Self-Reflection Protects Behavior from Volatile Beliefs Linked to Paranoia.

Processing uncertainty may be pathognomonic (characteristic of a disease) for some psychiatric conditions. Some people expect the world to change, even when it doesn’t. This tendency is central to paranoia, where individuals often anticipate threat or change without clear evidence. But what determines whether these beliefs translate into behavior? One possibility is that metacognitive structure – the coherence and depth with which one articulates their own thinking – acts as a buffer. An agent may endorse a belief but have sufficient accessory hypotheses to insulate it from action. To test this, we used metacognitive prompting in GPT-4 to score individual reflections on open-ended questions (e.g., did you use any particular strategy?) after completing a probabilistic reversal learning task. Individuals with higher paranoia demonstrate lower metacognitive structure (t = 5.98, p < 0.001), with metacognition moderating the relationship between volatility belief and switching behavior (β = -0.05, p = 0.001) even after controlling for reflection verbosity and general cognitive ability. These findings suggest that metacognition protects against uncertainty-driven instability, pointing to a key mechanism by which reflection protects against cognition under change. This work provides a novel framework to measure metacognition from behavioral task debrief questions.

In a probability learning task, participants estimate the probabilistic reward contingencies, and this task has been used extensively to study instrumental conditioning with partial reinforcement. In the probabilistic reversal learning task, the probabilistic reward contingencies are reversed between options in the middle of the experiment to measure how well people adapt to new contingency situations. In this work, we used the attention-gated reinforcement learning (AGREL) model (Roelfsema & Van Ooyen, 2005) to simulate how people learn the probabilistic relationship between stimulus-reward pairs in probability and reversal learning tasks. AGREL algorithm put forward two important aspects of a learning phenomenon together in a neural network scheme: (1) the effect of unexpected outcomes on learning and (2) the effect of top-down (selective) attention on updating weights. Contrary to its importance in the learning literature, AGREL has not yet been tested with these well known learning tasks. The results of the first simulation showed that in a binary choice probability learning experiment an AGREL model can simulate different learning strategies, such as probability matching and maximizing. Secondly, we simulated a probabilistic reversal learning experiment with the same AGREL model, and the results showed that the AGREL model dynamically adapted to new contingency situations. Furthermore, we also evaluated effects of learning rate on the model's adaption to reversal contingency by plotting the inter-phase dynamics. These results showed that AGREL model simulates the traditional findings observed in probability and reversal learning experiments, and it can be further developed to understand the role of dopamine in learning and it can be used in model-based fMRI research.

Advancing research on cognitive flexibility in eating disorders: The importance of distinguishing attentional set‐shifting and reversal learning

The exploration/exploitation tradeoff – pursuing a known reward vs. sampling from lesser known options in the hope of finding a better payoff – is a fundamental aspect of learning and decision making. In humans, this has been studied using multi-armed bandit tasks. The same processes have also been studied using simplified probabilistic reversal learning (PRL) tasks with binary choices. Our investigations suggest that protocols previously used to explore PRL in mice may prove beyond their cognitive capacities, with animals performing at a no-better-than-chance level. We sought a novel probabilistic learning task to improve behavioral responding in mice, whilst allowing the investigation of the exploration/exploitation tradeoff in decision making. To achieve this, we developed a two-lever operant chamber task with levers corresponding to different probabilities (high/low) of receiving a saccharin reward, reversing the reward contingencies associated with levers once animals reached a threshold of 80% responding at the high rewarding lever. We found that, unlike in existing PRL tasks, mice are able to learn and behave near optimally with 80% high/20% low reward probabilities. Altering the reward contingencies towards equality showed that some mice displayed preference for the high rewarding lever with probabilities as close as 60% high/40% low. Additionally, we show that animal choice behavior can be effectively modelled using reinforcement learning (RL) models incorporating learning rates for positive and negative prediction error, a perseveration parameter, and a noise parameter. This new decision task, coupled with RL analyses, advances access to investigate the neuroscience of the exploration/exploitation tradeoff in decision making.

