Biology subjects

Covelli, M.

Publications and source records attributed to Covelli, M..

3 recordsLinked to original sources

bioRxiv2026

The confidence code: parietal alpha oscillations turn expectations into beliefs

Metacognition, the ability to evaluate whether ones decisions are correct, is crucial for adaptive behavior. However, confidence judgments are not bias-free: they can be systematically modulated by predictive cues and beliefs, favoring expectation-consistent judgments, while leaving metacognitive precision unchanged. Here, we identify parietal alpha activity as the causal mechanism linking expectations to metacognitive bias. In Study 1, EEG was recorded while 75 participants performed a visual detection task with symbolic cues signaling target probability. Cues induced a metacognitive bias in confidence without altering metacognitive sensitivity, and cue-driven alpha modulation over the right parietal cortex predicted the magnitude of this bias. In Study 2 (N = 88), continuous theta-burst stimulation (cTBS) over the right parietal cortex abolished cue-induced alpha modulation, thereby selectively reducing metacognitive bias, while sham stimulation had no effect. Together, these findings demonstrate that parietal alpha-mediated gain control causally shapes metacognitive judgments, revealing an oscillatory code for predictive metacognition.

neuroscience↗

bioRxiv2026

Aging redirects prior expectations from perception to action

Aging profoundly reshapes how the brain integrates sensory evidence with prior expectations during perceptual decision-making. Combin ing behavioural modelling with high-density EEG, we show that older adults assign greater weight on prior information, resulting in stronger decisional bias and reduced sensory precision under probabilistic cues. Neural dynamics revealed a breakdown of anticipatory tuning in visual alpha rhythms, supporting sensory preparation in youth, alongside enhanced central beta activity, indicating a shift toward pre-emptive motor engagement aligned with expected responses. A data-driven decomposition further showed that, whereas prior-driven behaviour in younger adults emerges from a flexible interplay between sensory and executive oscillatory systems, it becomes dominated in aging by a rigid, response-centred mode. Together, these findings uncover a large-scale reconfiguration of predictive computations with age, marking a transition from adaptive sensory-executive coordination to a more automatized, action-oriented predictive strategy.

neuroscience↗

bioRxiv2024

Preparing to act follows Bayesian inference rules

Predictive brain theories suggest that human brain sets-up predictive models to anticipate incoming sensory evidence. Recent studies demonstrated these models to be integrated already in sensory areas, shaping even perceptual outcomes. Here, we hypothesized that this integration process informs the entire functional hierarchy, thus scaling all the way down to the motor system. Operationally, we propose that cue-oriented cortico-spinal excitability (CSE) modulation serves the pre-activation of motor representations aligned to prior-congruent decisions. To this end, 62 participants completed a probabilistic discrimination task while we delivered bilateral single-pulse TMS over the two primary motor cortices (M1s) and recorded motor-evoked potentials (MEPs) to assess motor excitability associated to prior-congruent vs. incongruent actions separately encoded by the two hands. Our findings revealed that prior expectations shaped CSE well before action execution, predominantly by inhibiting the M1 cortex coding for the prior-incongruent action. Importantly, this physiological modulation underpinned the prior-induced bias in participants choices, highlighting the link between motor preparatory modulation and actual decision-making. Furthermore, we observed significant interindividual variability in prior-driven CSE modulations, revealing two distinct predictive strategies: the believers style, who heavily rely on prior, and the empiricists one, who downplay its role, maintaining CSE level mostly unbiased. Crucially, autistic and schizotypal traits drove these differences in prior-driven motor strategy, with believers characterized by schizotypal traits, whereas empiricists displaying autistic-like features. These results demonstrate how predictive models are integrated into action representations and highlight the role of prior-driven CSE modulations as a potential marker able to intercept interindividual differences in predictive styles.

neuroscience↗