Prediction Error — How the Brain Learns
Prediction Error — How the Brain Learns
Prediction error is the difference between what the brain expects and what actually happens. Predictive coding theory proposes that the brain constantly updates its model of the world by minimizing these errors (Friston 2005).
When predictions match reality → processing becomes efficient. When predictions fail → the brain updates its internal model.
🔢 Bayesian Brain Model
The brain combines:
Priors (what it expects)
Likelihood (the current sensory input)
Posterior (updated understanding)
Perception = Prior × Evidence → Updated Belief
🌫 Why Illusions Cause Large Prediction Errors
Illusions exploit situations where sensory input is ambiguous or contradicts expectation. The brain “chooses” the most likely explanation, even when incorrect.