Hyperlearning Hypothesis: Network disruption and maladaptive learning in schizophrenia

Schizophrenia unfolds through a dynamic course in which perceptual instability, aberrant salience, negative symptoms, and cognitive impairment emerge and interact over time. Existing models have not fully explained how acute disturbances in perceptual inference develop into persistent dysfunction across large-scale neural systems. Here, we advance the Hyperlearning Hypothesis, a mechanistic account proposing that schizophrenia arises from a two-stage disruption in neural learning. First, NMDA/GABA abnormalities impair the precision of prediction-error signaling, leading to unstable perceptual inferences. Second, excessive hippocampal ripple activity drives maladaptive overlearning, which reinforces and stabilizes inaccurate internal models. This cascade links acute perceptual instability to long-term network disorganization and the emergence of chronic negative and cognitive symptoms. Importantly, this framework highlights how aberrant learning signals-rather than static structural deficits-shape the evolution of psychopathology over time. Electrophysiological evidence supports this framework, as individuals with schizophrenia exhibit elevated ripple frequency and power, delayed ripple-initiated network transitions, and reduced clustering and local efficiency within beta-band cortical networks. Together, these abnormalities suggest a progressive disintegration of large-scale templates that normally support stable cognitive function. By integrating computational psychiatry, electrophysiology, and network neuroscience, the Hyperlearning Hypothesis offers a coherent mechanistic account of schizophrenia's evolution and outlines potential avenues for clarifying how learning-related and network-level processes contribute to illness progression.