EPA-0643 – Effect of agomelatine on reward and punishment processing in a probabilistic reversal learning task in mice; role of its MT1/MT2 agonist and 5HT2C antagonist properties

Deficits in reward processing are a central feature of major depressive disorder with patients exhibiting decreased reward learning and altered feedback sensitivity in probabilistic reversal learning tasks. Methods to quantify probabilistic learning in both rodents and humans have been developed, providing translational paradigms for depression research. We have utilised a probabilistic reversal learning task to investigate potential differences between conventional and rapid-acting antidepressants on reward learning and feedback sensitivity. We trained 12 rats in a touchscreen probabilistic reversal learning task before investigating the effect of acute administration of citalopram, venlafaxine, reboxetine, ketamine or scopolamine. Data were also analysed using a Q-learning reinforcement learning model to understand the effects of antidepressant treatment on underlying reward processing parameters. Citalopram administration decreased trials taken to learn the first rule and increased win-stay probability. Reboxetine decreased win-stay behaviour while also decreasing the number of rule changes animals performed in a session. Venlafaxine had no effect. Ketamine and scopolamine both decreased win-stay probability, number of rule changes performed and motivation in the task. Insights from the reinforcement learning model suggested that reboxetine led animals to choose a less optimal strategy, while ketamine decreased the model-free learning rate. These results suggest that reward learning and feedback sensitivity are not differentially modulated by conventional and rapid-acting antidepressant treatment in the probabilistic reversal learning task.

Feasibility of use of probabilistic reversal learning and serial reaction time tasks in clinical trials of Parkinson's disease.

General cognitive and specific numerical abilities that underlay mathematical performance have been heavily investigated among hearing students; however, the inquiry has rarely been applied to students who are deaf or hard-of-hearing (d/Dhh). We examined whether general cognitive abilities (i.e. nonverbal IQ, processing speed, and spatial ability) and specific numerical abilities (i.e. symbolic and non-symbolic numerical magnitude processing) are related to mathematics achievement in 198 d/Dhh students in Grades 3 to 9. The results of our regression models indicated that, the three general cognitive abilities independently explained the variance in mathematics achievement when entered in one step; and spatial ability and processing speed had an independent contribution to mathematics achievement in the presence of the specific numerical abilities. The specific numerical abilities independently explained the variance in mathematics achievement when entered in one step; however, none of them had an independent contribution to mathematics achievement in the presence of the general cognitive abilities. These findings suggested that mathematics achievement in d/Dhh students depended more on general cognitive abilities, such as spatial ability and processing speed, than on specific numerical abilities.

Stage 2 sleep spindles have been previously viewed as useful markers for the development and integrity of the CNS and were more currently linked to 'offline re-processing' of implicit as well as explicit memory traces. Additionally, it had been discussed if spindles might be related to a more general learning or cognitive ability. In the present multicentre study we examined the relationship of automatically detected slow (< 13 Hz) and fast (> 13 Hz) stage 2 sleep spindles with: (i) the Raven's Advanced Progressive Matrices (testing 'general cognitive ability'); as well as (ii) the Wechsler Memory scale-revised (evaluating memory in various subdomains). Forty-eight healthy subjects slept three times (separated by 1 week) for a whole night in a sleep laboratory with complete polysomnographic montage. Whereas the first night only served adaptation and screening purposes, the two remaining nights were preceded either by an implicit mirror-tracing or an explicit word-pair association learning or (corresponding) control task. Robust relationships of slow and fast sleep spindles with both cognitive as well as memory abilities were found irrespectively of whether learning occurred before sleep. Based on the present findings we suggest that besides being involved in shaping neuronal networks after learning, sleep spindles do reflect important aspects of efficient cortical-subcortical connectivity, and are thereby linked to cognitive- and memory-related abilities alike.

We compared the relationship between general cognitive ability, social cognition and the ability to understand and make decisions about practical problems among groups of college age and older adults. Results indicated that both general cognitive functioning and social cognitive ability were related to practical problem solving ability among older adults. In contrast college age adults practical problem solving was related to only general cognitive ability. Results indicate that social cognitive ability may compensate for age related decline in general cognitive functions among older adults and allow for continued competence in practical problem solving as speed of processing and short term memory functions decline.

