A landmark paper by Randall C. O'Reilly at the University of Colorado Boulder proposes a unified framework for how the neocortex learns, meeting three stringent criteria: computational power comparable to human-level intelligence, algorithmic feasibility using known neural circuits, and detailed neurochemical implementation.
The Research
Published on arXiv (q-bio.NC) on June 7, 2026, the study titled "This is how the Neocortex Learns" outlines a framework called error-driven predictive learning via temporal derivatives. This mechanism is driven by corticothalamic circuits and relies on competitive kinase synaptic plasticity induction. O'Reilly implemented the model in the Axon neural simulation framework using spiking neurons and demonstrated its ability to learn across a wide range of challenging cognitively motivated tasks. The paper specifies that this is the only framework currently meeting all three criteria for a sufficient account of neocortical learning.
Why It Matters
Understanding how the neocortex learns has profound implications for cognitive enhancement. The framework suggests that our brains learn by predicting sensory input and correcting errors via temporal differences, a process that can be strengthened through targeted training. This aligns with evidence-based brain training that emphasizes pattern recognition and predictive tasks.
What You Can Do
To leverage this learning mechanism, engage in activities that require prediction and error correction, such as learning a new language, playing strategy games, or solving puzzles. These exercises enhance corticothalamic circuits and synaptic plasticity.
Source: arXiv q-bio.NC
Curious about your own brain? Take our free adaptive IQ test or try 306 brain training levels.
