Prototype updating, contextual renewal, and the role of reward responsiveness in trust learning | Synapse