No description available.

La dopamine du cortex préfrontal (PFC) est impliquée dans l’apprentissage et dans la prise de décision liée à l’effort. Comme l’apprentissage ne peut se faire sans effort, il n’est pas clair aujourd’hui si la dopamine est nécessaire pour l’apprentissage, ou pour l’engagement de l’effort pour apprendre. Dans ce travail, les rats apprenaient à pousser un levier pour obtenir de la nourriture, soit avec (apprentissage par observation, LeO) ou sans (essai-et-erreur, TE) observation préalable d’un congénère exécutant la tâche. TE et la phase d’exécution de LeO nécessitent l’effort physique (overt learning), l’observation dans LeO ne requiert pas d’effort physique (covert learning). Avant chaque session, les rats recevaient des injections de SCH23390 ou de la saline dans le cingulaire antérieur (ACC) ou l’orbitofrontal (OFC). Si la dopamine est nécessaire à l’apprentissage, le blocage des récepteurs D1 affecterait aussi bien l’apprentissage overt que covert. Si la dopamine n’est pas requise pour l’apprentissage mais pour l’engagement de l’effort, le blocage affecterait l’apprentissage overt, et non covert. Les résultats montrent que le blocage de la dopamine dans ACC ou OFC supprime l’apprentissage overt, laissant intact l’apprentissage covert. Une fois les injections arrêtées, les rats récupèrent la capacité d’apprendre, mais dans le cas de ACC, pas la tolérance à l’effort. Ces résultats suggèrent que la dopamine dans ACC et OFC n’est pas nécessaire pour l’apprentissage, et que les déficits d’apprentissage pourraient reflèter une réduction de la tolérance effort à l’effort liée au blocage de la dopamine. / Because prefrontal cortex (PFC) dopamine plays a pivotal role in associative learning and in effort-related decision making, it is not clear as of today whether PFC dopamine activity is required for learning per se, or rather for engaging the effort necessary to learn. In this work, we used observational learning (LeO) and trial-and-error (TE) learning to dissociate learning from physical effort. Both TE and the execution phase of LeO require physical effort (overt learning). Observation does not require physical effort (covert learning). Rats learned to push a lever for food rewards either with or without prior observation of an expert conspecific performing the same task. Before daily testing sessions, the rats received bilateral ACC or OFC microinfusions of SCH23390, or saline-control infusions. If dopamine activity is required for task acquisition, its blockade should impair both overt and covert learning. If dopamine is not required for task acquisition, but solely for regulating effort tolerance, blockade should impair overt learning but spare covert learning. We found that dopamine blockade in ACC or OFC suppressed overt learning selectively, leaving covert learning intact. In subsequent testing sessions without dopamine blockade, rats recovered their overt-learning capacity but, in ACC experiments, the animals did not recover their normal level of effort tolerance. These results suggest that ACC and OFC dopamine is not required for the acquisition of conditioned behaviours and that apparent learning impairments could instead reflect a reduced level of effort tolerance due to cortical dopamine blockade.

<div>While cognitive control enables the selection of goal-relevant responses, metacontrol enables the selection of context-appropriate control operations. In task switching, metacontrol modulates task-switching efficiency by retrieving the associations between a contextual cue and a particular cognitive control demand. While the automatic retrieval of cognitive control is appealing due to its time and energy efficiency, the effects of different contextual cues have been shown in separate studies and appear to have different characteristics. Here, we devised a single task-switching paradigm to test whether we can observe both list-wide and item-specific metacontrol within subjects. In two experiments, we demonstrated reduced switch costs in lists associated with a high probability of switching as compared with lists with a low probability of switching (i.e., a list-wide switch probability [LWSP] effect). Similarly, we observed an analogous item-specific switch probability (ISSP) effect such that items associated with a high probability of switching incurred smaller switch costs as compared with items associated with a low probability of switching. We also confirmed that both list-wide and item-specific switch probability effects were not dependent on lower-level stimulus-response associations. However, the LWSP and the ISSP effects were uncorrelated, suggesting a lack of dependence. Together, these findings suggest that there are two distinct modes of metacontrol that are deployed in a context-sensitive manner in order to adapt to specific cognitive demands.</div>

