Computational mechanisms of subjective confidence in human cortex
Year: 2022
Type: dissertation
Abstract: Confidence is a key component of almost all decisions we make. For example, we only cross a busy street if we are sufficiently certain that we will safely make it to the other side. But how does the brain compute the level of confidence in such decisions?\n\nSensory evidence is typically ambiguous and consistent with multiple interpretations. Bayesian theories of confidence propose that the brain represents such evidence as a probability distribution, and that subjective confidence should reflect the shape or width of that distribution. Alternatively, however, observers might use heuristic (cue-based) strategies to compute confidence.\n\nIn my thesis, I used a combination of computational modeling, psychophysics, and neuroimaging to test these hypotheses and investigate the neural mechanisms that underlie the computation of confidence. My findings suggest that subjective confidence is indeed based on a probabilistic cortical representation of visual information, and that areas in anterior cingulate, insular, and prefrontal cortex play a role in the computation of confidence from such probabilistic representations.
Author: Geurts, L.S.
Language: English
Subfield: Cognitive Neuroscience
Field: Neuroscience
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7036889019