What resolves this contention: Whether mouse V1 mismatch / prediction-error responses reflect a dedicated predictive-coding error signal (and a specific thalamo-cortical disinhibitory circuit) or instead emerge from the convergence of feature-selective sensory and motor inputs. / We show that violating animals' predictions by an unexpected visual stimulus preferentially boosts responses of the layer 2/3 V1 neurons that are most selective for that stimulus. / Our results indicate that prediction error signals can be explained by the convergence of known motor and sensory signals, providing a purely sensory and motor explanation for purported mismatch signals.
Replication
No replications yet
Discussion
No comments yet — be the first.