Cognition allows animals to acquire, process, and store sensory information from the environment and use it to adapt to their surroundings. A battery of behavioral tests was used to assess the cognitive abilities of black-handed spider monkeys (Ateles geoffroyi). Black and white cups were used to assess (1) object permanence by showing the animals under which cup the reward was placed, (2) associative learning by concealing where the reward was placed, and (3) long-term memory by repeating the second task after a 4-month break; petri dishes with varying amounts of food were used to assess (4) relative quantity discrimination; and boxes fitted with dotted cards were used to assess discrete number discrimination with (5) equallysized dots and (6) various-sized dots. For each task, one session comprised 10 trials (i.e. responses). All nine animals succeeded in all tests and, as a group, reached the learning criterion of 70% correct responses on session two in the object permanence and associative learning tasks; on session eleven in the quantity discrimination task; on session sixteen in the numerosity task with equally-sized dots; on session three in the numerosity task with various-sized dots; and averaged 84.4% correct responses in the long-term memory task. Their prompt high score in the numerosity task with various-sized dots suggests that the animals acknowledged the task for its numerical properties as opposed to the size or pattern of the dots. These cognitive abilities are thought to shape the necessary behaviors for the ecological and social needs of the species.

<div><div><div><p>The brain dynamically alters its production of flexible behavior: cognitive flexibility increases when demand is high. In task switching experiments, past exposure to a high demand for flexibility in conjunction with specific temporal contexts leads to learned switch readiness such that future exposures to those contexts will cue flexibility. According to a recent proposal (Dreisbach & Fröber, 2019), learned switch readiness following switch demands is supported by a concurrent activation (CA) cognitive mechanism whereby both sets of task rules are kept available in working memory despite only using one at a time. This can be differentiated from a competing candidate mechanism, working memory updating (WMU) thresholds which determine the ease of replacing one task’s rules with another. The WMU mechanism is expected to cause a global increase in flexibility while CA is conceptualized as limited to task-specific associations. To test whether learned switch readiness represents a global or limited change in the cognitive system, I conducted two experiments that both involved learning switch readiness in one context and generalizing it in another. In Experiment 1, I replicated and extended findings that switch probability manipulations can modulate voluntary switch rates (VSR), indicating one type of generalizability. However, in Experiment 2, I found that flexibility learned through switch probability manipulations did not transfer to new tasks when the task rules were changed but contextual cues remained the same, demonstrating a limit: learned switch readiness does not generalize across tasks. These findings together suggest that CA is likely the mechanism behind learned switch readiness.</p></div></div></div>

No description available.

No description available.

No description available.

L’apprentissage des relations entre événements dans notre environnement nous permetd’anticiper des futures cibles et guide notre comportement. Une partie de cet apprentissage alieu sans intention, i.e. implicitement. Notre capacité limitée de traitement, qui contraste avec larichesse de notre environnement, impose la sélection d’une partie des informations sensorielles.Quels stimuli sont donc sélectionnés quand on apprend des associations ? Dans quelle mesurel’apprentissage sollicite des ressources attentionnelles ? Cette thèse porte sur les interactionsentre l’attention et l’apprentissage associatif.Dans la première partie expérimentale, nous avons étudié si la capture attentionnellependant l’apprentissage associatif est modulée par la prédictibilité de la cible. Nous avonsconçu 2 études EEG dans lesquels nous avons manipulé la valeur de la contingence entre indiceet cible. Nous avons trouvé deux biais attentionnels différents. Dans la première expérience lescibles inattendues ont montré une priorité attentionnelle, tandis que dans la deuxième ce sontles cibles prédictibles qui ont été privilégiées, y compris quand l’apprentissage est implicite.Ceux deux biais attentionnels, qui ont déjà été décrits en référence aux cibles dans des modèlesattentionnels de l’apprentissage, pourraient être au service de buts comportementaux différents.Dans la deuxième partie, nous avons étudié si les ressources attentionnelles disponiblesaffectent la capacité à discriminer des associations entre un indice et une cible. Pour cela nousavons mesuré la sensibilité aux associations sous différentes contraintes attentionnelles, à l’aided’un paradigme de double tâche. Nos données montrent que la discrimination est diminué parune tâche de suppression articulatoire concurrente et abolie par une tâche de charge cognitiveélevée. Bien qu’il ait été suggéré que l’apprentissage associatif puisse être automatique, nosdonnées montrent qu’il sollicite des ressources attentionnelles considérables. / Learning relations between events in our environment allows us to anticipate futureoutcomes and guides our decisions. Part of this learning occurs without intention, implicitly.Given the enormous amount of information available, which contrasts with our limitedprocessing capacity, the selection of certain stimuli becomes crucial. So which stimuli do weselect when we learn associations? How do the available attentional resources modulatelearning? This thesis focuses on the intertwining between associative learning and attention.In the first experimental part, we investigated whether the deployment of attentionduring associative learning is modulated by expectations. In particular, we conducted twoEEG studies in which we manipulated the contingent relation between a cue and an outcome.We found two different attentional biases. In the first experiment, unexpected outcomescaptured attention preferentially whereas predictable outcomes were prioritized in the second,and importantly even when the learning of the associations was implicit. We argue that theseattentional biases, which have already been described in attentional models of associativelearning, likely serve different goals.In the second experimental part, we examined to what extent associative learningrequires attention. With that aim, we measured sensitivity to contingency in three studiesunder different attentional constraints. Our data show that the ability to assess associations isdiminished by an articulatory suppression secondary task and is abolished by a highlydemanding task. While it has been suggested that associative learning might be an automaticprocess, our findings demonstrate that attention is critical to contingency assessment.

