A graded neurocognitive model of responsibility that rejects metaphysical free will and reconceptualizes responsibility as a continuous function of cognitive control, affective integration, and predictive stability under contextual constraints. Drawing on neuroscience, neuropsychology, and predictive processing, the paper argues that responsibility is not binary but varies across individuals and states depending on the integrity of regulatory brain systems. The model offers a framework for rethinking responsibility in legal, ethical, and institutional contexts.