposers facial responses to participant decisions dynamically modulate how human decision-makers perceive self-other inequality, relaxing its otherwise negative influence on decision values. In iterative games, this cognitive model underlies how offers initially rejected can gradually become more acceptable under increasing affective load, and accurately predicts 86% of participant decisions. Activity of the central arousal systems, assessed by measuring pupil size, encode a key element of this model: proposers affective reactions in response to participant decisions. Taken together, our results demonstrate that, under affective load, participants aversion to inequality is a malleable cognitive process which is modulated by the activity of the pupil-linked central arousal systems.

neuroscience↗

bioRxiv2019

A nonlinear relationship between prediction errors and learning rates in human reinforcement learning

Reinforcement-learning (RL) models have been pivotal to our understanding of how agents perform learning-based adaptions in dynamically changing environments. However, the exact nature of the relationship (e.g. linear, logarithmic etc.) between key components of RL models such as prediction errors (PEs; the difference between the agents expectation and the actual outcome) and learning rates (LRs; a coefficient used by agents to update their beliefs about the environment) has not been studied in detail. Here, across (i) simulations, (ii) reanalyses of readily available datasets and (iii) a novel experiment, we demonstrate that the relationship between PEs and LRs is (i) nonlinear over the PE/LR space, and (ii) it can be better accounted for by an exponential-logarithmic function that can transform the magnitude of PEs instantaneously to LRs. In line with the temporal predictions of this model, we show that physiological correlates of LRs accumulate while learners observe the outcome of their choices and update their beliefs about the environment.

neuroscience↗

bioRxiv2018

The effects of the angiotensin II receptor antagonist losartan on appetitive versus aversive learning

Exposure therapy is a first-line treatment for anxiety disorders but remains ineffective in a large proportion of patients. A proposed mechanism of exposure involves a form of inhibitory learning where the association between a stimulus and an aversive outcome is suppressed by a new association with an appetitive or neutral outcome. The blood pressure medication losartan augments fear extinction in rodents and might have similar synergistic effects on human exposure therapy, but the exact cognitive mechanisms underlying these effects remain unknown. In this study, we used a reinforcement learning paradigm with compound rewards and punishments to test the prediction that losartan augments learning from appetitive relative to aversive outcomes. Healthy volunteers (N=53) were randomly assigned to single-dose losartan (50mg) versus placebo. Participants then performed a reinforcement learning task which simultaneously probes appetitive and aversive learning. Participant choice behaviour was analysed using both a standard reinforcement learning model and by analysis of choice switching behaviour. Losartan significantly reduced learning rates from aversive events (losses) when participants were first exposed to the novel task environment, while preserving learning from positive outcomes. The same effect was seen in choice switching behaviour. Losartan enhances learning from positive relative to negative events. This effect may represent a computationally defined neurocognitive mechanism by which the drug could enhance the effect of exposure in clinical populations.

neuroscience↗

bioRxiv2017

Characterising and Engaging a Computationally Defined Treatment Target for Depression

Affective bias, the tendency to prioritise the processing of negative relative to positive events, is causally linked to clinical depression. However, why such biases develop or how they may best be ameliorated is not known. Using a computational framework, we investigated whether affective biases may reflect an individuals estimates of the information content of negative and positive events. During a reinforcement learning task, the information content of positive and negative outcomes was manipulated independently by varying the volatility of their occurrence. Human participants altered the learning rates used for the outcomes selectively, preferentially learning from the most informative. This behaviour was associated with activity of the central norepinephrine system, estimated using pupilometry, for loss outcomes. Humans maintain independent estimates of the information content of positive and negative outcomes which bias their processing of affective events. Normalising affective biases using computationally inspired interventions may represent a novel treatment approach for depression.

neuroscience↗

bioRxiv2017

Computations underlying peoples risk-preferences in social interactions

Interacting with others to decide how finite resources should be allocated between parties which may have competing interests is an important part of social life. Considering that not all of our proposals to others are always accepted, the outcomes of such social interactions are, by their nature, probabilistic and risky. Here, we highlight cognitive processes related to value computations in human social interactions, based on mathematical modelling of the proposer behavior in the Ultimatum Game. Our results suggest that the perception of risk is an overarching process across non-social and social decision-making, whereas nonlinear weighting of others acceptance probabilities is unique to social interactions in which others valuation processes needs to be inferred. Despite the complexity of social decision-making, human participants make near-optimal decisions by dynamically adjusting their decision parameters to the changing social value orientation of their opponents through influence by multidimensional inferences they make about those opponents (e.g. how prosocial they think their opponent is relative to themselves).

neuroscience↗