Bei Untersuchungen zur klassischen Konditionierung mit mehreren prädiktiven Hinweisreizen konnte in den letzten Jahren gezeigt werden, daß Personen mit schizophrenen Spektrumsstörungen Besonderheiten beim Erlernen von Ereignisrelationen aufweisen. Bei den in dieser Arbeit vorgestellten Experimenten wird assoziatives Lernen bei Gesunden und bei Personen mit schizophrenen Spektrumsstörungen einer elektrophysiologischen Betrachtung zugänglich gemacht. Zu diesem Zweck wurde ein modifiziertes Reizinteraktionsparadigma verwendet, bei dem mit Hilfe zweier Hinweisreize ein Zielreiz vorhergesagt werden soll. Dabei wurde das prädiktive Potential eines Hinweisreizes konstant gehalten. Die Änderung der prädiktiven Validität des zweiten Hinweisreizes führte zu zwei Versuchsbedingungen: In einer war dieser Hinweisreiz ein guter Prädiktor für den Targetreiz, in der anderen ein schlechter Prädiktor. Die Kontingenzurteile der unselektierten und der niedrigschizotypischen Probanden sind durch eine kompetitive Beziehung der Hinweisreize charakterisiert. Die Kontingenzen eines Hinweisreizes werden unterschätzt, wenn ein zweiter besser zur Vorhersage des Targets geeignet ist. Die Kontingenzurteile der hochschizotypischen Probanden weisen dieses Muster nicht auf. Wird hier ein Hinweisreiz als ein valider Prädiktor für den Zielreiz erkannt, fällt auch die Beurteilung für die Kontingenzbeziehung zwischen dem zweiten Reiz und dem Zielreiz höher aus. Die elektrophysiologischen Daten unterstützen die Auffassung, daß nach jedem Lerndurchgang eine Fehlerkorrektur stattfindet. Normale und Schizotypiker unterscheiden sich diesbezüglich nicht. Gruppenunterschiede gibt es allerdings bei der durch die Hinweisreize ausgelösten N150, deren Amplitude bei den Schizotypikern an temporo-okzipitalen Ableitorten kleiner ist. Der Reizkompetitionseffekt konnte nur in der Versuchsbedingung induziert werden, die eine hohe Arbeitsgedächtnisauslastung erforderte. Die Bedeutung dieser Ergebnisse wird vor dem Hintergrund des aktuellen Forschungsstands diskutiert. / Several studies of classical conditioning with more than one predictive stimulus report that persons with schizophrenia spectrum disorders have deviances in learning how related to each other. I used a modified cue interaction paradigm to make associative learning accessible to an electrophysiological investigation. Participants had to predict a target stimulus based on two preceding cues, keeping the predictive potential of one cue constant. The predicitve validity of the second cue changed, yielding two different validity conditions: in one condition the second cue was a very good predictor of the target cue, in the other validity condition this cue predicted the target at random level. Judgements of contigencies of unselected probands and psychometric selected low schizotypals between target and each cue can be described as "competitive": Contingencies between one cue and target were judged lower if a second cue was a better predictor for the target, whereas contigency judgements of psychometric selected schizotypals can be described as "cooperative": If one cue is judged as being a relative valid predictor, the contingency between the second cue and target were also judged higher. Electrophysiogical dates supports the view of correcting errors after each learning trial. There seems to be no difference between high and low schizotypes. Differences on group-level appeared on cue-evoked event-related potentials, especially on the occipito-temporal N150-amplitudes, which were smaller in schizotypes. Cue-Competition could only induced in the low working memory load version of the cue competition paradigm. The relevance of these results are discussed with reference to recent research on associative learning and schizophrenia.