We provide a novel computational framework on how biological and artificial agents can learn to flexibly couple decouple neural task modules for cognitive processing. In this way, they address the stability-plasticity dilemma. For purpose, we combine two prominent neuroscience principles, namely Binding by Synchrony Reinforcement Learning. The model learns synchronize task-relevant modules, while also learning desynchronize currently task-irrelevant modules. As result, old (but...

Background: Student well-being is a multidimensional construct encompassing physical and mental health. However, its assessment in educational contexts remains fragmented, failing to comprehensively understand the diverse factors that shape students’ experiences. This gap particularly pronounced for cognitively strong students (CSS), whose unique intellectual, emotional, social challenges are frequently overlooked by traditional approaches.Objective: Current work introduces methodology -...

Humans are remarkably efficient at learning new tasks, in large part by relying on the integration of previously learned knowledge. However, research task typically focuses abstract rules minimalist stimuli, to study behavior independent history that humans come equipped with (i.e., semantic knowledge). In contrast, several theories suggest use knowledge and labels may help information. Here, we tested whether providing existing, semantically rich response allowed for more robust encoding...

Humans are remarkably efficient at learning new tasks, in large part by relying on the integration of previously learned knowledge. However, research task typically focuses abstract rules minimalist stimuli, to study behavior independent history that humans come equipped with (i.e., semantic knowledge). In contrast, several theories suggest use knowledge and labels may help information. Here, we tested whether providing existing, semantically rich response allowed for more robust encoding...

Testing enhances memory more than studying. Although numerous studies have demonstrated the robustness of this classic effect, its neural and computational origin remains debated. Predictive learning is a potential mechanism behind phenomenon: Because predictions prediction errors (mismatch between feedback) can only be generated in testing (and not studying), benefit from predictive learning. We shed light on effect multi-level analysis perspective via combination cognitive neuroscience...

In recent years, several hierarchical extensions of well-known learning algorithms have been proposed. For example, when stimulus-action mappings vary across time or context, the brain may learn two more in separate modules, and additionally (at a hierarchically higher level) to appropriately switch between those modules. However, how mechanistically coordinates neural communication implement such remains unknown. Therefore, current study tests computational model that proposed midfrontal...

The Rescorla-Wagner rule remains the most popular tool to describe human behavior in reinforcement learning tasks. Nevertheless, it cannot fit complex environments. Previous work proposed several hierarchical extensions of this rule. However, unclear when a flat (nonhierarchical) versus strategy is adaptive, or implemented by humans. To address question, current applies nested modeling approach evaluate multiple models environments both computationally (which performs best) and empirically...

ADVERTISEMENT RETURN TO ISSUEPREVArticleNEXTElectronic spectra of the d6 binary carbonyl complexes hexacarbonylmanganese(1+), hexacarbonylchromium and hexacarbonylvanadate(1-): an ab initio analysisK. Pierloot, J. Verhulst, P. Verbeke, L. G. VanquickenborneCite this: Inorg. Chem. 1989, 28, 15, 3059–3063Publication Date (Print):July 1, 1989Publication History Published online1 May 2002Published inissue 1 July...

Abstract Cognitive control is supported by theta band (4-7Hz) neural oscillations coordinating populations for task implementation. Task performance has been shown to depend on amplitude but a second critical aspect of oscillations, its peak frequency, mostly overlooked. Using modelling, behavioral and electrophysiological recordings, we show that adapt demands shifting towards the optimal frequency.

Theta and alpha frequency neural oscillations are important for learning cognitive control, but their exact role has remained obscure. In particular, it is unknown whether they operate at similar timescales, support different processes. We recorded EEG in 30 healthy human participants while performed a task containing both novel (block-unique) repeating stimuli. investigated behavior electrophysiology fast (i.e., within blocks) slow between timescales. Behaviorally, response time accuracy...

Abstract Cognitive control can be adaptive along several dimensions, including intensity (how intensely do signal influence bottom–up processing) and selectivity (what information is selected for further processing). Furthermore, exerted slow or fast time scales. Whereas on a scale used to proactively prepare upcoming challenges, also faster react unexpected events that require control. Importantly, systematic comparison of these dimensions scales remains lacking. Moreover, most current...

Why can’t we keep as many items want in working memory? It has long been debated whether this resource limitation is a bug (a downside of our fallible biological system) or instead feature (an optimal response to computational problem). We propose that the consequence useful feature. Specifically, flexible cognition requires time-based binding, and binding necessarily limits number (bound) memoranda can be stored simultaneously. Time-based most naturally instantiated via neural oscillations,...

Abstract The Rescorla-Wagner rule remains the most popular tool to describe human behavior in reinforcement learning tasks. Nevertheless, it cannot fit complex environments. Previous work proposed several hierarchical extensions of this rule. However, unclear when a flat (non-hierarchical) versus strategy is adaptive, or implemented by humans. To address question, current applies nested modelling approach evaluate multiple models environments both computationally (which performs best) and...

Abstract In recent years, several hierarchical extensions of well-known learning algorithms have been proposed. For example, when stimulus-action mappings vary across time or context, the brain may learn two more in separate modules, and additionally (at a hierarchically higher level) to appropriately switch between those modules. However, how mechanistically coordinates neural communication implement such learning, remains unknown. Therefore, current study tests computational model that...

Abstract Considerable evidence highlights the dorsolateral prefrontal cortex (DLPFC) as a key region for hierarchical (i.e. multilevel) learning. In previous electroencephalography (EEG) study, we found that low-level prediction errors were encoded by frontal theta oscillations (4–7 Hz), centered on right DLPFC (rDLPFC). However, causal relationship between and learning remains poorly understood. To investigate this question, in current participants received (6 Hz) sham high-definition...

A hallmark of human intelligence is the ability to flexibly adapt novel situations. This flexibility relies crucially on appropriate generalization previously acquired knowledge. One influential theoretical framework argues that humans organize their knowledge in a collection latent classes. Humans could then assign any situation one classes (or construct new if it too dissimilar), and thus generalize based older However, this not sufficiently flexible explain generalization. In particular,...

Humans are remarkably efficient at learning new tasks, in large part by relying on the integration of previously learned knowledge. However, research task typically focuses abstract rules minimalist stimuli, to study behavior independent history that humans come equipped with (i.e., semantic knowledge). In contrast, several theories suggest use knowledge and labels may help information. Here, we tested whether providing existing, semantically rich response allowed for more robust encoding...

Humans are remarkably efficient at learning new tasks, in large part by relying on the integration of previously learned knowledge. However, research task typically focuses abstract rules minimalist stimuli, to study behavior independent history that humans come equipped with (i.e., semantic knowledge). In contrast, several theories suggest use knowledge and labels may help information. Here, we tested whether providing existing, semantically rich response allowed for more robust encoding...

Cognitive control can be adaptive along several dimensions, including intensity (how intensely do signal influence bottom-up processing) and selectivity (what information is selected for further processing). Furthermore, exerted slow or fast time scales. While on a scale used to proactively prepare upcoming challenges, also faster react unexpected events that require control. Importantly, systematic comparison of these dimensions scales remains lacking. Moreover, most current models allow...