LAUSR

ABSTRACT Reward learning has the ability to bias both attention and behaviour. The current study presents behavioural and neural evidence that irrelevant responses evoked by previously reward‐associated stimuli are more robustly represented in the motor system using a combined go/no‐go and flankers task. Following a colour‐reward association training, participants were instructed to respond to a central target only in a response‐relevant context, while ignoring flankers that appeared either in a high‐value or low‐value colour. The motor cortex and cerebellum exhibited reduced activation to low‐value flankers in a response‐irrelevant context, consistent with goal‐directed response suppression. However, these same regions exhibited similar activation to high‐value flankers regardless of their response relevance, indicating less effective suppression, and the resulting interaction in motor cortex activation was strongly predicted by the influence of the flankers on behaviour. These findings suggest that associative reward learning produces a general approach bias, which is particularly evident when it conflicts with task goals, extending the principle of value‐driven attention to stimulus‐evoked responses in the motor system. HIGHLIGHTSResponse inhibition can be contingent on top‐down goals.This contingent inhibition is attenuated following associative reward learning.Such reduced inhibitory control is evident in the motor cortex and cerebellum.Stimuli associated with high value undermine inhibitory control in the motor system.The results reveal a mechanism for value‐driven biases in approach behaviour.