Cortical deviance detection represents a canonical difference signal.

Area of Science:

• Neuroscience
• Auditory Perception
• Computational Neuroscience

Background:

• Neural processing of sensory stimuli is modulated by context.
• Deviating stimuli typically elicit augmented neural responses, often termed "prediction errors."
• These prediction errors may enhance salience, direct attention, or support learning.

Purpose of the Study:

• To investigate whether neural responses to auditory deviants represent a true "prediction error" signal.
• To differentiate between prediction error signaling and generalized surprise or bottom-up signaling.
• To test interpretations within the predictive coding framework.

Main Methods:

• Comparison of neural responses to auditory stimuli across different oddball paradigm variants.
• Analysis of neuronal activity in the primary auditory cortex (A1).
• Utilizing a fixed-deviant oddball paradigm to isolate prediction changes from input changes.

Main Results:

• Neural responses in A1 to auditory deviants contained frequency change information.
• A memory trace of contextual information was observed in neural responses.
• In a fixed-deviant paradigm, neural patterns encoded the standard-to-deviant frequency difference, supporting a true difference signal.

Conclusions:

• Deviance detection in A1 can be interpreted as a sensory prediction error.
• This sensory prediction error reflects the difference between internal predictions and sensory input.
• Findings align with and support the predictive coding framework.