Stefano Palminteri

This study shows that reference point-dependent reinforcement learning, involving relative value encoding, is conserved in humans and rats despite some species-specific differences in behavioural and computational parameters.

This study identifies distinct roles of the parahippocampal and parietal cortex in multi-step reinforcement learning using fMRI data from 28 participants.

This study investigates the role of the thalamus in human reinforcement learning using intra-thalamic electrophysiological recordings and identifies temporally structured oscillations as encoding key learning signals.

This study shows that heartbeat-evoked responses in ventromedial prefrontal cortex influence value encoding during preference-based decisions, contributing to choice consistency.

This study investigates how feedback information affects outcome value coding in medial prefrontal and cingulate cortex, finding that complete feedback promotes relative coding over absolute coding.

The paper investigates how task-set learning relies on unsupervised temporal chunking of stimulus-response associations using behavioral and neural data.

The paper shows that computational noise in reward-guided learning causes behavioral variability in volatile environments rather than reflecting exploration.

This study tests three hypotheses about how imitation influences human reinforcement learning and finds that value shaping is the most supported model.

The paper tests whether learning bias stems from positivity or choice-confirmation, finding stronger positive prediction errors in free-choice conditions.

The paper warns about spurious results in inter-individual brain-behavior correlations due to scaling issues in neuroimaging analyses.

This study investigates how reward awareness affects reinforcement learning by combining computational modeling and EEG, showing that reduced reward visibility impairs learning but increases repetition of previous choices.

The paper investigates how conflict is perceived as a cost in value-based decisions, showing that conflict can influence decision-making and learning differently depending on its source.

This study investigates how depressive symptoms affect reward learning in social contexts, finding a negative audience effect rather than impaired integration of social information.

The paper shows advisers strategically modulate their confidence based on their influence level and relative merit, with distinct brain regions tracking these factors and influencing behavior.

The paper shows that valence biases factual and counterfactual learning in opposite directions, revealing a confirmation bias in human reinforcement learning.

This study investigates how aripiprazole affects counterfactual learning in Gilles de la Tourette syndrome patients, finding it preserves direct learning but impairs counterfactual learning.

The paper argues that model simulation is essential to support claims about behavioral function in computational cognitive neuroscience and proposes guidelines combining model comparison and simulation.

The paper shows that people update beliefs more strongly with positive outcomes than negative ones in a learning task, linking this to enhanced prediction error signaling in reward circuits.