Parental socioeconomic status (SES) is a well-established predictor of children's neurocognitive development. Several theories propose that specific cognitive skills are particularly vulnerable. However, this can be challenging to test, because cognitive assessments are not pure measures of distinct neurocognitive processes, and scores across different tests are often highly correlated. Aside from one previous study by Tucker-Drob, little research has tested if associations between SES and cognition are explained by differences in general cognitive ability rather than specific cognitive skills. Using structural equation modelling (SEM), we tested if parental SES is associated with individual cognitive test scores after controlling for latent general cognitive ability. Data from three large-scale cohorts totalling over 16,360 participants from the UK and USA (ages 6-19) were used. Associations between SES and cognitive test scores are mainly (but not entirely) explained through general cognitive ability. Socioeconomic advantage was associated with particularly strong vocabulary performance, unexplained by general ability. When controlling for general cognitive ability, socioeconomic disadvantage was associated with better executive functions. Better characterizing relationships between cognition and adversity is a crucial first step toward designing interventions to narrow socioeconomic gaps. RESEARCH HIGHLIGHTS: Understanding environmental influences on cognitive development is a crucial goal for developmental science-parental socioeconomic status (SES) is one of the strongest predictors. Several theories have proposed that specific cognitive skills, such as language or certain executive functions, are particularly susceptible to socioeconomic adversity. Using structural equation modelling, we tested whether SES predicts specific cognitive and academic tests after controlling for latent general cognitive ability across three large-scale cohorts. SES moderately predicted latent general cognitive ability, but associations with specific cognitive skills were mainly small, with a few exceptions.

This study used functional magnetic resonance imaging to examine the effects of acute tryptophan (TRP) depletion (ATD), a well-recognized method for inducing transient cerebral serotonin depletion, on brain activity during probabilistic reversal learning. Twelve healthy male volunteers received a TRP-depleting drink or a balanced amino-acid drink (placebo) in a double-blind crossover design. At 5 h after drink ingestion, subjects were scanned while performing a probabilistic reversal learning task and while viewing a flashing checkerboard. The probabilistic reversal learning task enabled the separate examination of the effects of ATD on behavioral reversal following negative feedback and negative feedback per se that was not followed by behavioral adaptation. Consistent with previous findings, behavioral reversal was accompanied by significant signal change in the right ventrolateral prefrontal cortex (PFC) and the dorsomedial prefrontal cortex. ATD enhanced reversal-related signal change in the dorsomedial PFC, but did not modulate the ventrolateral PFC response. The ATD-induced signal change in the dorsomedial PFC during behavioral reversal learning extended to trials where subjects received negative feedback but did not change their behavior. These data suggest that ATD affects reversal learning and the processing of aversive signals by modulation of the dorsomedial PFC.

Behavioral outcomes are rarely certain, requiring subjects to discriminate between available choices by using feedback to guide future decisions. Probabilistic reversal learning (PRL) tasks test subjects' ability to learn and flexibly adapt to changes in reward contingencies. Cortico-striatal circuitry has been broadly implicated in flexible decision-making-though what role these circuits play remains complicated. In this study we leveraged the fast temporal dynamics of local field potentials to precisely identify the role that cortico-striatal networks play during PRL reward feedback. We measured widespread (32-CH) local field potential activity of male Long-Evans rats during a PRL task wherein a target response delivered reward on 80% of trials while a non-target response delivered reward on 20% of trials. When subjects learned those reward probabilities, contingencies were reversed. We found that reward-evoked oscillations at beta (15-30 Hz) and high gamma (>70 Hz) frequencies marked positive reward valence and reflected probability of reward. Activity and connectivity at beta frequencies between orbitofrontal cortex, anterior insula, medial prefrontal cortex, and ventral striatum during expected rewards were correlated with behavioral performance and specific aspects of value/exploitative behavior as defined by a reinforcement learning computational model. Finally, we found that modulating beta activity in orbitofrontal cortex with optogenetic (20 Hz) stimulation promoted maladaptive behavior when stimulation was provided during non-target responses, consistent with our data and computational model predictions. Reward-evoked beta oscillations may reflect a crucial component underlying reward learning, and erroneous elevations in this physiological signal may contribute to maladaptive task performance and behavioral disruptions.

Is there G beyond g? (Is there genetic influence on specific cognitive abilities independent of genetic influence on general cognitive ability?)

In the first twin study of the old-old, individuals 80 years old and older, we examined the relationship between general and specific cognitive abilities from a genetic perspective. That is, we examined the extent to which genetic and environmental factors influence major group factors of cognitive abilities, independent of general cognitive ability. As part of the OctoTwin project in Sweden, general and specific cognitive abilities were assessed in 52 monozygotic and 65 same-sex dizygotic twin pairs 80 years old and older using a battery of seven tests that assess verbal, spatial, speed-of-processing, and memory performance. Results suggest that genetic effects associated with general cognitive ability (g) account for the correlation between g and verbal, spatial, and speed-of-processing abilities. No genetic influences were found for these specific cognitive abilities separate from g. In contrast, memory ability appears to be more distinct genetically from g than are other cognitive abilities. Comparison with younger samples suggests that cognitive abilities relating to speed of processing may be genetically dedifferentiated in the old-old.