A5087842270- Neural and Behavioral Psychology Studies
- Neural dynamics and brain function
- Motor Control and Adaptation
- Functional Brain Connectivity Studies
- Action Observation and Synchronization
- Explainable Artificial Intelligence (XAI)
- Topic Modeling
- Animal Vocal Communication and Behavior
- Memory Processes and Influences
- Language and cultural evolution
- EEG and Brain-Computer Interfaces
- Neuroscience and Music Perception
- Memory and Neural Mechanisms
- Meta-analysis and systematic reviews
- Mental Health Research Topics
- Human Pose and Action Recognition
- Control and Dynamics of Mobile Robots
- Innovative Teaching and Learning Methods
- Cognitive Functions and Memory
- Multimodal Machine Learning Applications
- Child and Animal Learning Development
- Decision-Making and Behavioral Economics
- Technology and Human Factors in Education and Health
- Advanced Data Processing Techniques
- Cognitive Abilities and Testing
Brown University
2021-2022
Ghent University
2018-2022
Allen Institute for Brain Science
2022
Politecnica Salesiana University
2021
Ghent University Hospital
2019
Université Libre de Bruxelles
2011-2018
Neuroscience Institute
2018
Neurosciences Institute
2017
Vrije Universiteit Brussel
2017
University of Amsterdam
2017
Summary Data analysis workflows in many scientific domains have become increasingly complex and flexible. To assess the impact of this flexibility on functional magnetic resonance imaging (fMRI) results, same dataset was independently analyzed by 70 teams, testing nine ex-ante hypotheses. The analytic approaches is exemplified fact that no two teams chose identical to analyze data. This resulted sizeable variation hypothesis test even for whose statistical maps were highly correlated at...
Investment of cognitive effort is required in everyday life and has received ample attention recent neurocognitive frameworks. The neural mechanism investment thought to be structured hierarchically, with dorsal anterior cingulate cortex (dACC) at the highest level, recruiting task-specific upstream areas. In current fMRI study, we tested whether dACC generally active when demand high across tasks different stimuli, connectivity between areas increased depending on task requirements level...
A prominent learning phenomenon is the testing effect, meaning that enhances retention more than studying. Emergent frameworks propose fundamental (Hebbian and predictive) principles as its basis. Predictive posits occurs based on contrast (error) between a prediction feedback (prediction error). Here, we in (but not studying) scenarios, participants predict potential answers, with subsequent yields error, which facilitates testing-based learning. To investigate this, developed an...
People adjust their use of feedback over time through a process referred to as adaptive learning. We have recently proposed that the underlying mechanisms learning are rooted in how brain organizes into similarly credited units, which we refer latent states. Here develop BG-thalamo-cortical circuit model this and show it captures both commonalities heterogeneity human behavior. Our learns incrementally synaptic plasticity PFC-BG connections, but upon observing discordant information,...
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...
Recent behavioral evidence implicates reward prediction errors (RPEs) as a key factor in the acquisition of episodic memory. Yet, important neural predictions related to role RPEs memory remain be tested. Humans (both sexes) performed novel variable-choice task where we experimentally manipulated and found support for with fMRI. Our results show that line previous observations, accuracy increases magnitude signed (i.e., better/worse-than-expected) (SRPEs). Neurally, observe SRPEs are encoded...
Adaptive sequential behavior is a hallmark of human cognition. In particular, humans can learn to produce precise spatiotemporal sequences given certain context. For instance, musicians not only reproduce learned action in context-dependent manner, they also quickly and flexibly reapply them any desired tempo or rhythm without overwriting previous learning. Existing neural network models fail account for these properties. We argue that this limitation emerges from the fact sequence...
For selecting an action, traditional theories suggest a cognitive architecture made of serial processing units.Others suggested that action selection emerges from the parallel implementation and competition between multiple plans.To disentangle these 2 hypotheses, we created reaching task assessing temporal dynamics selection.Crucially, our design did not force processes to operate in parallel, allowing informative comparison hypotheses.We manipulated probability congruence cue delayed reach...
Previous research attempted to explain how humans strategically adapt behavior in order achieve successful task performance.Recently, it has been suggested that 1 potential strategy is avoid tasks are too demanding.Here, we report 3 experiments investigate the empirically neglected role of metacognitive awareness this process.In these experiments, participants could freely choose between performing a either high-demand or low-demand context.Using subliminal priming, ensured were not aware...
Converging evidence has led to a consensus in favor of computational models behavior implementing continuous information flow and parallel processing between cognitive stages. Yet, such still typically implement discrete step the last stage motor implementation. This is implemented as fixed decision bound that activation needs cross before action can be initiated. Such an implementation questionable it cannot account for two important features behavior. First, does not allow select while...
The design of robot systems controlled by cables can be relatively difficult when it is approached from the mathematical model mechanism, considering that its approach involves non-linearities associated with different components, such as and pulleys. In this work, a simple practical decoupled control structure proposal requires practically no analysis was developed for position planar cable-driven parallel (CDPR). This implemented using non-linear fuzzy PID classic controllers, allowing...
Significance Many daily-life decisions consist of multiple steps (e.g., go outside, left, arrive at Italian restaurant). We distinguish four prominent models such multistep decision making. further propose a paradigm in two experiments to disentangle these models. Only the implementing additive integration from second- first-step choices were able account for track path movements. Specifically, we find that are initially based on sum/mean second-step future rewards. As information regarding...
Abstract A growing body of behavioral evidence implicates reward prediction errors (RPEs) as a key factor in the acquisition episodic memory. Yet, important neural predictions related to role RPE declarative memory remain be tested. Using novel variable-choice task, we experimentally manipulated RPEs and found support for on level with fMRI. Specifically, demonstrate that trial-specific responses ventral striatum (during learning) predict strength subsequent recollection). Furthermore,...
We propose a general method to break down main complex task into set of intermediary easier sub-tasks, which are formulated in natural language as binary questions related the final target task. Our allows for representing each example by vector consisting answers these questions. call this representation Natural Language Learned Features (NLLF). NLLF is generated small transformer model (e.g., BERT) that has been trained Inference (NLI) fashion, using weak labels automatically obtained from...
Monkey neurophysiology research supports the affordance competition hypothesis (ACH) proposing that cognitive information useful for action selection is integrated in sensorimotor areas. In this view, would emerge from simultaneous representation of competing plans, parallel biased by relevant task factors. This take place up to primary motor cortex (M1). Although ACH plausible environments affording choices between actions, its relevance human decision making less clear. To address issue,...
Currently, a large number of investigations are being carried out in the area robotics focused on proposing solutions field health, and many them have directed their efforts issues related to health emergency due COVID-19. Considering that one ways reduce risk contagion is by avoiding contact closeness between people when exchanging supplies such as food, medicine, clothing, etc., this work proposes use planar cable-driven parallel robot for transport hospitals whose room distribution has...
Large language models are based on machine learning. To be useful for educational purposes, they must also incorporate the field of study cognitive science: Human
Abstract Adaptive sequential behavior is a hallmark of human cognition. In particular, humans can learn to produce precise spatiotemporal sequences given certain context. For instance, musicians not only reproduce learned action in context-dependent manner, they also quickly and flexibly reapply them any desired tempo or rhythm without overwriting previous learning. Existing neural network models fail account for these properties. We argue that this limitation emerges from the fact sequence...